JeanneN.

Knutson
General Editor

PSYCHOLOGY

Jossey-Bass Publishers
San Francisco Washington London 1973
 The
Jossey -Bass
Behavioral Science Series

General Editors
HANDBOOK OF POLITICAL PSYCHOLOGY
Jeanne N. Knutson, General Editor WILLIAM E. HENRY, University of Chicago
Copyright @ 1973 by: Jossey-Bass, Inc., Publishers NEVITT SANFORD, Wright Institute, Berkeley
6 15 Montgomery Street
San Francisco, California 94111
and
Jossey-Bass Limited
3 Henrietta Street
London WC2E 8LU

Copyright under International, Pan American, and

Universal Copyright Conventions. All rights
resewed. No part of this book may be reproduced
in any form-except for brief quotation (not to
exceed 1,000 words) in a review or professional
work-without permission in writing from the publishers.

Library of Congress Catalogue Card Number LC 72-5893

International Standard Book Number ISBN 0-87589-174-8

' N~CRODNI'KNIHOVNA
Manufactured in the United States of America

JACKET DESIGN BY WIL_LTBAUM

1 11~11~~~~~~~11~~11II11~111~~~1~11~~
1000600435

FIRST EDITION

Code 7317
 PREFACE

Behavioral science today is heavily in-

luenced by widespread demands for its increased relevance to the problems of
ociety. The i-apidly growing interest in political psychology can be seen as an attempt,
~ybehavioral scientists in a number of disciplines, to accept that challenge. We have,
a the one hand, a rich and incrementally developed literature covering the social
nd intrapsychic determinants of human behavior. On the other hand, political life
presses our awareness with demands for solutions to intricate problems of violence,
alienation, racism, and despair-problems that crucially affect the quality and the
quantity of human life today, Political psychology, then, is an interdisciplinary effort:
knowledge gained by behavioral scientists is focused on critical human needs, so that
the good life-which is the ancient promise of the polity--can be better realized.
In the past three decades, political psychology has become the professional
interest of scholars in virtually all behavioral disciplines. This growing interest is
reflected in research reports and theoretical essays covering all aspects of political life
as seen through the lens of individual-level determinartts. The student of politics wish-
ing to grasp the major dimensions and accomplishments of this research has had
available no article, chapter, or separate book that conceptualizes the field of political
psychology as a whole and takes a close look at its major subdivisions in a rigorous

vii
viii Preface
Preface
and thorough fashion. He has had no intellectual guide to help him assess what is apprvdLh of the construct which is discussed in terms of available models, but addi-
generally understood about the determinants of political behavior and what large tionally attempts the difficult task of distinguishing the theoretical and empirical
gaps exist in our present knowledge.
usage of his consbruct from that of the other two constructs.
Increasingly, a pressing need has developed among scholars and students to -
From rl. !-
r ~ l l banalysis of basic psychological constructs, the book moves on to
examine political psychology in terms of its intellectual roots, alternative strategies of the manner in which stable orientations toward political belief and activity
consider
inquiry, available models, past research findings, and unmet research needs. From a led and maintained. In this part, three major orienting approaches-sociali-
are forrr
number of directions, it has been urged that those engaged in socially relevant re- :
zaLlulL, authoritarianism, and anomie-alienation-are assessed. These approaches,
I

search step back and examine the priorities, assumptions, and limitations both of which together are responsible for hundreds of research efforts, are summarized and
analytic frameworks and of "objective" methods of assessment. Such an inquiry their contributions are criticized by three contributors whose seminal work in these
requires a major intellectual commitment, and bringing a disparate group of noted areas well exemplifies the social relevance of political psychology.
scholars to make such a commitment is in itself a considerable political and psycho- 1'he Handbook then turns to a discussion of the critical areas in which the
logical feat. The obvious competence of the contributors is well reflected in the
nexus oj[ individual and polity has been assessed in theory and in the field. The
quality of the following chapters, in spite of the enormity of their task of summarizing
three arc3s-leadership; aggression, violence, revolution, and war; and international
and criticizing a subarea in which each has made original contributions and the politics-further reflect the social importance of work in political psychology, as well
lack of available guidelines to assist in this assignment. as its broadly interdisciplinary subject matter. Analysis here indicates additionally
As work on the Handbook of Political Psychology proceeded, the conceptual the dynamic, evolving nature of theory in political psychology and thus well illus-
and methodological inadequacies of much past work in political psychology became
trates the value of the integrative task on which our efforts here are predicated.
obvious, as did the need to gain new perspectives about the ongoing work in various 'he analysis of methods of inquiry in the following part of the book is of
subfields. In response to their dissatisfaction with much of what they reviewed, the Ir value to the student of political behavior because three of the methods
contributors to this volume have jointly made a greatly needed intellectual contribu-
alscussed (experimentation, simulation, and projective techniques) have been seldom
tion by creatively analyzing and critically assessing the issues, the models, and the
utilized in past research, and the remaining two (survey research and psychobiogra-
methods of inquiry employed in the sector of political psychology they discuss. While
phy) have been generally employed with less than optimal rigor. Each of the contri-
each chapter is necessarily written from the author's own vantage point, the reader
butions in this part of the Handbook emphasizes the growing importance of
will find the analysis comprehensive and the tenor-as Fred Greenstein's concluding
methodo logical sophistication in political research, the value of research which
chapter notes-pluralistic.
includes multimethod analysis, and the exciting opportunities for future research
This volume thus provides intellectually sophisticated, original-frequently
whicn.
. . a careful understanding of the potential of each method makes visible.
iconoclastic-analyses of pressing theoretical and empirical concerns. I t should there-
fore prove invaluable to the professional scholar who is making his own contribu- inally, the Handbook concludes with an incisive overview which well empha-
tion to the development of political psychology. The forcefully stated arguments- sizes the: interdisciplinary, broadly conceived approach to the study of political
L-L---!--
those with which he disagrees as well as those he finds congenial-will certainly stimu- aenavlur. which is stressed by each of the collaborators in this volume. The final
late his own intellectual process. The following chapters also make it possible for chapter emphasizes the contributors' consistent awareness of the dangers of psycho-
upper-division and graduate students to acquire a meaningful awareness of particular logical reductionism and intrapsychic overdeterminism and their continuing under-
areas of substantive and n~ethodologicalinterest, and provide a detailed, compre- standing- of the value of multivariate models and complementary methods of inquiry.
hensive bibliography to guide their further research. For those of us who have been In our definition, political psychology weds intrapsychic determinants to political
busily cultivating our own small gardens, the Handbook of Political Psychology r as these processes are intellectually experienced by the various disciplines.
presents an enlightening opportunity to become acquainted with other subdivisions of wnue each of the chapters addresses issues of particular interest to educators, political
political psychology. I t further offers us the chance to examine the methods of scientists,,sociologists, psychologists, and historians, the integrating focus remains a
inquiry to which we have become wedded and makes a concerted effort to move concern with individual-level determinants of political behavior.
research in various substantive areas away from complacent reliance on a single 1'oday, the students of political psychology are legion, and the field is becom-
mode by examining the special advantages and limitations of different methodologies ing inert-asingly international. A different group of these scholars, addressing them-
in the study of political behavior. selves to the same topics, would doubtless have had a number of other points to
The discussion in this book begins with a stimulating integration of the intel- make. T'he thoughtful and authoritative statements offered in the following chapters
lectual traditions which have flowed into political psychology and have provided the should stimulate this necessary further theoretical and empirical work. The success
basis for its distinctive interdisciplinary approach. The Handbook then turns to an of this Iiandbook will be seen in the furtherance of a sense of community, of intel-
examination of the three basic psychological constructs by which political behavior lectual I:Iurpose, and of methods of inquiry by those with a concern for the relevance
is most frequently analyzed: personality, attitudes, and beliefs. Each of the con-
- * .
ot the ~ndividualin the study of political behavior. The attainment of these goals has
wibutors to this part not only delimits in an original statement the special conceptual been already fostered: the interaction of the contributors to this book with one
 Preface

another and with a much wider group of peers who served to clarify and criticize
judgments by sharing knowledge arid viewpoints has begun a process that both re-
fines and elucidates. The collaborators in this first edition of the Handbook of Politi-
cal Psychology hope that it is but a beginning.

Los Angeles
June 1973

CONTENTS

Preface vii
1. Where from and Where to? 1
James Chowning Dauies

ONE : B A S E PSYCHOLOGICAL CONSTRUCTS

2. Personality in the Study of Politics

Jeanne N . Knutson

3. Political Attitudes
M . Brewster Smith

4. Patterns of Political Belief

Robert E. Lane

TWO : FORMING AND MAINTAINING STABLE ORIENTATIONS

5. Political Socialization 117

Richard G. Niemi
 Contents

.Authoritarian Personality in Contemporary Perspective 139

Nevitt Sanford

Anomie, Alienation, and Political Behavior 171

J. Milton Yinger

THREE: N E X U S OF INDIVIDUAL AND POLITY

Patterns of Leadership 203

Daniel Katz

Aggression, Violence, Revolution, and War 234

James Chowning Davies

Assumptive Frameworks in International Politics 261

Herbert C . Kelman, Alfred H . Bloom

FOUR: METHODS OF INQUIRY

Contributions of Psychobiography 296

Betty Glad

Surveys in the Study of Political Psychology 322

Herbert H . Hyman

Experimental Research 356 ALFREDH. BLOOMhas assisted Herbert C. Kelman on a book on the social psy-
John B. McConahay chology of international relations. Bloom is a graduate student in social
psychology at Harvard University, working in Hong Kong on his Ph.D.
Simulation : Attempts and Possibilities 383 dissertation in the area of individual conceptions of political authority. He
Rufus P. Browning graduated summa cum laude in romance languages from Princeton Uni-
versity in 1967, spent a year studying international politics in France on a
The New Frontier of Projective Techniques 413 Fulbright-Hays grant, and received his masters' in social psychology from
Jeanne N . Knutson Harvard University in 1971.
RUFUSP. BROWNING is a political scientist at the University of California, Davis.
He received hi doctorate from Yale University and has taught at the Uni-
CURRENT PERSPECTIVES versity of Wisconsin, the University of California, Berkeley, and Michigan
State University. His current research includes a computer simulation of
Political Psychology : A Pluralistic Universe 438 political bargaining and a study of decision-making in experimental schools.
Fred I . Greenstein
JAMES CHOWNING DAVIES is professor of political science at the University of
471 Oregon and has taught at the California Institute of Technology and the
References University of California, Berkeley and Los Angeles. He has held fellowships
from the Rockefeller Foundation, the Social Science Research Council, and
Author Index 526 the Carnegie Foundation. His interests in political psychology range widely,

Subject Index 538

xiv Contributors Contributors xv

with major emphasis on socialization and revolution. PIis current interest is Study of Social Issues and of Division Eight of the American Psychological
in constructing a political theory concerning the initiation, growth, stabiliza- Association, is vice-president of the International Studies Association, and has
tion, decay, and overthrow of governments. served on numerous editorial boards, including those of the Journal of Per-
sonality and Social Psychology, the International Studies Quarterly, Psy-
BETTYGLADis associate professor and head of the Department of Political chiatry, and Human Relations. He was the recipient of the Kurt Lewin
Science at the University of Illinois at Urbana. She is on the council of the Memorial Award for 1973. His research interests include social influence
Midwest Political Science Association and is vice-president of the Interna- and attitude change, nationalism and political ideology, international rela-
tional Studies Association, Midwest. She has served on the editorial board tions, and ethics in social science.
of PS and chaired the Committee on the Status of Women of the Midwest
Political Science Association. Her primary research interest has been the im- JEANNE N. KNUTSONis senior research associate in psychology and political
pact of modal psychological and ideological structures on American foreign science at The Wright Institute in Berkeley and member, Psychiatry Depart-
policy-making. ment, Ross-Loos Medical Group, Los Angeles. She has a Ph.D in political
science and is completing a Ph.D. in psychology (social psychology and psy-
FREDI. GREENSTEIN is Henry Luce Professor of Politics, Law and Society at chopathology). She has had internships in clinical psychology at Cedars-
Princeton University; he was previously professor and chairman of the De- Sinai Medical Center, Los Angeles, and at the Los Angeles Psychiatric
partment of Government at Wesleyan University. He is on the editorial Service. She has held pre- and post-doctoral research fellowships from the
board of the British Journal of Political Science and the Journal of Youth National Institute of Mental Health and is presently studying black and white
and Adolescence and was previously on the editorial board of the American patterns of political socialization under a research grant from the Society for
Political Science Review. He is a former member of the Political Science the PsychoIogical Study of Social Issues. Her current work includes an anal-
Advisory Panel of the National Science Foundation and a member of the ysis of the long-term effects of personality on political belief and activity, the
Committee on Pre-Collegiate Education of the American Political Science use of projective techniques in the study of political socialization, psycholog-
Association. His primary interests within political psychology have been ical models of political recruitment, and a case study of visualization in the
socialization and personality research. treatment of hysteria.

HERBERT H. HYMAN is professor of sociology at Wesleyan University; previously, ROBERTE. LANEis professor of political science at Yale University and past
he was on the faculty of Columbia University for many years. His long and president of the American Political Science Association. He has received
varied experience in survey methods includes the position of associate director fellowships from the Fund for Advancement of Education, the Social Science
of the Columbia Bureau of Applied Social Research, membership in the Research Council, and the Center for Advanced Study in the Behavioral
National Opinion Research Center at the University of Chicago, and the Sciences, and has been given awards by the Ford Foundation, the Carnegie
position of program director of surveys for the United Nations Research Foundation, the Social Science Research Council, and the Woodrow Wilson
Institute. He is on the editorial board of the Public Opinion Quarterly and International Center for Scholars. He was a senior Fulbright-Hays scholar in
is associate editor of Sociological Methodology. He is past president of the 1972-1973. He has published numerous works in political psychology, with
American Association for Public Opinion Research. special emphasis on ideology, beliefs, and opinions.

DANIEL KATZis professor of psycholcgy and program director of the Survey JOHN B. MCCONAHAY is director of the Psychology and Politics Program in the
Research Center at the University of Michigan. He is on the editorial board graduate school and assistant professor of political science and psychology at
of the Journal o f Cross-Cultural Psychology and has served on the editorial Yale University. He has been a National Institute of Mental Health research
boards of numerous other journals, including Human Relations, the Journal fellow and was a research assistant on the Los Angeles Riot Study carried
of Social Issues, and the Pournal of Conjlict Resolution, as well as being editor out by the University of California, Los Angeles. He has published articles
of the ]ournal of Abn,ormal and Social Psychology and of the Journal of and a book on urban violence, and he is on the editorial board of the Journal
Personality and Social Psychology. He is a past president of the Society for of Conflict Resolution. His current research and writing are in the areas of
the Psychological Study. of Social Issues and of Divisions One and Eight of political ideology and the effects of human crowding upon aggression and
the American Psychological Association. Among his numerous honors is the political socialization.
1965 Kurt Lewin Memorial Award for outstanding contributions to the
integration of social research and social action. He has made major con- RICHARD
G. NIEMIis associate professor of political science at the University of
tributions in the areas of social psychology, personality, and attitudes. Rochester. He has received fellowships and research grants from the National
Science Foundation, the National Institute of Mental Health, and the Ford
HERBERT
C. KELMANis Richard Clarke Cabot Professor of Social Ethics at Foundation. His major research interests are probability models of collective
Harvard University and was previously professor of psychology and research decision-making and political socialization. In the latter area he is curreiltly
psychologist at the Center for Research on Conflict Resolution at the Uni- working on a panel study of political change and development over the eight-
versity of Michigan. He is past president of the Society for the Psychologicall year period after high school.
xvi Contributors

N E ~ SANFORD
T is scientific director of The Wright Institute in Berkeley and was
previously professor of psychology and director of the Institute for the Study
of Human Problems at Stanford University. Before that, he was professor of
psychology at the University of California, Berkeley. He is past president of
the Society for the Psychological Study of Social Issues and of Division Eight
of the American Psychological Association. He received the 1970 Kurt Lewin
Memorial Award for outstanding contributions to the integration of social
HANDBOOK
research and social action and has been on the editorial boards of numerous
journals, including the Journal of Social Issues and the Journal of Abnormal
and Social Psychology. His major interests include socialization and per-
sonality development during the college years, authoritarianism, and social
action research.

M. BREWSTER SMITHis profe~sorof psychology and vice-chancellor of the Divi-

sion of Social Sciences at the University of California, Smta Cruz. He was
formerly the chairman of the Psychology Department at the University of
Chicago and was director of the Institute of Human Development and pro-
fessor of psychology at the University of California, Berkeley. He is past pres-
ident of the Society for the Psychological Study of Social Issues and was
editor of the Journal of Abnormal and Social Psychology and the Journal of
Social Issues. His major research interests currently lie at the boundary of
personality and social psychology, concerning the self in experience and action.

J. MILTONYINGERis professor and chairman of the Department of Sociology and

Anthropology at Oberlin College. He has been on the editorial boards of
Social Problems, the American Sociological Review, and the Journal of Con-
flict Resolution. He is the secretary of the American Sociological Association.
From 1965-1970, he held a rqearch grant from the Office of Education to
study the effects of educational intervention on the culturally disprivileged.
He has been a Guggenheim Fellow and, in 1958, received (with G. E. Simp-
son) the Anisfield-Wolf Award in Race Relations for the best work of non-
fiction in race relations. He has made major contributions to the sociology
of religion, to the study of race and minority group relations, and to develop-
ing a field view of personality, areas of interest in which he continues to work.
 WHERE FROM AND
WHERE TO?
James Chowning Daaies

]EDlitical science in America started in the

late nineteenth century. It began as the study of institutions, with such influential
works as Woodrow Wilson's Congressional Government, published in 1885. In its
early decades the new discipline largely ignored two basic phenomena: the effect of
people on political institutions (and vice versa) and the political effect, on both
people and institutions, of rapid social change and instability.
After World War I, however, while most American political scientists re-
flected the public mood of turning away from the unsettling and the unpleasant,
some young professionals began to seek understanding of the forces inside people's
heads that help to establish, maintain, or disestablish political institutions. In the
1920s, these young political scientists ventured into what their respectable peers and
elders looked down on as the ill-reputed house of psychology. Companionships
developed. A few progeny appeared. Then, after World War 11, in the 1950s, the
liaisons and their issue became accepted, institutionalized, and even sanctified. There
was a baby boom: hundreds of studies in the behavior of the public-in stable
democracies.
 James Chowning Daviei Where From and Where To? 3

Studies of political behavior were typically legitimated only when they ana- ~
b rh
luy ,
n' -n- t ~however,
r the basic premise is that every one of the enormously important
lyzed voters and lobbyists and legislators and administrators who behaved as rational environmental forces depends, for its effect on behavior, on the existence of a
bargainers in the familiar political marketplaces. Unions between political science selectiursly responsive organism. Because of its enormously intricate nervous and
and psychology that produced analyses of political misbehavior-when people "vote" endocrine systems, the human organism interacts in very precise ways with specific
with their feet, their lungs, and even with sit-ins and guns-remained illegitimate. aspects of the environment. If it did not, it would behave in a randomized, entropic
Even in the 1960s and 1 9 7 0 ~after~ decades of wars, revolutions, rebellions, and nonpattern comparable to that of marbles being shaken in a box or to the nonpattern
riots, the predominant trend remained the study of stable, safely observable, and
'
of molecules of gas bouncing against each other inside a closed container.
easily quantifiable phenomena. Political behaviorists still largely ignored the causes The work of the many investigators who have contributed to the development
and effects of deep changes that alter the behavior of people in turbulent times. of politically relevant psychology is divided into three main categories in this chapter:
The long-range prospect of continuing this neglect is that citizens and leaders (1) theory and research emphasizing the organism as the major determinant; (2)
will continue to face new problems with the same shock of ignorance that numbed theory iand research emphasizing the close social environment in its direct, proximal
their reactions to World War I and to revolution in Russia, Latin America, and Asia influenc:es on the behavior of individuals and groups; and (3) theory and research
after both world wars. The short-range prospect is that political science will lose emphasizing the broad social environment in its indirect, mediated, distal influences
both its readers and its followers. One of the major purposes of this volume and this nn sucn -7
behavior. In addition, there is a summary of some of the work on the central
chapter is to demonstrate the enormous potential fruitfulness of more frequent and systems (the nerves and the endocrines) that has been done by physiologists,
deeply involved and meaningful relationships between psychology and political )ter 9. This latter work will some day be directly relevant to an explanation of
science. ,. ,,,,or aspects of political behavior, but at present it is only directly relevant to
Natural scientists have learned to control floods and epidemics by under- ing violence.
standing their causes. When social scientists pay less exclusive attention to healthy The basic distinction between kinds of environmental influences is significant
bodies politic, they too will be able to diminish the suffering and destruction that arlu ~ a l for s clarification before we go farther. Proximal influences, including group
accompany profound and rapid social change. The contributors to this book seek to influences, are those in which the influencing person is directly in contact with the
call attention to more fundamental causes, to more generally significant political selectively responding human organism, the human individual. The actions-verbal
phenomena, than just those that occur in calm times in calm countries. I t is the and otherwise--of parents, spouses, friends, teachers, preachers, and speakers at
minds of men that-either with calmness or agitation-resist change or welcome it. political rallies are proximal influences. Each of the persons with whom the
We can fruitfully look for the fundamental causes in the minds of men. individual is in direct contact is a proximal influencer. Distal influences are those
which are carried to the individual by a proximal influencer. Churches, trade unions,
B = f (OE) social classes, economic and political institutions provide distal influences, and they
are always mediated by a person who is in direct, proximal contact with the indi-
The discussion in this and many later chapters is organized around a basic vidual. A workingman, for example, may identify with the working class generally,
behavioral. equation: all observable behavior, including political behavior, is a but in every case the identification is mediated by people in circumstances with whom
function (or product) of the interaction between the organism and the environment. he (more exactly, -his highly selective and-by the time of early adulthood-highly
B = f (OE) . The equation is by no means novel and would indeed be a banality if conditioned sensory, memory, and response system) is in direct contact. Nonpersonal
it had not been neglected or garbled. The present form of the equation differs from influences fall similarly into the proximal and distal categories. An individual who
earlier ones (like that of Lasswell and Kaplan, 1950, pp. 4-6) in that it explicitly sees a beautiful view while hiking in the mountains is being proximally influenced.
locates nonenvironmental determinants in the human organism. For a consideration An individual who reads a book about mountains is being distally influenced. The
of the basic equation and its political implications, see Davies, 1963, ch. 1; in the major preoccupation in this chapter is with those proximal and distal-those direct
present chapter the term situation is replaced with the more inclusive environment. and 11iediated-influences that are between people.
In some writings, the organic part of the equation is left unnecessarily vague,
by such terms as unconditioned response, predisposition to respond, or nonenviron- Human Organism as a Determinant of Behavior
mental determinants. These terms avoid saying what these responses, predispositions,
and nonenvironmental determinants are and where they come from. Some writers General interest in the social significance of the human organism developed
altogether ignore the organic part in their preoccupation with environmental in- somewhat earlier than did political science as a discipline. The major starting point
fluences under such terms as stimulus, training, conditioning, socialization, politiciza- was the publication in 1859 of Charles Darwin's Origin of Species, which sought to
tion, and personality formation. In many analyses, even such self-evident organic demonstrate the organic continuity between men and other species. This argument,
needs as food, rest, sex, and stimulation become undifferentiated parts of the activity like the prior argument that the earth is not the center of the universe, was a shock
of human beings, who thereby become indistinguishable from other animals. Tn this to human egocentricity. Only a generation after this shock to men's egocentricity came
 James Chowning Davies Vhere I7rom and Where To? 5
nner, ro their rationality. This was the assertion that meds behavior is almost fore, if social harmony is to prevail, a careful balance must be maintained between
ally controlled by forces of which they are unaware. That is, most of the conscious overspecifying and overgeneralizing the sense of empathy.
sons men give to explain their behavior are rationalizations of unconscious forces. T'he most striking difference between Freud and Adler is their opposing view
Sigmund Freud (1856-1939) remains the great pioneer in this effort to place ,-
01 ~nrenlalconflict. Freud described the conflict as being fundamentally between
men mentally in their natural organic context, as Darwin has placed them physically internal forces within the individual-the life instinct (Eros) and the death instinct,
in that context. The work of most of the psychologists discussed in this section- which we will consider in Chapter Nine as it relates to Freud's analysis of war. Adler
including those who stressed the environment-derives from the genetic psychology argued
-
that internal conflict results from the failure of the environment to properly
tha )ased in the human organism, notably in the glands and the brain. nurture the individual's social feeling. Properly nourished, this feeling can cause
1 conceptualized mental processes as a threefold complex: the id, the everyone to "feel himself bound to his fellow man" and so to internalize as part of
ego : superego. The id (the libido or Eros) is the force that impels an in- himself
-- "the welfare of humanity."
divldual to create and procreate, the instinct that seeks life and is the energy for life's 0In the genetic, psychoanalytic grounds established by Freud, Erich F r o m
development and perpetuation (Freud, 1933). These libidinal, erotic forces emerge (b. 1900) developed and expanded the analysis of broad social and political phe-
whether or not the individual tries to deny them or is even conscious of them. The nomena that Freud (1922, 1930) had undertaken only in his sixties (long after he
ego is the force that enables him to become aware of the libidinal forces and exercise - .
had established his major theories) and that Adler (1927) had only begun to sys-
some control over them. tematize. However, Fromrn sidestepped a consideration of innate determinants. He
This potential for control, however small, is crucial. Because it exists, men can .
said that "although there are certain needs . . . common to man, . . man's na-
live in some kind of harmony with themselves and with their fellow men. Indeed, the ture, his passions, and anxieties are a cultural product" (1941, pp. 12-13). Influenc6d
libido is the basis for not only the sexual ties between people but also-when the possibly as much by Marx as by Freud, Fromm focused on the socia1 and cultural
sexual goal is inhibited and controlled-for social and political ties. Although matrix that produces mental crises in entire peoples, entire civilizations.
civilization produces deep anxieties when it frustrates the natural desire for sex, In Fromm's view (1941), the central crisis facing Western civilization results
civilization also performs the necessary function of so inhibiting that desire by de- from the interaction of two tendencies within a developing individual: "growth of
flecting it from sex to other activities as to make society and culture possible (Freud, self-strength" and "growing aloneness," which together produce the dilemma of
1930). Civilization is a necessary external source of control. independence and interdependence. This dilemma within the minds of individuals
This external control gradually is internalized by the individual on whom his is improperly managed in the modern world with its intense frustrations, and tends to
parents, as his ideal figures, exert their influence aild that of civilization, whose agents produce what Fromm called the authoritarian personality. Such a personality type,
the parents are-in our terms, the proximal influencers who mediate social influences. which he said became very prevalent, for historic and contemporary reasons, in the
The mother and the father give affection when the growing child acts in accordance German lower middle class, exhibits strong feelings of "powerlessness, anxiety, and
with parental (and also, invariably, social) restrictions and demands; they withdraw isolation" and as a result develops a sado-masochistic "love for the powerful" and
affection when the child does not. This internalization, which Freud called the "hatred for the powerless." Fromm also included within this personality type some
superego and equated with conscience, is a necessary part of the establishment, less central characteristics: pettiness, hostility, and asceticism ("thriftiness with feel-
maintenance, and advancement of civilization. In a continuous interpersonal, inter- ings as well as money") (pp. 21 1-2 12) .
generational process of interaction both individuals and civilization develop and William McDougall ( 1871-1 938) was a British contemporary of Freud, but
perpetuate themselves. was influenced by the earlier American psychologist William Jam& (1842-1910),
One of Freud's original disciples, Alfred Adler (1870-19371, broke with particularly by his Principles of Psychology, which appeared in 1890. Like Freud and
Freud because of what Adler regarded as Freud's total emphasis on sex and his James, McDougall began his professional career on a base of rigorous laboratory
rather tardy concern for the social influences on behavior. In Understanding Human training in physiology. But his major contribution, again like that of Freud and
Nature (1927), Adler insisted that sex is only one of many manifestations of men's James, was no direct product of the laboratory and also bore no evident relationship
desire for superiority, which includes a 'cstriving for recognition." When this desire to any influence of the Vienna psychoanalytic school. McDougall's Introduction to
is frustrated, individuals develop a neurosis, an "inferiority complex." They can rid Social Psychology (1908), a mainly theoretical work, became enormousIy influential
themselves of this inferiority feeling by developing their "social feeling," their sense in both England and America.
of community with their fellow men. This social feeling and the empathic identifica- McDougall listed seven major instincts, to each of which he conjoined an
tion with others are innate tendencies not necessarily related, in Adler's emphatic emotion. The combinations are these: flight and fear, repulsion and disgust, curiosity
view, to sexual impulses. Empathy makes it possible for people to be influenced by and wonder, pugnacity and anger, self-abasement (or self-assertion) and negative (or
others,, sometimes excessively so, particularly during childhood by parents. This positive) self-feeling, and finally the parental instinct and the tender emotion. In
"authoritative influence" can instill in individuals a habit of unreasoned obedience, what seems more like an afterthought, he added other instincts to the seven major
even to public authority figures. Empathy, authority, and thus unreasoned obedience ones: reproduction (sex), eating, gregariousness, acquisition, and construction.
can give rise to race hatred (ethnocentrism), capital punishment, and war. There- The innovative aspect of McDougall's list is its considerable length. He
 6 James Chowning Davies Where From and Where To? 7
avoided, as had James, reducing innate predispositions either to the self-preservation the hard work, the frugality, the confident future orientation toward production of
instinct, which is the solitary instinct lying at the base of Social Darwinism as it material wealth, the social exclusiveness of the elect--occurs not just in Protestant
derives from Darwinian biology, or to the sexual instinct, which Freud said was the but also in Catholic, Communist, Occidental, and Oriental nations. The basic com-
one basic life urge-although Freud broadened its definition to make sex mean the mon phenomenon is the need for achievement, and McClelland relates this need to
same thing as all the meanings of love. One can reduce innate drives to Eros (and the process of modernization. His behavioral research shows the inadequacy of
the death instinct) as Freud did, or leave the question vague, as Fromm did. But institutional analysis. It is not capitalism or socialism or even protestantism that
either reducing or evading the issue of innate drives makes it more difficult to explain spontaneously activates the need for achievement. This organically based need, latent
such profound, earth-shaking phenomena as revolution and such seemingly self- in a premodern society, becomes active whenever a favorable environment develops,
evident phenomena as political stability. whatever ideology or institutional structure may accompany its development.
An even more extended listing of basic needs was developed by Henry Alex- Abraham Maslow (1908-1970) generated a list of basic drives (194313) that
ander Murray (b. 1893) in his Explorations in Personality (1938). Borrowing and is some 1;ind of midpoint between reductionism and elaborationism. I t is more appli-
integrating ,ideas from Freud, Jung, Adler, James, McDougall, and others, the book cable to political analysis than Freud's very brief or Murray's very long list. Maslow
marks some kind of ultimate in specification of innate behavioral tendencies. Murray specified five categories of needs; physical, security, social-affectional, self-esteem, and
defined need as "an organic potentiality or readiness to respond in a certain way self-actualization needs. More important, however, for social and political analysis
under given conditions. [It is] a latent attribute of an organismyy(p. 61). He states is his assertion that the basic needs are hierarchically ordered. When an individual
that all needs are complex electrochemical processes taking place within the central is deprived of a need in the first category (of physical needs), this condition will
nervous and endocrine systems. But he separates needs into two main classes: primary activate the central control systems so that he will pursue its satisfaction before re-
or "viscerogenic" and secondary or 'cpsychogenic," the latter having "no localizable turning to any other (nonphysical) activity. If an individual is playing a game,
bodily origins." In regard to this latter category he implicitly acknowledges the non- reading, or working at a job that is either physically or mentally tiring, and gets
existence of physiological research that can help establish the organic loci of perhaps, tired, he will stop the play or the work and rest. Furthermore, not until the physical,
for our purposes, the most important of the organic tendencies to respond. security, and social-affectional needs are at least minimally satisfied will the individual
The viscerogenic needs are inspiration (of air), water, food, sex, lactation, seek satisfaction of his self-esteem needs. A person will suffer a variety of degradations
expiration (of air), urination, defecation, harmavoidance, heatavoidance, cold- rather than go without food, clothing, and shelter or without the company of those
avoidance, noxavoidance, and sentience (for sensuous stimuli). The psychogenic needs close to him. And only when there is a relatively steady satisfaction of at least some
are acquisition, conservance, order, retention, construction, superiority, achievement, of his self-esteem needs will he be able easily to pursue his self-actualization needs.
recognition, exhibition, inviolacy, infavoidance (need to avoid failure, ridicule), The idea of priority, of hierarchy, makes it possible to explain patterns of
defendance, counteraction, dominance, deference, similance (need to identify with political behavior that otherwise seem irrational. To give a rudimentary example,
others), autonomy, contrarience (need to be unique), aggression, abasement, blam- an individual who is very hungry is unable to turn his attention to political concerns
avoidance, affiliation, rejection, nurturance, succorance (need to receive aid), play, at the very time when his political action may get him some food. An individual who
cognizance (need to explore, to satisfy curiosity), and exposition (need to explain, feels unaccepted and unloved will be too preoccupied with his social-affectional needs
interpret) (pp. f 7-83). to be able to develop a concern for politics. If he does turn to politics, he quite likely
Murray considers all these needs to be specific "pushes" from within that are will become excessively demanding of affection, either from his followers (on Hitler,
directed in an orderly way toward something whose achievement is pleasant or away see Langer,,1972) or from his beloved leader. Empirical support for the hierarchy
from something that is unpleasant. Correlative to the organic need is the environ- will be touched on in discussing the research of Harlow, Spitz, and Bowlby on emo-
mental "press," which is "a threat of harm or promise of benefit to the organism" tionally deprived monkeys and children. The political implications of the hierarchy,
(PP. 40-41 ) both theoretically and empirically, have been explored by Davies (1963, ch. 1-2;
If Freud may be accused of reductionism-of oversimplifying the distinguish- 1972), Aronoff (1967), Inglehart (1971b), and Knutson (1972a).
able basic driving forces-Murray may be charged with elaborationism. The scheme There does appear to be an incongruity in the appearance of security as one
is really to elaborate to be readily usable as a coherent set of drives, at least to explain of the need categories along with the physical, love, esteem, and actualization needs.
political behavior. Its boldness lies in denying oversimplification, in diminishing the Security seems to be related to the other needs in an instrumental way: one seeks to
tendency to reduce everything to a simple, general (but, neverthless, directional) gain a sense of security in the pursuit of his physical, love, esteem, and actualization
energy state, as in Freud's concept of the libido. needs; one does not normally pursue security for its own sake (Davies, 1963, ch. 1;
Starting with research on one of the needs on Murray's list, David McClelland 1970, pp. 617-618; 1972; Kagan, 1972, p. 55). Furthermore, the need system becomes
'(b. 1917) produced the first psychologically empirical attack (1961) on a problem more coherent if two other instrumental needs-knowledge and power-are listed
that had hitherto been analyzed sociologically and for the most part only theoretically with security. So conceived, the instrumental needs for security, knowledge, and
by Max Weber (1904) and R. H. Tawney (1926). This problem is to explain what power are akin to these in Murray's list: superiority, defendance, counteraction, domi-
Weber called the Protestant Ethic. According to McClelland, the Protestant ethic- nance, aggression, cognizance, and exposition. While the instrumental needs do not
8 James Chowning Davies Where From and Where To? 9
as such appear in Maslow's need hierarchy, not only security but also knowledge and ipatory response to food, by salivating, when ondy the conditioned stimulus was
power are convenient and probably necessary units of analysis for both theory and presented.
research in political behavior (Davies, 1972). Pavlov's research-which was well endorsed by the postrevolutionary Soviet
From Freud to Murray and Maslow, the primary emphasis among motiva- reglme, with whose environmental determinism Pavlov's views were compatible-
tional psychologists as here described has been on the organism. But Adler and laid the groundwork for much modern learning theory and for a broad "objective"
Frornm, as we have noted, strongly emphasized the environment. Another psychiatrist, line of research that has argued the great manipulability-and the eventual emanci-
Harry Stack Sullivan (1892-1949)' stressed the environment even more, at times pation--of all mankind. His presuppositions about manipulability were c~nsistent
arguing that the individual must adjust to society instead of expecting society to with the optimistic view associated with environmental determinists-from John
adjust to him. In Sullivan's view, the neurotic is fundamentally someone who has Locke (for whom the human mind was a blank sheet, on which society wrote what
failed to achieve satisfactory interpersonal relations. The mentally healthy person, it chose to write) to Karl Marx (who argued that the human prsducts of the immiser-
according to Sullivan (1940), is the one who has adjusted his behavior to the norms ating capitalist system could be emancipated quite simply if that system were
of the society he lives in. The "radical," in contrast, has adjusted only to his ingroup;
changed).
he and his group are paranoid with respect to the outside world, and the ingroup John B. Watson (1878-1958), an American successor to Pavlov, virtually
norms serve to justify and organize the destructive impulses of the radical. eliminated central nervous activity from Pavlov's sequence of stimulus, neural
This ends our discussion of psychologists-from Freud to Sullivan-who processes, and response. Watson became almost totally preoccupied with the observ-
heavily emphasized the central nervous system as the entity that, in various ways and able. In his view, hunger is a set of stomach contractions and thought is a series of
amounts 'influences-or must adapt to-the environment. In the next section we muscle movements. He said at one point, "Give me a dozen healthy infants . . . and
move to a discussion of work that most heavily emphasizes the environment and its I'll guarantee to take any one at random and train him to become any type of spe-
ability to control the organism. cialist I might select-doctor, lawyer, artist, merchant-chief, and yes, even beggar-
man and thief, regardIess of his talents, penchants, tendencies, abilities, vocations,
Environment as Determinant: Direct, Proximal Influence and race of his ancestors" (Watson [1924], 1961, p. 104). Although he did recognize
some "unlearned original responses," including fear, rage, and love, he preferred those
Among the studies of direct, proximal influences we here include the work of
that require no abstract classification because they are readiIy observable-responses
stimulus-response psychologists, situationalists, and developmentalists. such as sneezing, hiccoughing, crying, penis erection, urinating, defecating, eye move-
Stimulus-Response Psychology. The grandsire of all environmental deter- ments, limb and torso movements, smiling, feeding, crawling; walking. Edward
minists in psychology is Ivan Petrovitch Pavlov (1849-1936). He was an exact con- Chace Tolman, the distinguished motivational-behavioral psycholoeist, called this
temporary of Freud, the grandsire of' all genetic, organic determinists. Freud approach "muscle-twitch psychology."
emphasized the role of organic (mainly sexual) determinants but by no means The irony is that Watson was as much at home within American capitalism
neglected the role of-the environment (in the form of socially derived repressions as Pavlov was within Soviet socialism. Without some criteria-whether derived from
and suppressions). Pavlov heavily emphasized the environment but did not neglect a structured view of human nature or a set of political ethics-it is manifestly possible
the organism. For him, behavioral events proceed from stimulus to (organic) neural to condition stimulus-response psychology to any ideology, any social and political
processes to response. system.
In Pavlov's view, instincts are simply reflexes, reactions to internal and ex- The most recent theorist among stimulus-response psychologists is somewhat
ternal stimuli. In addition to the more basic instinctual reflex, the hunger drive, less like Watson in his dogmatism and more like Pavlov in his distinction as a scien-
however, he also observed the "freedom" and "investigatory" reflexes in his dogs- tist. And he is perhaps more polemically inclined than either. B. F. Skinner (b. 1904)
their dislike of the harness that kept them in the experimental apparatus and their designed what has become known as the Skinner box, the container in which an
alertness to a variety of stimuli that had nothing to do with the hunger drive. In this experimental subject (a pigeon or a rat) becomes an active participant in gaining
respect, Pavlov evidently anticipated two of the needs in Murray's list-the need for its reward. Instead of learning (as Pavlov's dogs did) that he will be fed when a bell
autonomy and the need for cognizance. rings, a subject learns that if he does what he is expected to do (like pushing a bar)
Pavlov's great contribution, summarized in his Conditioned Reflexes (1927); he will be rewarded. The technique is called operant conditioning and tends to
lies in his demonstration that higher vertebrates (and by implication men) can learn integrate positive action on the part of a subject with positive reward, thereby
-be trained, conditioned-to associate a symbolic stimulus ( a conditioned stimulus) producing the response which both experimenter and subject desire.
with a nonsymbolic stimulus (an unconditioned stimulus). He trained dogs first to Skinner frankly faced the social and political implications of his stimulus-
expect a food reward (the unconditioned, natural, real stimulus) whenever they response psychology. In Walden T w o (1948), he argues that people generally do not
were placed in the laboratory apparatus. He then trained them to anticipate the want to make their own decisions and prefer to be governed by those who understand
reward by activating; in advance of the reward, a bell, buzzer, metronome, or mild them. These governors, in turn, must be skilled enough to be able successfully to
shock (the conditioned stimulus). When fully trained, the dogs would make an antic- condition people to accept the social harmony and good life that they really want
 10 James Chowning Davies 'here Fro]m and Where To? 11

but do not know how to get. In Beyond Freedom and Dignity (1971), Skinner backs .. ljehaviorists focus on the effects of uncomplicated signaling stimuli that
,nsnlps. 7,.

off, at least in emphasis, from the extreme elitism of Walden T w o . He nevertheless n condition behavior when the subject learns to associate a reward with the signal.
insists that people think they are free when in fact they are only unbuttoned-which euationalists consider the complex and powerful influences of those small chunks
is quite different from saying they think they are free when they are conditioned to society that we call face-to-face or proximal groups, in contacts that are of rela-
social control. They would do well, in his view, to rid themselves of the notion that titrely brief Iduration (Smith, 1971) .
freedom and dignity, as they have beeiz written about, really have much to do with One of the most innovative pieces of research in this category was done on
the good life. -- e basis oi[ the autokinetic effect, the innate tendency of the brain to perceive a
thc
Skinner (1971) depreciates any psychology that uses such concepts as goals, stationary point of light as moving, in a totally dark room, when there is no other
intentions, or purposes. "We find the purpose of the skilled movement of the hand stimulus to serve as a reference to the point of light. Sherif (1936, 1947) added a
in the consequences that follow it." And "we do, indeed, feel things inside our own social dimension. He placed an innocent subject in a darkened room, together with
skin, but we do not feel the things [like purposej we have invented to explain behav- a stooge of the experimenter, the stooge being told to say that the point of light had
ior. . . . We do feel certain states of our bodies associated with behavior, but . . . mcwed a certain number of inches. The amount of movement that innocent subjects
they are by-products and not to be mistaken for causesJJ (pp. 12-14). But after thc:n reported seeing tended to approximate the amount mentioned by the stooge.
vociferously kicking intent and purpose out the back door, Skinner welcomes them S h e ~ *I.-Ll L~L C I C ~ ~demonstrated that social reference groups can influence perceptions.
quietly at the front door. "The intentional design of a culture and the control of The enormous power of a group to distort perceptions was explicated in
human behavior it implies are essential if the human species is to continue to developJ' ex1periments undertaken by Asch (1952). College students, who were presumably
(p. 167). Two pages later he uses the word again, but in quotation marks. "What is rather mature and surely above average in autonomy, were placed, usually alone,
needed is more 'intentional' control, not less, and this is an important engineering with a group of students who were the experimenter's stooges. The stimulus in this
problem" (p. 169). case was not a point of light but a set of lines whose length the innocent subjects were
However, changing from his Walden T w o position, Skinner in his second asked to compare with a standard, criterion line. When quite alone, subjects were
polemic does not consistently advocate control by an elite group but rather by the le to judge almost perfectly which of a set of comparison lines was the same length
individual himself. His criteria for control are what is "naturaI" for developing the criterion line. When they were in the group of stooges, however, and the
human beings, but he does not explain how anything can be natural to a being with loges unanimously agreed that a line-which in fact was not the same length as the
no definable nature. We do not know very much about the organism, he says, but criterion line was the same length, innocent subjects made errors in about a third of
eventually physiology will tell us more about what goes on under the skin. In any all their judgments-errors in the direction of their conniving fellow group members.
event, it is evident from his manner of writing that Skinner likes people and dislikes (See Chapter Three in this volume.)
frustrating environments-aversively conditioning environments, as he calls them. Extensive studies by some Yale psychologists (Hovland, Janis, and Kelley,
This is sophisticated and benign behaviorism, having as its purpose the 53) spell out a variety of relationships involving the interaction between communi-
eventual removal of external controls, as the individual moves toward autonomy :on and communications on the one hand and opinion change in communicatees
(Hall, 1972). Such behaviorism is far removed from Watson's incredible belief in on the other. One of these studies examines the effect of the communicator on ,the
the near-total manipulability of humans. Supposedly Skinner looks forward to the chiange in opinions held by the communicatee. When the communicator was a person
withering away of the state. At least Marxism does. The problem is what controls wilth whom the communicatee could positively identify, he tended to change his
to establish and who are to establish them in the interim. In Walden T w o , psycholo- ..-< inion in the direction argued by the communicator; the communicatee was less
gists are in control, but later -Skinner backs away from according power to them ely to change his opinion when he disliked or mistrusted the communicator. This
or any other elite group. Lenin did not back away from the issue: the vanguard of ding is less of a surprise than the finding that, after a lapse of time, opinions
the proletariat are the custodians of liberation. ---. tded to change in the direction of the argument of the disliked communicator-
Behaviorists, as well as Leninists, express honestly their compassion and respon- ancd then to reverse again when the communicatee was reminded of who said that
sible concern for humanity. What behaviorists and Leninists alike lack is any very wh ich he was now coming to accept. The implications are potentially enormous and
specific set of standards by which they will appraise the appropriateness of control. rel;ate to the success of prophets and polemicists (and housewives) in changing the
At the same time they both presume to have remarkable confidence in their own views of people who despise or depreciate them. Radical change, violent and non-
ability to lead the way. They lack knowledge but want power to act on the basis of violent, has typically been advocated by social rejects, from Socrates to Christ, from
the vaguest criteria, the most prescientific premises of their scientism. Mahomet to Marx. The continuing difficulty is that, consistent with the Yale findings,
Situationalism. A substantial category of proximally oriented psychological the:se prophets finally get accepted and their teachings established-and then finally
work does not neatly fall into the behaviorist classification. This residual category is ignlored.
composed of situationalists (mainly the Gestaltists), who analyze behavior in tenns A significant integration of a variety of seemingly discontinuous but related
of the present (rather than the historical) environmental antecedents of actions. earch has been made by Leon Festinger in A Theory of Cognitive Dissonance
Freud and his descendants concentrate on the patient's history of interpersonal rela- 257). Without establishing the loci in the brain, he argues that man
 12 James Chowning Davies Where From and Where To? 13
. irrepressible need to establish consistency in his opinions-to order and rationalize A' . study of the physical effects of deprivation on brain tissue (undertaken by
them. In quite different language, Festinger is talking about what Freud called the David Krech, Mark Rosenzweig, and others at the University of California, Berkeley)
reality principle. He emphasizes the need to eliminate dissonance not only between has demconstrated the inextricable relationship between the developing organism and
one belief and another but also between belief and act. He elaborates on the social . er~vlronment.
me _--.1..
In this study, genetically identical laboratory rats were divided into
circumstances in which dissonance is diminished and intensified. two groups. In one, the rats were raised in what the experimenterscalied an enriched
Two major statements within the situationalist category (Dollard and others, environment: their cages were placed in brightly lighted rooms, and they had abun-
1939; Berkowitz, 1962) deal with a very basic problem: the causes and circum- dant handling by experimenters and many toys and other rats to play with. In the
stances of frustration and aggression. Discussion of these statements is contained in other grc)up, the rats were well fed and housed, but in darkened rooms with minimal
Chapter Nine in this volume. handing by experimenters and neither toys nor other rats to play with. After about
Developmenta2ism. In the theory and research thus far considered under L
..-
- ---
two ~lullthsof these different environments, the enriched rats were alert and sociable;
the environmental determinant category ( f r ~ mPavlov to Hovland and others), the deprived rats were apathetic and dull. Then the two groups of rats were killed
little attention has been paid to the organic part of the equation B = f (OE). In and their brains compared. The enriched group had larger brain weights than the
contrast, developmentalists symmetrically emphasize the necessity of interaction deprived ones and statistically significant differences in amounts of two chemicals
between organism and environment for normal development of the individual. (acetylch~olinesteraseand cholinesterase) that are involved in the passing of electrical
In the 1950s Hany Harlow and his associates began to publish reports of their signals a(:ross nen7esynapses (Krech, 1968; Quay and others, 1969).
research with monkeys that had been taken from their mothers at birth and 'hised" M Torking with normal children in normal environments, the Swiss psycholo-
by artificial mothers-wire-cloth and terry-cloth mothers with monkey-like3aces and gist Jean Piaget (b. 1896) observed the process by which children began, at the age
one rubber nipple, from which milk could be drawn. The effects were dramatic. of six or seven, to develop rules and judgments of right and wrong. This period is of
When afforded contact with the artificial mother, the infant monkeys clung to it, course much later than the critical postnatal period (the period studied by ~arlo;,
even when blasts of air frightened or suddenly emerging spikes actually hurt the little Spitz, and Krech and Rosenzweig). By age six or seven the children whom Piaget
monkeys. As soon as they could, they returned to their offensive fake mothers. Other (1932) studied had already passed through the "egocentric" stage of confusing self
monkeys, raised in isolation from any contact, became generally withdrawn in their and non-self and had entered the "cooperative" stage, when they began to change
posture, varying apathy with rage at environment and at themselves. Sometimes they from accepting rules as parentally or divinely established and became able to delib-
would bite at their hands or arms till they reached bone. As they grew, they developed erate and decide ori their own rules for play. Again, as in the earlier stages studied
little ability to play with their n&nally raised agemates and when they were physically by Harlow and others, there is apparently a crucial interaction between developing
mature had no interest in sex. .
organism and environment that makes possible the establishment of cooperation, of
When mother-deprived female monkeys were involuntarily impregnated, they rule making rather than rule obeying.
became frightfully cruel mothers, sometimes killing their unwanted offspring by biting Some theorists have attempted integrations of what is known about successive
their skulls. But the severely mistreated infant monkeys nevertheless clung to their stages of development. Erik Erikson (b. 1902) conceptualized "eight ages of man,"
emotionally deprived mothers-and in time, in some kind of elemental group therapy, starting in infancy and culminating in maturity: basic trust, autonomy, initiative,
changed their mothers into rather normally affectionate creatures (Harlow, 1953; industry, identity, intimacy, generativity, and ego integrity. Each stage develops on
Harlow and Zimmerman, 1958; Harlow and Harlow, 1962; Harlow and Suomi, its predecessor-in a process that Erikson, using a term from embryology, calls
1970). "epigenesis." If failure in development should occur at any stage, the individual
Similar phenomena were observed by Ren&Spitz (b. 1887), a child psychia- probably will be unable to develop normally through the stages that follow (Erikson,
trist trained by Freud. Spitz studied what happened to infants raised in a foundling 1950, ch. 7; 1963, ch. 7 and p. 273).
home. These children had had normal contact with their own or surrogate mothers A further refinement of the stages in the interaction process in a developing
during the fist three months of life and then were deprived of contact with them human being has been made by Kohlberg (1964, esp. p. 400; 1968). He describes
for at least three months. The infants were given adequate physical attention by six stages in moral developme.nt. A growing child successively changes the bases for
nurses, who fed, bathed, and changed them. But the nurses averaged eight infants conformity to rules: ( 1) to avoid punishment, (2) to obtain rewards, (3) to avoid
each to care for and so could spend only a fraction of the time of a normal mother disapproval, (4) "to avoid censure by legitimate authorities and resultant guilt," (5)
with a single child. The effects were similar to those in Harlow's monkeys. Of the to maintain the respect of the impartial spectator judging in terms of community
ninety-one children raised in this well-run foundling home, thirty-four died before welfare, and (6) "to avoid self-condemnation."
their second,birthday. Those who survived were "human wrecks who behaved either These writers and the experimenters discussed before them are working on the
in the manner of agitated or apathetic idiots" (Spitz, 1949, 1959). For a careful basic building blocks whereby the initial interactions between infants and their
specification af the step-by-step sequence of development when there is normal maternal environment gradually form the structure of society. This interplay between
interaction between child and mother, see Bowlby (1969, ch. 14-16). individuals and this interdependence of organism and environment are what lead to
14 James Chowning Davies Where From and Where To? 15

the development of social and political institutions whose growth, establishment, and vallows his pride in the new society when the boss calls him a peasant, a Wop, a
disestablishment are inextricably related to the growth of individual human beings, ohunk, or a Nigger. After achieving some job security, he gets together with xeal-
from infancy to maturity. le compatriots or fellow landsmen to begin to act, in a real movement of people who
It is not yet possible to generate a broad theory of political development on assionately oppose their common degradation.
the empirical psychological basis of the work done by such as Erikson, Spitz, and HIowever, it is sometimes inconvenient to reduce analysis to the immediate
Kohlberg. But some of the foundations for such work are already in existence. Normal terpersc~ n a llevel; furthermore, some influences may affect large bodies of people
organic growth processes have, as Spitz put it (1959, p. 43), "a life of their own." mucn the same way. Socioeconomic status plays an enormous role, in both develop-
1

But this life will not develop unless the environment is supportive of the stage of g and developed industrial societies, in establishing patterns of political behavior.
) do such phenomena as social mobility and religious affiliation. Some writers using
development for which the organism is ready. Spitz refers to this interaction of or-
ganism and environment as "dependent development'' (meaning the same thing that ch units of analysis have reduced men to undifferentiated particles in a homoge-
Erikson means by "epigenesis"). An infant cannot reach Erikson's "ego integrity" :ous mass. T o avoid, the opposite error of supposing that every environmenta1
or Kohlberg's "internalized conscience" without having gone through a continuous fluence is uniquely personal, we need to consider forces that are of broad and in-
process of building and interacting as his organism make its demands and as an rect origin.
environment responds in the epigenetic process. A
Karl Marx ( 181&1883) was a developmentalist before that genre developed
One implication of developmental analysis, of the finding of the necessity for nong psychologists. But Marx conceived the great determinants of development to
effective continuous interaction between organism and environment, is not men- not within the individual or in the interaction processes that Harlow and Spitz
tioned by any of these investigators. Maslow's hierarchy reminds us that the physicd mined. The grand determinants for Marx were the interactions of vast classes
needs precede all others. An organism that lacks food and good health is ill prepared thin vast systems: between the feudal lords and system against which the bour-
to undertake any developmental interactions if the environment fails to feed it. Unfed oisie revolted and between the bourgeoisie and the capitalist system against which
people die. Unfortunately there are perhaps billions of people on earth so mal- e proletariat was beginning to revolt. This last revolt was to be the product of the
nourished that bodily needs absorb almost their entire attention. gregation of vast "masses" in wretched urban slums, often worked quite literally
The political involvement of emotionally deprived people may be one of basic death by the machines and system which they did not control. What this system
indifference (unconcern for anything beyond rudimentary social interaction) inter- J, however, was to unite the wage-enslaved "masses" in some consciousness of their
rupted by an occasional outburst of rage against society, against humanity, against self. mmon class interest (Marx and Engels, 1848).
But for those people whose deprivation is physical and therefore even more funda- On a common-sense level, the analysis makes much psychological sense. Peo-
mental, political involvement is perhaps even nearer to absolute zero. When it does : who are oppressed do tend to form a unified group. But there are a t least two
appear, it is likely to take the form of totally asocial concern of each individual for his 3blems with the analysis. One is that people who are so poor that their entire
own survival. -occupation is with staying alive and striving for some minimal nuclear family
idarity are not going to revolt: they will get food and shelter for themselves, their
Environment as Determinant: Mediated, Distal Influences )uses, and chiIdreq by whatever means necessary. Their '6moraIity" is individual and
unitive social survival rather than class survival. Only when their basic material
Studies of mediated, distal influences emphasize the role of the great society, ~ d are
s met (so that they no longer need to struggle each against all others or at
the vast groupings in which the individual appears to have an undifferentiated and least all other families) can they afford to become conscious of their common class
relatively passive role and in which the emotional considerations are supposedly mini- interest. The other problem is that, once they have achieved a modicum of control
mal, They deal, that is, with such processes as modernization, industrialization, and over the productive forces by such means as collective bargaining and the strike, they
immigration--and with such influences as regional, religious, and political party arc! apt for a time to become conservative, opposed to further change and hostile to
as these affect the individual's political behavior. thc)se who advocate it. McCarthyism-the era of communist baiting in America in
These processes generally take place in intensely personal contexts, with an the, 1950s-was more a populist protest than a capitalist inquisition (see Stouffer,
often strongly emotional reaction on the part of the person influenced by such seem- 55).
ingly impersonal forces. That is, they generally are mediated by direct, affect-laden The need hierarchy helps us explain these events. The process of development
contacts with specific events and specific individuals that serve as carriers of the eeking satisfaction of the physical, social, dignity, and finally self-actualization
impersonal influence. A person who is compelled to increase his output in order to needs-takes place a step at a time, in the epigenetic manner that both Spitz and
increase the enterprise's profit, in order to provide private or public capital for Erikson described. The interest of working people in free speech, in the creative arts,
expansion, is made to do this not by a book or a concept but by a boss. A indeed in individuation and individual autonomy, requires the prior secure satisfaction
person who lives through the numerous traumata of leaving an old and defined com- of the social-affectional and dignity needs. In their vision of the good society, writers
munity for a new, exciting, and unpredictable society does so in no impersonal way. who objectively were in the bourgeois class like Marx, Engels, and Lenin ignored
He cries when he leaves his parents at the bus station, the pier, or the airport. He the need of poor people to go through the stages of development that their own
16 James Chowning Davies Where From and Where To? 17

parents (even their ancestors) had already passed through. Their sympathy with the he did note that the wish for response is related to the love instinct (including sex).
poor (that is, with factory workers and most peasants) began after these pioneers ind he described the wish for recognition in words that make it sound much like
among the intellectuals of their time had emerged out of a bourgeois class identifica- Haslow's needs for self-esteem and self-actualization, Altogether it is a remarkable
tion to a classless, human identification. They envisioned the good society that was ist, and it became the basis, as noted in Chapter Nine in this volume, for a major
appropriate to themselves as pioneers who had subjectively rejected class identifica- heoretic:a1 work by Lyford Edwards on revolution.
tion, but they were not yet able to rid themselves of a doubtless unconscious identity 'wo other sociologists, Robert and Helen Lynd, made a major advance when
'I
as members of the intelligentsia. hey intc:rviewed a sample of a city population and produced a massive study of the
Such a good, classless society was not yet then, in the mid-nineteenth century citizens' manner of living and their social (including political) interactions-more
-if it is even in the late twentieth century-within the realm of real possibility for specifically, their pattern of work, family life, education, leisure, church activities, and
unemancipated factory workers and poor peasants whose major and almost constant government. The study demonstrated the wide range of phenomena that can be
day-to-day concern was with sheer survival and who could not share the human, per- examined when systematic and direct contact is established with ordinary people.
sonal basis (or even universally apply the ideological basis) of the identifications of Using some of the techniques of cultural anthropologists plus quantitative analysis of
humanitarian intellectuals. The work-not just in words but also in deeds--of Marx, questionnaire data, the Lynds (1929, 1937) established trends that were broadened
Engels, and Lenin was enormously influential in developing a sense of common and deepened in future decades of basic social and political research. They did
interest among poor people; I am arguing, because of their moral identification with factually what in various ways had been done fictionally by Charles Dickens in
poor people and not because of the intellectual validity of their arpment or their England, fimile Zola in France, and Upton Sinclair in America: they showed what
evidence. Visions of the good life, as poor people saw or see it, are inescapably based life is like in an industriaIized community.
more solidly on material abundance than on the possibility of identifying with other In 1950 appeared a book that grew out of Erich Fromm's (1941) description
people not as poor people who also need material goods more than anything else but the (Nazi) authoritarian character. The Authoritarian Personality, in which four
as human beings whose differences are in their stages of development rather than in L~sychologists(Theodore Adorno, Else Frenkel-Brunswik, Daniel Levinson, and
their nature. Nevitt Sanford) collaborated, is a broadly based quantitative study undertaken at
If Marx was psychologically rather innocent, the French sociologist Durkheim the University of California, Berkeley. It examines where socially the authoritarian
(1858-1917) was not. His intimate analysis of the effects of industrialization on personality type appears and what its relationships are to such political phenomena
individuals, spelled out in The Division of Labor (1893), remains a sophisticated s fascism and democracy. The book is crude in its methodology and naYve in its
original source of explanation, nearly a century after he wrote it. Rural people com- olitical concepts. Yet it is one of the most profoundly influential studies to appear
mence their involvement in the industrializing modern world by entering the factory ince World War I1 and one of the first explicit efforts to link psychological and
system, which severs their original social ties, binds them to the machines they tend, olitical concepts. One of the original collaborators in the study, Sanford, has written
and dissects their coherent skills into tiny identical, repetitive pieces. Urban and new appraisal for this volume (see Chapter Six). Analysis at this point would be
factory influences produce in them what Durkheim called anomie, the sense of root- edundant.
lessness and normlessness that alienates industrial workers from past and present Backtracking a bit, we turn to the development of research in political atti-
associations and from themselves as individuals. (See also Chapter Seven in this ldes as these relate to voting behavior. The pioneer in this research, Paul Lazarsfeld
volume.) At the same time, Durkheim emphasized, the division of labor becomes the b. 1901), studied patterns of persistent and changing social characteristics (notably
principal source of social solidarity. An individual in primitive rural society is not so )cioeconomic status, religion, and residence), the factors that he said determine
much free as he is stunted. Cities, factories, the specialization of jol? function do olitical preference. His ground-breaking study (Lazarsfeld, Berelson, and Gaudet,
alienate, routinize, and even regiment the lives of people; but they also make possible 944.) was of the 1940 election and was done in a single city, as the Lynds' work was.
a more total emancipation of man, who is otherwise stunted by his poverty and thus 'he trend of political analysis of voting based on a sampling of the entire nation was
limited to concerning himself with only survival, with satisfying his material needs. eveloped to some kind of methodological ultimate in the work done under the
W. I. Thomas (1863-1947), nominally a sociologist like Durkheim, developed irection of Angus Campbell (b. 1910) at the University of Michigan Survey Re-
over several decades a rather systematic psychological system in which he concen- :arch Center. The Survey Research Center provided probably the most intensive
trated on analysis of the great adjustment problems of poor people who emigrated or at least extensive) analysis of the process of public opinion formation that has
from rural Poland to urban Chicago. He concluded that the individual and the situa- ver been undertaken in a stable polity. The center seems less prepared and inclined
tion define each other and that the individual modifies the environment by defining get at the roots of change or of intense political instability. I t has ventured into
and solving problems according to his "wishes" and tendencies. By way of John B. nalysis of public attitudes toward political disturbances that produce violence (Blu-
Watson's "unconditioned responses" of fear, rage, and love, Thomas (1923) devel- lenthal and others, 1972) but not the tendencies within individuals and societies
oped a list of "wishes," "the forces which impel action." These are the desires for lat produce political violence.
new experience, security, response, and recognition. Curiously, he gave but slight Two authors have concerned themselves not with public opinion in stable
mention to a couple of rather basic forces that impel action-hunger and sex; but democracies but with various social-psychological aspects of political development.
 James Chowning Davies Where From and Where To?

Daniel Lerner (b. 1917) used data gathered in several Middle Eastern nations by nining behavior but avoided a direct consideration of these "nonenvironmental"
the Bureau of Applied Social Research at Columbia University and produced some Jeterminants of response. He wrote at a time when the stimulus-response Watsonian
concepts and findings that are applicable to developing nations all over the world. ~ehavioristpsychology was still a dominant influence, and so it is understandable
He emphasized, in different language, the need for what Adler called the community hat he shied away from these nonenvironmental (that is, organic) forces. Instead,
sense, and he asserted that empathy-in this context the feeling of a citizen toward le chose to list as a related set of basics a system of "pyramids," as he first called them
his fellow nationals-is a crucial antecedent of modernization (1958). Walt Whitman (Lasswell, 1935), or "values," as he came to call them (Lasswell and Kaplan, 1950).
Rostow (b. 1916) is a broad-gauged economic historian who looked into con- The "pyramids" are safety, income, and deference; the "values7' are divided into two
temporary developmental processes and thus into the future (1952); His novel road categories, welfare and deference. Since values are objects sought in the en-
phrase, "the revolution of rising expectations," became a clich6. that has concealed vironment, they correspond more to what Murray called "presses" than to organically
the significance of his findings about the "propensitiesy' that, in certain favorable based needs.
circumstances, impel people toward material development. Without adequate elab- The values that Lasswell considers are the more or less visible objects that
oration, he emphasized the psychological foundations of his theory. all people pursue, in ways that vary with individuals and cultures. The welfare values
This summary of some notabIe contributions to political psychology has been include well-being (health), wealth (income in goods and services), various kinds
too brief, but it does suggest the very varied origins of relevant ideas and research. )f skills, and enlightenment. The deference values include power, respect, rectitude,
In the range of intelIectual disciplines it has included psychoanalysts and social psy- and affection. The kinship of these valued objects to basic needs is evident, but Lass-
chologists. I t has also included sociologists and economists, who might turn in their vvell avoided establishing the relationship. He does not say what nonenvironmental
graves at having psychology imputed to them. But the wide range of contributions :haracteristics predispose people to want these things, because "no generalizations
only demonstrates the impossibility of classification on any other basis than what :an be made a priori concerning the scale of values of all groups and individuals"
each has added to our understanding of the roots of political action. Broadly con- (Lasswell and Kaplan, 1950, p. 56). By saying that it is the values-and, by *in-
strued, that is what this book is about. I t is not about psychology or political science ference, the predispositions-that make objects valuable, Lasswell avoids the organic
but about political behavior, and there are no a priori disciplinary barriers to its ialf of his basic equation, which neither Freud, McDougall, Murray, Maslow, nor
understanding. Now we can turn to have a look at some more explicitly political :ven the sociologist W. I. Thomas avoided. This avoidance has made it more difficult
writings, having by now some foreknowledge of the ancestry of these later works. to make political inferences from Lasswell's writings'that are of cross-cultural validity.
One of Lasswell's early main interests was analysis of symbols and their use
Some Progeny of the Union of Psychology and Political Science by political elites to elicit "blood, work, taxes, applause from the masses" (Lasswell,
1951, p. 31 1). Propaganda-that is, symbol manipulation by the elite that is in power
The first notable liaisons between psychology and political science developed -succeeds when it handles "aggressiveness, guilt, weakness, affection" (Lasswell,
at the University of Chicago under the encouragement of the political scientist 1951, p. 317). Propaganda generated by the counterelite, including a revolutionary
Charles Merriam. He saw the exciting possibilities of psychological (and even phys- elite, "selects symbols which are calculated to detach the affections of the masses
iological) attachments (Merriam, 1934, pp. 20, 43-44) but was never himself able from the existing symbols of authority, to attach these affections to challenging sym-
to exploit them. bols, and to direct hostilities toward existing symbols of authority" (p. 322).
T h e Work of Lasswell. One of Merriam's numerous intellectual progeny, Lasswell's preoccupation with analysis of the elite, those who get the most of
Harold Dwight Lasswell (b. 1902) was the first to enter boldly into the psychological what there is to get, at times seems to spill over into the assumption that the elite
house of political ill repute, establish a liaison, and sire a set of ideas and influences are more important than the "mass," that it is important to learn the skills of symbol
of great vitality. And, like Merriam, he not only produced progeny but encouraged manipulation in order to keep the "masses" compliant: "The consensus on which
others to do likewise. He thus became not only the intellectual father but also a kind srder is based is nccessarily nonrational; the world myth must be taken for granted
of grandfather of many of the political psychologists who developed after him. by most of the population. The capacity for the generality of mankind to disembarrass
Perhaps the most cited of Lasswell's succinct formulas appears in his Psycho- themselves of the dominant legends of their early years is negligibleJy(Lasswell, 1935,
pathology and Politics (1930, pp. 75-76) : Political man is the product of private p. 181). In sum: "The study of politics is the study of influence and the influential"
motives, displaced on a public object, and rationalized in terms of public interest. (Lasswell, 1951, p. 295) . Even in his study of revolutions, the major preoccupation
In Power and Society (1950), which he coauthored with Abraham Kaplan, there is is with the elites who lead their publics and with their use of symbols to incite them
a more fundamental and symmetrical statement on the "fundamental units" of (Lasswell and Lerner, 1965). This preoccupation continued at least to 1956, when
political behavior. They are specified as "response, environment, and predisposition" he suggested as an alternative to the murderous destruction of nuclear weapons a
(p. 3 ) . This becomes the fundamental equation: response is "a function of E[nviron- minimization of coercion by the gentler device of the "paralysis" bomb or a "paralyz-
ment] and P[redispositionl" (p. 6 ) . This is substantially the same as the equation ing" beam of sound or some other high-energy source which could immobilize a city
with which this chapter started. (Lasswell, 1956, p. 968).
Lasswell formally recognized the basic part that predispositions play in deter- Another preoccupation is his concern with equalizing the amount of values
20 James Chowning Davies Where From and Where To? 21

that each person should get. In Democratic Character (1951), he emphasized the universal applicability, in stable and unstable circumstances, in democracies and
necessity-for sharing power, respect, and other values if a democratic society is to nondemocracies.
come into being-and he presumed that such a society should come into being. In Some Research Developments Since Lasswell. Research in political behavior
1956 he reiterated his growingly explicit and equalitarian values and argued for the slnce rhe end of World War I1 has taken four somewhat different directions along
universalization of what had been deemed only the white man's heritage: "the dig- lines that are not quite parallel and extend to differing lengths. The first of these is
nity of freedom." the study of voting behavior in stable democracies. The second is cross-national com-
There nevertheless remained a kind of stewardly willingness to manipulate parative research, in polities that are relatively stable and democratic (that is, the
people, not in the brutal manner of punishment but by techniques which Skinner nations studied are at least not dictatorships and not undergoing revolution or wide-
called positive reinforcement. "Since the basic postulate of behavior is the maxirniza- spread violence). The third line of research is in the genesis, the origins, of behavioral
tion of indulgences over deprivations, our task is to consolidate democratic conduct patterns that become established in childhood, the starting point generally being set
by directing the indulgences toward those who act democratically, and the depriva- at the age of about six, when children begin their schooling. The first of the trends
tions toward those who do not" (Lasswell, 1951, p. 513). The democratic task, has been the dominant one. The second and third have provided relief from the
ironically, is one that must be shouldered by the elite. rather static study of behavior under stable circumstances, a study that has become
The dilemma that has faced elitists from Plato to Marx and Lenin, with increasingly dull, repetitious, and a precious picking of nits. The fourth is systemat-
eighteenth-century libertarians (including Rousseau) in between, is also expressed ically psychological political biography.
by Lasswell. None of these individuals successfully resolved the dilemma involved in The principal systemic frame of reference of each of these major trends has
the asymmetrical possession and exercise of influence in every real society in every continued to be political stability-behavior shaped by stable environmental condi-
actual era on this earth. The task of increasingly equalizing power and dignity, as a tions among relatively contented people. This limitation, particularly in the first two
task of those who have more of both, is manifestly as hard to specify on paper as it lines, has amounted to settling for half a loaf of what was fresh in the 1940s but
is to accomplish in reality. Lasswell avoided the extravagant rhetoric of those so already stale in the 1960s. The limitation was not established by the wide range of
attached to an ideal that they lose contact with the realities and the responsibilities Lasswell's concerns or by the vast breadth and depth of widely known psychological
of power in any society, however democratic. theory and research, only some of which has been recounted in the earlier part of this
The theoretical contributions of Lasswell have been truly large. His profes- chapter. But this limited research has nevertheless produced some work of enduring
sional contribution has been truly enormous: he was the first to insist successfully on utility, and the limitation is gradually being removed, most notabIy in the third line
the need to bring psychology into political analysis. He more than any other person of research, in what has come to be called political socialization (see Chapter Five
insisted that unconscious forces which so significantly determine how people behave in this volume) and finally in research in political instability, a trend that now
politically must be subjected to analysis. His ideas were not always consistent, but he promises to become established (see Chapter Nine).
stated his concepts and his orientation so forcefully that they were at last listened The first line of research is exemplified by the opinion and attitude studies that
to. The large body of theory and research mentioned in the previous parts of this have emerged from the University of Michigan Survey Research Center (A. Camp-
chapter had for the most part been around before Lasswell wrote. And political bell and others, 1960, 1966). This work is lineally descended from the studies of
science had continued to ignore psychology. such pioneers as the Lynds and Lazarsfeld, but with refinements and innovations.
Unfortunately, the very broad range of Lasswell's interests was ignored or Voting behavior has been analyzed as a function of the relationship between such
narrowly viewed by most of the political scientists-many of them his disciples-who social characteristics as rural-urban immigration, urban-rural residence and socio-
subsequently studied political behavior, Lasswell was at last listened to, but with economic status on the one hand and orientations toward party, candidates, and
ears that were sensitive to only a narrow range of his ideas, which resonated so issues on the other. Most innovative has been the study of the correlates of the sense
broadly with political reality under conditions of stability and instability. For decades of political efficacy, the sense that taking part in politics makes a difference for the
after Lasswell wrote his World Politics and Personctl Insecurity in 1935, political single individual among scores of millions. And the research has ventured beyond
scientists continued to study political behavior under stable circumstances, when social national boundaries, to study voting behavior in such nations as Norway, England,
and political developments proceed slowly and institutions operate effectively. Writ- and France-which, of course, are relatively stable democracies.
ing sometimes as a technologist on symbol manipulation, he nevertheless saw the Various traditions and other irrational (but psychologically explicable) forces
ineffectiveness of symbol manipulation when people are not symbolically but really have inhibited the use of the great instrument of survey research to achieve the
deprived and when symbols do not relate to their real problems, their changing value comprehensiveness and depth that other techniques and writers outside the University
expectations (their basic needs as they emerge) (Lasswell, 1951, p. 435). Many of of Michigan have attempted. Two examples of stepping beyond secure methodologies
Lasswell's heirs-evident, in their concentration on the low volatility of political and frames of reference may be mentioned. Both remain within the framework of
behavior in s'table democracies, have failed to appreciate that they themselves are analyzing stable polities, but both range broadly. Lester Milbrath (b. 1925) produced
not meeting the knowledge expectations of students and publics and elites. There is a remarkable synthesis in a study of political participation (1965). I t is so nicely
an increasingly evident need, expectation, and valuation of explanations of more ordered and develops such a logical sequence of statements and propositions as to
22 James Chowning Davies Where From and Where To? 23
"how and why" people get involved in politics that it reads a bit like a geometry a higher degree of participation and political self-confidenceamong people of higher
text. But it is far richer in data: it is a summation of the state of a well-matured socioeconomic status-possibly because the parents in such families encourage partic-
subdiscipline. And it notes, without elaboration, that there is a lack of research on ipation in the making of family decisions. Almond and Verba note that in those
relationships between individual needs (both physical and mental) and political countries where there is a greater participation by all family members in family
behavior. decision making (the United States and Britain), there is also a greater amount of
The other study lacks the elegantly ordered development of Milbrath's syn- participation in intermediary groups of a religious, political, or yther social sort. In
thesis but probes more deeply. Robert Lane (b. 1917) in Political Life (1959b) con- these countries, they found, people feel politically more efficacious ( 1963, ch. 11, 7) .
siders as wide a range of behavioral relationships as Milbrat11 does, but he boldly In the late 1950s a new development started in political psychology, on re-
introduces a listing of human needs (economic, affectional, understanding, relief of search beginnings of the 1940s and theoretical foundations that are as new or old
tension, power, and self-esteem). It is even more significant that Lane spends at as Erikson, Freud, Rousseau, and even Plato in his concern for the training of the
least a fourth of the book in considering the political consequences of such needs. To youth. The new development retains the continuing emphasis on behavih- in stable
be sure, he remains within the framework of research data in the United States, but polities, but it has undertaken to examine the most fundamental dynamic process of
the data he uses had been largely ignored by behaviorally oriented political scientists. all: the origins and growth of political tendencies in the individual as he moves from
The synthesis that Lane was striving for in Political Life remained psychologically birth and familial nurturance as a child to political maturity as a citizen. (See
unstructured, but the volume was a pioneering effort-less extremely confident than Chapter Five in this volume.)
Lasswell and more sedulously aware of the necessity of gathering evidence to evaluate Herbert Hyman, in his pioneering Political Socialization (1959), ordered
theoretical propositions. and summarized a substantial body of theory and research on how political tendencies
In Political Ideology (1962), Lane achieved unprecedented depth and a much first get established. A notable work cited is that of Theodore Newcomb (1943, 1947),
more psychologically concerted analysis of the origins of political beliefs and values studying the changes in political outlook of girls of generally conservative upper-
of fifteen people, all of them in what may loosely be called the upper-working class middle-class background, as they lived through the liberal atmosphere of four college
or lower-middle class. His observations of the fifteen are applicable to perhaps a years. But Hyman went back earlier than the influences of late adolescence. He
majority of citizens in a stable society undergoing rapid internal change, but survey compared the socialization process in working-class and middle-class families and
research has not yet undertaken to use the concepts around which Lane built his covered the age range from grade school up to middle age and the kind of group
research. Political Ideology is less a study of ideology than of the roots of political influence from the family to age-peer groups in school.
belief systems, of political attitudes. Lane's fifteen people have a sincere but shallow In a related effort, Sidney Verba did not emphasize the developmental se-
belief in the values of democracy, of hard work and reward for it only among the quence but nevertheless considered proximal social influences on the establishment
rich and powerful when they have indeed worked hard. They also share a sense of of basic political predispositions. And in his Small Groups and Political Behavior: A
futility at trying to control any major aspect of political life; at the same time, they Study of Leadership (1961) he synthesized the extant research. Much of the research
have a fear of change. They both respect and resent those who have wealth and focus of T h e Ciuic Culture (which, as we have noted, Verba coauthored with
power; but they resent and do not respect those who radicaIly criticize the existing Almond) was established in Verba's early work, where he noted the need for under-
(American) political system, perhaps as poor people within the Russian imperial taking comparative analysis in different societies.
and Soviet systems have resented internal radical criticisms of their government. The two major initial research studies of the early stages of political (as
Lane presents a rather frightening picture of the ambivalence toward authority and distinct from total) socialization appeared more or less simultaneously and quite
toward stable systems themselves. Such beliefs are perhaps prevalent in any greatly independently. In 1957 Fred I. Greenstein completed the planning and began the
diversified society among individuals who are less than successful by that society's interviewing of grade school and high school students who provided the data first for
standards. They feel unconnected and powerless within a colossal system that they his doctoral dissertation at Yale University (1959) and then an article on "The
fear-and fear to change. Benevolent Leader : Children's Images of Political Authorityy' ( 1960) and a book,
Some of the roots of assent to stability have been analyzed in a study of five Children and Politics (1965). That same year two others, David Easton and Robert
nations. Again, the broad matrix is of basic stability, but there is an impressive new Hess at the University of Chicago, began to publish the results of their studies of
use of survey techniques to ascertain the incidence of political dispositions among grade and high school students (Hess and Easton, 1960; Easton and Hess, 1962).
major segments of the population of the United States, Great Britain, Germany, The findings of these more or less simultaneous and quite separated sources
Italy, and Mexico. Gabriel Almond (b. 1911) and Sidney Verba (b. 1932) in T h e are not the same but generally confirm each other. Consistently with Hyrnan (and
Ciuic Culture (1963) found varying and generally diminishing degrees of both Lane's Political Ideology), Greenstein compared differences in the basic interactions
political participation and sense of the efficacy of participation in these five nations. between parents and children in working-class and middle-class families, noting the
The diminutions generally were in the order of the nations listed, from the United more frequent tendency of children in the former class to show feelings of inade-
States as the highest to Mexico as the lowest. Among numerous other relationships, quacy, limited imagination, and deference. He noted that working-class boys are
they found something consistent with Lane's study of fifteen lower-status Americans: relatively more interested in politics than working-class girls are, and that working-
24 James Chowning Davies Where From and Where To? 25
class women are less interested in politics than middle-class women are. But in general invariably fragmentary. Even less is known or knowable about the reasons that leaders
working-class boys and girls showed less interest than middle-class boys and girls. became such while their brothers, sisters, parents, and children usually are individuals
In short, studying the same epiphenomena as Almond and Verba (1963, ch. 7, 12) of no great distinction. We can trace back from the broad, fast-running river of the
but with a deeper probe, Greenstein found a more pronounced tendency among leader's public career to the headwaters of his childhood, but we cannot yet explain
middle-class than among working-class families to socialize children in the direction why most such headwaters do not become such big rivers. More abstractly, how much
of equal deference and involvement in decision making. It adds up to evidence of of the behavior of leaders is a function of the development of their complex or-
the increasingly obvious irony that people of middle-class origin are more equalitarian ganisms, over decades, and how much is the product of their complex environments,
and on that dimension more democratic than people of working-class origin. For a over decades, is perhaps impossible to know.
more direct comparison of what they call "elites and masses" in democracies, see We are likely to be able to predict fairly well where the broad class of
Dye and Zeigler (1970, esp. ch. 5 ) . political elitists comes from but to predict quite badly which of the few individual
Hess and Easton (1960) emphasize the relationship between attitudes toward members of this elite will become chiefs of state. Members of the elite, including
father and attitudes toward the president. For children in the early years of grade modern revolutionary leaders, below the very topmost level, are more likely to emerge
school, the image of the father and the image of the president tend to be much alike. from the middle class or from some other above-average status than from the working
By the time children finish grade school, however, they have developed more distinctly class or from rural backgrounds (Brinton, 1965, pp. 95-100; Strauss, 1973)-a1-
separate images of father and president, in the direction of seeing the latter as work- though occasionally the sons of carpenters, poor farmers, and factory workers do
ing harder and knowing more than father does. become even immortal agitators and chiefs of state. Such findings remind us to be
Two more recent books, by Dawson and Prewitt (1969) and by Langton less surprised at the clergy background of Woodrow Wilson, Malcolm X, and Martin
(19691, both entitled Political Socialization, show the considerable strides that have Luther King; the lawyer background of Karl Marx; or the high government official
been made in research, in the decade since Hyman delimited the field. But with background of Nikolai Lenin and Mohandas Gandhi-and more surprised at the poor
rare exceptions (see Davies, 1965), no direct effort has been made to consider the farmer background of Abraham Lincoln. They remind us to be most surprised at
fireschool influences of parents on political socialization. If Freud-and such major any present scheme for predicting who among those now living will be chiefs of state
institutions as the Catholic Church and totalitarian parties of left and right-are in any land in the year 2001.
correct in saying that the first six years are the most crucial ones, then the largely un- The' conceptual tools and measuring devices with which we can ascertain
explored political consequences of the training during those years are most crucial. adequately why some men become political leaders indeed remain rather rudimentary.
The political implications of the research of Harlow and Spitz and Krech and But some analysis explaining the style of men who become leaders has become
associates, as described above, remain unexamined. Langton (1969, ch. 2) found increasingly sensitive and psychologically sophisticated. The study of Wilson by the
marked differences in political attitudes among children raised in families where Georges (1956) and of the first American Secretary of Defense, James Forrestal, by
both mother and father were present and those raised in father-absent families. This Rogow (1963) indeed show the limitations imposed by scant knowledge of childhood
research describes the political product of interactions that do commence at or near influences that Freud and-Erikson have psychoanalytically emphasized (and Harlow
birth but does not, perhaps could not, examine the process that produces the product. and Krech behaviorally emphasized). Yet the work of the Georges and of Rogow
Another line of research is the political biography on a psychological base. succeeds. The Georges, however, do not discuss the relevance of the fact that Wilson,
Again, as in political socialization, this volume has a separate chapter on psycho- 3 wartime president whom the war tore apart, was eight and living in Georgia during
biography (see Chapter Eleven). It is enough here for me to mention works that the Civil War when the violent march of Sherman's army laid much of Georgia waste.
seem particularly promising or successful. One of the most brilliantly intuitive psycho- A very bold and successful systematization has been made by James David
biographies appeared well before there was such a word. It is Lord Charnwood's Barber (1972a), the boldest step yet in establishing a typology applicable to all the
Abraham Lincoln, first published in 1916. Edgar Snow did a similar and classic 4merican presidents from Taft to Nixon. The study establishes personality types and
study of Mao Tse-tung in 1938. Intuition of the same special sort that Charnwood
makes a successful case for the predictability of style of presidential rule, partly on
and Snow possessed became happily joined with thorough and systematic knowledge
the ba'sis of tendencies that began in childhood. Although the book does not say who
in Erik Erikson. On the foundation of his Childhood and Society (1950-second
ill become president, it cogently argues how presidents will act. And Barber does
edition, 1963), he wrote his landmark studies of Martin Luther (1958) and of
bake note of the fact that Wilson experienced war, firsthand, at a very tender age.
Gandhi (1969). The work of Erikson (and before him Henry Murray and Sigmund
Freud) has inspired not only later psychobiographies but also theory and research
Toward General Political Theory on a Psychological Base
in the background of political elites and nations undergoing development (Pye,
1962). In the previous sections of this chapter I have skimmed selectively-necessarily
The problems in psychobiography are quite special. It seeks to explain the .oo selectively to be encyclopedic-the surface of a vast and rapidly growing body of
personalities of leaders who seem to share little in common other than being leaders.
~ritingthat bears on the manifestations and causes of political action. Those writings
Knowledge about the early childhood influences in the lives of political leaders is hat have been discussed are by no means a cross section of those that have been
26 James Chowning Davies Where From and Where To? .27

written. I have sought to compensate for what I regard as a hypertrophy of research The fourth book, Jeanne Knutson's Human Basis of the Polity (1972a), is
in epiphenomena, phenotypes, and other evidences of the sometimes self-evident or not the first effort to test Maslow's concept of a hierarchy of basic needs but the
long since demonstrated, by ordering the discussion in accordance with a theoretical first effort empirically to apply it in its political context.
rather than epiphenomenal basis-B = f (0E)-and by discussing first the organic The concept of basic needs, mentioned by all four of these writers, has become
emphases. There is not a dystrophy but a kind of atrophy of theory and research that increasingly central to the thinking of each of them. The statement of what needs
can help us link observable acts with their deeply and generally antecedent causes are indeed basic-in varying degrees and at different times common to the entire
in the human organism, notably the nervous and endocrine systems. Aristotle sought human species-is different in each of the four writers, but each emphasizes them:
such relationships. So did.Hobbes, whose Leuiathan (1651) founded its analysis of whatever behavioral characteristics human beings do have in common, these writers
political institutions on a theory of human nature. And likewise Lasswell has sought assert, they surely include some needs. In my Human Nature in Politics they are the
to relate fundamental determinants to observable effects-and vice versa. subject matter of the first two chapters. Lane mentions them somewhat unsystemati-
The concern in American studies of political behavior for precisely validating cally in his Political Life and then very centrally in his Political Thinking and Con-
the self-evident, in each of its infinite nuances, seems almost an industrial phenom- sciousness.
enon, consistent with the ability to turn out identical and shiny automobiles that Another characteristic of all four of the books is that each shows a sense of
speedily move nowhere and everywhere. The ability to produce printouts on the awareness of the evolution of political theory. In an era of political behaviorists who,
computer has displaced the ability to think and postulate more fundamental causal as Arthur Koestler might have put it, are a generation without an umbilical cord,
relationships, While Mendelian genetics has been transformed into microbiology and these four writers have not sat and mystically contemplated their navels. They are
traditional somatic physiology has been transformed into neurology and endocrinol- research oriented. But they show an awareness of their intellectual heritage in political
ogy, the study of political behavior for the most part has continued to pursue that theory.
which is no longer novel. In turbulent times it has studied stability. In a time when All of these works and any works in political psychology are only beginnings,
systems are under intense pressure, political science as a total discipline has studied but in an age when publics and individual citizens are asking fundamental questions
not the pressure so much as the container-the behavior not of people but of institu- and when political "apathetics" and other rebels have threatened to tear down gov-
tions. ernments unresponsive to their demands, these works and some others have not been
There are three exceptions to the heavy emphasis on the epiphenomenal, on limited to ascertaining whether the shift in political loyalties means that the new
the study of manifest effects-three efforts to establish the fundamental linkages majority is Democratic, Republican, Undecided, or Not Ascertained. They have
between people and institutions, in stable and in turbulent times, that Hobbes sought to assemble ideas and research that can have more enduring value.
attempted. None of them is a finished product, an ultimate statement, but each has In such directions political psychology can go. In such directions, toward
reflected an effort to link things more fundamentally and each has reflected the more fundamental explanations, political science cannot go without psychology.
scientific commitment to seek empirical validations for assertions that of necessity Systematic knowledge of it is as basic to political science as organic chemistry is to
remain untested hypotheses. And a fourth effort has involved the empirical testing biology and as physics is to chemistry. There is no other way to find answers to the
of phenomena about which there has been some light laughter or solemn scholarly fundamental questions of why governments, as Jefferson put it, are instituted among
headshaking, about supposedly vague notions that were deemed really untestable. men and why men assert the right to alter or abolish government that does not serve
The earliest of these broad-gauged efforts was undertaken by Christian Bay their needs. If indeed that is fundamentally why Americans and other people have
in his Structure of Freedom (1958). Frankly avowing a reality which affects everyone revolted, the study of what men want is a rather appropriate place to begin to study
but which most social scientists are either prudish or schizophrenic about-a set of the stability and instability of their political institutions.
values-Bay has attempted to examine such empirical foundations as exist for
distinguishing the human purposes that government should serve and those it should
stay away from.
The second of these efforts, growing out of a doctoral dissertation (Davies,
1952), was undertaken by me, in a book (Davies, 1963) whose title was filched from
Graham Wallas (1858-1932) on Human Nature in Politics (1908). In my study of
the relationships, I attempted to order the presentation of ideas and research that
bear-less explicitly in a normative context than in Bay's writing---on some of the
basic psychological questions.
The third of these efforts is Robert Lane's Political Thinking and Conscious-
ness (1969). In this work, Lane examines the ways in which people become not just
knowledgeable about politics but aware of themselves as self-conscious participants
in the political process.
 Personality in the Study of Politics 29

However, a second-and still impressionistic-examination of personal evi-

dence forces us to conclude that personality stereotypes are a blend of veracity and
mendacity and therefore do not accurately represent any given group. In fact or
fiction, we all are acquainted with informal and humanitarian teachers, spendthrift
peasants, bourgeois revolutionaries, and other common "aberrations." Thus, we are
initially forced to acknowledge that the relationship between personality and social-
political behavior is one of contingency: what is true is so only in a certain degree, in
some cases, and on certain occasions. At this point the professional student of political
behavior, when faced with a research problem in which personality is clearly a

PERSONALITY IN relevant dimension, turns to a standard volume on personality theory. Unfortunately,

he quickly discovers that there is no one personality theory, but rather a veritable
smorgasbord of theoretical systems. Thus, the scholar who is interested in studying
the effect of personality on political behavior is forced to become familiar with the
THE STUDY OF major theoretical systems in order to acquire an understanding of useful models and
constructs. Also (as the following section attempts to make clear), the scholar who
wishes to employ intrapsychic determinants in his research design must do more than
simply pick a congenial theoretical framework within which to work.
Additionally, since each personality theory is stated in general terms without
clear behavioral referents, the problem of operationalizing the selected personality
theory becomes paramount and leads the political psychologist into the thorny path-
ways of methodological acrimony and ambiguous research findings. The thoughtful
Jeanne JV. Knutson researcher, therefore, must be wary lest the availability of a particular personality
theory or frequently employed personality scale overdetermine and deleteriously
shape his research strategies and goals.

Conceptualizing Personality

Both historically and currently, personality as a concept has lacked behavioraI

referents and objective specificity and therefore has eluded rigorous analysis and
assessment. Historically, the existence of religious, moral, and legal norms seemed to
obviate the need for scientific examination. Currently, the concept of personality re-
mains difficult to objectify largely because of its theoretical development in the
I n discussions of political behavior, clinical tradition (Hall and Lindzey, 1957)-a tradition in which intuition has been
scholars and laymen alike make widespread use of personality concepts, because in- valued as highly as empiricism. Additionally, the increasingly popular social learning
tuitively and impressionistically we all sense that the type of person who occupies a
theories have retarded the analysis of personality as a trans-situational construct
certain role has a good deal to do with the way the role is performed. Furthermore, (see Alker, 1972). As the following discussion will suggest, equally deleterious to
although again it is more observational than evidential, certain types of people do the scientific usage of personality have been the antiempiricist assumptions upon
seem to occupy certain roles. Hence, we tend to stereotype the personality aspects of
which the concept has usually been based. It is thus initially necessary to examine
role performance: the businessman's Babbitry, the pedantic nature of a professor or those assumptions and to delimit personality in terms which allow its scientific study.
preacher, the hoarding tenacity of the peasant-plus myriad racial and cultural To begin with, in its scientific usage today, the concept personality has no
stereotypes which have also dogmatically and at times existentially circumscribed the common currency. Indeed, its meaning is very much determined by the theoretical
lives of those the labels seem to fit. persuasion of the scholar who is using the term. Although at least ten such viewpoints
For their thoughtful critiques of a draft version and for a good deal of the clarity of exist (Misiak and Sexton, 1966, pp. 169-170), two main assumptions are general
the present chapter, I would like to express appreciation to Alexander L. George, Daniel to most of these viewpoints. In the first place, personality is generally conceded to
Katz, and M. Brewster Smith. refer to the "internal dispositions" or "stable attributes" which an individual brings to
a situation. As part of this assumption, most personality theories posit some organiza-
tion by which intrapsychic attributes are grouped. This organization constrains the
30 Jeanne N. Knutson Personality in the Study of Politics 31

effectsof intrapsychic attributes and outer (extrapsychic) stimuli. Second, as Lazarus nonspecific). Like other concepts, it exists only as a theoretical construct (Kelly, 1955)
(1963, p. 37) notes: "One hallmark of personality is consistency, or stability. If we and thus requires specification in terms of entities which can be directly observed.
had no consistent personal qualities, we could not conceive of personality, since we But in spite of these obstacles, a basis can be provided for the empirical use of
would all be continually changing so much that we would scarcely be recognizable." personality in the study of behavior. This basis necessitates (a) conceptualizing per-
Thus, common definitions of personality assume two meanings: (a) organized sonality in terms less encompassing than the overarching concept itself and (b)
internal dispositions and (b) stability or consistency over time. To Allport (1937, specifying behavioral referents, thus providing "decision rules" by which the presence
p. 48), who recorded over fifty then current definitions: "Personality is the dynamic Dr absence of a particular aspect of personality may be objectively judged. Hence,
organization within the individual of those psychophysical systems that determine his though personality exists as a glibal (and as such unmeasured) concept, by con-
unique adjustments to his environment." Murray's (1968, p. 6) statement is more vention different personaIity theorists have specified their meanings of personality in
definitive: "A personality at any designated moment of its history (in middle life, terms of subconcepts with more or less clear objective referents (although these may
for example) is the then-existing brain-located, imperceptible and problematical be specified to vary with the situation). These subconcepts, which the global concept
hierarchical constitution of an individual's entire complex stock of interrelated sub- personality encompasses, are such entities as traits, values, needs, drives, and habits-
stance-dependent and structure-dependent psychological properties (elementary, subconcepts stemming from the general psychological theory that is intellectually
association, and organizational)." compatible to the theorist developing his view of personality (Sanford, 196813, p. 589).
As! the above definitions imply, personality in some unspecified way is seen to However (as Sanford points out), methods of "dividing" personality often involve
refer to organized, stable internal predispositions which each individual brings to a "arbitrary cuts" in the interest of some practical aim or "abstracting" features which
situation. These dispositions orient his behavior, and they vary among individuals. A appear of topical concern, proclivities which have provided the basis for most trait
major confounding factor, however, is recognized by all personality theorists (though measurement efforts.
acknowledged in idiosyncratic terms). That factor is the situational specificity of the Perhaps the best-known and certainly the largest (as well as perhaps the most
relationship between any one behavior and the personality construct to which it instructive) attempt at an inclusive definition of personality traits was begun by All-
refers. port and Odbert (1936) and carried out by Cattell (1957), whose search in a
standard English dictionary netted some 18,000 trait terms; from this initial list,
Even when there is little or no consistency at the level of behavioral acts, Cattell compiled a list of 4504 terms and then further reduced c this list to 171
ccsynonym groups," which presumably spanned the "personality sphere." Ratings of
there may be great stability or consistency in the hypothetical structures
and processes that determine these surface acts. To borrow an attractive these elements were submitted to cluster and then factor analysis, from which
expression from [Allport, 1937, p. 3511, "the same heat that melts the appeared five distinct, independent (and statistically orthogonal) factors. (These
butter hardens the egg." The same structure, when reflected in different Factors are Extroversion, Agreeableness, Conscientiousness, Emotional Stability, and
circumstances, may have superficially different, even opposite effects. This Culture. See Norman, 1963, p. 577). Yet, after all this effort to objectify traits, the
kind of consistency of determining structures and processes is of the utmost best evidence to date suggests that these five independent factors are commonly used
importance in personality theory [Lazarus, 1963, p. 391. English-language constructs that exist in the minds of the raters employing them-
leaving unanswered the more pressing question of whether these concepts also have
An empirically useful (that is, rigorously scientific) personality theory, then,
intrapsychic meaning for the subjects so rated. This conclusion was forced by the
must specify behavioral referents for its constructs; it must also delimit the situational
ingenious work of Passini and Norman (1966), who illustrated through a study in
specificity of these behaviors and/or the interaction pattern beween personality con-
which the same five factors emerged from judges1 ratings of the personalities of com-
structs and situational determinants. Thus, a third assumption underlying definitions plete strangers that the existence of these traits was located in the minds of the raters.
of personality is that behavior is related to a system of intrapsychic determinants in (For another attempt to build a trait system, see Guilford, 1959.) Similar to the
knowable ways and that these ways vary with the parameters of the external situation. work of trait theorists in substance, if not in terminology, are efforts to develop
For example, in the Berkeley study of authoritarianism (see Chapter Six in this inclusive Iists of needs which serve to motivate behavior (see Cronbach, 1963;
volume), a person could be viewed as possessing the trait of even Fromm, 1955; Linton, 1945; Murray, 1938).
though he acted aggressively only toward physically weaker persons and was com- Personality theorists who employ a trait model adhere to certain assumptions
pliant when challenged by a physical or psychological superior. 3bout intrapsychic functioning. The trait model, as Mischel (1968, p. 6) notes, is
At least two major obstacles, however, confound attempts to measure per- :umulative, in that indicators are related additively to the inferred underlying dis-
sonality according to the assumptions discussed. First, personality is unmeasurable position. (Through summing the frequency of behavioral indications of a trait, one
because it is global or general; that is, it is inclusive of the organization of all internal arrives at the intensity or quality of the trait the person possesses.) Trait theorists not
traits. Second, personality is a mediating construct; that is, it cannot be measured ! only measure individual traits in an additive manner; they also summate groups of
directly but can only be inferred from instances of behavior (which themselves are traits possessed by a person on the assumption that including more traits in their
 Jeanne N. Knutson Personality in the Study of Politics 33

model of personality allows a more complete understanding of the personality of the however, "the issue concerning uniqueness is settled in terms of predictive power,
subject. Thus, for example, if a subject's behavior is scored on the needs for creativity, which is greater when idiographic knowledge is employed" (Allport, in Smith,
endurance, intraception/extraception, and change (Murray, 1938), it is assumed 1971). Such a holistic viewpoint of course characterizes the psychoanalytic theories
that one will know more about the subject's personality than if his behavior is scored (Hall and Lindzey, 1957; Sarnoff, 1962) in which the understanding and predicti-
for only one of these needs. bility gained by such holistic labels as "obsessive compulsive" and "hysterical, con-
As should be clear from the above, a trait thus refers to a theoretically limited version type" need to be optimally supplemented by in-depth study of the person to
and empirically isolable aspect of a person's total personality. A trait focuses on whom the label is being applied.
individual (behaviorallyrelevant) differences on a single dimension. Trait statements While most nomothetic analyses of personality to date have been based on
are inherently unconcerned with the question of intrapsychic organization (relations trait approaches and holistic models have generally been reserved for clinically
between traits) and do not generally speak predictively regarding the existence and oriented case studies, such a distinction has been based more on convenience than
intensity of other hypothesized traits. necessity. As was exemplified in a recent study (Knutson, 1972a), it is feasible to
Other personality theorists attempt to make a global or holistic estimate which employ a holistic model on a nomothetic level. For discussions of the interrelated
is suggestive of a person's total psychic functioning. This group emphasizes the (rather than dichotomous) nature of nomothetic and idiographic levels of personality
necessity of first understanding the organizing principle of intrapsychic functioning analysis, see Falk ( 1956) and Beck ( 1953).
in order t6 assess, for example, whether passivity implies a dependent character or Explicitly, holistic theorists consider each personality as an open system, with
is a reaction formation (as in a passive-aggressive) against the expression of aggres- its own logical imperatives and organizing principle. Theorists with this perspective
sive impulses. Maslow ( 1943c, 1954), for example, has conceptualized human be- emphasize the necessity of knowing the function (D. Katz, 1960) that a particular be-
havior in terms of a hierarchy of five basic need areas: physiological, safety or security, havior serves for a person in order to understand what the behavior implies about
affection and belongingness, esteem (both self and social), and self-actualization. his personality. This functional principle is seen as the key to unlocking the manner
Because of the contingent influence of personality on behavior and the theoretical in which outer stimuli and intrapsychic constraints are organized. From this viewpoint,
complexity of these need areas, however, such holistic personality constructs can only it is meaningless to state that a man who works for many political causes possesses
with difficulty be defined in ways that allow their empirical validation and the assess- the trait of "active orientation." A relevant statement (which necessarily implies a
ment of their effect on political behavior, (For works that attempt such linkage, see phenomenological stance) would rather label his personality as self-actualizing, other-
Knutson, 1972a; Simpson, 1971; Davies, 1963.) directed, hysterical, and so forth. Such a label would not only subsume the meaning
Holistic theorists assume that the personality concepts which they have ab- of the observed behavior but would also be predictive of the individual's functioning
stracted address themselves to the total manner of a person's functioning-to the along dimensions yet unstudied.
structure, dynamic organization, and operation of any one personality. In this tradi- Holism in personality theory is not only grounded in a consideration of the
tion, Maslow (1943b, pp. 528-529) speaks of discrete "syndromes" of personality: individual psyche, but-adding to the difficulty of employing such concepts in
rigorous empirical research-holism is inclusive of every isolable subconcept in the
In our definition of the syndrome, the main quality which characterizes personality area. For example, Murray (1968, pp. 6-7) declares (in a statement
the whole ("meaning," "flavor," or aim) can be seen in any of its parts if which classically summarizes the indistinct and qualitative nature of much holistic
these parts are understood not reductively, but holistically. Of course this theory) :
is a theoretical statement and we may expect to find operational difficulties
with it. Most of the time we shall be able to discover the flavor or aim of In this PS [personological system], "personality," the most comprehensive
the specific behavior only by understanding the whole of which it is a term we have in psychology, is given a functional meaning embracing
part. And yet there are enough exceptions to this rule to convince us that everything from basic temperamental variables-for example, energy level,
the aim or "flavor" inheres in the part as well as in the whole. hedonic level, affective state of being-to such higher mental processes as
may be devoted to superpersonal (cultural) endeavors-for example,
This holistic viewpoint is thus grounded in understanding the organizing principle artistic, historical, scientific, philosophical. Consequently, even by restrict-
(for example, the dominant need area) of each personality and in the belief that an ing one's attention (as one inevitably must do and should do) to the most
adequate measure of a single personality's functioning is predictive of all aspects of important properties, a personality cannot yet be adequately represented
the personality. as a functional and temporal whole in less than 5000 words, let us say;
While holistic theorists agree that constructs such as traits, narrowly defined certainly not by a short list of traits.
needs, and drives have a large degree of commonality (thus allowing nomothetic
statements), they stress that any one personality can be completely understood (and From the above distinctions, it is apparent that a holistic personality construct
thus completely predicted) only in relationship to itself. Nomothetic analysis, then, would be much more difficult to employ in political research than would a trait
can bring some general understanding of the operation of personality in behavior; theory, in which specified, disparate traits (with presumably clearer behavioral
34 Jeanne N. Knutson Personality in the Study of Politics 35

referents) are used to stand for the concept of personality. Yet ease of measurement attributes of personality which can be subjected to scientific study (that is, quantified
is perhaps deceptive, for (at least from a holistic perspective) when an individual's or in other ways objectively assessed) as well as to specify how these attributes, which
functioning is assessed in terms of several discrete areas, most of his predispositions stand for the concept personality, relate to the concept as a whole. (To these concerns
may be left unmeasured-and an understanding of the interactions among his pre- we turn in the pages that follow.) MacIverYs(1937, p. 26) pointed statement about
dispositions in the various, discrete dimensions may be obviated as well. For example, the measurement of attitudes aptly summarizes these issues: "In their zeal for mea-
a trait theorist who sets out to measure aggressiveness and conformity does not surement [psychologists] fail to ask, what in the attitude is it that we are undertaking
necessarily make any assumptions about the traits of intellectuality and creativity as to measure? We do not measure things, but only certain quantitative aspects of things.
well (nor about the relationships between these traits-although he may hypothesize We do not measure a table, but its length and breadth and height and weight. We
the existence of certain patterns of covariance), but a holistic theorist who makes a do not measure the sun, but its radiation, the composition of its light, its size, its
global personality-relevant statement clearly does make such assumptions. weight, its apparent motion among the stars, and so forth. What aspects then of an
While Murray, Freud, Erikson, and others who employ a holistic approach to attitude do we set out to measure?" And how, we must add, do these aspects relate
personality discuss this viewpoint in different terms from those used by Maslow, the to the personality and behavioral dimensions which we have defined as our areas of
basic belief is remarkably similar: Each behavior in which a person is psychically concern?
engaged provides an expression of some aspects of his total personality structure;
that is, his basic needs, drives, motivations, as well as his idiosyncratic manner of Defining and Assessing Personality
coping, his response patterns, his habits. The obverse of this belief is extremely
important in understanding the holistic viewpoint. Organization is more basic than The area of personality measurement has proved to be a difficult domain for
analytic elements. A person's total personality forms a gestalt, the parts of which can those concerned with objective assessment. After the great flush of success accompany-
be separated only theoretically and then with peril to the meaning of the part in the ing the Woodworth Personal Data Sheet in World War I (Holtzman, 1968), there
person's psychic economy (Katz, 1960). Thus, the employment of the holistic view followed the sobering realization that a good deal of the early promise was due to
stems from dissatisfaction with the usefulness of knowing only that a person has a artifacts of the measurement situation and that "much of what is called behavior
certain intelligence quotient or dogmatism score, or uses a certain number of cogni- rating in fact involves higher-order behavioral interpretation, and research on the
tive categories of certain width through which to process the information about his correlations between diverse behavior ratings actually correlates diverse behavioral
world, or tends to register a certain degree of anxiety (masculinity, introspection, and interpretations or constructions" (Mischel, 1968, pp. 68-69) .
so forth) in relation to a sample or standardized population. Basically, the difficulties in measuring personality center around the inevitable
Both trait and holistic definitions of personality remain hypothetical con- levels of inference involved. One level of inference is added as objective and measur-
structions, with an explicit existence limited to the eye of the beholder. Between able phenomena must-in every personality system-stand for the personality concept.
these constructs and measurable phenomena lie levels of inference which greatly Personality is understandable only in behavioral terms: "Personality measurements,
compound the difficulty of measuring personality. Yet personality can be understood if they discriminate at all, always express, in a pure or impure form, explicitly or
only as an inference from behavior. As Murray (in Smith, 1971, p. 356) has noted, implicitly, a personality concept in terms of which the differential behavior may be
it is imperative to "distinguish between the facts and manifestations of personality, understood. Personality concepts, if they are not hopelessly vague or inconsistent,
on the one hand, and the formulation or conception of personality on the other. always imply specifiable behavioral differences in people" (Block, 1968, p. 30). Yet-
. . . The distinction is the same as that between symptoms and signs versus diagnosis. to add additional layers of inference-we have seen that a discrete item of observable
Personality is a diagnosis." Further compounding the problem of measurement, we behavior cannot be tied to a specific meaning: behavioral correlates of personality
have seen that a discrete, observable behavior (that is, a somatic or verbal response) are dependent upon situational parameters. Conversely, the same behavior may be
cannot be tied to a specific functional or dynamic meaning: behavioral correlates of expressive of different intrapsychic meanings for different individuals, as well as
personality are dependent upon situational parameters. having different meanings for the same individual at different times. An inadequate
For many research problems, personality may not be a useful variable to appreciation of these factors has been responsible for much unfounded criticism of
include-in spite of the Siren's call to employ easily available and easily administered progress in personality research. (See Alker, 1972.)
scales. In a number of politically important areas, situational, role, and demographic Thus, the first requirement of a system of personality measurement is to
factors offer the most parsimonious avenue to predicting political behavior. Thus, the specify, in terms of a trait or holistic theoretical perspective, the behavioral correlates
first job of the political psychologist interested in employing personality variables in by which personality is defined and to make these definitions situation specific. Other-
his research design is to make sure that he has explicated an adequate theoretical wise, as Mischel (1968) rightly notes, so-called "objective" tests lose their objectivity
basis for the inclusion of the dimension of personality. because intuitive interpretation is needed to assess the meaning of standard scores.
In order to operationalize a research design in which personality is hypothe- Such specification of the meaning of behavior includes a definition of the role of
sized to have an integral part, the political psychologist must first concept~alize "moderator variables" (Wallach, 1962; Kogan and Wallach, 1964)-that is, inter-
clearly what he means by the concept personality. The next step is to delineate actions among several variables which shape the influence between any one variable
 Jeanne N. Knutson Personality in the Study of Politics 37

and the rest of the data. (For example, as we noted in the case of authoritarianism, A second step to the measurement of personality involves a choice of method-
the presence of psychological or physical authority mediates or moderates the ex- ology. For example, if one wishes to measure anxiety, Block (1968, p. 31) notes that
pression of aggression.) there are at least four basic ways of proceeding:
Let us consider this requirement in the light of Lane's (1962) excellent in-
depth study of the roots of the political ideology of fifteen working-class citizens. (1) A subject may be observed in his everyday life and actions (presum-
Much of Lane's discussion focuses on the personality traits that distinguish a democrat ably without his actions' being affected by these observations) and from
from an undemocrat. His conclusions in this regard are based on a wide-ranging his behaviors a judgment is made as to whether he is anxious; (2) a sub-
ject may be asked, by means of a questionnaire, to state directIy (or in
analysis of democratic theory and his largely impressionistic evidence from these
ways he may not completely understand) whether he is anxious; (3) a
fifteen men. On the basis of Lane's data and additional items derived from subject may be placed in a controlled or test situation designed to elicit
democratic theory, it would be possible to conceptualize a cluster of personality special behaviors or projects relevant to anxiety; (4) a subject's physio-
correlates of democratic behavior (see Knutson, 1972a, ch. 5 ) , such as open-minded- logical reactions may be assessed by various instruments, to determine
ness, tolerance of others and of ambiguity, and low anxiety level, and then to define whether he shows certain responses or changes presumed to be indicants of
these variables in terms of measurable, situation-specific behavior (including, but not anxiety.
limited to, performance on questionnaire items). For example, "tolerance of am-
biguity" could be defined, as it was in a study by Block and Block (1951), as the As will be discussed (also see Knutson, 1972c), political psychologists have unneces-
lack of development of a frame of reference in an experiment employing the auto- sarily limited the methods they have employed to study personality. Since each method
kinetic effect of apparent light movement. nposes inherent limitations on the types of data that may be acquired, some methods
Next, the political variables need also to be made behaviorally specific. System- re more appropriate for certain purposes than are others. (The special value of
supportive ''democratic behavior" could perhaps minimally be defined as referring to ertain major approaches is discussed in the methodology section of this book.)
such variables as voting in the last four presidential elections, paying taxes regularly, In addition, the measurement of personality makes a third requirement-'
possessing a certain degree of political information, and obeying laws protecting the that of cross-situational measurement, an area which Mischel (1968, pp. 86, 78) has
person and property of others. (Democratic behavior could of course be defined in summarized as being fraught with peril for the quantifier: "The correlations between
a number of other ways-for instance, by specific outcomes of interpersonal relations yeasures by themselves cannot be interpreted as evidence for the associations between
as defined in a laboratory experiment or by the prevalence of certain story themes on le labeled traits because diverse trait names often cover highly similar operations
TAT cards, as discussed in Chapter Fifteen in this volume. In this case, however, the
use of such methods would run the very real risk of measuring personality traits a
lat require subjects to do similar things... [and] the correlations obtained among
ersonality measures to some extent may simply reflect their common associations
second time, under the guise of political outcomes.) 7ith intelligence and education." As Campbell and Fiske (1959) note, it is necessary
At this point, it is possible to offer a series of testable hypotheses: "Personality ) ascertain construct validity through the employment of a multitrait-multimethod
variables x, y, and z will be predictive of (positively correlate with) a cluster of un- iatrix, in which the designated trait is more highly correlated with itself across
democratic behaviors, defined as a, b, and c." We could then replicate Lane's inten- methods than it is correlated with other traits within any one method. I t is also
sive analysis of fifteen working-class citizens and supplement our data gathering with necessary to show what is not being measured, as well as what is being measured-
an anaIysis of fifteen convicts at the state prison who have identical demographic and thus to achieve what Campbell and Fiske (1959; Campbell, 1960) have labeled
profiles but have never engaged in any of the activities which we have defined as discriminant validation."
supportive of democratic systems. Objective scoring methods, plus a blind analysis of , When one ascertains that a subject's score is indicatively similar across methods
our open-ended interviews by two coders unfamiliar with our assumptions, would
nd o v ~ time,
r he may feel some assurance that he is tapping a dimension with some
then allow us to test whether it is possible to separate, on the basis of personality
stability and consistency and (if the scientific requirements of the above methods are
traits, subjects whose behavior conforms to our minimal requirements of a democrat
observed) some validity for the construct being measured. One may not, however,
from those whose behavior does not conform.
assume that the subject's "personality" is being measured--except in the limited sense
Unless political psychologists proceed through such steps as outlined above,
of one dimension, which has been isolated for the purpose of a specified research
continual impressionistic work with personality variables becomes intellectually limit-
design.
ing on several counts. First, we run into the very real danger that our model of the
. However, when-in terms of Maslow's (1954)' need hierarchy, for example-
personality correlates of democratic behavior has developed along with our data
one aligns subjects on a dimension of self-actualization, one makes a different set of
analysis and has been subjectively tailored to fit our data, thus offering no testable
assumptions-assumptions that need to be specified in advance. In terms of holistic
analysis of any point which we may set out to study. Second, we obviate the addi-
theory, one is measuring an intrapsychic organizing principle, which, by a pre-
tional opportunities which more systematic research provides to extend and revise our
determined theoretical stance, specifies the position of the subject on a host of other
theories. Such opportunities come not only when our hypotheses are supported but,
dimensions, allowing for the process that Cronbach and Meehl (1955) have carefully
perhaps even more when they are not-and, continually, when we take the rewarding
defined as "construct validity." Thus, a person categorized as "self-actualizing" is
time to engage in deviant case analysis (Peak, 1953).
38 Jeanne N. Knutson Personality in the Study of Politics 39

assumed in holistic terms also to be categorized as (relative to other subjects) creative, that political behavior results from intrapsychic predispositions being displaced on
flexible, empathic, open-minded, and so forth (Knutson, 1972a) and as not dogmatic, public objects. Yet, in all the years since Lasswell's early work, the assumption that
rigid, extrapunitive, and so forth. (For a classic discussion of the demands made by personality at least partially determines political beliefs and political behavior has
a hierarchical model employing stages or "milestones," see Loevinger, 1966.) Hence, received inadequate critical analysis.
while holistic personality theory allows the concurrent measurement of a number of This statement should not be taken to imply that a relationship between the
dimensions, it makes much heavier demands upon the accuracy of the measuring in- , intrapsychic and the political is totally unsupported. On the contrary, a wide variety
strument. of correlational studies have successfully predicted politically relevant attitudes, be-
Part of the difficulty in personality research to date appears to stem from lack liefs, and behaviors from personality variables. Both theoretically and empirically,
of rigor in areas of conceptualization, definition, and assessment. In addition, in- however, such correlational studies have failed to support Lasswell's contention.
adequate attention has been paid to the criterion problem of validating the meaning Theoretically, the studies have generally avoided the difficult task of clearly expli-
of the measures employed (as was done in the classic study of authoritarianism; see cating the specific linkages by which personality influences political behavior and,
Chapter Six). Only by attending to the criterion problem can newly developed in turn, is influenced by political behavior. Empirically, these studies have generally
measures become the basis for cumulative knowledge'; others who employ them then failed to illustrate the stability or coherence (or both) of personality over time as
have some confidence in their meaning. Another difficulty rests on a fundamental well as its constant (or predictably variable) relationship to matters political. (The
conceptual difference between those personality researchers who see promise in their best evidence for the consistency bf personality over time appears in Block and Haan,
work and those who feel that the utility of personality constructs has been disproved. 1971; Jones, Bayley, MacFarlane, and Honzik, 1971; Kagan and Moss, 1962.)
Hunt ( 1965, p. 81) clearly states this issue : "If one takes the square of the coefficient The theoretical view that personality is a stable and/or consistent attribute
of correlation as a rough, 'rule-of-thumb' index of the proportion of the variance that shapes an individual's political beliefs and activity retains a powerful intellectual
attributable to persons, it would appear to be limited to somewhere between 4 and 25 appeal, an appeal grounded in psychiatric experience and in clinical data fed into
percent of the total. This is incredibly small for any source which is considered to be the literature of politics (Greenstein, 1969; Greenstein and Lerner, 1971; Knutson,
the basis of behavioral variation, but we personologists have blamed our instruments 1972a). Perhaps this view is most clearly expressed in the frequently made distinction
rather than our belief in the importance of static dimensional traits." between indirect and direct political learning (Dawson and Prewitt, 1969) ; accord-
Personality-in what could be labeled "the fallacy of reductionism1'-fre- ing to this view, certain hypothetical levels and processes of learning account for both
quently and mistakenly has been considered synonymous with behavior. Consequently intrapsychic stability and the constraints of experience, societal values, and situation
(as suggested by the quotation above), personality measures which in standardized (Smith, 1968c; Greenstein, 1969).
situations at best account for 25 to 30 percent of the variance are rejected as but In a seminal analysis, Almond (1960b, p. 28) has discussed the difference
tangentially related to the end goal of predicting and understanding BEHAVIOR. The between "manifest" and "latent" socialization. Manifest socialization "takes the form
alternative view-that intrapsychic factors do account for only about one third of of an explicit transmission of information, values, or feelings vis-8-vis the roles,
the variance in most overt behavior (realizing of course that some behaviors are inputs, and outputs of the political -system." I t is the most obvious and frequently
totally determined by inner or outer constraints)-is given little currency. As discussed studied means of transmitting cultural values and approved norms of behavior and
in more detail below, I would suggest that social and cultural predispositions are includes the variables of adult reference-group constraints, formal learning in school
responsible for another 25 to 30 percent of the variance, leaving approximately half and church, and other structured life experiences. Yet, in Almond's view, the
of the variance to Ise accounted for, under usual conditions, by the actual field individual has much earlier begun a process of "latent" socialization, which "takes
situation ( as manipulated, for example, by the laboratory experiment) . the form of a transmission of information, values, or feelings vis-a-vis the roles, inputs,
Most writers today (as this volume attests) agree that a multivariate approach and outputs of social systems such as the family which affects attitudes toward
which encompasses all the above factors is necessary for the analysis of political analogous roles, inputs, and outputs of the political system." Such factors as personal
behavior. In employing personality constructs as part of such an approach, certain values, self-concept, unfulfilled needs, and sense of competence (or ego strength)
critical issues involving the linkages between personality and political behavior need tap the personality dimension of such a latent, personally consistent manner of
to be delimited. I t is to this task that we now turn. responding.
Theoretically, such a model can be expressed as shown in Figure 1. In this
Influence of Personality on Behavior model, basic personality needs (such as the need for security, in MaslowJs theory;
the need for trust, in Erikson's model; or a thwarted need for aggression, in Freud's
In political psychology, it has become a truism that personality-in some analysis) are seen as developing at an early stage in the child's life cycle, so that the
unspecified way-affects political beliefs and political activity. This assumption can school-age child's personality can be characterized holistically in terms of certain
be traced back to Plato, who expressed a concern with the promotion of personality ascendant needs, which are organized in a dynamic pattern. (See, for example, the
growth supportive of the polity. I t received general professional acceptance through excellent analysis of esteem needs in elementary children by Coopersmith, 1967.)
the seminal work of Harold Lasswell, whose books (1930, 1948) advanced the thesis The above model-moving one step back from Rokeach's (1960) assertion that an
 40 Jeanne N. Knutson Personality in the Study of Politics 41

FIGURE
1. LINKAGES
BETWEEN PERSONALITY
AND POLITICS analysis. For some purposes, a knowledge of psychogenesis and total personality
configuration may be essential; for others, it will be superfluous.)
A study by Browning and Jacob (1964) of the frequently held view in political
LEVEL I Basic personality needs 4
- psychology (see below) that power motivation is a major determinant of political
activity provides an excellent example of an adequate Level I1 analysis. Drawing a

LEVEL I1
I
Idiosyncratic belief system regarding self, others, and
causality; and idiosyncratic modes of relating
representative sample (N = 50) of the elected officials in two parishes in Louisiana
and, in an eastern city, a random sample (N = 23) of businessmen-politicians, these
investigators used a projective technique to measure the needs for power, achievement,
and for affiliation. The study produced two important findings: (1) None of the
individual motivations discriminated between politicians and nonpoliticans of similar
occupation and status; but (2) "politicians in high-potential positions scored much
higher in both achievement and power motivation than their matched sample." Thus,
the authors concluded that the effect of these personality traits is mediated by situa-
tional constraints. Avoiding the knotty, costly, and clouded search for total personality

v i
v
configurations and developmental understanding, Browning and Jacob present a
Level I1 analysis which inventively delimits one area where individual predispositions
interact with environmental constraints; and they adequately address the question of
why certain types of roles may be peopled by certain types of actors.
LEVEL I11 Social-cultural learning Situational and By contrast, a recent analysis of presidential character (Barber, 1972a) s u g
(values, roles, beliefs experiential factors
which may or may not gests some of the limitations of an examination that remains at Level 11. Barber
(Informal)
become internalized) amply and in fascinating detail documents the contention that personality does indeed
(Formal) shape presidential actions and presents (pp. 445-454) a valuable analysis of likely

J
relationships between presidential character and the situational and psychic needs
of the constituency. Turning to the weightier question of the linkages by which
personality influences political behavior, Barber offers two dimensions (active-passive;
‘y v
positive-negative), from which he develops a fourfold categorization of the presidents
whose character he analyzes. Unfortunately, his model of personality styles is too
LEVEL IV POLITICAL IDEOLOGY AND POLITICAL ACTIVITY
general to be predictive of the behavior of the men whom he studies, and the reader
is left with the impression that a person's psychic energy level and affective orienta-
tion are an inadequate basis by which to judge his future performance in specified
individual's belief system is built on (cadiates out from) certain basic, "central," or situations. For example, in Barber's "active-negative" category Wilson, Hoover,
6C
primitive" (that is, unquestioned) beliefs-suggests that a person's unquestioned Johnson, and Nixon become uneasy bedfellows. When the idealistic, humanitarian,
views of himself, of human nature in general, and of the nature of causality are socially sensitive Hoover is linked with the pragmatic, power-oriented, and often
determined by his basic personality needs. Supportive of this viewpoint, as I have consciously ruthless Johnson, one becomes aware that intrapsychic needs, goal states,
illustrated and reviewed elsewhere (Knutson, 1971; 1972a), is the fact that certain and internalized values are as important as coping strategies in "predicting per-
need-related attitudes almost invariably cohere. If, for example, other people are formance."
seen as trustworthy, control tends to be internal (see Rotter, 1966), and the viewpoint Because one looks in vain for clear decision rules by which a future president's
is optimistic, tolerant, and efficacious. specific behavior could be predicted or which place him in this rough categorization,
Our earlier discussion of the ways in which personality is conceptualized is one is unfortunately forced to conclude that while Barber's argument is a useful
directly related to Levels I and 11. When a political psychologist speaks in holistic coarse division of presidential character, it needs to be supplemented by Level I
terms, he is of course referring to Level I. When he speaks, on the other hand, of analysis, which specifies the needs the individual seeks to gratify as well as the
such traits as dogmatism, machiavellianism, low self-concept, intrapunitiveness, and organizing principle of those needs. (For example, Barber's evidence suggests that
aggressiveness, he is making dynamic statements about Level I1 as it relates to politi- three of his "active-negative" subjects-Wilson, Johnson, and Nixon-were motivated
cal behavior-rather than dynamic and genetic statements about the nomothetic by the often interwoven needs to express aggression and enhance self-esteem, while
or idiographic origin and organization of the stuff of which politics is made. (This Hoover lacked libidinal gratification and was motivated by needs for affection and
does not imply a value judgment about the advisability of employing either level of belongingness.) Finally, since the influence of personality is contingent, it would be
42 Jeanne N. Knutson Personality in the Study of Politics 43

necessary (after assessing basic needs) to understand the multidetermined ways in of factors that shape Level I11 elements; because of our focus on personality, such
which these basic personality needs influence political behavior, given specific personal factors are omitted here. A more important, substantive difference is that our
values and political constraints. model makes divisions in Smith's "Personality Processes and Dispositionsy'level. With
Let us look again at our model of intrapsychic linkages. Once established, our concern here for the importance of intrapsychic dynamics, our model stresses the
basic personality needs (usually mediated by the person's belief system and standard- tentative nature of social learning and the necessity (in order to successfully predict
ized modes of relating) interact with direct political learning and situational and behavior over time) that there be a close fit between Levels I and 11, on the one
experiental factors (Level 111)-which, in turn, interact with (and thus may hand, and Level 111, on the other. Much of the socialization research (see Chapter
modify) basic personality needs. For example, a person who is self-actualizkg and Five in this volume), for example, has focused on the means by which social agents
who thus has a personalized belief system characterized by empathy, flexibility, and instill social values and attitudes and on counting what these values and attitudes
tolerance may nevertheless-through direct learning of social values in an intolerant (and, presumably, behaviors) are. The above model emphasizes the necessity of
society (Maslow, 1957; Pettigrew, 1958)-accept intolerance as a major belief. Yet, determining the consistency and stability of Levels I11 and IV in terms of their
through personal experiences (such as interpersonal relationships or a university relevance for the psychic needs and perspectives of the individual (Levels I and 11).
education) his values and attitudes may be modified so that they more closely fit his Finally, turning to Level IV, our model illustrates that personality predisposi-
idiosyncratic belief system (which includes the belief that human beings are intrin- tions, social and cultural values, and situational and experiential constraints (as well
sically valuable) ; and this changed orientation may be sustained in later years. Thus, as the interaction between these three factors) lead to a political ideology and to a
it is important to note, socia1,changes in values and beliefs come about because direct level of political activity. Under this model, even though peripheral attitudes (that
socialization, until verified by experience and consonant with inner needs, is tentative is, attitudes in which the person has no psychic investment) may be incongruent, on
-as studies of major shifts in college years have shown (Newcomb, 1943; Sanford, a deeper and more meaningful level the person holds a consistent political ideology.
196213). Finally, "growth" experiences may fulfill psychic needs and allow the person When a survey analyst declares that on a set of basic political issues a respondent's
to function on a more competent intrapsychic level; or, on the darker side, psychic attitudes are inconsistent, the personality theorist must point out the need to find a
trauma-as exemplified in the life of Kurt Schumacher (Edinger, 1965)-may cause deeper level, on which consistency may rest securely and meaningfully (Sears, 1969),
a basically healthy person to become, in Maslow's terms, psychically deprived. Such and this can be done only by attending to the psychological meaning of the attitudes
changes in basic personality needs would then initi'ate correlative changes on Level for the person in question. Thus, only when basic personality needs are uncovered is
I1 and the formal belief area of Level 111, it possible to make "psychological sense" of an activist who advocates both extending
Anrinstructive example of the importance of distinguishing among Levels I, numan rights of ethnic groups and sterilization of "welfare mothers" in order to
11, and I11 is found in a study of political socialization among black children :ontrol population, or a citizen who abhorred the organized inhumanity in Vietnam
(Liebschutz and Niemi, 1972). Using as baseline data a 1969 survey of 886 children but supported his president's perseverative hostility in that part of the world. (See
in grades 2 .through 8, the authors examined the effects on the children's politically Zellman and Sears, 1971, in this connection.)
relevant attitudes of a curriculum designed to improve the self-concept of disad- The above model offers a further caveat, which has been a constant theme of
vantaged black children. (These attitudes included a sense of political efficacy and this chapter: that is, the multidetermined nature of the behavior and beliefs in which
evaluations of various political authority figures and of the fairness of laws). In results 3olitical psychologists are interested. As Maslow stated in an early article (1943c,
that are both psychologically and politically important, the authors found (a) that 3.384) : "Motivation theory is not synonymous with behavior theory. The motivations
there were some changes during the experimental intervention toward increased are only one class of determinants of behavior. While behavior is almost always
realism in political evaluations and a greater sense of political efficacy-although the notivated, it is also almost always biologically, culturally, and situationally deter-
changes are generally small and may not be statistically significant; and (b) that this nined as well." Indeed, Kagan (1972) posits the view that there is no necessary
concerted intervention apparently had no long-term effects on political attitudes. relation between motive and either action or affect.
Unfortunately, the concept of the self (the independent variable) is not clearly As others have also noted, behavior is a result of the interaction of personality
specified, nor do we learn whether a regression effect occurred in this area also. What predispositions, social and cultural precipitants, and the field situation. I would sug-
does seem apparent, however, is the difficulty of changing personality predispositions gest that a major reason for the lack of generality of political and of personality re-
on Level I1 and the formal side of Level I11 once they are established-especially search is that too many researchers explore one of these dimensions as if it operated
when the target attitudes must represent a departure from the reality experienced by in a vacuum. Yinger (1965, p. 587) caustically sketches the relevance of such work:
the child (the informal side of Level 111) and the unmet basic needs which can be \

hypothesized to motivate ghetto children (Knutson, 1972a). If one lives in an environment that never falls below O0 C. or goes above
The careful reader will note, at this point, some differences between the above 100° C. he can afford to predict, solely on the basis of the "traits" of
model and Brewster Smith's well-known "map for analysis of personality and politics" H,O, that it is a liquid. This is an adequate way of saying that H,O, a
(1968~;also see Chapter Three in this volume). Most obvious is Smith's inclusion compound with certain potentialities, is liquid under certain conditions
44 Jeanne N. Knutson 'ersonality in the Study of Politics 45
which, being constant, can be disregarded. That does not mean, however, nergy level-which is predictive of the person's ability to engage in all forms of
that the conditions are not always involved in producing the results. )cia1 behavior, including that categorized as political.
Thus, it is useful to assess the personality and politics literature from the re-
Research on Political Relevance of Personality tio on ship between personality and two dimensions of political involvement: activity
and ideology. Under the first category belong questions as to the amount (apathetic,
One can employ a number of conceptual schemes by which to categorize the conforming, activist), type (leader, follower, isolate, joiner, decision-maker or im-
personality-relevant research in political psychology. In the two major, summative plementer) and quality (flexible, rigid, creative) of activity. In the second category
works in this area (Greenstein, 1969; Greenstein and Lerner, 1971), Greenstein has fall questions as to the direction and ideological intensity which political activity is
utilized a tripartite division of the literature into studies of individual political actors, likely to take. ( I t is important to note that little research has been done on the critical
types of political actors, and aggregates of political actors. Sigel (1970, pp. 232-236), relational area between the activity and ideology dimensions.)
on the other hand, classified the literature on personality and politics into analyses of Throughout the accumulating research, a basic principle which appears is
populations and biographies of individuals-with the latter subdivided into the "hero" iat the inpuenct of personality is directly related to the specificity of the politically
school and the motivational or "antihero" school. Lane (1968) discussed the rele- !levant behavior. For broadly popular acts, in other words, social and cultural norms
vance of personality to ideology, as well as to legislative, judicial, and electoral behav- ften "carry the action" so that behavior does not engage deep levels of intrapsychic
ior; Levinson ( 1968) covered "ideology-personality constellations," using such mctioning. Hence, to become a legislator appears to require. either considerably
categories as conservatism, liberalism, authoritarianism, and alienation. (For other lore or less esteem (Barber, 1965) than to be a member of the citizenry who elect
major discussions, see Milbrath, 1965; Lane, 1959b.) ie legislators. Similarly, it seems to require more psychic competence to adopt a
In this chapter, I shall follow a somewhat different approach--one dictated eviant political stance than to work for a major political party (Knutson, 1971) .'
by what I see as the relevance of personality for the study of political behavior and . second major principle (discussed above but worth keeping in inind as we review
belief (Knutson, 1972a). In considering broadly popular behaviors, such as joining sample of the literature) is that the effect of personality is mediated by and
a major political party or voting in presidential elections, we now recognize that ~ediatessocial, cultural and experiential factors.
personality variables are unlikely to account for a significant amount of the variance, Level of Activity. A number of personality traits relating to individual
because of the well-studied importance of social and cultural factors in determining olitical activity have been isolated. One of the most frequently studied is the trait
a person's basic political orientation (Campbell and others, 1960; Flacks, 1967) and F efficacy or competence (Douvan and Walker, 1956). Interest in this concept
his conformance with modal political activity. Milbrath and Klein (1962, p. 54) have ?came widespread with the development.by the University of Michigan's Survey
noted another constraint on the impact of personality on political behavior: "Political , .esearch Center of a scale measuring "political efficacy" (Berelson, Lazarsfeld, and
participation seems to Ee a special case of a general social participation pattern. icPhee, 1954; Campbell, Gurin, and Miller, 1954). As a recent analytic discussion
Personality factors requisite for general social participation are also requisite for e this dimension makes clear (Renshon, 1972), both the unidimensionality of this
political participation, but their presence does not necessarily produce political :ale and its theoretical basis have been accepted without a thoughtful critique, which
activity. We are not aware of any study which has isolated a personality trait which gitimization required. Nevertheless, a multitude of studies, in addition to the voting
drives people specifically into politics; even the much discussed "drive for power" udies carried out by the SRC itself, have demonstrated that there is a relationship
finds many alternative modes of expression." :tween an inner feeling of potency and the degree of political activity in which one
The literature of political psychology, however, suggests that personality varia- igages (Eulau and Schneider, 1956; Janowitz and Marvick, 1953; Milbrath, 1965;
bles are selectively felt at a number of points in the political process. One point, for gger, Goldstein, and Pearl, 1961; Almond and Verba, 1963; Rosenberg, 1954;
example, involves the selection process for leadership roles; another is the inner pre- arber, 1972a). In personality theory, a sense of efficacy has been related to the
dispositions which lead individuals to engage in deviant forms of political activity sgree of psychic competence a person possesses (Smith, 1968b).
or to adhere to deviant political beliefs (although it must be underscored that no A second dimension of personality which has been related to political activity
intrapsychic commonality accounts for all forms of deviancy). Personality factors are authoritarianism (see also Chapter Six in this volume). As I have summarized
also selectively felt in the process of political learning because values and patterns sewhere (Knutson, 1972a), research to date has presented a conflicting view of the
of behavior are inculcated which may or may not find the psychic resonance necessary ccmtribution of authoritarianism to the level of political activity. This lack of clarity
for their behavioral actualization. Again, personality is clearly important in under- is;, in no small measure, due to the fact that authoritarianism is clearly compounded
standing the manner in which a person carries out the roles to which (often because of : intrapsychic as well as social (demographic) factors (Greenstein, 1969). In terms
of cultural and societal constraints) he has been assigned. In this regard, personality
I the model of personality linkages presented earlier in this chapter, it seems likely
at authoritarianism (in general use, a Level I1 variable) influences political
serves to define the use of coping or defensive mechanisms (Kroeber, 1966)-that
:tivity in ways dependent upon the personality needs (Level I ) of the individual
vital distinction which precludes or advances the overall attainment of political goals.
who scores high on this trait and the values (Level 111) into which he has been
In a similar way, personality variables allow the political psychologist to gauge psychic
socialized. Thus, authoritarians with affiliative needs will be likely to join social and
46 Jeanne N. Knutson Personality in the Study of Politics 47

political groups, a relationship which Lane (1955) found to occur in one set of data. view political leaders (or anybody else) as driven by one narrowly defined need.
But, as a study by Janowitz and Marvick (1953) suggests, security needs may account or example, Maslow (1954) has analyzed the need for self-esteem as composed of
for the correlations found between authoritarianism and age; lack of education, and number of separate subneeds; the gratification of each presumably can bring a
lower socioeconomic status, as well as for the relationship between authoritarianism :nse of fulfillment to the esteem need area. I would suggest that research here should
and isolationist attitudes, lack of political efficacy, and nonvoting in the 1948 ~cuson the idiosyncratic ordering of politically relevant needs for which any one
election. That there are different predispositions to authoritarianism is clear from past olitical leader strives-an ordering which, as George (1968) notes, is sensitive to
research on the F scale (see the interesting work by Pettigrew, 1958), but how such le person's perception of the situational opportunities for their gratification. For
differences relate to the effect of authoritarianism on political activity level has ample, the person who strives for power but also places a high value on human
never been adequately studied. In terms of the model discussed in this chapter, fe (as Robert Kennedy appeared to do), is quite different from the leader (such as
authoritarianism appears to present an example of the need for an holistic approach [itler) who was motivated by unfulfilled power needs but devalued human 1ife.I
to the study of personality and politics. For work in the area of value ordering, see Rokeach, 1968.)
Another area in which a personality trait has proved useful in the under- Each of the above personality traits, as the reader might imagine, has been
standing and prediction of political activity is that of anomie and alienation (see .udied-not only to determine its relation to political participation but also to assess
Chapten Seven in this volume for a more detailed discussion; see also Knutson, ze quality and quantity of role performance by individuals who have successfully
1972a, ch. 3). This dimension has been conceptualized and measured in rather assed through the political selection (or self-selection) process. For example, a
different ways, including a sense of powerlessness or internal-external control (Rotter, umber of studies (for instance, Nie, Powell, and Prewitt, 1969a, 1969b) have sug-
1966; De Charms, 1968), a generalized sense of despair or meaninglessness defined gested that individuals lacking such traits as efficacy, if they are members of organiza-
as "the breakdown of the individual's sense of attachment to society'' (Srole, 1956; tions that value political participation, may nevertheless be politically active, though
see also McClosky and Schaar, 1965) or more focused types of alienation (Seeman, lacking in political knowledge or concern. Fragmentary evidence suggests, however,
1966, 1967aJ 196713, 1971, 1972). A sense of powerlessness, as Seeman has shown, is that participation without a sense of psychic competence shapes political activity in
related to a lack of political knowledge and knowledge in other specified areas in a negative stance (Thompson and Horton, 1960) and that to the extent to which
which control-relevant information is not of value to the subject; equally important, participation is correlated with lack of competence, a democratic system is inherently
a sense of powerlessness i s clearly separable from a sense (in Marxian terms) of weak (Cnudde and Neubauer, 1969).
alienation from work. The sense of powerlessness (as opposed to other types of Another psychological trait shown to affect the quality of a person's political
alienation) is rather clearly related to a lack of engagement in social and political ctivity is the attribute of dogmatism (Rokeach, 1960), a trait referring theoretically
activity, but because this dimension has been repeatedly correlated with lower socio- I the structure (as opposed to the content) of a person's belief system. In a lengthy
economic status and has been conceptually indistinct in terms of psychic and societal :ries of laboratory experiments, Rokeach showed that subjects can be placed along
roots (Knutson, 1972a), its meaning in personality terms calls for careful assessment. continuum going from closed-mindedness to open-mindedness and that individuals
Another personality trait in which a good deal of theoretical work has been t the closed-minded end of the continuum differ in a number of politically important
invested is that of power motivation. Since Lasswell's (1930) early statements about rays from those at the open-minded end of the continuum. First of all, closed-minded
the importance of political activity as a compensation for unfulfilled needs for esteem, idividuals are typified by anxiety and have difficulty assimilating new information.
it has been a largely unexamined truism that power motivation, as a psychologically lonsequently, they may reject information outright if it conflicts with their beliefs;
distinct and unitary dimension, is exceedingly important in explaining political r they may perceptually distort the information so that they can accept it (for
activity levels. This is true even though Lasswell (1954) later qualified his earlier lstance, a voter who approves of Eugene McCarthy can also support George
dictum in major ways, and even though,the differential importance of power motiva- Vallace because "both are for the individual and opposed to the system which denies
tion in political behavior has been carefully delimited in a number of places (George, idividuality") ; or they may accept-but not integrate-the new information if it
1968; Knutson, 1972a; Lane, 195913). 3mes-gram a valued authority ("I am totally opposed to the war in Vietnam, but
The limited empirical work here to date (for instance, Browning and Jacob, resident Nixon feels that we should continue to fight until the Vietnamization policy
1964) has generally failed to substantiate this motivation as a simplistic indicator forks and he must be right, because he's president") ; or-if the change in beliefs
of political involvement. Thus, the intuitively important variable effect of power
:quired by information from a positive authority is too great-they may devalue
striving remains a major area for future research and presents two major caveats. le authority.
First, as Horney ('1937) pointed out a number of decades ago, any need-including Further, closed-minded subjects are likely, without analysis, to accept and act
the need for power--can be neuroticized or healthy and adaptive and, within these
"n new information which appeals to them. Consequently, as Rokeach (1960, p. 240)
strictures, may .serve various functions. The possibility of healthy, creative needs for
power has been given relatively little attention to date in theory and even less in 1 The reader should note that a construct such as power can be considered either an
research. (For ardiscussionof the varying needs which power seeking and attainment intrapsychic need or the goal objective of such a need, depending on its function in the
may serve, see George, 1968; Knutson, 1972a, p. 69.) Second, it is quite simplistic to individual's psychic economy.
48 Jeanne N. Knutson Personality in the Study of Politics 49

showed experimentally, open-minded individuals take considerably longer to solve leadership roles in Italy. Using a sample drawn from the Chamber of Deputies of the
problems when presented with a new, conflicting piece of evidence, because of their Third Republican Parliament of Italy and a nonrandom control group, DiRenzo
need to analyze before assimilating or rejecting the information. On the other hand, found a significant difference between the mean dogmatism scores of the two groups,
"problem solving proceeds more smoothly in closed persons when new beliefs are with the political leaders being more closed-minded. Because of his inadequate con-
presented all at once than when presented gradually because of their proclivity to trol group, more convincing is the relation that he found between dogmatism and
easily accept without assimilation 'packaged' information" (pp. 287-288). While political ideology: the two groups lowest in dogmatism were on the left extreme,
there have been few applications of this personality trait to the study of political role and the two highest were on the right extreme (p. 123). The author (pp. 152-153)
performance and activity level, its relevance in such study is obvious. Suggestive in comments: "The more open-minded parties in our political sample are those that
this respect is the frequent finding of voting studies that "the most partisan people have gained progressively in strength; the more closed-minded parties are those that
protect themselves from the disturbing experience presented by opposition arguments .
have lost strength, even to the point of passing out of existence. . . These data
by paying little attention to them" (Lazarsfeld, Berelson, and Gaudet, 1944, pp. 89- support the explanation of dogmatism correlating with threatened existence." (For
go), offering the testable hypothesis that undeviating partisanship may be associated a critical review of DiRenzo's book, see Greenstein, 1970b.)
with an intrapsychic trait which closes off new information. The above evidence also The answers provided by personality variables -to the question of who is
suggests that the individual who changes his beliefs and joins a mass movement- avaiIable for leadership roles have been, as Sigel (1970) points out, varied indeed.
accepting wholly, as a matter of faith, dogma stemming from positively valued Senerally, two schools of thought have arisen which speak to this phenomenon (also
authoritieemay also be typified by this personality trait. (The reader should note lee Chapter Eight in this volume). First, Lasswell (1930, 1948)-in an analysis of
that thus far in this literature review, we have been able to discuss psychodynamics 2olitical leaders, a number of whom had been hospitalized for psychiatric reasons-
without needing to take up the more difficult problem of psychogenesis.) persuasively argued that a political leader projects his neurotic needs into the pubIic
In the area of leadership, dogmatism offers a useful concept for studying sphere and enters political life in search of their fulfillment. (DiRenzo's study can
people who close off new information in a manner dysfunctional to their announced be seen as supportive of this view.) Lasswell (1951) later modified this thesis to
political goals. This phenomenon has been illustrated in the detailed study by account for his observation that in a democracy the successful political leader is not
Edinger (1965.) of the German Social Democratic leader Kurt Schumacher. I t can likely to be driven by compensatory personality needs; his initial argument, however,
also be seen to operate in the recurring, self-made crises of Woodrow Wilson (George has continued to have widespread appeal. A second theoretical stance (Lane, 1959b;
and George, 1956). Indeed, as the Georges illustrate by the supportive function served Knutson, 1972a, forthcoming) is summarized by Rutherford (1966, p. 390) in, these
by Colonel House and others, dogmatic individuals frequently arrange their life ;erms:
space so that conflicting information is filtered out before it reaches their input
channels. For example, at one point in his career, Wilson's problems clearly stemmed If an individual suffers from intrapsychic conflict, so much energy will be
from "his deliberate refusal to make use of the channels of information that were consumed by the struggle within the person that no.surplus will remain to
available to him. [In another instance] ... Wilson did not care to listen to opinions cope with conact in the political arena. Interpersonal relations and ego
strivings will thus suffer to a great extent. Those experiencing intrapsychic
he did not welcomeJy(Janis, 1959, p. 13). This phenomenon can also be observed in
the tacit understanding among Hitler's staff: "For God's sake, don't excite the conflict, then, would be expected to withdraw from political participation
rather than project upon political objects. We would have this proposi-
Fuehrer-which means do not tell him bad news-dwnot mention things which are
tion: the higher the level of political participation, the greater the psychic
not as he conceives them to be" (Von Wiegand, in Langer, 1972, p. 76). A similar, energy need, the less the intrapsychic conflict, and, hence, the more ra-
less extreme, example of information contsol has been observed to operate in Nixon's tional the participant.
milieu (Barber, 1972aJp. 439).
An interesting discussion of the dysfunctional nature of closed-mindedness in Rutherford, in the same article, then attempts to refute this position by presenting
political leaders is found in an article by Janis (1959), which thoughtfully traces a evidence from a patient-governed ward of a psychiatric hospital. He reports that
number of sources of decisional conflicts. Janis concludes that a leader's decisions these patients before they were hospitalized, had reported participating in politics
may improve in quality (and his "decisional conflicts" may be reduced) if he is as much as the average citizen in upper socioeconomic groups and more than would
given more accurate information and learns, in advance, the possible consequences )e expected for lower socioeconomic groups. Furthermore, he found only a small
of his acts. In view of Rokeach's work, this appears an overoptimistic conclusion, '-294) correlation between clinical judgment of mental health and participation on
since the dysfunctional nature of at least part of the decision making which Janis he patient council; in fact, manic-depressive and paranoid schizophrenic types were
uses as evidence can be attributed to something other than inadequate or inaccurate overrepresented among the patient council leaders, compared with their number in
information; namely, to the inability of closed-minded persons to avail themselves of, the ward population.
to assess, and to integrate information because of the "unconscious affective charge" The Rutherford article, by overstating the position he is attempting to refute,
(to Janis's phrase) which conflicting, nonsupportive information entails. gives us an opportunity to look closely at the theoretical stances involved. One school
The trait of dogmatism was also employed in a study (DiRenzo, 1967) of (which we will label here the Lasswell-Rutherford position) advances the view that
50 Jeanne N. Knutson Personality in the Study of Politics 51

unfulfilled personality needs, far from being a barrier to participation, may actually Table 1.
motivate psychically deprived citizens to seek fulfillment by engaging in political
CONTINUUM
OF MENTAL
HEALTH
VERSUSPOLITICAL
PARTICIPATION
activity because of the special need gratifications available through political activity. -
Thus psychically deprived individuals will tend to be overrepresented among the
population of political leaders. The other school of thought (which we will call the Low Middle High Percent
Lane-Knutson viewpoint) holds (1) that psychically deprived individuals are less
likely to seek active political roles than are more psychically fulfilled persons- Psychically deprived 36.6 37.4 26.1 100
particularly when the role requirements are stressful; and (2) that unfulfilIed psychic Low self-actualizers 33.3 40.8 25.8 100
needs are eventually dysfunctional to role performance in terms of the individual's High self-actualizers 23.5 32.3 44.1 100
own political goals.2
Supporting our everyday observations, a number of studies have pointed out Source: Knutson, 1972a, p. 236.
that persons who lack psychic competence do serve important functions in political
systems (as Tucker, 1965, argues in the case of the dictator) and that political activity Note: Psychically deprived refers to those subjects categorized as primarily
motivated by physiological, safety, affiliative or esteem needs, with the
also serves to meet the particular needs of these individuals. See, for example, the
balance of the subjects categorized as either low (less fulfilled) or high
psychobiographic studies of Civil War political leader Thaddeus Stevens (Brodie, (more fulfilled) self-actualizers, depending on their scores on the indices of
1966) ; of Abraham Lincoln, a president with severely unmet basic needs (Clark, psychic deprivation. Low, middle, and high participation were ascertained by
1933) ; of Secretary of State James Forrestal (Rogow, 1963) ; of Chicago mayor a modification of the Woodward-Roper (1950) Political Participation Index,
Anton Cermak (Gottfried, 1962) ; of Secretary of State Charles Evans Hughes which sums reported political activities, such as voting and attending politi-
cal meetings.
(Glad, 1966) ; of Reformation leader Martin Luther (Erikson, 1958) ; and the
previously mentioned studies of Kurt Schumacher (Edinger, 1965) and Woodrow than by objective, consciously articulated, political parameters. This process is con-
Wilson (George and George, 1956). Moreover-as studies of Hitler (Langer, 1972) cisely discussed by Langer (1972, pp. 74-75) as follows:
and Stalin (H. Salisbury, 1969; Deutscher, 1949) and Lasswell's (1930) early
analyses of political leaders indicate-severely pathological responses may occasionally This is a very fundamental trait in Hitler's character structure. He does
be instrumental in achieving and solidifying a position of political importance. not think things out in a logical and consistent fashion, gathering all avail-
In aggregate terms, however, the Lane-Knutson position begins by noting that able information pertinent to the problem, mapping out alternative
a person will be likely to engage in social and political activity to the extent that his courses of action, and then weighing the evidence pro and con for each of
basic psychic needs have been met. As I have illustrated in terms of Maslow's need them before reaching a decision. His mental processes operate in reverse.
hierarchy with a nonrandom sample of 495 ordinary citizens, there is a significant Instead of studying the problem, as an intellectual would do, he avoids it
correlation between political participation and fulfillment of psychic needs (Knutson, and occupies himself with other things until unconscious processes furnish
1972a). The correlation, while significant, certainly is not large enough to disallow him with a solution. Having the solution he then begins to look for facts
the view that unmet psychic needs may produce political activity ( r = .14), although that will prove that he is correct. In this procedure he is very clever, and
by the time he presents it to his associates, it has the appearance of a ra-
an examination of the relationships (see Table 1) indicates that the most fulfilled
tional judgment. Nevertheless, his thought processes proceed from the
citizens are twice as likely to engage in active political roles as are those with unful- emotional to the factual instead of starting with the facts as an intellectual
filled psychic needs. (It is also possible that gratification of psychic needs is related normally does. It is this characteristic of his thinking process that makes
to the concomitants of participation, such as social status and education.) it difficult for ordinary people to understand Hitler or to predict his
The second part of the Lane-Knutson position interprets the low correlation .
future actions. . . p h i s method] is not without its shortcomings. He
and the evidence that unmet needs are not an absolute barrier to participation. becomes dependent on his inner guide, which makes for unpredictability
According to this position, persons who engage in political activity as a means of on the one hand and rigidity on the other. The result is that he cannot
compensating for unresolved needs will eventually face critical situations; at such modify his course in the face of unexpected developments or firm oppo-
times, the necessary (intrapsychically compulsive) fulfillment of unmet psychic needs sition.
will disallow the attainment of valued political goals. In such a psychically rigidified
I'he frequency of episodes in which intrapsychic considerations operate virtually
situation (discussed above as episodic in the careers of Wilson and others), the
~nchallengeddepends on the individual's ego strength and on the degree to which
individual's behavior is shaped by unconscious or preconscious psychic needs rather
t functions in a coping rather than a defensive manner (Kroeber, 1966). Thus, level
)f ego functioning (which, in turn, is related to severity of psychic deprivation)
2As Alexander George notes (in a personal communication), compensatory need
gratification and dysfunctional role performance do not necessarily go together. determines a person's ability to find an adaptive situation which meets both inner
 Jeanne N. Knutson Personality in the Study of Politics 53

and outer constraints. Hence, the Lane-Knutson position also addresses the degree enriches our understanding of the dimension of dogmatism, and of personality factors
of successful role performance and, given the severity of the leader's unmet needs, related to intensity of political activity. In a study based on in-depth interviews with
the likelihood that he will be able to choose his time and manner of exit from public ninety-three members of the British House of Commons and eighty-three members
life. of the Italian Chamber of Deputies, a factor called "ideological style" was derived
Any attempt to compose a fourfold table of health and political activity is from the data. Subjects rated high in ideological style tended to be "generalizers."
doomed to failure, because persodality is too complex and various to be neatly Generalizers are those who deduce their analysis from a general, abstract theory
categorized; if one attempts such abstracting; he risks losing the essential heuristic instead of dwelling on particular, technical questions. Those high in ideological style
and predictive powers which legitimate a model. In assessing the relation between also tend to refer to a specific ideology or doctrine to buttress their arguments, and
psychic competence and political activity level or role-performance ability, we need they employ utopias as the basis for judging policy. They "also tend more frequently
to look at what the unrnet need is and the manner in which the person character- to be motivated by ideological satisfactions, to interpret political phenomena such
istically deals with this need (sublimination, in the case of Abraham Lincoln's need as parties in terms of more abstract schema, and to reject a merely 'possibilist' ap-
for affection, produced very different results politically than repression and denial of proach to politics" (Putnam, 1971, p. 665). In both leadership samples, such "ideo-
the same need did in the case of Thaddeus Stevens). This point has been beautifully logical style" was related to extreme (left and right) positions on the political
illustrated by Erik Erikson (1958, 1969), whose works have analyzed how Martin spectrum. Ideological style was also correlated (in both samples) with alienation
Luther and Mahatma Gandhi were able to integrate inner and outer reality in ways from existing political and social institutions. Most important for our discussion,
that fulfilled both their own unmet needs and the requirements of the history into however, these ideologues are not distinguished by a dogmatic uqwillingness to com-
which they were born and which they helped to make. promise, by valuing ends more than means, or by their scores on moral absolutism,
Direction of Ideology. The previous section has provided a number of clues authoritarianism, political distrust, or social distrust.
to the ways in which personality variables may assist in our understanding of the Putnam (1971) did find, however, that his index of partisan hostility, while
direction (for example, conformist or deviant, left or right) and intensity which notm significantly positively correlated with his index of ideological style, was also
political activity may take. To begin with, DiRenzo (1967) found dogmatism asso- correlated with political extremism. He concludes: "Some extremists are 'ideological'
ciated with a right-wing political stance, although Rokeach (1960) had presented and some extremists are 'hostile,' but they are rarely the same extremists" (p. 672).
evidence (derived from very small samples) that his measure is unrelated to direction Finally, he notes that partisan hostility is, in both samples, inversely correlated with
of political ideology. Studies such as that by DiRenzo have, however, failed to sub- age, which he sees as a reflection of shifts in societal values as well as (age) cohort-
stantiate the lack of ideological relevance of the dogmatism scale and have generally specific experiences.
found a conservative bias in this measure of cognitive functioning (Parrott and Putnam's analysis underlines the impossibility of our finding any simple rela-
Brown, 1972). Barker (1963), for example also found dogmatism associated with tionship between political behavior and personality variables. Here, the same be-
rightist activity, although the correlation was not as strong as that between author- havior-political extremism-is clearly correlated with at least two different (and
itarianism and rightist activity. On the other hand, in a study of political activists largely unrelated) intrapsychic traits, both of which appear to be Level I1 variables.
ranging from left-wing radicals to members of the American National Socialist White Some extremists were characterized by an abstract, goal-oriented, and somewhat
People's party (Knutson, 1971), a significant correlation occurred between authori- messianic and idealistic cognitive style. Other extremists were identified by their
tarianism and left-wing political activity and ideology and also between dogmatism hostility toward political outgroups, their unwillingness to compromise, their author-
and left-wing political activity and ideology. (Correlations between dogmatism and itarianism, and their moral absolutism. Uncovering such contingent relationships is
left direction ranged from .46 to .69; those between authoritarianism and left intellectually demanding, but such relationships are more psychically accurate-and
direction ranged from .67 to .80.) As this sample was distinguished from the others hence more meaningful-than earlier, more simplified models of political behavior.
in this area (being composed of political leaders, instead of college students, and of This distinction between complex and contingent theses versus less satisfying
subjects whose political beliefs reflected a much wider range of ideology than student but simpler theses is well illustrated by two studies of political conservatism. McClosky
famples offer), its results are worth replicating. (1958), in a well-known study, looked at a large number of personality variables in
The original Berkeley study (Adorno and others, 1950) found a relationship relation to what he defined as conservatism and found, on all counts, a dear relation-
between authoritarianism and a rightist political orientation. This finding was repli- ship between traditional conservatism and various measures of psychic deprivation.
cated by Leventhal, Jacobs, and Kudirka (1964), but the relationship held only From the items of his conservatism scale made available, it appears that McClosky
when the hypothetical choice offered the subjects placed the conservative candidate was correlating personality variables with a punitive, negativistic view of human
in the Republican party. These various studies suggest that authoritarianism (and nature, rather than with the view of conservatism as dedicated to the conservation
its variapt, dogmatism) may be related to the preference for closed-minded candi- and gradual change of things established. A later study (Schoenberger, 1968) illus-
dates or dogmatic ideologies-which are at times, but not always, found on the right. trated that if conservatives are defined by dedication to furthering social and
A recent study (Putnam, 19'71) identifies a dimension of cognitive style which economic viewpoints rather than by agreement with a negative view of human nature,
54 Jeanhe N. Knutson Personality in the Study of Politics 55

conservative party affiliation is inversely related to authoritarianism and positively situational constraints which interact with personality to produce ideological stances
related to support of civil liberties. As in the study by Putnam (1971), these two in that nation.
studies of conservatism taken together suggest again that there may be at least two
types of extremists: ego-defensive and ideological (in Putnam's sense). A number of -
A Case of Relative Deprivation
other studies (Anderson and others, 1965; Simmons, 1965; Nettler and Huffman,
1957) have shown that there is no general conservatism or liberalism factor. , AS the above literature review illustrates, the uses of personality in the study
On the basis of research to date, we can conclude only that, personality var- of political behavior have been manifold. Indeed, applications of personality to
iables undoubtedly are related to the direction of political ideology. In order to political concerns can probably be found in all the chapters of this book. Like the
understand what linkages exist, a good deal of prior work is required to conceptuaIize man on the street, the professional student also sees great utility in employing per-
an ideological stance as consistent and to empirically demonstrate that it differen- sonality as a means of conceptualizing the political phenomena with which he deals.
tiates those who are active in various political groups so-labeled before one can take As the above discussion also suggests, however, much of the work in the area of.
up the weightier question of what personality characteristics relate to what ideo- personality remains tentative; carefully constructed, replicated designs are few indeed.
logical stance. Further, we scholars have sometimes repeated the sins of our forefathers, who
As with activity level, the direction of political activity appears to be multi- used personality as a conveniently simplistic way of categorizing political activity.
determined. DiPalma and McClosky (1970) found psychic fulfillment consistently When we try to understand those political beings whom our research interests neces-
correlated with a conformist, majoritarian political view, although thisirelationship sitate examining, we are able to see with the wisdom of distance that Machiavelli's
was much weaker in higher educational than in noncollege levels. On the other hand, view of personality as action-oriented and manipulative did not adequately describe
Block and her associates (Block, Haan, and Smith, 1969; Block, 1972) have illustrated human nature, that Hobbes's opinion of human nature as overdetermined by the
-with careful studies of student radicals-that there are meaningful, statistically need for security did not fully describe reality, and even that Marx's belief that
significant personality differences between student deviants whose political activity human nature had one major determinant (Fromm's and Freud's viewpoints as well)
is an a h a t i o n of parental values and those for whom it represents generational is not too helpful. We are less able to see, I think, that theoretical viewpoints which
discontinuity. Thus, work in this area seems particularly suitable for the employ- categorize political behavior of the right or left in terms of one set of intrapsychic
ment of a holistic personality model, enabling the researcher to understand the determinants; that see leadership behavior either as stemming from the achievement
function of the ideological stance adopted from the phenomenological view of his of mental health or as a reaction to unfulfilled needs; or that divide all political
subject's personality. actors in terms of ad hoc fourfold tables are also likely to be of little use in the study
One further, impressive development in the study of personality and ideology of political behavior.
needs to be mentioned; namely, the exciting work of Inglehart (1971a, 1971b) on The analysis of personality and political behavior in cumulative terms re-
intergenerational value changes in Western Europe. Using Maslow's need hierarchy, quires first of all a delimitation of research areas in which personality is likely to be
Inglehart examined political preferences specific to different generations or age co- a meaningful dimension (see Knutson, 1972a, ch. 5 for a longer discussion of what
horts. He found, in line with the differential growth of affluence in the West European these major areas of poIitica1 behavior appear to be). when such a research focus
countries, major differential shifts in political preferences. The older generations tend has been selected, it is then essential to adequately conceptualize personality and
to select acquisitive values (characteristic of unfulfilled physiological and safety further define it as measurable behavior. Next, as specific behaviors (for example,
needs). But the younger generations of West Europeans express preferences for what voting, making successful decisions, becoming a student activist) are repeatedly
Inglehart calls "postacquisitive" or "postbourgeois" values-values which, he per- shown to correlate with separable personality traits, these traits need to be related to
suasively argues, have grown in political importance as lower psychic needs have each other in terms of a conceptual framework which sees an individual in holistic
been gratified through decades of continued economic prosperity. terms (but allows for nomothetic statements). Finally, it is necessary to consistently
Inglehart's work offers a substantive base for the generally polemical assertions engage in deviant case analysis-so that, for example, we can understand why un-
of Reich (1970) about a sweeping value change in America today. It also provides dogmatic individuals adopt a right-wing viewpoint or why individuals with severely
additional understanding of the differences in the "liberalism" of working-class and upmet personality needs become successful leaders.
middle-class citizens (Lipset, 1960) and elucidates the psychic basis of the increas- Given these priorities, it is necessary to move back and forth from methods
ingly clearly explicated causal chain in democratic development (Lerner, 1958; which focus in depth on a single actor and methods which are suitable for aggregate
Knutson, 1972a, pp. 276-281 ; Cnudde and Neubauer, 1969; Aronoff, 1967, 1970). analysis. (See Beck, 1953.) The p~~chobiography, for example, has added a good
If one assumes (as this chapter clearly does) that psychic health, like physical health, deal to our knowledge of the deviant political leader, whose personality departs
is universally based on the satisfaction of certain needs, personality-related differences markedly from our expectations of the requirements for political success. In-depth
in ideology across cultures become a matter of investigating the ratio and social interviews and projective techniques also increase our understanding of deviant types
distribution of personality types in a specific polity, as well as the cultural values and and enable us to define new personality constraints which require laboratory experi-
56 Jeanne N. Knutson

mentation and the scaling abilities of a psychometrician to delineate. In addition,

survey research is available to allow us--once our understanding is refined-to ascer-
tain the patterns of variance and the range of attributes with which we deal.
By working back and forth between the individual and the aggregate level,
between attempts to uncover and analyze separate traits and efforts to amalgamate
them into a holistic view, we will arrive at an understanding of the relevance of per-
sonality for the study of political behavior which is richer, more complex, and more
satisfying than present knowledge makes possible. Such a process is far from novel.
As Freud's theoretical and clinical work brilliantly illustrates, the study of personality
continues to require that its practitioners be open-minded, creative, intellectually
tentative, and theoretically well grounded as they seek, through observation and
introspection, to understand the complex interaction between personality and be-
havior.

M. Brew ster Smith

l n the initial, classical review of the topic

of attitudes, written at a time that from present perspectives seems almost antedilu-
vian, Gordon W. Allport (1935) began by saying, "The concept of attitude is
probably the most distinctive and indispensable concept in contemporary American
social psychology." Through many vicissitudes during the intervening years, attitudes
have remained a central topic in the psychologist's social psychology; they have also
figured prominently in the sociologist's variant of this interdisciplinary field. Although
the social psychologists of whatever persuasion who have used the term seem to
have known what they were talking about, the vast literature concerning attitudes
includes a substantial controversial literature about their definition. In this case as
in so many others, definitional controversy is a tedious, but not wholly sterile, enter-
prise. As we shall see, fundamental issues of conceptual strategy are involved.
Conception of Attitudes
For present purposes, a working definition that is as unpretentious as I can
make it seems desirable: "The concept of attitude, although variously defined, is
most commonly employed to designate inferred dispositions, attributed to an indi-
58 M. Brewster S d t h Pdfitical Attitudes 59

vidual, according to which his thoughts, feelings, and perhaps action tendencies are iorist does not stay consistently within it. Henceforth, therefore, I will take for
organized with respect to a psychological object" (Smith, 1968a, p. 458). Even such granted a dispositional view of attitudes.
an informal definition embraces a controversial theoretical commitment to a strategy To return to my working definition, a few additional comments are in order.
that accounts for observed social behavior by extracting analytically two classes of Definitions of attitudes often specify that attitudes are relatively enduring learned
inferred, reconstructed determinants: features of the situation of action, and inferred dispositions. But it seems unnecessary to build duration or, for that matter, learned
dispositions or properties of the behaving person (see Lewin, 1951). This strategy, acquisition into the definition. The crux of the definition, and also the point on
which is one that I feel at home with and prefer, is by no means necessary for social which virtually all proposed definitions agree, is that attitudes are "organized with
psychology and behavioral science. I t has never been attractive to behaviorists and respect to a psychological object." Attitudes are toward something: a thing, a con-
other positivists, who like to stay as close as possible to observables. In fact, the truly cept, a policy, a person, a political party--or even the self. They also involve organi-
radical behaviorist (Skinner, 1957; Bem, 1967) has no use at all for the concept of zation-a structured set of beliefs held in readiness; a readiness to feel and to act
attitude, which he dissolves into particular behaviors, verbal and otherwise, viewed toward the object in different but coherently specifiable ways depending upon features
as under the control of particular discriminative stimuli and contingencies of rein- of the immediate situation in which the object is encountered. One of the founders
forcement. Various compromise positions-for instance, merging attitudes with of social psychology, William McDougall (1921), a great and presently underrated
habits (Doob, 1947) or treating attitudes not as dispositions but as a special class of theorist, adapted this concept from Shand (1914) as the basis upon which mental
behaviors-have been taken by theorists influenced by positivism. (See D. T. Camp- life is organized, though he and Shand used the term sentiment instead of attitude.
bell, 1963, for a view that holds cognitive and behavioral concepts of attitudinal dis- The dispositions with which we are concerned govern both thoughts (beliefs)
positions to be essentially equivalent.) and feelings (affect), which are hardly separable. As Osgood, Suci, and Tannenbaum
The 'choice between a dispositional or a positivistic-behavioral commitment (1957) have shown with the Semantic Differential, the primary, preemptive dimen-
can hardly be based on established fact. I t is not a matter of right or wrong. Cer- sion for any concept tends to be evaluative, pro or con. The main tradition of psy-
tainly, the Skinnerians have shown that they are able to push their approach sur- chological research in attitude measurement has overwhelmingly stressed this pro-con
prisingly far into the precincts of human action-though I think only by letting a dimension (see Scott, 1968). If one were to attend only to this tradition, a working
good deal of common-sense "dispositionaly' psychology in through the back door definition of attitude might restrict it to dispositions of favorableness or unfavorable-
(Smith, 1973). My own commitment to a dispositional orientation rests on two ness toward objects. But survey research and qualitative studies, including studies of
grounds. particular relevance to politics, are equally concerned with the cognitive content of
One has to do with a broad view of the "image of man" (Chein, 1972) that beliefs.
is compatible with human freedom and dignity, and therefore with politics, and also My working definition fudges in its reference to "action tendencies,'' because
with empirical, self-corrective, and thus on the whole cumulative science. The reduc- I do not want to take the route of definition to decide empirical-theoretical issues
tive, positivistic view of Skinner (1971) has no meaningful place for politics, which that are very much alive. In political psychology, we are of course ultimately in-
concerns the allocation of power among human actors in a polity. The values of terested in behavior-what the person does (which includes what he says on a given
liberty and justice, traditional to normative political theory, require a concept of occasion). The complexities involved in the relation between attitudes and behavior
human agency. And a view of man as an actor, politically and otherwise, needs con- will be touched upon below. It is still not clear, however, whether behavioral con-
cepts to formulate man's more enduring dispositions. Such an analysis of variable sequences are more profitably viewed as following from a given set of beliefs and
social behavior into more or less stable personal attributes and variable properties of feelings, given the person's appraisal of his immediate situation, or whether a "behav-
the situation of action corresponds with Heider's (1958) analysis of "common-sense ioral component" needs to be built into the conception of attitude itself (Katz and
psychology," and I take that as an advantage if our aim is to give a systematic account Stotland, 1959).
of human action in the world of living human concerns. In this chapter I will not take seriously the specification of $olitical attitudes.
The second is pragmatic. Because human beings as we know them grow up This term could obviously be given either a narrow reading (of attitudes toward
in language communities in which, however imperfectly and mysteriously, commmi- political objects such as issues, parties, candidates, or leaders) or a broader one (to
cation takes place, we do have approximate access to one another's thoughts and include whatever attitudes of political actors are thought to be relevant to their
feelings, from which we can make approximate inferences.about each other's more political actions). I will generally have the broader meaning in mind, with no con-
enduring orientations. Neither as na'ive human actors nor as social scientists are we cern about boundaries. I will be concerned rather with interpreting the social-psy-
restricted to the Skinnerian world of fragmented behaviors. We can listen to one chological literature on attitudes for its relevance to political psychology-focusing
another; we can even conduct systematic interviews and administer verbal scales. mainly on the nature of attitudes and the processes involved in their development
Although inferences from the data of communication can mislead as well as inform, and change. (See also Chapters Four and Five in this volume.)
we have good warrant to depend on them in.socia1 science as in everyday life. A My review will perforce be selective and personal. There are forty-three pages
conceptual framework that excludes inferred psychological dispositions from any of references in McGuirels (1968) definitive chapter on attitudes and attitude change
legitimate place is so severely restrictive that in practice even the committed behav- in the Handbook of Social Psychology. Two excellent books (Insko, 1967; Kiesler,
60 M. Brewster Sniith Political Attitudes 61

Collins, and Miller, 1969) have been devoted to a critical examination of theories of it lacks a psychological object. As a trait, however, it is defined by consistencies among
attitude change, the main focus of recent psychoIogica1 attention; and a very fat one attitudes toward a variety of social objects.) It was popular to correlate radicalism-
(Abelson and others, 1968) treats the single topic of consistency theories. Sears' conservatism with personality traits like introversion-extraversion-also being newly
(1969) chapter on political behavior for the Handbook of Social Psychology is in and crudely measured. There was little by way of theory to guide the venture, and
fact mainly a substantive treatment of political attitudes as they bear on voting in little by way of theory resulted from it. More recently, Eysenck (1954a) has pursued
national American politics; it has sixteen pages of references. Clearly, then, I must a similar strategy with explicit theory and more sophisticated methods, but his study
aim more at perspective than at summary. The serious student will need to refer to is badly faulted and its conclusions cannot be accepted (see Christie, 1956).
the sources just mentioned. About the time of World War 11, then, the psychology of attitudes was at
low ebb. Maybe it was still the central topic of social psychology, but it did not attract
, Some History the bright young investigators, some of whom were finding exciting new directions
in the experimental group dynamics of Kurt Lewin. Several developments in the
The scholarly history of the concept of attitudes is shorter than Allport's late 1930s and the 1940s radically changed the complexion of the field. As a result,
(1935) review suggests. In his day, Allport felt the need for legitimizing the concept during the 1950s and 1960s attitudes at last really earned a central place at the heart
by tracing.ambiguous linkages to the history of experimental psychology. Given the of a burgeoning, primarily experimental social psychology.
solid social-psychological use of the concept since his chapter was written, we can One development, which began in the late 1930s with the invention of pubIic
forgo such legitimizing pseudohistory. I have already mentioned a more germane opinion polls (later rechristened for scientific purposes as sample surveys), carried
theoretical antecedent in McDougall's (1921) concept of sentiment. The term atti- the study of attitudes off the campus into the territory of real political life. The major
tude was launched in social psychology (sociological version), however, by Thomas thrust was under sociological auspices. Paul Lazarsfeld in particular introduced
and Znaniecki (1918) in the methodological preface to their classic monographic analytic methods for treating successive interviews with "panels" of respondents, and
study The Polish Peasant. For Thomas and Znaniecki, attitude and value are com- employed this approach in studying the determinants of voting decisions in presi-
plementary terms. Attitude refers to any disposition of a person toward an object, dential campaigns (see, for example, Lazarsfeld, Berelson, and Gaudet, 1944).
while any object becomes a value by 'virtue of being target of a person's attitude. The emphasis was on demographic predictors of the voting decision, such as ruraI-
Thus launched conceptually, attitudes became fully established in social psychology urban residence, socioeconomic status, and religious affiliation. Survey methodology
when Thurstone (1928) directed his sophistication in psychophysics to the problem in the study of attitudes (though not directly political ones) was substantially ad-
of their measurement. What could be more legitimizing? In fact, a sophisticated vanced in the wartime studies, directed by the sociologist Stouffer, that resulted in
theory of attitude measurement has developed from the beginnings launched by the volumes of The American Soldier (Stouffer and others, 1949).
Thurstone (see Scott, 1968), a line of development that has thus far mainly served But psychologists too were employing the new survey technology-initially,
the legitimizing function. For better or worse, the theoretically interesting experi- under the leadership of Rensis Likert and Angus Campbell, in the Bureau of Pro-
mental studies of attitude of the last two decades have mostly used very simple, even gram Surveys of the United States Department of Agriculture and later at the Survey
nayve indices of attitude: single items presented for agreement or disagreement, or Research Center of the University of Michigan. When these psychologists turned
ad hoc and a priori "scales" assembled without regard for psychometric nicety. The their attention to the study of voting decisions, they placed focal emphasis on atti-
methodological inventiveness of the experimenters has been invested almost wholly in tudes-such as toward candidate, toward party, and toward campaign issues. (For
the manipulations of experimentation, not in the mapping and measurement of atti- a classic later example in this tradition, see A. Campbell and others, 1960; see also
tudinal variables. As for descriptive analyses of public attitudes, particularly relevant the review by Sears, 1969.) They typically paid much more attention than sociolo-
to political psychology, these have mainly drawn upon the techniques of survey gists to the beliefs-cognitive structures--embedded in these attitudes. Whether under
design and analysis, not of attitude scaling (see Hyrnan, 1955; and Chapter Twelve psychological or sociological auspices, survey research has provided the major de-
in this volume). scriptive substantive content of the psychology of political attitudes. More than any
The research of the early period-reviewed by Allport (1935) and by Murphy, other research tradition, it has brought their study into contact with the realities of
Murphy, and Newcomb (1937), who each declared attitudes to be a central topic of political life.
social psychology-was in fact not very interesting or substantial. The characteristic A second development (reviewed in detail by Sanford in Chapter Six) fol-
approach was descriptive and correlational (mainly on college sophomores, at that), lowed from the surge of academic interest in psychoanalytic theory that, in the im-
with little sustained attention to the conditions ulider which attitudes are formed mediate postwar years, accompanied the launching of a self-conscious clinical
and modified and little effort to connect the psychology of attitudes with more psychology. The studies brought together in The Authoritarian Personality (Adorno
general explanatory principles. and others, 1950) drew upon a liberalized psychoanalytic theory (in keeping with
One of the more substantial studies of this period (Murphy and Likert, the emerging dominance of "ego psychology" in psychoanalytic circles) to give a
1938) came up with little that was more solid or reliable than the finding that among motivational interpretation of anti-Semitic and more generalized ethnocentric atti-
college studekts radicalism was somehow associated with dissatisfaction. (Strictly tudes in terms of defensive personality processes. En route, the authors extended their
speaking, radicalism-conservatism is a personality trait rather than an attitude, since interpretation to antidemocratic or "fascist" orientations generally. The F scale,
62 M. Brewster Smith Political Attitudes 63

devised to measure these orientations, staked a claim for itself in an ambiguous com- (In Chapter Twelve, Hyman comments on the negligible contribution of experi-
mon ground overlapping attitudes, ideology, and personality. Over the ensuing mental methods to political psychology. For a more broadly based and positive view
decade, its availability led to a large amount of research, much of which was un- of the role of the experiment in political psychology, see McConahay's discussion in
fortunately consumed in exploring the scale's technical defects. The tide of research Chapter Thirteen.)
interest in the topic unfortunately waned at about the point when more adequate Toward the end of the decade, disillusionment began to set in (Ring, 1967).
measures were becoming available (Kirscht and Dillehay, 1967). In the present There were a variety of grounds for discontent. I t became apparent to those who did
context, The Authoritarian Personality is important because it reintroduced interest not already know it that experimentation. is no royal road to truth: ambiguities
in the psychodynamic basis of political attitudes, emphasizingmore systematic methods abound in the experimental realization of theoretical concepts. Although the heavy
and more explicitly delimited and elaborated theory than was characteristic of Lass- social investment in experimental social psychology produced gains in the sophisti-
well's ( 1930) initial ground-breaking foray. cation of experimental design and in alertness to alternative possibilities of interpre-
The third major development began with wartime experimental studies of tation, experimentation (which came to involve increasingly intricate stage mafiage-
the impact of army films for internal~propaganda,and emerged full blown shortly ment) did not seem to be paying off as much as hoped in replicable results. And the
after the war as a new laboratory-based experimental social psychology, in which issues to which the experiments were being addressed were getting increasingly
attitude change was the focal topic of experiment and theory. Carl Hovland, a detached from the real world of social-and political-behavior.
brilliant experimental psychologist who was drawn in as a consultant to direct a The preciousness of much work in experimental social psychdlogy also ran
program of research on army-orientation films (Hovland, Lumsdaine, and Sheffield, I afoul the tide of social concern and discontent that was sweeping the campuses with
1949), carried forward the research at Yale after the war in a vigorous collaborative its call for "relevance." Newly aroused ethical sensitivities were disturbed by the
program on persuasive communication. Unlike the earlier studies of propaganda and manipulative cast of social-psychological experimentation, especially by its heavy
attitude change (Murphy, Murphy, and Newcomb, 1937), the Yale studies were reliance upon deception to bring social reality into the laboratory. And questions
guided by theories drawn from general psychology. The predilection of Hovland and were raised, too, as to whether the experimenters themselves were being deceived
his closest collaborators was for the kind of learning theory identified with Yale and (Rosenthal and Rosnow, 1969). Meanwhile, cold winds from Washington added
the names of Clark Hull and Kenneth Spence, but the program was actually guided their force to these intrinsic trends. "Basic" research no longer could enjoy a growth
by a theoretically neutral empirical scheme that mapped the effects of persuasive economy, and the carrot of funds began to point in the direction of problem-focused
communication into those attributable to the source of the communication, to the studies.
content of the communication, to the predispositions of the audience, and to responses Concommitantly with these developments, social psychologists who had se-
induced to the communication. (This scheme can be traced to Lasswell.) Different questered themselves in .the laboratory were becoming interested in field studies.
theoretical models could be drawn upon, and were, for hypotheses. A programmatic Ingenuity was being directed to stage-managing field experiments that tested social-
statement that presented an impressive initial series of experiments (Hovland, Janis, psychological hypotheses in real-life settings with unwitting participants. (Ethical
and Kelley, 1953) was followed by a series of monographs that filled in much of the problems here, too!) An influential early example, which displayed the value and
picture sketched in the initial volume. (See Hovland and Janis, 1959; Rosenberg and the great appeal of such an approach, was the exploration of the "unresponsive by-
Abelson, 1960; M. Sherif and Hovland, 1961.) stander" problem-why he so often remains uninvolved when help is urgently needed
Added to this manifestly fruitful research program, which seemed to be yield- (Latan6 and Darley, 1970). D. T. Campbell (1969a) was providing a rationale for
ing a new "experimental rhetoric," was a simple, new, imperialistic theory pushed social psychologists to regard "reforms as experiments" and contributing to the devel-
with proselytizing enthusiasm by an incomparably ingenious experimentalist. The opment of techniques for doing so. Field experimentation was thus added to survey
proselytizing experimentalist was Leon Festinger, the theory concerned "cognitive research as an alternative--or a complement-to the laboratory.
dissonance" (Festinger, 1957), and the heyday of experimental social psychology (the O As I write this chapter, it is too early to predict the extent to which the
late 1950s and the 1960s) followed. In other respects the timing was right: through pendulum will swing. I hope that a better balance in research strategies will result,
the National Institute of Mental Health and the National Science Foundation, the not just a change of fashion (Smith, 1972). I t is already clear, however, that the
federal government released unprecedented resources to train researchers in the substantial recent reviews of theoretically oriented experimental studies of attitudes
behavioral sciences and to support their basic research. That the new laboratory (Abelson and others, 1968; Insko, 1967;Kiesler, Collins, and Miller, 1969; McGuire,
studies of attitude change came to concern themselves with abstruse theoretical and 1968) adopted a more sanguine view of past and future achievements than comes
methodological issues was for the time being an advantage, not a handicap. Under easily to us today.
the system of research grants by peer review, the new field expanded under its own I
directives, with few external brakes applied. The Yield of Attitude Research
The study of attitude change, then, suddenly became integrally linked with I
the competitive extension and elaboration (not just the application) of general My own appraisal, as the reader surely understands, will be cast in the emerg-
theories in psychology. But these process-oriented developments proceeded in virtual ing, more skeptical mode. There have been gains from two decades of eager experi-
isolation from the content-oriented tradition of field research using survey methods. mentation, but their contribution to political psychology is surely far less than that
64 M. Brewster Smith Political Attitudes

of the steady, relatively untheoretical progress toward describing and in part under- 5. Somerimes emotional appeals are more influential, sometimes factual
standing the attitudinal determinants of political behavior, especially voting, by ones. I t all depends on the kind of audience.
means of sampling surveys (see the review by Sears, 1969). I will discuss the yield 6. Fear appeals: The findings generally show a positive relationship be-
of experimental attitude research for political psychology under two headings: em- tween intensity of fear arousal and amount of attitude change, if recom-
pirical generalizations and the relevance of theory. mendations for action are explicit and possible, but a negative reaction
Empirical Generalizations. Experimental studies of persuasive communica- otherwise.
tion have yielded a number of generalizations that can be fitted into an untheoretical
7. The fewer the extrinsic justifications provided in the communication
for engaging in counternorm behavior, the greater the attitude change
map, like the one with which Hovland, Janis, and Kelley (1953) began. For a read-
after actual compliance [if it occurs-MBS].
able summary of this "scientific rhetoric" I quote at length from Zimbardo and 8. No final conclusion can be drawn about whether the opening or closing
Ebbeson's (1969, pp. 20-23) introduction to the psychology of attitude change. parts of the communication should contain the more important material.
(Zimbardo and Ebbeson draw, in turn, from Karlins and Abelson, 1970.) The 9. Cues which forewarn the audience of the manipulative intent of the
backup for these dogmatically phrased statements is most fully available in McGuire's communication increase !resistance to it, while the presence of distractors
(1968) compendious chapter in the Handbook of Social Psychology, simultaneously presented with the message decreases resistance.
C. The Audience as Individuals
A. The Persuader 1. The people you want most in your audience are often least likely to be
1. There will be more opinion change in the desired direction if the com- there. There is evidence for selective seeking and exposure to information
municator has high credibility. consonant with one's position, but not for selective avoidance of informa-
Credibility is : , tion dissonant with one's position.
a. Expertise (ability to know correct stand on issue). , 2. The level of intelligence of an audience determines the effectiveness of ,

b. Trustworthiness (motivation to communicate knowledge without some kinds of appeals.

bias). 3. .Successful persuasion takes into account the reasons underlying atti-
2. The credibility of the persuader is less of a factor in opinion change tudes as well as the attitudes themselves. That is, the techniques used must
later on than it is immediately after exposure. be tailored to the basis for developing the attitude.
3. A communicator's effectiveness is increased if he initially expresses 4. The individual's personality traits affect his susceptability to persua-
some views that are also held by his audience. sion; he is more easily influenced when his self-esteem is low.
4. What [the members of] an audience [think] of a persuader may be 5. There are individuals who are highly persuasible and who will be
directly influenced by what they think of his message. easily changed by any influence attempt, but who are then equally in-
5. The more extreme the opinion change that the communicator asks for, fluenceable when faced with countercommunications.
the more actual change he is likely to get. 6. Ego involvement with the content of the communication (its relation
a. The greater the discrepancy (between communication and recip- to ideological values of the audience) decreases the acceptance of its con-
ient's initial position), the greater the attitude change, up to ex- clusions. Involvement with the consequences of one's response increases
tremely discrepant points. the probability of change and does so more when source-audience discrep-
b. With extreme discrepancy, and with low-credibility sources, there ancy is greater.
is a falling off in attitude change. 7. Actively role-playing a previousIy unacceptable position increases its
acceptability.
6. Communicator characteristics irrelevant to the topic of his message can
D. The Influence of Groups
influence acceptance of its conclusion.
1. A person's opinions and attitudes are strongly influenced by groups to
B. How to Present the Issues which he belongs and wants to belong.
1. Present one side of the argument when the audience is generally 2. A person is rewarded for conforming to the standards of the group and
friendly, or when your position is the only one that will be presented, or punished for deviating from them,
when you want immediate, though temporary, opinion change. 3. People who are most attached to the group are probably least influ-
2. Present both sides of the argument when the audience starts out dis- enced by communications which conflict with group norms.
agreeing with you, or when it is probable that the audience will hear the 4. Opinions which people make known to others are harder to change
other side from someone else. than opinions which people hold privately.
3. When opposite views are presented one after another, the one pre- 5. Audience participation (group discussion and decision making) helps
sented last will probably be more effective. . . . to overcome resistance.
4. There will probably be more opinion change in the direction you want 6. Resistance to a counternorm communication increases with the salience
if you explicitly &ate your conclusions than if you let the audience draw of one's group identification.
their own, except when they are rather intelligent. Then implicit conclu- 7. The support of even one other person weakens the powerful effect of
sion drawing is better. a majority opinion on an individual.
 66 M. Brewster Smith Political Attitudes 67

8. A minority of two people can influence the majority if they are con- resistance makes sense from the perspective of adaptation, unless or until compelling
sistent in their deviant responses. [The last two propositions pertain to evidence or social pressure is introduced. The fact of resistance, in persuasive com-
small face-to-face groups-MBS.] , munication as well as in psychotherapy, makes the claims of would-be persuaders less
E. The Persistence of Opinion Change menacing than they would otherwise be if taken at face value.
1. In time, the effects of a persuasive communication tend to wear off. More seriously, this experimental rhetoric is incomplete. I t utterly ignores the
a. A communication from a positive source leads to more rapid impact of rational argument as such: cogent versus irrelevant considerations, weighty
decay of attitude change over time than one from a negative source.
versus trivial arguments, logical versus illogical reasoning. If the set of propositions
b. A complex or subtle message produces slower decay of attitude
change. gives the global impression of manipulativeness and disrespect for the political actor,
c. Attitude change is more.persistent over time if the receiver actively the impression is fair. Although, to be sure, most political actors do operate at a
participates in, rather than passively receives, the communication. discouragingly low level of rationality (Sears, 1969), the fact'is that researchers have
2. Repeating a communication tends to prolong its influence. not brought the intrinsic features of rational argument under experimental scrutiny.
3. More of the desired opinion change may be found some time after ex- The manipulative flavor of the conclusions drawn has to be understood in this con-
posure to the communication than right after exposure (sleeper effect) text.
[which is greater for communications from less trustworthy or negatively Relevance of Theory. Since attitudes are learned in the first place, and
evaluated comunicators-MBS]. enduring change in attitudes is a matter of further learning, theories of learning- are
one major approach to the conceptualization of attitude change. Attitudes can also
How the reader reacts to such a list must~bea virtual Rorschach test of his be seen as embodying the results of information processing and in turn affect the way
orientation to social research. Most of the points seem obvious, yet the contraries of that a person conceives and judges aspects of his world. Theories of the cognitive
many of them could reasonably be maintained. The passage quoted is a fair rendition processes are therefore a second source of hypotheses about attitude change. And as
of the empirical outcome of an enormous amount of research-some of it directed organized dispositions toward psychological objects, attitudes are important com-
primarily to more esoteric theoretical issues. As a rhetoric of political communication, ponents of personality. A third group of theoretical orientations to attitude change
it surely contributes to political psychology. Yet qualifications must be introduced thus have their roots in personality theory. These differing theoretical orientations
about the evidential status of these assertions and about the adequacy of this account highlight different research questions that involve different variables, so they are
of persuasive communication for political psychology. perhaps best regarded as complementary rather than competitive.
Some of the propositions quoted (for instance, A-2, B-1, E-1, 2, 3) concern Let us first look briefly at learning theories. We have already noted that
the longer-run consequences of exposure to a communication. But most of the re- Skinner's radically behavioristic treatment of learning bypasses attitudes, along with
search on which this summary is based tests the immediate effects of very brief and all other dispositional concepts, and focuses on the stimulus control of discrete re-
therefore, in the long run, inconsequential communications. A confident scientific sponses as it is established and maintained by specified contingencies of "reinforce-
rhetoric would have to be based upon research that followed the effects of more ment" (Skinner, 1957). This is a program that can be argued in principle, and
extensive exposure to persuasive communication through longer time. illustrated, but hardly carried out in adequate detail to meet the needs of students of
As Proposition C-l recognizes implicitly, research on attitude change pertains political behavior. The analyst of political behavior can nevertheless learn from the
mostly to what can be done to captive audiences. Except under totalitarian circum- Skinnerians to attend closely to the payoff contingencies of reward and punishment
stances, political communication in the real world faces conditions quite other than ("reinforcement") under which particular politically relevant behaviors are "ac-
those studied in the laboratory. (See Hovland, 1959, for a discussion that attempts quired" and to'the current contingencies under which they are maintained. Such
to reconcile the contrasting results of experimental studies of attitude change in attention to significant detail can be a healthy corrective to the sloppy habits that a
response to attempts at persuasive influence, which typically find change to spare, global functionalism may seem to countenance. (The functionalist perspective is
and sample survey studies in the field, which typically do not.) Only McGuire (1964) presented at some length below.) As a critical perspective, moreover, Skinnerian
has made a concerted attempt to study the conditions under which communication behaviorism usefully questions the assumption that attitudes are intrinsically struc-
can induce resistance to counterpersuasion, and he limited his research to the special tured. (The degree to which public beliefs and attitudes are structured in the sense of
case of cultural truisms (like the precepts of good hygiene) that are not normally batterns of correlations across persons is readily exaggerated; see Converse, 1964.) To
,exposed to attack in the ordinary arena of communication. he extent that the coherence of beliefs, attitudes, and behavior can be accounted for
A more important empirical reservation is that the propositions listed do not ~yexternaI contingencies, the concept of intrinsic structure can be misleading. What
give sufficientprominence to the underlying fact of psychological inertia-resistance people say and what they do are surely governed in part by different contingencies,
to change-which, after all, is the sensible response to any isolated new fact that different positive and negative sanctions. The Skinnerians usefully remind us of this,
challenges a person's existing attitudes (Janis and Smith, 1965). Particularly when though I do not believe that their theoretical strategy can be extended to deal ade-
the issue is important and therefore engages with well-established systems of belief quately with the domain of political psychology.
and attitude, or with the norms of groups with which the person is identified, Psychologists have drawn on the broader framework of stimulus-response
68 M. Brewster Smith Political Attitudes

(S-R) learning theory for a conceptual language to apply in the study of attitude During the heyday of experimental research on attitude change, the main
change, which they analyze in terms of the familiar categories of stimulus ahd focus of attention concerned theories of cognitive consistency (Abelson and others,
response, incentives, reinforcement, generalization, discrimination, and conflict. AS 1968). The topic became elaborated to such an extreme that it defies responsible
illustrated in the work of. the Yale group (Hovland, Janis, and Kelley, 1953), the summary in my allotted space. I will therefore restrict myself to identifying a few of
categories of learning theory serve heuristically to set the terms of empirical problems the main contenders in the barest terms, preparatory to venturing some judgments
and to suggest lines of interpretation that give direction to subsequent investigation- about the contribution of consistency theories to political psychology.
rather than t'o generate specific hypotheses. Campbell (1963) suggests that in this One kind of consistency that seems firmly established but because of its ob-
kind of use, the S-R and cognitive languages may well be intertranslatable and viousness (once established) has not attracted a great deal of research is that between
equivalent for most purposes. The student of political psychology need feel under no beliefs and feelings (Rosenberg, 1960) ;that is, there is a strong tendency for people
obligation to learn the S-R language, I think, unless he wishes to pursue some of the to bring their beliefs and feelings into. line. The tendency of voters to support the
technical psychological literature. The recent stream of interest in the experimentaI issue stands of their preferred party has been interpreted by Sears (1969) as reflect-
psychology of verbal learning has yet to influence attitude research and theory, to ing such a rationalizing trend toward consistency. Along somewhat similar lines,
which it might be thought to be especially relevant. Rosenberg and Abelson (1960) developed a formal model for assessing the degree
As mentioned, cognitive theories are another source of hypotheses about atti- of "balance" in attitudinal structures, conceived as networks in which affectively
tude change. There has always been a strong cognitive emphasis in the psychology of evaluated cognitive elements are linked by positive, negative, orenull relations. Thus,
attitudes. As the philosophical basis for dogmatic positivism lost its legitimacy and to draw on one of their examples, a state of imbalance is illustrated by a student who
psychologists ceased to be cowed by its dictates (see Koch, 1959), cognitively oriented favors coeds at Yale, wants good grades, yet believes that having coeds at Yale would
theories of attitudes came into greater prominence. interfere with getting good grades. The imbalance motivates a search for a balance:
In the traditional alignment of controversy in systematic psychology, learning producing resolution, which might be attained by changing the affective sign (+ or -)
theories were linked with behaviorism and cognitive theories with Gestalt psychology. of "coeds at Yaleyyor of "good grades" or of the imputed relation between the terms,
During the period of behavioristic predominance, the Gestalt psychologist Asch or by introducing further cognitive differentiation in one or the other term. Attitude
attacked the stimulus-response account of "prestige suggestionyy(a traditional topic in change is accounted for by pressures to restore balance when new information has
the earlier literature on attitude change) .in a series of studies that are described in upset it. The formal model, which develops widely influential ideas suggested by
his excellent textbook (Asch, 1952). The controversy centered on how to interpret Heider ( 1958), has not been extensively tested.
the empirical fact that people tend to evaluate a slogan or literary passage more The ability of cognitive-affective consistency theory, taken rather loosely, to
highly when it is attributed to a prestigious source than when it is attributed to a suggest political strategies is illustrated by Rosenberg (1967) in regard to peace-
source toward whom they feel less favorable. The S-R account asserted an essentially oriented politics of the Cold War era (Rosenberg also draws on the broader gamut
mechanical process of associative learning, in which the positive or negative affect of empirical generalizations and theories about attitude change) and by Rosenberg,
aroused by the source. adheres to the message (on the model of classical condition- Verba, and Converse (1970) in regard to peace advocacy during the Vietnam war.
ing). Asch's Gestalt-oriented view held, to the contrary, that attribution serves rather In a handbook for "doves," the latter authors urge that "theagreat persuasion task is
to provide a new context of meaning; this new context induces changes in the cog- to show Americans that some of the doubts they already feel are indeed very legiti-
nitive object, so that changed evaluuative judgments and accompanying affect are mate ones, and that the ['hawkish'] Southeast Asian policy which they still endorse
then appropriate. At least two issues apparently are involved. One has to do with and mildly favor does, in fact, violate their positive values and advance negative
the priority of cognitive as compared with affective factors in attitude change. Do ones" (p. 97). The relative lack of firmly crystallized, consistent structure in "dovishy'
people change their feelings about an object because,they have come to see it differ- 2r "hawkish" attitudes (Converse and Schumann, 1970) suggested to the authors
ently, or do they change their beliefs about it to fit prior alterations in their feelings? that the appropriate cultivation of inconsistencies could change public attitudes in
The evidence now seems clear that both processes occur; what may be primary is a the direction they favored.
tendency to bring beliefs and feelings into congruence (Rosenberg, 1960). From By far the greatest experimental attention has been devoted to Festinger's
present perspectives, the second issue also seems less clearly drawn than it did in (1957) theory of cognitive dissonance. The theory itself is simple (too simple, its
1950: Are the processes of influence to be interpreted in associative or in meaningful
critics maintain). Any two cognitive elements-beliefs or bits of knowledge-may be
terms? Recent elaborations of associative theory, which emphasize central media-
either consonant, dissonant, or irrelevant to one another..Dissonance occurs when one
tional processes intervening between stimulus and response, tend to converge with
dement follows psychologically from the contrary of the other. Thus, my knowledge
the older cognitive theories. Heat has dissipated from controversy as theorists so-
that I have irretrievably purchased a new house is dissonant with my awareness of
cialized to feel at home with stimulus-response or with cognitive terminologies came
zach of the disdavantages entailed in my purchase. The total amount of dissonance
to see ,their differences as more a matter of linguisic preference and conceptual
that a person experiences is a function of the importance of the elements in a disso-
strategy and less a question of truth versus fa1sity.l
nant relationship and of the pro$ortion.of relevant relations that are dissonant.
1The discussion of Asch is adapted from Smith (1968a),p. 461. People tend to attempt to reduce dissonance when it arises: states of dissonance have
70 M. Brewster Smith Poliitical Attitudes 71

motivational properties. Dissonance may be.reduced in three major ways: by chang- which may be ccobvious"but needs to be formulated and specified [Smith,
ing one or more of the elements involved in dissonant relations, by adding new 1968a, p. 4641.
cognitive elements that- are consonant with already existing cognitions, and by de-
DW feel comfortable in standing by this assessment. There is some solid residue
creasing the importance of the dissonant elements. A general tendency for cognitions
trom the flood of research oriented to dissonance theory, but it is clearly no longer
to be brought into correspondence S t h impinging reality is a s s ~ m e d . ~
In many respects this capsule summary is misleading, though accurate. One at the "cutting edge" of apparent advance.in social psychology. Attention has turned
elsewhere.
could never guess from it the main preoccupations of dissonance research or the
Paradoxically, the consistency theories, dissonance theory among them, are
nature of its contribution: I n the first place, dissonance theory has been applied by
pole:s apart from the cognitive theory of Asch (1952), which, Gestalt-wise, exag-
Festinger and others to a much narrower range of phenomena than was implied by
gerzrted the rationality of human judgments. (Asch claimed that even if people have
Festinger's initial general statement: mainly to the case in which a person's knowl-
edge of what he has just done is dissonant with his awareness of grounds for not
.- i-lecide and act on inadequate or distorted information, they still can be counted
tn

on 1:o behave rationally in the world as they see it. The kind of irrationality brought
having done it-a special case indeed. Thus, the smaller the bribe or threat used to
lo liight by Freud embarrassed him, and he could only treat it residually.) Consistency
induce a person to take a public stand contrary to his private attitude, the more
thec)ry has turned out to be mainly a theory of rationalization, not of rationality. In
likely he is subsequently to change his private attitude in the direction of his public
its alternative versions, it gives a good account of the strain toward a simple com-
action. Knowledge that one has sold out for a small price is more dissonant with
placent view of oneself and the world, which, we can agree, is all too characteristic
knowledge of one's discrepant private attitude than is knowledge that the "price is
of political attitudes. I t has much less to say about the circumstances in which some
right," and it therefore results in greater pressure to attitude change (when other
people, some of the time, come to attend to uncomfortably insistent facts and even
routes for the reduction of dissonance are excluded experimentally). Such "non- to seek them out-though both Rosenberg and Abelson (1960) and Heider (1958)
cbvious" predictions have been the stock in trade of dissonance theorists. before them had the promising thought that we move toward views that are more
Nowhere in research on dissonance theory is the amount of dissonance in- complex cognitively because simpler ones make trouble for us, either in our commerce
dependently assessed, and the impression that "cognitive elements" can be isolated, with reality or in keeping our own inner peace.
weighed, and counted is wrong. Festinger and his followers compensate for the weak- An additional cognitively oriented approach to attitude change remains to
ness of the theory in this respect (the lack of coordinating definitions to link it with be considered. This approach-the application of psychological theories of judgment
social reality) by great ingenuity in the creation of special experimental situations (Sherif and Hovland, 1961; Sherif, Sherif, and Nebergall, 1965)-has, I think, more
where the presumed balance of elements is such that predictions can be generated. direct applicability to political attitudes and behavior than the consistency theories,
The theory is thus ill adapted to application in the interpretation of social behavior though it is less fully developed and tested in research. According to this approach,
outside the laboratory. a person's attitude on a controversial issue corresponds to the range of discriminable
Much of the ingenuity of Festingerian experimentation goes into "conning" opinion positions that he finds acceptable. The person's latitude of acceptance is
the experimental subject into taking a stand or engaging in a specified behavior typically narrower than the accompanying latitude of rejection when he is highly
while under the impression that he has a free choice. The predictions of the theory ego-involved with the issue or when his position is extreme. In responding to a per-
mostly concern what happens after this behavior because of the cognitive discrep- suasive communication that advocates a particular position on the issue, he first
ancies that it gives rise to. The elaborate manipulations employed to set up situations locates it on a subjective pro-con scale of favorability with respect to the issue. The
in which the theory can be tested are themselves more a matter of artistry than of persuasive effect of the communication on the recipient will depend heavily on the
theory. This whole line of research has tended to substitute experimental ingenuity distance between the recipient's own stand and where he locates the position advocated
for theoretical explicitness. by the communication on his scale of judgment. Maximal persuasive effects are to be
Several years ago I was taken to task by some of my colleagues for the follow- expc:cted when the position advocated in the communication falls within the recip-
ing appraisal of dissonance theory : ient''s latitude of acceptance but near its boundary. (If the position advocated falls
centrally in the person's latitude of acceptance, he agrees with it, but there is no
The conjecture may be ventured that, in the long run, dissonance theory occasion for persuasion.) Under these conditions, the recipient is likely to minimize
will turn out to have made sense of certain paradoxical feedback effects of its judged distance from his own position (assimilation effect) and to be open to its
a person's behavior upon his attitudes but to have said little that is im- influence. When the position of the communication falls within his latitude of
portant about the main themes governing the formation of attitudes and rejection, he is likely to exaggerate its judged distance from his own stand (contrast
the direction of behavior. Or it may become incorporated in a more com- eflect) and to resist influence. On issues that do not "ego-involve" the recipient,
prehensive theory that deals with these themes. The lure of the paradox-
ical "nonobvious prediction" can deflect attention from the main story, relatively broad latitudes of acceptance may be expected. Under these circumstances
the more the position of the communication differs from that of the recipient, the
2 The foregoing paragraph is adapted from Smith (1968a),p. 463. greater the persuasive effect, within broad limits. High ego involvement, however, is
72 M. Brewster Smith Political Attitudes 73

accompanied by narrow latitudes of acceptance. With high ego involvement, therefore, From this perspective, attitudes themselves are important components of personality
contrast effects are likely to magnify the perceived discrepancy between the positions and also may be regarded as influenced by the encompassing personality system in
of the communication and the recipient. A more discrepant communication would which they are embedded. Social psychologists have focused mostly on relatively
thus be expected to be less effecti~e.~ superficial attitudes, leaving the more central ones to clinicians (like attitudes toward
An additional concept drawn by Sherif from the psychology of judgment con- the self and complementary attitudes toward the world, such as hopefulness and
cerns the anchoring of judgments to particularly strong or salient reference points. trust or suspiciousness and despair). But there is nothing inherent to the concept of
If I know that a particular weight is "heavy," I am likely to judge any lesser weight attitude that pushes it outside the realm of personality or even toward its periphery.
as "light." So with opinions: depending on which ones define a "radical1' position on Researchers of many persuasions have drawn upon psychoanalytic theory for
my subjective scale, my judgments of what is "liberal" will vary. The extreme or end hypotheses about the motivational sources of attitude formation and change. Al-
stimuli in any continuum of judgment are especially likely to serve as anchors when though Sarnoff (1960) has attempted the most direct and explicit application to the
a person has had little experience with an ordered stimulus series on the particular theory of social attitudes, the most substantial impact has undoubtedly come via the
dimension, when the potential range of stimulus values is unknown, or when no theory of the authoritarian personality (Adorno and others, 1950; see Chapter Six
explicit standards for judgment are provided (Kiesler, Collins, and Miller, 1969, pp. in this volume). The prejudiced person, and by implication the proto-fascist, uses his
241-242). attitudes to maintain a rigid and defensive posture; bolsters his vulnerable self-esteem
There are problems with the theory, and data from the fairly substantial body by identifying with the strong and rejecting the weak; and resolves his own uncer-
of research that it has stimulated are not entirely consistent with it. On the whole, tainties and keeps his unacceptable impulses in check (while giving them covert
however, principles of social judgment are directly suggestive about important phe- expression) by cleaving moralistically to a world of clear-cut alternatives, one in
nomena of political life. In observing the radical protest politics of the late 1 9 6 0 ~ ~ which the safe areas of conventional respectability seem bounded by unknown dangers
for example, I often thought that a constructive political consequence of dramatizing and conspiracies. This is a special theory of a particular type of political-social orien-
an uncompromisingly pure and extreme position-say, on the Vietnam war-is to tation that I would myself place in a broader framework of functional theory (see
establish a new extreme "end anchor" in people's scales of political judgment. The below). The theory is better than much of the research that has been adduced to
extreme position will not initially attract many supporters, and when it is promoted support or to refute it. I t is still very much alive and relevant.
without benefit of invitation to reasonable compromise, it is hardly designed to do Christie's concept of Machiavellianism (Christie and Geis, 1970) occupies
so. But to the extent that it catches the public eye sufficiently to anchor a new end similar position-intermediate between attitudes, ideology, and personality-though
point on people's subjective scales, it may lead people to redefine what they regard .s sources are not psychoanalytic. Scores on the "Mach scale," drawn in part from
as a "moderate" position and thus contribute to shifting the grounds of political dachiavelli's writings, tap an orientation of manipulativeness and cynicism about
debate. uman nature that turns out to have behavioral consequences in accord with ex-
Another group of theories, personality-oriented theories of attitudes, were low . ectations in contrived experimental situations. The dimension has obvious relevance
in prestige during the boom years of laboratory experimentalism, if only because > political behavior.
individual differences in personality cannot themselves be manipulated as inde- A theoretical approach to attitudes and attitude change that has yet to be
pendent variables. (But the empirically oriented Yale program nonetheless paid laborated would draw upon conceptions of the self. The data upon which such a
attention to personality variables as facilitating, limiting, or interacting with the leory could build come not from the laboratory with its pallid "ego involvements"
independent variables of persuasive communication; see Hovland and Janis, 1959.) Sherif and Cantril, 1947) but from the engagements of self in life crises-including
Before surveying the theories that guided such research as did go on, we need to clear le crises artificially provoked by the Chinese communists in "brainwashing" and
up a confusion that I think prevails in references to personality factors in political thought reform" (Lifton, 1961; Schein, Schneier, and Barker, 1961) and in psycho-
psychology. lerapy. Laboratory studies of attitude change have perforce been confined to atti-
For most writers outside of psychology, "personality'' has come to connote the ides that do not make too much difference to people, and the theories to which they
realm of deep motivational dynamics typified by Freudian psychoanalysis. When ave given rise may well be correspondingly limited.
researchers announce, with disappointment or with satisfaction, that they have tried I n successful cases of "brainwashing," a person's deep-seated convictions are
and failed to show the relevance of "personality factors'' to some class of outcomes attacked and their social supports withdrawn to achieve a profound reorientation of
in political and social life, attempts to show the effects of variables in this realm are . attitudes. The metaphor of "death and rebirth" comes naturally here, as in the case
usually at issue. A less restrictive conception of personality is also possible, and I of religious conversion. If attitudes that have become central constituents of the self
think has much to recommend it. re to be changed, the person's very sense of identity must be challenged; guilt is
According to this view, personality is the overall organization of a person's voked, confessed, and expiated. The person is then given practice in performing
dispositions-a view congruent with that elaborated by Knutson in Chapter Two. le roles of his prescribed new identity and is supported in this identity by a new set
f social relationships. Erikson (1958) has given a strikingly similar account of Martin
3 This paragraph is adapted from Smith (1968a), p. 462. ,utherls transformation from a young layman into a monk. The parallels between
74 M. Brewster Smith 'olitical Attitudes

attitude change in self-involving life situations and in psychotherapy (in faith healing If we are concerned with understanding the personal basis of political attitudes
and brainwashing as well) have been noted in a provocative and readable book and their contribution to political behavior, some of the features of the functional
(Frank, 1961) that has in no way been superseded. approach that put off the experimentalists may even be appealing. This may explain
The most systematic recent attempt to formulate the common ingredients of why the schema advanced in Opinions and Personality (Smith, Bruner, and White,
these "self-reconstitution" processes is by Sarbin and Adler (1970-1971), whose con- 1956) may have had more currency among political scientists than it has had in psy-
ception of the self is embedded in the more general perspective of social role theory. chology. Greenstein's Personality and Politics (1969), which featured it saliently,
They note the following common themes: the enactment of symbolic death and surely has much to do with its currency in his discipline.
rebirth, the importance of the group or the "other" as a source of role demands and
a model for identification, the use of ritual behavior, reliance on proprioceptive Functional Approach
stimuli to manipulate "core anchorages" of the self, and the prominence of "trigger"
events that enhance the process of conversion. In the cases they examined, they ob- The' functionalist conceptual map presented in Figure 1 is an outgrowth of
serve that "three central processes appear to be at work: (1) a physical and/or my attempts to apply the approach developed in Opinions and Personality to the
psychological assault (symbolic death) ; a developing confusion about self and other analysis of various problems involving social attitudes and behavior-in particular,
beliefs (the bridge between death and rebirth) ; (2) surrender and despair (becoming McCarthyism, civil liberties, and anti-Semitism. I have also found it useful in teach-
a nonperson) ; and ( 3 ) a working through, active mastery, reeducation or adapta- . ~is not a theory that can be confirmed or falsified, but rather a declaration of
i r ~ gIt
tion process (the rebirth experience)" pp. 614-615. Unfortunately, largely because intellectual strategy to be judged as profitable or sterile rather than as true or false.
of the emphasis on laboratory studies of attitudes, these important processes have not 4s a framework, it provides a roost for verifiable assertions. My main claim for it is
been treated in most formal presentations of attitude theory. Among the academically hat it may help to counterbalance the normal human tendency to stress the exclu-
based contributions that can be drawn upon to extend this view of fundamental, self- ive importance of the variables or theories that one is momentarily captured by. A
involving attitude change is Rokeach's (1960, 1968) structural treatment of personal :cond claim is that such an armchair "path analysis" of relationships between types
belief and value systems, in which (among other distinctions) he contrasts a central of variables can undercut prevalent disputes between partial accounts regarded by
region of primitive beliefs about the self and world with a peripheral region com- their proponents as mutually exclusive but perhaps better regarded as complementary.
prising beliefs received on authority. For the moment, disregard the tangle of arrows in the fine structure of the
A final set of personality-oriented theories to be considered has come to be lap, and look only at the five major panels identified by Roman numerals. In keep-
known as the functional approach, because their angle of access is concerned with ~g with the psychological focus of the map, Personality Processes and Dispositions
the functions that opinions and attitudes serve in the ongoing economy of personality. Panel 111) occupy the center of the stage. Because we are used to reading them from
During the era dominated by laboratory experiments, functional theory did not enjoy left to right, I have put the payoff in Political Behavior (Panel V) at the extreme
much prestige among experimental social psychologists, although the variant pro- right. Imagine that panel to include any politically relevant actions that we may be
posed by Katz and Stotland (1959) generated its share of research. We have already interested in: voting, information seeking, influence attempts, administrative decision
noted that personality-oriented theories did not fit the experimentalist mold of focus- making, or even question answering. Our observational data come from this panel
ing on independent variables that could be manipulated. The laboratory era was also and the other peripheral ones; onIy by reconstruction and inference do we arrive at
a time of formalization, of miniature theories that seemed to lend themselves to the contents of Panel 111.
definitive test. In contrast, functional theory attempted to create a coherently eclectic Starting with Political Behavior (Panel V), the arrows (marked A and B)
framework in which, for example, the defensive processes highlighted by psycho- that link it with Personality Processes and Dispositions (Panel 111) and with The
analysis have an appropriate place-but only a place, not an exclusive billing. Situation as Immediate Antecedent of Action (Panel IV) represent Lewin's method-
From the standpoint of aspirants to hard science, another drawback inheres in ological premise: All social behavior is to be analyzed as a joint resultant of char-
any functional approach: it is admittedly only a way station en route to a more acteristics of the person, on the one hand, and of his psychological situation, on the
detailed causal analysis. Only from a dubious teleological standpoint can functions other. To specify the contribution of either requires taking the other into account.
be regarded as explanations. In his focus on the adaptive transactions of the human To take this feature of the map seriously is to regard the old quarrel between psy-
organism, the functional theorist rests his pragmatic case on the de facto teleology of chologists and sociologists about the relative importance of personal dispositions
the adaptive process. He insists that although functions are not causes, an analysis of ' versus situations as silly and outmoded: the two classes of determinants are jointly
how a person's opinions and atitudes are useful to him is bound to be relevant to a indispensable. Depending on the behavioral outcome in question and historical
causal understanding of their acquisition, maintenance, and change. In the social contingencies, one or the other may control more of the variance; one or the other
psychology of attitudes, he would argue, we may need to go through the naturalistic may also be strategically more accessible if we are interested in influencing the
functionalism of Darwin before it makes sense for us to aspire to the causal models behavior. Thus, chaAging the law (Panel IV) may be a better way of reducing racist
of Watson and Crick. Psychology has a bad habit of aping the most advanced models
of the hard sciences, whether it is ready for them or not. 4 This section is adpated from Smith (1968d).
 1
Political Attitudes 77
-------------
_--------- behavior than attempting to change personality-even if one grants the correctness
of the theory of authoritarianism.
Causal antecedents can be traced back from each of the two panels that
show immediate determinants of action. From Personality (Panel 111) a cluster of
mows leads us to Panel 11, Social Environment as Context for the Development
f Personality and Acquisition of Attitudes. Both the environment of socialization
Panel 11) and the immediate situation of action (Panel IV) have their own more
istal antecedents, represented in Panel I. Historical, economic, and institutional
kctors have their impact on individual behavior-both by shaping the contexts in
rhich socialization occurs and attitudes are learned (Arrow D) and as sources of
le features of the immediate situations in which action takes place (Arrow E) .
The broken arrows from Panel V reflect the consequences of political behavior,
rhich may alter the situation in which it occurs (Arrow F), and cumulate across the
iany actions of many persons to modify the social environments that shape and
lpport the attitudes of each (Arrow H), in the longer run constituting history and
iaping institutions (Arrow I ) . Arrow G, leading back from behavior to personal
,ispositions in Panel 111, represents the effects that self-committing behavior can have
on attitudes. This is the phenomenon so much emphasized by Festinger (1957),
which stands in its own right independently of dissonance theory. A political actor
who adopts a position for expedient reasons may be convinced by his own rhetoric.
With the broad framework of the map clarified, we can now turn to the
details. Consider first Panel 111, personality processes and dispositions, working from
right to left within the panel. We are concerned here with the inferred dispositions
that the person brings to any situation that he encounters, and with their motiva-
tional basis. The problem is a dual one: to formulate how a person's attitudes come
to bear on his political behavior and how these attitudes arise and are sustained
in relation to their part in his ongoing psychological economy.
A first point that the map suggests is that we cannot take for granted just
which of a person's attitudes will become engaged as a codeterminant of his behavior
in a political situation. Political scientists are probably less naYve than psychologists
about this. A citizen's vote for one or another presidential candidate depends, as we
know (Campbell and others, 1960), not orily on his focal attitude toward that
candidate but also on attitudes toward alternative candidates, toward party, and
toward issues. As Arrows M, N, and 0 are intended to indicate, the situation plays
a dual role. It engages with certain of the person's attitudes and leaves in abeyance.
others that might potentially be engaged; and it serves as a codeterminant of be-
havior, together with the engaged attitudes. An example: On the floor of Congress,
certain of a congressman's attitudes become engaged with the issue under discussion-
I different ones, very likely, from those that would be engaged in his discussion of the
m e issue with an important constituent. But what he says in either situation (and
lying is behaving) will depend not only on his engaged attitudes but on what seems
ppropriate and instrumentally effective given the norms and contingencies of each
situation.
These complex relationships give us no reason to suppose that people's political
behavior should correspond to their attitudes on a single focal issue in any simple
I
, I way. There are other technical aspects to the problem of the relation between atti-
I tudes and behavior (Fishbein, 1967) ; if the foregoing analysis makes sense, however,
-,
----------------,-------
-------0
 M. Brewster Smith Political Attitudes

much of the puzzlement of sociologists and psychologists about the lack of clear tngenial to his needs" (p. 53). This formulation closely parallels my conception
correspondence between attitudes and behavior is pointless. object appraisal. As Kelman notes, the conditions and consequences of social
Moving to the left of Panel 111, we turn to the problem of how attitudes are fluence are describably different for the three cases. The function of externalization
formed and sustained. Three functions of attitudes are proposed, slightly modified and ego defense understandably falls outside his scheme, since it does not lend itself
from Smith, Bruner, and White (1956). so obviously to deliberate social influence (his scheme really pertains more germanely
Under object appraisal we recognize the ways in which a person's attitudes to social influence than to opinion change). Attitudes strongly grounded in this
serve him by "sizing up" significant aspects of the world in terms of their relevance function are likely to be held rigidly, but if they do change, the change is likely to be
to his motives, interests, and values. All attitudes, not just prejudice, involve an saltatory and irrational, to another attitude that is dynamically equivalent for the
element of "prejudgment" : they are useful to the person in part because they prepare person, as in the conversion of a true-believing communist to a true-believing
him for his encounters with reality, enabling him to avoid the confusion and in- Catholic.
efficiency of appraising each new situation afresh. , To return to the map, Arrows S and T, near the bottom of Panel 111, reflect
A person's attitudes not only embody a provisional appraisal of what for h i a different kind of relationship, which falls outside a strictly functional analysis. A
is significant reality; they also serve to mediate the kinds of relationships with others person's attitudes and the way they engage with particular political situations bear
and the kind of conception of self that he wants to maintain. Is it important to the the mark of his stylistic traits of personality as well as of the purposes that they serve
decision maker to think of himself as a liberal Democrat? Then his adopting a liberal him. Intelligence or stupidity, incisiveness or vagueness, zest or apathy, optimism or
position on an issue may contribute to his self-regard, as well as maintaining his pessimism, decisiveness or hesitation-cognitive, temperamental, and behavioral
standing with his political fellows. traits like these have their own history and may perhaps partly be residues of the
Finally comes the class of functions to which psychoanalytic depth psychology person's previous motivational conflicts, but their immediate relevance for his political
has given the closest attention, here labeled externalization and ego defense, which attitudes and behavior is hardly motivational.
1
h
underlies the theory of the authoritarian personality (Adorno and others, 1950). j
As we turn to Panel I1 at the left of the map, we now have a basis for identify-
Here there is a covert agenda: the person's attitude is really less concerned with the ing aspects of the person's social environment that are relevant to the development,
avowed object than with containing some inner conflict that is analogically linked maintenance, and change of his political attitudes. To the extent that object appraisal
with it. McClosky's (1967a) data suggest a substantial ingredient of externalization - involved, he should be responsive to the information that his environment provides
in isolationist attitudes. Not only do isolationists tend to score high on scales of bout the attitudinal object or issue (Arrow U ) . The actual facts about it will be
hostility, paranoia, misanthropy, and authoritarianism-content that suggests un- nportant in this connection only insofar as they affect the information that is socially
!
finished intrapsychic business-but they also show more contempt than others~for vailable to him. The quantity and quality of this information will obviously vary
weakness and more intolerance of human frailty. As McClosky notes, it may be ~idelyfrom issue to issue and across the various niches that people occupy in society.
logically inconsistent for an isolationist simultaneously to fear the demonic power of The information about a topic that reaches a person has a dual relevance, as
others and to scorn them for their weaknesses. The psychological consistency lies in ie internal arrows in Panel I1 are intended to suggest: it feeds into his processes of
the realm of externalization. bject appraisal, and it also carries a second-order message about the social norms
The arrows P, Q, and R raise the functional question about the motivation iat prevail. When discussion of birth control began to percolate through Catholic
underIying any attitude. Different topics may be biased toward one or another func- hannels, new grist was provided for object appraisal; and, in addition, the important
tion (think of the fluoridation issue, for example), and on the same topic people will ews was conveyed that these previously taboo topics had become moot and dis-
differ in the balance of the functional mix on which their attitudes are based. ussable. As Arrow V indicates, the second motivational basis of attitudes-the medi-
We have by now traced through enough of the map to place on it the dis- tion of self-other relationships-may then lead to a state of affairs in which it
tinctions drawn by Kelman (1958) in a widely cited article on three processes of ecomes safe to think in new ways. These relationships link reference-group theory
opinion change. Compliance in overt behavior for the sake of rewards or punishments Merton, 1957, pp. 255-386; Kelly, 1952) to the functional analysis of attitude
represents the impact of Arrow B- the predominance of situational pressures over ~rmation.and change.
one's own attitude in the resultant behavior. This is really not a case of opinion A person's life situation and socialization experiences may also predispose him
change, but of social influence on behavior. Identification, which "can be said to -in general or in a particular topical domain-toward one or another of the
occur when an individual accepts influence because he wants to establish or maintain lnctional bases of attitudes (Arrows W, X, and Y). What makes the rational man,
a satisfyingself-defining relationship to another person or group'' (p. 53)' corresponds in which the first function predominates? A good guess is that part of the story is
in my terms to influence governed by mediation of self-other relationships. Internali- rearing by loving and confident parents who give reasons for their discipline. In the
zation "can be said to occur when an individual accepts influence because the content shorter run, environments that sustain self-esteem and allay anxiety should also favor
of the induced behavior-the ideas and actions of which it is composed-is intrin- object appraisal. Research by the Witkin group (1962) on field dependence-inde-
sically rewarding. He adopts the induced behavior because it is congruent with his pendence and Miller and Swanson's work (1958, 1960) on child rearing and per-
value system. He may consider it useful for the solution of .a problem or find it sonality in entrepreneurial and bureaucratic families contain suggestions about the
80 M. Brewster Smith Political Attitudes 81
/

sources of primary orientation to the second function, mediation of self-other rela- who switch [by crossing party lines 'from one presidential election to
tionships. As for externalization and ego defense, I would guess that conditions that another] do so to support governmental policies or outlooks with which
subject the developing person to arbitrary authority, that deflate self-esteem, that they agree, not because of subtle psychological or sociological peculiarities
arouse vague anxiety, that provoke hostility but block its direct expression are likely [P 1041.
sources.
I am in no position to assess the extent to which Key's assertion at the end of the
To provide a useful simplification, any map leaves out complexities that it
first quotation accords with the full data available, although there is apparently
does not attempt to handle. This one assumes the basic processes of motivation, per- I room for legitimate disagreement (see Sears, 1969). My concern is rather with Key's
ception, and learning rather than spelling them out. Thus, the threefold functional
assumption that any account of political behavior in terms of social or psychological
classification sorts out the ways in which a person's attitudes are rooted in his under-
determinants is incompatible with the respectful treatment of "civic decisions." This
lying motives, whatever they may be, without spelling out a conception of human
seems to me to misunderstand the purport of sociological and psychological analysis
motivation. As for perception, the map ignores the perceptual screening process that 1 (see Smith, 1968d).
intervenes between the environmental facts (Panel 11) and what the person makes of
Nobody has claimed that "sociological peculiarities" or "social determinants"
them (Panel 111); likewise, between the immediate situation as it might appear to
an objective observer (Panel IV) and how the person defines it for himself. In regard
1 exert their influence by any mysterious process of "straitjacketing'' the electorate.
Their influence lies rather in the fact that common experience and common social
to learning, the present formulation makes the broad functionalist assumption (com-
patible with either a reinforcement or a cognitive theory of learning) that in general
i position yield similar perceptions of interest, and over time lead to the emergence of
norms that reflect these perceptions. "Civic decision" is channeled, not abolished. I t
people acquire attitudes that are useful to them. But it ignores the details of the
would be the less rational if it did not reflect social determinants.
learning process and the persistence of learned structures beyond their original point
Similarly with psychological determination, which is not to be equated solely
of usefulness. It also ignores incidental learning, according to which a person may
with "subconscious ~rges"-the realm of externalization and ego defense. (Note,
acquire much of the content of his political attitudes in an unfocused, only mildly
'however, that political rationality in the aggregate is entirely compatible with the
attentive effort to make sense of his world. At the time of learning there may be little
influence of "subconscious urges" as long as they are divergent or randomly dis-
or no real payoff in object appraisal or mediation of self-other relations; yet should
tributed in the electorate. It is when blocs of voters start externalizing in tandem that
the occasion arise, the basis for resonance to certain political positions rather than
a democracy is in trouble-for instance, from McCarthyism.) Decisions governed by
others has been laid.
I
object appraisal and the mediation of self-other relations remain decisions, potentially
Political Attitudes and Model of Political Man rational ones, even when their psychological determinants are displayed.
Skinner (1971), the humanists, and on this occasion Key (1966) appear to
The functional view of political attitudes, as I have just illustrated it, fits a
agree on philosophical or "metapsychologicaly' premise; according to which the
view of man as a political actor who is guided in part by rational considerations of
scientific search for determinafits of human behavior is incompatible with human
how to advance his values and interests according to his lights, in part by the social
freedom, dignity, and responsibility. These premises .have recently been given a
imperative to locate himself in relation to significant others, and in part by irrational
searching examination by Chein (f9f2), who makes a strong case that human
by-products of his symbolic ways of handling his inner conflicts. The quasi-teleologi-
freedom as we know it (of which political freedom is a special instance) depends
cal phrasing embedded in the functionalist approach is in keeping with the view of
on causal relations of the sort that science explicates, far from 'being incompatible
man as an actor or agent, not as a mechanism or "robot," which it seems to me is
with them. These abstract considerations need concern us here only insofar as they
essential if the realm of politics is to retain its legitimate human meaning. But is the
'legitimize the scientific treatment of man as an agent.
search for determinants of political attitudes and behavior, for which my map pro- Empirically, it is obvious that some people enjoy more freedom than others,
vides a scaffolding, compatible with the conception of man as a political actor?
whether we are concerned with political freedom (as cherished in normative political
V. 0.Key (1966), the salty master analyst of American politics, thought not.
&eory) or with inner freedom (to which psychological analysis is appropriate).
He saw the new empirical analyses of electoral behavior as adding up "to a con- Political man may 'be more an agent or less. A ~oliticalpsychology that seeks to
ception of voting not as a civic decision but as an almost purely deterministic act''
contribute, as did Key, to a democratic politics that enhances human values should
(p. 6 ) , and summarized his own counteracting view thus:
be particularly interested in the psychological determinants of political agency. Recent
research and theory have made a promising start in identifying one set of these
In American presidential campaigns of recent decades the portrait
that develops from the data is not one of an electorate straitjacketed by determinants, the sense of political effectiveness-a concept that lies at the boundary
social determinants or moved by subconscious urges triggered by devilishly of the psychology of attitudes and of broader features of personality as treated by
skillful propagandists. I t is rather one of of an electorate moved by con- Kilutson in Chapter Two. I cross this boundary with the justification of "hot pur-
cern about central and relevant questions of public policy, of govern- suit," because my own theoretical interests have veered sharply in this direction from
mental performance, and of executive personality [p. 81. . . . the traditional area of attitude psychology.
[There is] at least a modicum of evidence for the view that those Almond and Verba (1963) showed that a sense of political efficacy on the
 M. Brewster Smith

part of the citizens appears as a central aspect of the "civic culture" of more effective
democracies. Somewhat earlier, Douvan and Walker (1956) provided survey evi-
dence that the sense of effectiveness in public affairs is at least loosely related to
people's general feelings of competence. We appear to be dealing here with highly
generalized attitudes toward self and world that are important partly because they
tend to be self-confirming. A person who feels ineffective in politics guarantees his
ineffectiveness by failing to participate. Conversely, a person who feels effective is
more likely to participate in ways that actuaIly make him so. The extent to which
people's sense of effectiveness is generalized, the extent to which it is specific to
particular types of situations or realms of concern, is of course an important empiri-
cal issue.
The work on political effectiveness seems to be dealing with a special case of
much the same variable that Rotter (1966) studied under the concept "locus of con-
trol." With a rather crude pencil-and-paper scale that has since come to be widely
used (called the I-E scale, for internal-external), Rotter found that people show
consistent differences in whether they regard their significant outcomes as under
POLITICAL BELIEF
internal control (resulting from their own skill and ability) or under external control
(resulting from fate or chance). Their standing on this variable is related to whether
they take an active or a passive stance in various contrived situations. From a differ- Robert E. Lane
ent theoretical starting point, De Charms (1968) arrived at approximately the same
distinction, between people who regard themselves as origins of social causation and
those who regard themselves as pawns-and by virtue of this attitude act in ways to
perpetuate their being so in fact. He is currently working with schoolchildren on
methods for increasing their sense of being an origin.
Much work remains to be done to clarify concepts and measures in this area.
Psychological research has unfortunately tended to freeze on the I-E scale because
of its availability, much as it did on the F scale of authoritarianism two decades ago.
Nevertheless, the line of investigation opened up seems likely to add to the empirical
base for understanding, in a causal framework, how people can become more self-
determining personaIly and politically. Since fundamental attitudes toward the self I n the analysis of political belief systems
are involved, this research focus makes contact with the subtype of attitude-change and ideologies, the methods of many disciplines are germane and fruitful; this paper
theory that, in an earlier section, I noted as especially needing further elaboration draws upon several of them: history and biography; psychology, anthropology, and
and development: theories hinging on the self as a psychological construct. I am sociology; and, more specifically, sociology of knowledge, phenomenology, and sys-
currently very much interested in both matters, so I am keenly aware that the tems theory. Two special problems will be considered: the problem of constraint
functionalist map presented in Figure 1 has no convenient place to represent them. (that is, what causes beliefs to cluster in certain ways) and the problem of economiz-
A next agenda item, then, might calI for refocusing the functionalist map to ing (that is, the mangement of a belief system). Finally, I will discuss the ingredients
relate the psychology of attitudes more cogently to the acts of political actors (in of a core belief system.
keeping, in this respect, with the proposals for the study of politics made long ago
by Lasswell and Kaplan, 1950). In the new focus, a kind of personality variable Terms and Modes of Political Analysis
would be highlighted that differs from the unconscious motives of orthodox psycho-
analytic theory and the stylistic variables of psychoanalytic ego psychology: attitudi- The language used to refer to the ideas under examination has such rich
rial orientations that are constitutive of the self as actor and as reflexive object. variety, such extreme individual variation, that the casual observer would hardly
Origin versus pawn is a good example. Perhaps in this new conceptual setting, the expect these terms to be dealing with the same kinds of things; yet they often are.
artificial barriers between a social psychology of attitudes and a personality psychology A sampling of the various terms used to refer to political beliefs includes the follow-
of motives and mechanisms may be dissolved. Such a development is fervently to be ing words : analysis (Dahl), attitude (Allport), belief system (Converse), belief-
desired if we aspire to a scientifically more powerful and humanly more relevant
political psychology.
84 Robert E. Lane Patterns of Political Belief 85
I
disbelief system (Rokeach), character (Fromm, Inkeles, andmLevinson), codes ihcluded here. The list is intended to illustrate the richness of the field and the use-
(Leites), counter-ideology (Johnson), creeds (Sutton, ~ ~ r d a lculture
), (Almond, fulness of variety, a usefulness that will be enhanced by current efforts at synthesis.
Easton), doctrine (Oakeshott) , dogma (Rokeach, Adams) , ethos (Lewis), heritage
(Elliott and McDonald), idea systems (Levinson, Lane), ideals (Chandler), identity Ideological Production
(Erikson), ideology (Adorno, Mannheim, Lane), image (Barghoorn), "isms"
(Ebenstein), knowledge (Mannheim) , lore (Becker and Barnes), mental products What we have just seen is an inductive classification of the literature on
(Merton), mind (Cash, Rokeach, Tucker), miranda and credenda (Merriam), myth political thought and belief systems; other, more systematic, approaches have more
(Cassirer, Sorel), opinion (Key, Smith, Lippmann) ,personality (Kardiner, Allport), to do with ideology than with thought. One synoptic paradigm is suggested by the
perspectives (Lasswell), philosophies (Coker) , reflections (Barker), spirit (Lynd, following (Lane, 1962, pp. 4 1 5 4 1 6 : "For any society, an existential base creating
I certain common experiences interpreted through certain cultural premises by men
Beard), theory (Brecht, Hacker), thought (Coker, McIllwain, Rossiter) , tradition I
(Hartz, Hofstadter) , value orientation (Kluckhohn and Strodtbeck), values (Morris, with certain personal qualities in the light of certain social conflicts produces certain
Jacob), view (Owen), utopia (Mannheim, Lasswell), weltanschauung, zeitgeist. I political ideologies." The terms, of course, are ambiguous, and controversy inevitably
At the same time that there is much overlap and confusion in these terms, attaches to whether ideologies imply action, veridicality, conflict, and so forth. I have
there is also an intentional designation of different properties of ideas: cognitive set forth some of these controversies in Political Man (1972, pp. 170-172) and take
emphasis (credenda, knowledge, belief system) ; evaluative, normative emphasis the liberty of abbreviating and reorganizing that discussion here. The paradigm has
(miranda, ideals, values) ; distributional emphasis (culture, ethos, spirit) ; person- to do with shared experiences, thus referring to social, not indiqidual thought.
alistic emphasis (personality, identity, character) ; implication of breadth and scope Let us look first at the final term in. the quotation-political ideology. Most
(philosophy, weltanschauung) ; implication of organized and coherent thought (code, analysts think of ideologies as mixtures of empirical observation and my tho lo^
belief-disbelief system) ; implication of probable invalidity (myth, lore) ; implication (MacIver, 1948). Although some use the word ideology to refer only to ideas that
of official quality (doctrine, creed) ; historical emphasis (tradition, heritage). are embodied in action, such as those employed in social movements, most think of
The uses of the terms do not always reflect the indicated treatment, but they ideology as selectively guiding action of all kinds (Bell, 1960; Sutton and others,
often do. I t is the terms thought, mind, theory, and ideology that are particularly 1956). Ideologies and belief systems are generally thought to be partly conscious and
unhelpful in this.respect. What is missing in this terminology are the words argument, partly unconscious; all ideologies have evaluative components, and most imply a
reasoning, thinking-important concepts to be developed later. "telos" or goal reference (Billy, 1953; Parsons, 1951). Most analysts agree that in any
But instead of looking at the terms employed, with their special emphases, I organization or society there is some compulsion to believe certain core elements of
let us look more closely at the kinds of subject matter treated. Twelve such,classes the belief system (Parsons, 1951) .
of subject matter (with considerable overlap) emerge: studies of historical periods Let us now look at the other italicized terms in the quotation, taking them in
(Coker, 1934; Gierke, 1958) ; studies of political institutions such as political parties order. An existential base, a subject and term explicated by the sociology of knowl-
or the Supreme Court (McIllwain, 1910; Ostrogorski, 1902);close textual analysis of edge, is the carrier or social vehicle of the ideology in question. For the ethos
a particular work (Strauss, 1952) ; studies of a.politica1 philosopher such as Thomas literature-as reflected, for example, in Cash's T h e Mind of the South (1941)-a
Hobbes or a statesman such as James Harrington (Blitzer, 1960; Strauss, 1952) ; people, community, or region is the vehicle of transmission and the source of nur-
studies focusing on a particular issue or concept, such as justice or equality (Friedrich turance. But for belief patterns more precisely termed ideologies, the vehicles are
and Chapman, 1963; Tawney, 1926) ; studies of a particular political philosophy or variously said to be political parties, social classes, or occupations (Bell, 1960; Lowen-
ideology, such as liberalism or communism (Laski, 1927; Ruggiero, 1927) ; studies stein, 1953a, 195313; Mannheim, 1949).
of events such as the American revolution or revolutions in general (Brinton, 1938; Common experiences reflect thoughts-just as thought guides experience. The
Rossiter, 1953) ; studies of political thought in an entire nation or in a community term that perhaps best captures this relationship is function. As we shall see, com-
(Banfield,, 1958; D. M. Brown, 1959; Merriam, 1929; Warner, 1949) ; theories of mon belief patterns are thought to be "functional" to organizations and societies (as
personality, such as the authoritarian or the psychopathic personality (Adorno and well as to individuals) in helping the members work together, in providing them
others, 1950; Lasswell, 1930) ; studies of a set of experiences (Lifton, 1961) ; studies with rationales for their sacrifices and their relative social positions, and in main-
in the sociology of knowledge (Mannheim, 1949; Merton, 1957) ; studies in the taining their morale (Parsons, 1951) .
psychology of belief (Harvey, Hunt, and Schroder, 1961; Rokeach, 1960) ; analysis Cultural premises are the common fund of values, epistemologies, and beliefs
of opinion structures or belief systems (Converse, 1964; Lane, 1962.) of any particular culture; political belief patterns and ideologies draw upon these
This is a sampling; it omits some studies of specific kinds of influencing and premises. All ideologies have moral components, all imply epistemologies, all rely
'
origins (as in the socialization studies), the communcation and persuasion studies, upon causal theories available in the culture. The counterideology must selectively
and some of the psychoanalytic studies (including biographies). Studies of belief employ the cultural beliefs of the established order to persuade members of a society
systems inevitably overlap anthropological studies of culture, but these are not to criticize that order (Lane, 1972, p. 171).
86 Robert E. Lane Patterns of Political Belief 87

Personal qualities are so intertwined with cultural premises that analytical Other candidates for meaning change are the words consciousness (individual
separation is difficult, although personality and culture are separate entities. The self-awareness versus collective or shared consciousness of group problems, group
importance of the separation for the analysis of belief patterns is emphasized later, in norms, and traditions) and integration (individual assimilation of ideas into a single
the discussion of agency. Without this separation, the analysis tends to be tautological belief structure versus collective assimilation of persons, with their various beliefs,
or circular because the belief systems that are to be explained (explicandum) tend into a single social unit). The meanings of some terms applied to individuals (for
to turn up in the explanation (explicans). instance, extremist, follower, other-directed) are significant only wth reference to
Finally, without social conflicts, there is no change, no tension in the system, others, Finally, certain transformations of perspectives take place as one moves from
no consciousness of the beliefs in question. The beliefs are just "there." The important analysis of individuals to analysis of collectives: individual experience becomes col-
concepts of rationale; diffusion, persuasion, and other means of learning-even lective hltory; individual thinking becomes collective policy making; individual choos-
ideology itself-lose their meaning without the presence of conflict (Coser, 1956). ing becomes collective choice; individual self-actualization becomes a free society;
individual self-discipline becomes social control or order; individual conscience
Sociology-of-Knowledge Paradigms becomes collective moral norms.
If we are interested in the sources of political belief systems, we might %be Where are the ideas located? In answering this question, we would have to
guided by a sociology-of-knowledge approach, as reflected in Mertonys (1957, pp. consider (1) geography--distance and proximity, important because of communica-
456-488) paradigm. Merton puts his discussion in the form of questions about the tion nets; (2) societal location-roles, institutions, strata, communities, generational
where, what,'how, why, and when of the mental productions that result in belief positions, social structures, nations; (3) ideational contexts-ideas placed in relation
systems. His questions are useful but, as he himself goes on to show, require careful to other ideas, as suggested by the terms modal, deviant, avant-garde, marginal, elite.
explication. Departing from his analysis here, we may note some implications of these What are the properties of ideas and their relationship to other ideas? For
questions. analytical purposes, the most important properties of ideas are relational. Specifically,
Who produces the political ideas? Since ideas are mental products, the minds we would want to know the relationship and the logical connections between one
that produced them should be located. Among the important questions for analysis idea and another; the empirical ground on which the idea stands; the logical struc-
are the following. Are the ideas to be read in the contexts of lives and actions (states- ture of premises, deductions, and inferences; the psychological basis for an idea
men, politicians, agitators) or in the context of other ideas (authors and their (prejudice, dogma, obsession) ; the social basis for an idea (norm, convention, de-
works) ? Are the men important for their representativeness or for their deviance or viance) ; the style of thought reflected in the idea (fantasy, reason) ; the tenacity or
because of the roles they occupy in society--or some combination of these? An stability of the idea; the evaluative component of the idea; the affective component
important distinction in analytical modes is to be found in the analysis of the masses of the idea (preference, desire) ; the mood component of the idea (hope, wish) ; the
compared to the analysis of elites, with their superior information and, above all, future (expectancy) reference of an idea (promise, threat) ; the demand component
their superior consciousness. If the distributed or shared quality of the ideas is im- of an idea (request, order) ; and the conative component (will, plan, intention).
portant, analysis of the collectivity of which the "respondentsy' are members is This suggestive list deals with the properties of a political idea, but we still
do not know their actual referents, what they deal with. They have to be about some-
essential.
thing. Viewed traditionalIy, they are about the state; they answer the questions
When we talk about an individual's pattern of beliefs, compared to our dis-
"Who shall rule?" and "By what principles shall the rulers rule?" From a distributive
cussion of group belief $patterns,the meanings of the terms we use often change, and
point of view, they deal ethically with distributive justice and practically with the
there is a shift in language and in the questions asked. Look, for instance, at the
answer to the Lasswellian (1936) question "Who gets what, when, how?" Institu-
notion of the stability of political beliefs. Speaking of individuals, we might ask:
tionally, political ideas deal with certain structures and processes: executives, legisla-
Under what conditions will an element of a belief pattern change? In a hierarchically
tures, judiciaries and the law, bureaucracies, parties and political cultures, and, of
ordered system, what ideas dominate others? What are the functional equivalents of
course, with ideologies. But viewed from another point of view, they deal with some-
a given idea for a given person? Speaking of the stability of group beliefs, that is,
thing much closer to the individual believer-namely, self in society (Lane, ,1972,
the beliefs shared by the individual members of groups, we would ask: How will the
pp. 170-190).
pattern of individual changes affect the net support in various social positions for a
given element of a belief pattern? What is the pattern of opinion changes in the Phenomena and Their Contexts
collectivity (random, cyclical, constant) ? Similarly, if we were exploring the idea of In some ways the paradigms that we have discussed fail to bring forth the
individual conflict, we would ask how an individual handles ambivalence, disso- basic framework of the questions that scholars and others are interested in. For the
nance, logical inconsistency, psychic conflict. If we were concerned with social con- analysis of political beliefs, it seems to me that there are five basic questions: What
flict, we would ask: How will persons with incompatible interests, needs, and are they? What caused them? What are their consequences (implications)? How
preferences express and deal with these incompatibilities in different social and p'oliti- shall we evaluate them (Are they true? Are they useful? Are they good?)? What
cal contexts? shall we do about them?
 Robert E. Lane Patterns of Political Belief
I
I
I n addition, there are four basic contexts, each of which adds meaning or new suficient condition, one of the most important concepts in science. F'inally, the
I dimensions of understanding to belief systems: comparative (to set standards, give reader will see the ambiguity of the term certainly in cases where objective and
alternatives, reveal commonalities and differences) ;historical (to reveal change; give subjective certainty may be confused, as in the sentence "I am certain that the
time perspectives; indicate antecedents; reveal standards, alternatives, commonalities, probabilities are equal that the electorate will choose the Republican or the Demo-
and differences) ; societal (to show institutional, cultural, and social settings and cratic candidate." The point is that descriptive clarity is impeded dot only by ob-
implications) ; personological (to show the relations of ideas to a man's experience, server bias but also by the very language of description.
conflicts, behavior, and to his other ideas; Greenstein, 1969; Knutson, 1972a). A third confusion often arises between what was in the author's mind when
The description of a political belief system sometimes seems to be a straight- he wrote his work and what a reader sees in the work (text) itself. Minds and texts
forward description of "what is there," that is, the analyst looks at the beliefs he are not the same thing, especially when one has an oral text, such as is provided by an
wishes to describe and simply reports what he sees. But the processes of description interview. Understanding this point clarifies the problem of separating the author's
and interpretation are more complicated than that. They involve three possible mis- meaning from the "plain meaning" of the text, usually through looking at the text in
understandings: those dealing with "objectivity"; those dealing with the very meaning the light of other things the author has written or said. This is especially important
of the term ir there, that is the problem of the meaning of the term is or is not; and in cross-cultural interpretation or interpretation of works written a long time ago
the misunderstanding of the relationship between an author of a work and the work since words change their meanings in translation, in different cultural contexts, and
itself, the text. These are difficult matters, but they are important because of the over time.
widespread confusion arising from a failure to understand them. Now we shall deal with the first question mentioned above: What are they?
We said that the analyst may believe that he is simply looking at the beliefs That is, we shall deal with the explication and analysis of the belief patterns them-
and describing or analyzing them, just as a Rankian historian believes that written selves.
history is a record of "what actually happened" or the New Critic believes that he In this connection, let us look briefly at various critical analytic methods often
is simply analyzing "what is in the poem." But we know that perception is always eniployed in phenomeriological examinations of political beliefs.
selective and biased by what a person expects or wishes to see. This is especially true Explication. Explication seeks to clarify ambiguities, confusions, or incom-
of Tocqueville (1856), Cash (1941), Banfield (1958), and others who describe pleteness in a textual argument; point out implications; and so forth. Although by
whole cultural or community bdief systems, however valuable their insights may be. no means the only method so employed, linguistic analysis such as that used by the
The careful author seeks to correct for this bias first by an awareness of his own British analytical school represents a form of explication. The tell-tale phrase is
preferences, and second by reliance on evidence that is public and "objective," in .
"What do we mean when we say. . . ? Meaning is found not only in word and
the sense that others looking at the same evidence agree upon its character. The phrase usage but also in reference to historical events and in the conventions of the
reader corrects for this bias by informed interpretation of the belief system of the I author's culture as well as statements in his other texts (see Weldon, 1953).
author writing about the belief system of a given culture or community. Exegesis. An exegesis is an analysis of the terms, metaphors, paradoxes, tensions
The second source of misunderstanding lies in the ambiguities of the language in a work, especia~lyto reveal hidden meanings, to compare one text with another,
I
of description, especially the word is or more generally the verb t o be or more to show how the elements combine to state a message (Lane, 1961). Psychoanalysts
generally still the concept of existence, a concept treated by metaphysics and logic. (like the New Critics in literature) employ exegetical skills to discover latent mean-
When an author says, for example, that "American belief systems are authoritarian," ings in a patient's oral texts.
he may mean "all American belief systems," "some American belief systems," or Thematic Analysis. This type of analysis uncovers the latent themes or ideas
I
"these American belief systems." "All" is clear enough, but "some'' can mean "some in ambiguous material; finds an "ethos" in discrete opinions, an argument for capi-
and perhaps all" or "some but not all"; and "these," without precise specification, is talism in Calvin's Institutes or elements of Christian morality in Das Kapital. In
notoriously ambiguous. Similarly, the problem of opposites or negations poses ques- I contrast to the exegetical discovery of fresh paradoxes in each "poem" by the New
tions, for they include contradictions ("American belief systems are not authori- Criticism, thematic analysis may find Jungian archetypes in every character. In the-
tarian"), contrapositives ("Non-American belief systems are not authoritarian"), matic analysis there is an assumption that in the "ground" the figure or theme is
and contraries ("American belief systems are equalitarian"). Logic is important in there.
the analysis of ideas, as we shall see, because this analysis often deals with the logical Classification and Typology. Sometimes there is a gain in understanding if a
relationships betwen two concepts, especially when one idea is said to follow from text is assigned to a "class" of things; the literary critic, for example, often asks
another. In the same way, the analyst of ideas must employ the terms either . . or . whether a work is a good example of a particular genre and hence must answer
with care since the statement "American beliefs are either ones of dominance or of I the question "What is it?" To classify is, in one sense, to give genus and species or
submission" seems to (but may not) rule out the possibility that the same belief pat- differentia; it helps to place a work. Thus, one asks, for example, whether a work is
tern might include both-which is often the case, as where a person is dominant 1 "fascist1' or "authoritarian"; or whether a given constellation of ideas calls for a
toward the weak and submissive toward the strong. The same ambiguity inheres in new concept, a new class, of which it is an exemplar. Typologies, often created in
the term if, which can, but need not, mean 'Ff and only if," that is, a necessary and social science by the intersection of two or more variables (and sometimes by "sub-
 Robert E. Lane Patterns of Political Belief 91

struction" of literary treatments), do not have hierarchical arrangements. They are some difficulties. Here we wish to set forth some of the possibilities of a systems-theory
useful as names for syndromes (properties that occur together). Where a belief system paradigm, also in abbreviated form.
is ambiguous, explication will help to locate it in an established or novel set, class, The current popularity of systems theory is based upon its capacity to incor-
or type. porate multiple considerations simultaneously; it lends itself to multidisciplinary use,
Structural Analysis. Belief systems have forms; idea constellations have as in ecological analysis; and it is compatible with complex computer simulation. But
architectures. In Political Ideology (1962) I observed that the common men of there are important differences between the application of systems theory to certain
Eastport developed their ideas in a relatively amorphous manner; there were few living systems (organisms, organizations, societies) and nonliving systems (markets,
.
"stages," "points one, two . . ," setting forth of premises. I n structural analysis, watersheds, communication networks). A systems analysis is applicable to both; but
constraint is an important concept, for it establishes "what goes with what" in a the living systems have what we may call the property of agency-that is, decisional
habitual fashion; it gives idea clusters and sometimes idea hierarchies. units, or "regnant centersn-while the nonliving systems (and lower-order living
Linguistic Analysis. Since language expresses meaning, the analysis of systems like cells and organs) must rely upon programmed "decision rules" of some
language often gives additional clues to meanings, lexically through word choice; kind.
stylistically through examination of mood, tense, and such rhetorical devices as A belief system independent of the believers or some other unit of decision
hyperbole, irony, or metaphor; and through such literary devices as personalization, is a nonliving system. Although, as Converse (1964) has shown, there are some ad-
climax, and paradox. vantages in an analysis of belief systems independent of persons, personalities, or
Logical Analysis. Syllogisms, chain arguments, truth tables are difficult to other "regnant centers," it seems to me that this mode of analysis gives up too much.
apply to belief systems for a number .of reasons. Because referents are imprecise, What it gives up is the complex intrapersonal considerations that help to explain
escape from the requirements of logic is possible in most instances by the lawyer's the resolution of conffict in an individual, and the play of forces in an organization
trick of distinguishing cases. Because identities are often not clear, because implica- or in a society that account for the resolution of social conflict. In the case of the
tion is vague, and because premises and propositions are almost never fully explicated, individual, the analysis of beliefs apart from the individual's personality fails to take
logical constraints are elusive. Thus, invalid arguments may be frequent (it does not into account the individual's predispositions, his private "decision rules," the per-
follow that .. .), but the conclusions may be true and may be saved logically by sonal functions of a belief for his ongoing life strivings. In the case of mass belief
new premises. The most important utility of logical analysis, then, lies in pointing out systems, an analysis of beliefs that does not include social structure, decisional centers,
lacunae in the arguments, propositions that loosely (if not logically) imply opposing leadership, and communicators, as well as communication nets, loses explanatory
inferences, and various kinds of fallacies: fallacies of composition (what is true of power. Social forces such as technological and economic change operate through
the part is true of the whole), fallacies of division (what is true of the whole is true people, and the people should be included in the analysis. The rule is not a dogma;
of the part), fallacies of accident (misapplication of a rule, often because of failure it is an analytical algorithm designed to aid explanation. If followed, it would make
to specify quantities or degrees), equivocation (deliberate use of a term in more than beliefs the products of a system, with feedback into the system; or, alternatively, it
one sense), ambiguity (unintentional equivocation), non sequitur, begging the would regard beliefs as elements of a system but not themselves a sufficient system.
question or arguing in a circle, ad horninem, and appeals to pity or prejudice or Perhaps the term belief patterns would be more appropriate, although in this chapter
vanity or the like (Black, 1946, pp. 2 11-2 17) . we occasionally slip into more common and looser usage.
Reconstruction. Since not every man produces a text, and since most literate With these considerations in mind, let us consider a paradigm for the analysis
societies provide for the scholar plural texts of ambiguous import and nonliterate of belief patterns (or "systems"), both individual and collective; and discuss its uses
societies provide only inadequately articulated conventions, inferences about beliefs for the analysis of ~olitical(and perhaps other) beliefs. The paradigm-an amalgam
and values must rely on behavior and artifactual evidence serving as text substitutes. of various organization and systems theories (see Blau and Scott, 1962; March and
From such patterned behavior as suspicion of strangers, propitiating prayer, child Simon, 1958; Miller, 1972; Parsons and Shils, 1951)--employs the concept of
abuse, competitive social norms, the analyst must reconstruct the premises of a belief I(regnant centers," giving the system the characteristic of agency. Such a paradigm
system. The anthropologist and the psychiatrist are familiar with this method; but may be outlined in the following way:
imputation of analytical constructs to "natives1' and patients depends on the analyst's
imagination as much as on native behavior. 1. Differentiated elements
2. Related to each other in patterned ways with
Systems Theory 3. Some regnant centers of control and
coordination and with
So far, we have introduced three possible paradigms for consideration in the
analysis of beliefs : the ideological-production paradigm, the sociology-of-knowledge B. External Functions and 4. Goals, problem-soIving strategies, functions,
paradigm, and the phenomenological paradigm. While the explication of these Production Processes needs, requiring
schema has been abbreviated, each has promised both some analytical results and 5. Resources (external and internal) to
92 Robert E. Lane Patterns of Political Belief 93
6. Process inputs to produce outputs with of "coercion" when she thought of government. Imagery is sometimes "tied"-that
7. Informational exchange (feedback) with is, conventional; but sometimes it is intensely personal or idiosyncratic (Richards,
the environment. 1954). Thus, unless the differentiated element in the analysis of political beliefs allows
C. Internal Functions 8. Integrative (not task) mechanisms, for variation created by personal experience and private meaning, the analyst may
for Pattern Maintenance combined with be misled by the standardized codes or words employed.
and Adaptation When we say these differentiated elements are related to each other in
9. Selective admission, perception, and contact
with the environment for patterned ways, we refer mainly to the relationships involved in constraint and
in hierarchy (or centrality). Constraint refers to "what goes with what," and to the
10. SystemLrewarding pattern maintenance and sense of necessity experienced by the thinker (implying agency). Hierarchy deals
adaptation to environmental change.
with "what idea dominates or subsumes or forces change in other ideas under condi-
D. Location and Limits 11. Boundaries distinguishing system and tions of stress." Both of these .concepts imply the agency of the thinking person;
environment (marked, again, by effective without him, ideas are inert. We shall deal with constraint in a later section; here
interchange), let us discuss some problems of establishing hierarchy. We shall take an example
12. Located in time-space-cultural settings. from the ethos literature, as illustrated by Oscar Lewis (1960) in his account of the
E. Lawful, Routinized 13. Rules, laws, habits, conventions to guide
belief pattern in Tepoztlan. In this example we can see a variety of ideas or con-
Behavior element interaction, goal striving, product cepts intertwined, including the basic ones: I know (belief) ; I feel (attitude) ; I
processing, and pattern maintenance and value-desire or consider desirable (value) ; I shall (conation-goal direction) ;You
adaptation. will or please do (demand).
One way to think of the hierarchical organization of the belief system of the
The paradigm has certain uses for the analysis of political (and perhaps Tepoztecans is to examine the interrelationships of the several elements. For example,
other) beliefs which we now discuss. the belief that people and the Gods are hostile and dangerous leads to an attitude
First of all, in its organization this paradigm is made up of differentiated. of suspicion and to a value of independent self-reliance and work, which become
elements related to one another in patterned ways, with some regnant centers of moralized and thus' embody both aspects of value (the desired and the ought) ; to
control and coordination. goals of "accumulating property in land and animals"; and to demands that others
The differentiated elements of a belief system might be considered to be the not impose upon a person-but, more interestingly, lack of demands.from government
separate beliefs; but that might imply a cognitive emphasis, which would exclude for aid and support. In this sequence of thinking, the belief element, the perception
attitudes, values, and other less cognitive terms. Converse (1964) employs the term of the world as dangerous, seems to organize the other elements in the sense that
''idea-element"-partly, no doubt, to avoid this confusion. More simply, one could these latter elements "follow" from the beliefs or are constrained by them (in a sense
use the term idea. In order to embrace the thinker and the thought, it is necessary to we shall discuss later). But belief and attitude are so intertwined that it is often
conceive of these terms in a special way: they must be considered to be emergent from difficult to assume priority., With equal plausibility one might say that early experi-
the thinker's experience; they must mediate that experience, express it, and prepare ence creates a hostility toward others (attitude) and thus leads, through various
for additional experiences. In mass belief systems, or in culture itself, the units of projective styles of thought, to a belief that others are hostile, that suspicion is
analysis must likewise be considered to be linked by thought to what people have justified, and therefore that self-reliance is required. Or one who values property-
experienced and are trying to do. Harvey, Hunt, and Schroder (1961) employ the and moralizes it, as the Calvinists and others have done-might rationalize expro-
term concept for this purpose: "A concept is a system of ordering that serves as the priative behavior with a perception of the world not unlike that of the Tepoztecans.
mediating linkage between the input side (stimuli) and the output side (response)" Only a thorough knowledge of the society that produced and employs these beliefs
(p. 1). These authors develop their theory of Conceptual Systems and Personality will inform the analyst of hierarchical chains among such fundamental ideas. Just
Organization in terms of "the nature and development of these subject-object ties as the belief systems of individuals are best explicated in the context of their person-
and .. . facets and effects of variations in the kind of conceptual linkages between alities, so cultural systems are accountable in terms of their host societies (Kardiner,
the individual and his world" (p. 1). There is no difficulty in also employing this 1945).
mediation concept for mass belief systems. . We have said also that our paradigm emphasizes regnant centers of control
There are some advantages in making the differentiated elements of the hnd coordination. In certain levels of living systems, such as cells and organs, these
system do this work of mediation. All terms and their related public concepts have regnant centers may be difficult to locate; in nonliving systems, like markets or eco-
private meanings. In Eastport (Lane, 1962), the private meaning of "communist" logical areas, it might be said that the control is in "the program." But in individuals
for a packing house checker was "people who seek to bind me"; a young female and societies, there are regnant centers, although their location is not always easy.
clerk, reported in The Authoritarian Personality (Adorno and others, 1950) thought For the individual human organism, certain terms suggest the properties of
this' regnant center. The term that has sometimes been used to suggest the agent of
94 Robert E. Lane Patterns of Political Belief

political thinking, the mind, may be too heavily loaded with cognition. Murray and A systems theory of mass opinion may consider, for any society, that the prime
Kluckhohn (1949, p. 9), use the term personality, which, they say, "is the organiza- mover is something called "technologyy' or cultural diffusion or, perhaps, economic
tion of all the integrative (regnant) processes in the brain." In the Freudian system, relationships; but in specifying the mechanisms whereby this "force" is translated into
ego has the power of conscious control and direction, but ego may not be regnant over new values and new beliefs, the persons or the persons-in-roles who are responsible
the impulse life of the individual or the monitoring minatory conscience. This line of will be accounted for. Thus, two of the current important controversies dealing with
thought leads quickly into two difficult puzzles, the mind-body problem and the states of opinion-the rise of consciousness in modern society and the decline of
problem of free will, both of them dealing with the regnancy of the mind, the self, ideological thinking (Bell, 1960; Lane, 1966)-would be improved by discussions
the ego. More fruitfully, it raises other questions that are important because they of whose consciousness has risen and whose ideological thinking has declined-and
affect the concept of agency in a functioning belief system. why certain people and not others have been affected.
The questions concern consciousness and control. Murray and Kluckhohn say In addition to a specific organization (with differentiated elements, patterned
that "regnant processes . . . have the property of consciousness"; but that explains relationships, and regnant centers), our systems paradigm has specific external and
little. Others use the metaphor of one computer monitoring another (Deutsch, 1963), internal functions. The external functions and production processes involve goals,
but that does not explain selective consciousness--or unconsciousness. Freud gives resources, process inputs, and informational exchange (feedback) with the environ-
some explanations in the concepts of repression and sublimation-unconsciousness ment.
for a purpose. The point is that the conscious mind may on occasion, like the queen The common analysis of belief systems in terms of the functions they serve
of England, reign but not rule over political thinking. The point is further made by
1 in the life of the individual treats these external functions primarily as those of reality
the concept of ego strength, representing the notion of a conscious mind with a guidance and social adjustment (Smith, Bruner, and White, 1956). In light of the
somewhat larger domain in the personality, and a somewhat greater sense of mastery individual's goals, purposes, or needs, how do his opinions help him to satisfy his
over the environment. needs and achieve his purposes and goals? As we shall point out in a later discussion,
Fromm (1941) argues that a t certain historical moments men find the burden we can judge the "rationality" of an opinion only after we discover what the indi-
of free choice too much for them and they Escape from Freedom; it has been said ldual is trying to do. It may seem paradoxical, but opinions that do not follow the
that historically the "impulsive Russiansyylook for outside controls to help them with ules of logic can be "rationalJyin this sense. And, in a similar sense, ignorance may
their problems of self-control (Mead, 1951) ; Riesman (1950), among many others, le more rational than knowledge.
has pointed to the abdication of the self in favor of control by group opinion. In my The thinking that men do is here considered as the processing of information
1
investigation (1962) of the political ideas of the common men of Eastport, an and other stimuli to produce outputs in the form of belief patterns. The resources
American urban community, I found that those with the weakest egos (the least are not only the information a man has at his command but his cognitive capacities;
I
control over impulsive drinking, eating, spending, and sexual expression) were most his capacity for moral reasoning, including empathic powers; and his capacity to
frightened by the idea of an extension of freedom. Further, they were the most likely eschew short-run gain at the expense of long-run disadvantages and costs. While
to project a system of conspiratorial or "cabalistic" control on the world-largely, I this may be obvious enough, consider the utility of the "feedback" item in the para-
thought, because they themselves had no experience of internal control and hence no digm. Illustratively, we may take the case of the common men of Eastport (Lane,
understanding of pluralistic democratic controlling mechanisms (pp. 54-56, 124-26) . 1962), framing their opinions about. demdcracy. O'Hara, a factory mechanic, in
Knutson (1972a, pp. 71-80) reports other findings with similar import. Thus, it answer to a question of what he thinks are the major problems facing America today,
seems to me, without the concept of agency and control, an analysis of belief systems urns to foreign policy and indicates that the (then recent) Russian advances in tech-
or patterns is incomplete; with it, a whole new set of analytical problems, of which lology and power threaten the United States, especially at a time when "everything
the above are only illustrative, emerges. tarted getting tight in this country." He blames the Republicans for the recession and
Societies, like organisms, are "living systems" (Miller, 1972) ;they have system 1 ,herefore, in some measure, for the relative disadvantage of this country vis 21 vis the
in part because they too have regnant centers. To explain change in the Russians. Under these circumstances, what kind of feedback will O'Hara get to
opinions of mass publics, or in cultural values, or in policy positions, analysts employ correct or confirm such an opinion? His opinions on matters at the shop-and he has
a variety of theories of social change, but they cannot leave out reference to in- many such opinions-are subject to some kind of correction from the observation of
fluential people in specified social positions. Thus, the "executive committee of the the effects of different policies, effects on things that matter to him. But on foreign
ruling class" or "the intellectuals" are assigned these deciding or influencing capacities policy, or even on tax policy (his second major problem facing America today), the
in the sociology of knowledge (of .whatever political persuasion). In the systems correctives from any kind of feedback are minimal. In discussions of the nature of
theories of Parsons ( 1951) , Parsons and Shils ( 1951) ,and Easton ( 1965), the officers 1 pinion in mass publics, the limitations of feedback (designation of a policy, ob-
of political institutions serve as deciders under the influence of popular supports and erved effects,and corrected policy) must be taken into account.
demands. Controversies over the nature of community power-pyramidal or plural- 1 The external functions served by belief systems for organizations or societies
reflect important schools of thought on the place and levels where values are imple- usually embrace the following things. The belief system or ideology explains to its
mented, policies formed, and agendas set forth. members why they are where they are and do what they do; it provides common
96 Robert E. Lane Patterns of Political Belief

goals for striving, reasons for collective success, and rationales for defeats or failures. morale and sense of well-being, but beliefs dealing with group loyalty or patriotism
I t gives common codes for coordinate effort and common justifications for the dis- do double duty in these respects.
tribution of rewards and punishments. It enlists morality in the service of necessity As with functional arguments dealing with external functions, the specifica-
or at least of practice. In these ways it makes possible the use of resources to process tion of gainers and losers, of needs gratified or group purposes served, gives clues to
matter/energy from inputs to outputs, and to process information into decisions and why beliefs are adopted, eroded, changed.
communications. Our belief-system paradigm also includes location and limits: boundaries
The functional arguments have difficulties, notably their implicit teleological distinguishing system and environment (marked, again, by effective interchange)
themes, but the questions they seek to answer are important. Among others, these and time-space-cultural settings.
, The individual's identity includes a social identity-a sense of group place-
are: Of what use to the group are these beliefs? What rewards sustain these beliefs
in the face of challenge? Who benefits and who loses through the adoption of these ment, of allies, of "people like me," in contrast to strangers and enemies. Here, then,
beliefs? What are the principal contending alternative beliefs that might serve similar we include group reference, which seems to guide opinion formation as much as
purposes? How would they fit into other features of the belief system? Who would anything does. It is this group reference that saves the individual from egoism;
gain and who would lose what through change in the specified directions? beyond this reference lies a broader human empathy; without group reference an
From such a functionalist perspective, both ignorance and knowledge, both individual is literally sick (Lane and Sears, 1964, pp. 90-91). Thus, for the indi-
orthodoxy and heterodoxy, derive meaning; that is, from analysis of their effects, vidual as for the society, the belief system must help him in this placing function,
1
inferences can be made about their "purposes" and functions and therefore about defining roles and role behavior, status and status expectations, norms that are
the forces that sustain them. Again the feedback problem is important, since orga- appropriate to the individual in these times, in those places. For most people, uni-
nizations that do not receive information about their environments misjudge the versality is a disadvantage, relativism, when a person needs guidance, a failure of
reception of their efforts and risk decline. directives.
Internal functions include integrative (not task) mechanisms, combined For the organization and the society, for building a movement as'for building
with selective admission, perception, and contact with the environment, for system- a nation, the definition of boundaries is crucial; organizations and societies need
rewarding pattern maintenance and adaptation to environmental change. both a "we" and a "they." Languages, customs, rituals, codes, histories, and myths
For individuals, the internal functions of belief systems are various need help provide demarcations and, equally important, rationales for these boundary
gratifications, sometimes called ccexternalization'~ (Smith, Bruner, and White, 1956) ; lines. Two codsiderations are most important here: the permeability between bound-
that is, the treatment of the outside world in terms of the requirements of internal aries (for example, social class versus caste, or "the circulation of elites") and
conflicts, self-justifications, and so forth. On the basis of a set of student ideological interchangeability of elements within boundaries (role specificity, division of labor).
self-analyses, I (1969) tried to relate these students' beliefs to their need to be liked Inevitably the belief systems must provide guidance in these matters.
and to validate their likability, their need to make themselves seem important or To analyze the dynamics of belief systems is to deal with the agencies of
strong (in their own eyes), their need to express and to control their own aggressive change within and between boundaries, to identify the persons or roles where deci-
feelings, their need for a sense of their own morality, and their need to be autono- ions on boundaries are made and moralized, and to observe the traffic across bound-
mous from their own families or, alternatively, to identify with and carry on the ~ries,especially as this may be captured by the concept of "cultural diffusion."
family's political and social traditions. (Of course, their beliefs also had reality func- Finally, our paradigm provides for lawful, routinized behavior-in the form
tions, especially to provide them with views that would help rather than hinder their uf rules, laws, habits, and conventions to guide element interaction, goal striving,
careers, but also to express positions that would, if adopted, advance their own indi- product processing, and pattern maintenance and adaptation.
vidual or group interests.) To these ends, from their own course work or, more What is lawful for the individual is that which does not violate his conscience
important, from their reading and discussions, they heard and remembered and or, more usually, his routinized habits and expectations. Although for every act there
used those things that reinforced their beliefs, but with some tolerance for challeng- is an implicit or explicit belief to justify the act, the reverse is not true; not every
ing, dissonant ideas. From the various theories that identify self with beliefs and thought, fantasy, image, or argument is reflected in behavior, especially since
personality with ideology, the uses of beliefs in maintaining identity may be inferred. thoughts often rehearse alternative lines of behavior. The world of behavior, there-
The model alerts us to this with its reference to pattern maintenance, a useful way fore, is smaller than the world of thought; the two worlds are not isomorphic. Fur-
of thinking of identity. ther, as one moves along the concrete-abstract continuum, both for individuals and
For organizations and societies, the internal functions served by belief systems for societies, the lack of correspondence between thought and action increases.
are those of integration, coordination, morale building, leadership legitimization, Studying behavior patterns, one infers thinking and reasoning; studying
defining equity and justice within the system, and providing formulas for conflict thought patterns, one may infer behavioral consequences. Studying individuals,
resolution. With reference to the question of agency, in managing belief systems as especially at the level of personalitfr (conflict resolution, learning strategies, uncon-
in other matters there is often differentiation between beliefs associated with the siousness), one can synthesize behavior patterns and thought patterns-a synthesis
tasks or the external functions and those concerned with the internal functions of made possible by the introduction of the powers of agency. Lawful, routinized be-
98 Robert E. Lane Patterns of Political Belief 99

havior (codes, habits, and mechanisms) is the property of living systems at this level those of the analyst. I t is a nice question whether the concept of constraint as it is
of the organism or person. employed by Converse applies to the subject of analysis or to the analyst. With
At the level of the organization or the society, the lawful regularities of the reference to the analyst he refers to "the success we would have in predicting [what
system that bear on its political beliefs are institutional practices; communication idea-elements go together]." With reference to the subject of analysis Converse refers
conventions; behavioral patterns, encapsulated in mores; and the cultural codes, to the "probability that a change in the perceived [by the subject] status . . . of one
often based on premises serving as answers to the questions "What is real?'(meta- idea-element would psychologically require [a change in the status of other idea-
physics), "What is true?" (epistemology), and "What is good?" (ethics). The inter- elementsy (p. 207). This concept of constraint on the capacity of the analyst to
section between systems of social behavior and social beliefs comes at the level of predict takes Bridgman (1928) literally: A concept is defined by the operations that
culture, but culture without society lacks agency; it is just there to be described measure the phenomena to be examined. But there is a looseness here that needs to
(Benedict, 1934; Kroeber, 1944). Change acts through the agency of living men; be questioned. The phenomena to be examined are not merely the patterns of asso-
thus; the dynamics of culture (value change, techn~logicalchange) implies the inter- ciation of idea-elements and patterns of change; for, by definition, they include a
action of society with culture. Here, then, lie the great questions: What kinds of social subject's sense of necessary relationship ("psychological requirement") and thus a
changes produce major shifts in belief systems? How does the change in property subject's experience of constraint, the sense that this idea should go with that one.
relationships alter moral values? What is the effect of the great increase in technology This confusion between reference to the subject and reference to the analyst
in modern society upon the epistemologies and values of mass publics? would not be important if it did not have implications for the research findings that
flow from the use of the concept. The most important of these implications is the
Reasoning and Constraint priority this confusion gives to the analyst's role in setting forth the idea-elements he
thinks are important, developing the conceptual framework that the analyst regards
In this section I will deal in more detail with the idea of constraint, briefly as most likely to "govern" the more specific beliefs (exemplars), and thus providing
touched upon in the preceding discussion. I shall rely upon, and criticize, the irn- a guided opportunity for measuring association and change. Equally important, this
portant discussion by Philip Converse (1964) in "The Nature of Belief Systems in focus on the analyst gives him, but not the subject, an opportunity for talking about
Mass Publics" and shall contrast the concept of constraint with the concept of the patterns of idea-element association. This is a substitution of correlation for cause
reasoning. Converse defines a belief system as a configuration of ideas and attitudes and, as the analysis progresses, of cause for reason. Put another way, if one thinks of
in which the elements are bound together by some form of constraint or functional the term constraint in the less specialized sense of "to confiney'or "to bring into nar-
interdependence. The constraint he has in mind is measured in the static case by rower compass," the term should apply to the subject's experienced choice among
predictability about other idea-elements, given any one or group of them, and in the idea-elements. He can explain why, for example, he feels constrained to say "yes" tb
dynamic case by the probability that a change in the status of any one idea-element the question :'Are you in favor of leaving things like electric power and housing for
would psychologically require some compensating changes in the status of others private businessmen to handle?" His answer will be only part of such an explanation,
(pp. 207-208). One can quite easily measure constraints of this kind, given survey but it is an important part.
data and panel data; but, as Converse points out, the inquiry into why certain idea- The point can be made more specific. Among the working-class men I inter-
elements go together is more difficult to conduct and offers less satisfactory answers. viewed in 1957 (Lane, 1962), O'Hara, a factory maintenance mechanic, like every
Through the analysis of survey data, Converse shows that little use is made of other person in the group of fifteen respondents, chose the statement "It would be
broad economizing concepts such as liberal-conservative in the mass public; that better if mines and factories remained under private ownership" over the statement
political and policy information is rare among the 88.5 percent of the population who "It would be better if the government owned the mines and factories." Nevertheless,
do not offer evidence of a liberal-conservative ideology; that such information is on the four discriminating items in Centers' (1949) liberalism-conservatismscale, he
important in providing constraints of the kind mentioned; and that opinion change chose the more liberal answers: he generally sides with workers in labor disputes; he
on a policy item in which the public has little interest is nearly random. Looking at believes that management sometimes takes advantage of workers; he wants the
"what goes with what," Converse shows that congressmen have much more pre- government to guarantee jobs rather than opportunities; and he believes that workers
dictable belief systems than the public. And the belief systems make more sense should have more power. From his perspective, what are the constraints that led him
according to the liberal-conservative organizing principles. For the masses, belief to oppose the nationalization of the factories and mines? Some of them emerge from
systems are more likely to be organized according to attitudes toward social groups. his discussion of big business. As a worker in a huge factory, he feels his own liberty
Any inquiry such as Converse's (that is, an inquiry about the relationship more threatened by union decisions than by big business; he believes that big business
between various political ideas) is in my view, at least partly an inquiry into political should be credited for investing money into research; he fears monopoly (apparently
reasoning and therefore requires an analysis of the ways in which people think about including government monopoly) ; and he thinks an excess-profits tax will take care
In order to do this analysis, one must examine the thinking and reasoning of the tendency of big business to cut back jobs and to grow even bigger through
processes as well as the patterns of association which these produce. mergers. These beliefs, individual to O'Hara, constrain his views on the nationaliza-
But the processes we have in mind are those of the subjects of the study, not tion of industry; that is, they are the ideas that go with his opposition to nationali-
 Robert E. Lane Patterns of Political Belief

zation, that "prevent him" from supporting that policy, that would have to change if by no means settle the case, for several reasons-among them the dubious value of
he were to change his views. In a wider sense, there are, of course, cultural forces, economy, the fragile cohesion of ideology, and the weakness of logical constraints in
group pressures, information deficiencies, media discussions, authoritative statements ideological thinking.
by respected leaders, and socialization limits that channel his thinking. He is not Economy of thought is a value of uncertain priority; conformity and stereo-
aware of most of them; but as we examine "what goes with what" and why, his typy also economize thought but are too costly of other values. Moreover, judgmental
reasoning is important. Even though he does not refer to a liberal-conservative con- dimensions or yardsticks are often made up of disparate elements that cluster in
tinuum or use it as a ''judgmental yardstick," and even though he is conservative on people's minds. These elements may have no connection except that they are derived
"nationalization of industry" but is one of the two most liberal men in the group of - )m a common intellectual source. An obsolescent tradition may be all that cements
fifteen on the liberalism-conservatism scale, he has not thereby demonstrated a lack ideology together.
of capacity to organize ideas. If, within a year or two, O'Hara were to find himself Finally, it is not enough to say that elites "experience" ideologies as logically
unemployed, to blame this on a merger that he regards as monopolistic, and to change constrained clusters, if these bodies of thought are not, in fact, closely constrained
his views about government ownership (perhaps at the same time that someone else by the rules of logic-which they are not. The next set of elites may have a quite
changes his views in the opposite direction), it would make little sense to call different experience of logical constraint (the conservative of one era may believe
O'Hara's changes "random3-that is, the product of many small forces of opposing that social security is incompatible with the premises of capitalism, while the con-
tendencies in society,even though the change process were to reveal to the analyst servative of the next era will find social security and capitalism quite compatible).
a pattern resembling a random distribution. O'Hara would be then, as he is now, 1 While ideology is flexible over time, logical relationships should not be.
basing his judgments on what he has known and experienced. There are political More specifically, one may agree with Converse (1964, p. 220) that conserva-
philosophers, like Edmund Burke ( 1790), who would applaud this tendency. tism consists of ( 1) caution in responding to new problems and reluctance to change,
Political reasoning is a term we may employ to refer to the ways in which (2) resistance to welfare state programs and support for "free enterprise," (3)
people justify or argue about their political beliefs. I t bears the same relationship to resistance to the expanding power of government, and (4) support for individualism
belief system that moral reasoning bears to morality and is discovered in the same and individualized solutions to social problems. This seems reasonable; it is com-
way: 'by finding out the grounds on which people base their beliefs. I n the work of Pa'tible with the definitions of others (Huntington, 1957). But is there anything
Piaget (1965) and Kohlberg (1964), the ethical decision in moral dilemmas is not illc)gical or unreasonable in the following "nonideological" combinations: slow and
so important as the ways in 'which decisions are made. Similarly, in logic, the im- cai~tiouschange in the direction of the welfare state; governmental intervention in
I soc:iety to give the individual more freedom from corporate control; accelerated
portant thing is the process, not the conclusion (which may be true but invalid) ;
fallacies are processual. Political reasoning is different from opinion formation, which chiange to restore capitalism to its former autonomy; enhanced use of governmental
usually refers 'to external or unconscious forces that affect an individual's reasoning regulatory powers to make capitalism work.
process. Reasoning is not eliminated by stimulus-response sets or by conformity to Burke and Adam Smith do not necessarily go together. violating conservative
group pressure; these simply add new, if usually inarticulate, premises, such as "It is doctrine, a conservative president proposed a program of governmentally financed
better to agree than disagree" or "I have more to gain by making friends than by income maintenance; he also instituted price and wage controls. Experience makes
defihing new principles." These premises, it is true, impede the unraveling of the clear what logical analysis could explicate: the reconstruction of arguments so that
reasoning by an observer, but an unconscious premise need not make the reasoning a chain of plausible (true?) propositions in proper form, with a set of distributed
illogical or ineffective or the object of forces properly described as "random." I t is 'ddle terms fulfilling the requirements of a logically valid chain argument, might
just as well that it should be so, since we ull have unconscious premises and for most formed-with outcomes (inferences) that would embrace at least three of the
of us the most enduring political (or other values) are r6oted in unconscious premises. ir policies mentioned above. Yet they would be unlikely to be included together
l
Much reasoning which seems illogical is seen to be logical as soon as the unconscious a formal statement of ideology. As experience indicates, one can be quite logical
premise is stated. In some ways, therefore, the injunction of logicians "Make certain violating even conservative 'or liberal ideologies.
that all unexpressed premises have been included" (Black, 1946) is quite congruent If we substitute values for ideology, referring then to equality, community,
with the rules and practices of psychotherapy. Y ""lty, order, and matters of that nature, do we discover a better set of constraints on
The view that there is something deficient about any political thinking that inclividual policy choices? Operationally, if we know a person's values, can we predict
does not employ the efficiencies of a "highly constrained system of multiple elements" his policy preferences? There are logical problems, psychological problems, and
(Converse, 1964, p. 214) deserves examination. I t may be true that "the idea orga- :ial problems in such prediction. One logical problem deals with the problem of
nization that leads to constraint permits [the individual] to locate and make sense of Tree. Thus, in the argument
a wider range of information from a particular domain than he would find possible
without such organization." I t may also be true that the liberal-conservative con- The welfare of mankind is served by equality.
tinuum bffers'lcdnomies of thought (and time and attention) and helps a person to
Equality is promoted by a confiscatory estate tax.
locate a policy and its implications in a set of ideas and values. But these advantages
 Robert E. Lane Patterns of Political Belief

Therefore, the welfare of mankind is promoted by a confiscatory income very largest ones-such as Social Darwinist theories of the survival of the fittest,
tax. Marxist theories on the necessary stages of history, and racial theories based mainly
on genetics-all seem to dictate certain policy choices. Yet spokesmen for the National
the conclusion does not follow from the premises because it is not clear how much Association of Manufacturers, who adopt Darwinist views on domestic policies, also
equality is included in the major premise. This kind of looseness is called the "fallacy argue for protectionism in international trade; Chinese "Marxists" argue for abolish-
of accident," and it is bound to plague any arguments involving values of unspecified ing intermediate stages and for moving from peasant (Asiatic) 'forms directly to
degree. i communist forms; and racial theorists, upon finding Negro aptitude scores lower
A second logical problem, related to the problem of degree, stems from the than Caucasian scores, may argue for differential education for the two groups but
plurality of values. Thus, it is plausible for a man who adopts the major premise 1 do not adopt their own "logic" upon finding that Japanese mathematical aptitudes
above to argue as follows: exceed Americans'. Theories at this level guide policy choices, but one cannot say
that they determine them; the constraint, therefore, is weakened by intruding con-
The welfare of mankind is served by high individual achievement. siderations. Nor could one say that these intruding considerations weakened the
High individual achievement is promoted by monetary rewards for rationality (as distinct from logic) of the policy preferences. On the contrary-at
achievement. this level of theory, at least-societies are better served by employing them as
heuristics rather than algorithms.
Monetary rewards for achievement are effective only if they accrue to the 1 The mistake underlying reliance on the constraints implied by statistical
achiever's family.
clustering, scalar ordering, or acceptance of an idea cluster by an authoritative ,elite
Accrual to the family is denied by a confiscatory estate tax. based on the fallacious view that if some people see idea-elements properly cluster-
Therefore, the welfare of mankind is not served by a confiscatory estate ,g in a certain way, others should too. Such "constraints" or clustering refer to
tax. :ither logic nor rationality. Here it is important to repeat two points made earlier:
I ) The uses of logic to constrain political thought may be weak (a) because mass
Reasonable men do value both equality and the achievement motive; some also lblics do not understand or do not use the rules of logic and (b) because ideologies
believe in the relationship of inheritable rewards as a stimulus to achievement. The ,e rarely stated so as to permit close logical constraint. But in revealing the weak-
apparent contradiction in the policy outcomes flowing from belief in these values is :sses of an argument, the employment of logical (critical) reasoning is crucial-
not likely to be resolved by abandoning one value for another. Instead, one might I pecially for the critic of mass belief systems.l (2) Rationality is impossible to assess
address the problem of degree ("great inequalities," not "modest inequalities," vio- itil the critic knows the ends pursued by the subject; only then may he judge
I
late one's values) or adopt a belief that permits escape from the contradiction hether means-ends rationality has been employed.
(inherited wealth is likely to be dissipated and hence does not lead to permanent Connected to the concept of constraints is the tendency toward contextualiz-
inequalities; confiscatory estate taxes do not prevent gifts during the donor's life- g. Most members of mass publics tend to fragment their views and to respond to
time). Values do guide policy conclusions; but as long as the degrees are not specified Issues or policies without regard to many other relevant considerations. Elsewhere I
and the introduction of new causal connections is permited, they do not, through the have. argued that "one of the features of what is sometimes called 'understanding' is
instrumentalities of logic, force a choice between values. (The usual escape of the to grasp the context of an event; that is, temporally to know what went before and
equalitarians in such situations is to redefine equality to mean "equality of oppor- -,hat is likely to follow, spatially to know the terrain, in human terms to see the
tunity," but this was ruled out by our example. Nevertheless, redefinition of values ay of the many motives involved. To understand an event in this way is to con-
to avoid self-contradiction is a standard escape from illogicality.) xtualize it; not to do this is to rnorselize it, to see it isolated from the surrounding
The psychological grounds for evading logical constraints are reflected in the --atures that give it additional 'meaning"' (Lane, 1962, p. 350). In examining this
above illustration. Just as the person who is bound entirely by his ideology, so that problem, I looked at three men representing three points between the morselized and
experience does not have the power to change its elements, is an ideologue and unfree, the contextualized extremes of my sample. The issue was the discussion of Soviet-
so the person who governs all his policy recommendations by a single value is close to American relations-something rather distant from group interests that help many
"obsession1' or borders on "fixation" and is similarly unfree. The healthly person has 1 Even the classic case of the violation of logical constraint turns out not to have that
multiple values, and he finds them often in conflict; his health is revealed in his property. McGuire (1960) performed some experiments, cited by Converse, in which persons
toleration of the conflict and the means he chooses to reconcile the conflict, not in who believe simultaneousIy in ( 1 ) expansion of welfare services, (2) cutting taxes, and ( 3 )
the way he makes all policy recommendations serve a single value, however econom- I lancing the budget are deemed illogical. But the logic is not constraining for two reasons:
ical that might be for him. ch such person can argue for cutting sodething he does not care about and expanding
slfare services and cutting taxes. Or, alternatively, he can argue, as did O'Hara, for a non-
Since ideologies necessarily imply large causal theories, one may isolate causal I
K device to increase revenue-namely, a lottery. The constraint is in the middle-class
theories for a brief examination. In what sense do they constrain idea-elements? The eology that does not think of lotteries as alternative government revenue devices. ,
 Robert E. Lane Patterns of Political Belief

people to locate themselves on an issue, but also not clearly guided by liberal-con- years in the Peace Corps expresses graphically what many others felt,: ''Trying to
servative organizing ideas, at least not any more. write this paper has proven to be a real bitch. I found that I don't really have any
Perhaps it might be said that another standard organizing "ideology" or definite opinions, any logically constructed philosophy of life, not one of the way
quasi-ideology, internationalism-isolationism, might be helpful here. But, I believe, that society should be constructed or administered" (Lane, 1972, p. 127). One
such a judgmental yardstick was as foreign to these men as was conservatism- wonders how many "ideologues" captured in a survey net would, given the chance,
liberalism. Nevertheless, there were distinct differences in the way they thought about have responded in the some way. And if that is true, it is problematic whether a
Soviet-American relations-differences representing clear stages of development.
These differences might be indicated by listing what the best contextualists do, in
i judgmental conservative-liberal yardstick would prove to be an efficient economizing
I device for policy evaluation, however well it might predict %hat goes with what" in
contrast to the morselizers. The contextualists include a time dimension (Russia I a set of questions. One would, I think, prefer to see the political reasoning that
"then," "now," and possible future changes) ; relate domestic and foreign policy; employs these and other economizing central ideas, even if they are relatively un-
include various aspects of foreign policy, especially trade and military policy; look at I
conscious, to arrive at policy decisions.
'the issue from the Russian side, especially possible Russian responses to American
initiatives; refer to a theory or philosophy, such as the "philosophy of disarmament";
1 One alternative to the use of an ideology, like liberalism or conservatism, to
guide a person in adopting an opinion is the use of reference to a group, like work-
consider the effects of Soviet-American diploma~yon other nations (Lane, 1962, ing class or Italian-Americans, for such guidance. A person using this group reference
pp. 350-35 1) . may believe that what is good for the group is good for him and substitute this
Here the common man has gone beyond easy reference to his own experience, reference for any further thinking about the matter. A generation of political scien-
does not make any link to group interests, is not limited (by the interview situation, tists has developed the view that a group-referenceor group-interest form of political
for example) to vague references to "the nature of the times," and does not clearly guidance is not only the most common form but also the.most fruitful and useful
draw on ideologies or near ideologies. Yet, perhaps because he can be observed (Herring, 1940). They believe that the play of group interests in a free marketplace
thinking and not merely responding, he cannot be said to be responding randomly of interest groups produces a political balance benign for society. With qualifications,
and must be said to be constrained by values, information, and a modest under- there is much to be said for this position. Since members of mass publics have only a
standing of other related events and issues. I t is true that most of the men did not , dim and blunt perception of any political ideology that might inform their decisions,
employ all of the contextualizing varieties of thought mentioned above, but almost the group-reference solution may, indeed, not only be easier for them, but also
no one employed none of them. If there was any one common judgmental yardstick
employed, it was the degree of force or threat of force to be employed; by analogy
it might be said that this way of looking at things did enlist personal experience
1
I
guide them better toward decisions that are in their long-term interests. For these
reasons one might differ from Converse's argument that reference to ideology makes
for better opinions than does reference to group interest,
(except that Russia is not easily personified as the man next door) ;but that particular
yardstick (attitudes toward force, violence, war) escapes the subdimensions of the Economizing
conservative-liberal yardstick the academic is tempted to employ. Something different
from reference to an ideology, then, offers understanding and policy guidance; with- The concepts that help a person organize his political ideas and that determine
out standard ideological referents, these men were contextualizing a policy problem. what ideas go properly together are not drawn primarily from some major judg-
In contrast, it might be argued, reference to an ideological posture would not only mental dimension, such as liberalism-conservatism, but from more intimate sources,
"constrain" policy thinking but would confine it. There are meanings of the term ' xer to the individual and the world he knows. Inevitably, if he has spent much
ideology that suggest defensive, postures (Rokeach, 1960) such that the main ne thinking about political-philosophical matters, he will employ the ready-made
objective of ideological policy thinking is to defend an ideological commitment, not ganizing concepts to be found therein. This practice is economical in the sense that
IL saves time and effort; it is efficient in the sense that these organizing concepts are
to explore alternative policies as they affect values.
Finally some reference to the self-consciousness of an ideology is in order. I t designed to direct attention to policy implications. But most people have not spent
much time on political-philosophical concepts; economy and efficiency for them
is true that most people can locate themselves on this dimension (only 4 percent of
a national sample had "no opinion" regarding their own position on a scale running
-r~quire
-. the employment of experience and reference to "contexts" closer to their own
liv.es. We shall examine these contexts under the heading of a "core belief system"
from "very conservative" to "very liberal"-American Institute Public Opinion poll,
below, but first let us look more closely at the term economizing.
April 20, 1972), but many who do so have confused the two meanings of liberal:
'C
Among the dictionary definitions of economy are these: "the management or
generous," in one sense; "opposed to conservative," in another. I (1969) asked
or( dering of parts, functions, etc., in an organic or organized system; organization;
some sixty Yale students to explain in an ideological self-analysis what advantages
ale ;o, a system or body so managed or ordered"; "thrifty administration"; "manage-
they receive from their liberalism or their conservatism. In the process they had to
set forth what they did believe. Typically they had great trouble in setting forth a
mt:nt . . . of an establishment . .. directly concerned with maintenance or produc-
tlveness." What, then, is economical political thinking? And what is the economy of
coherent ideology. One of them, an extremely able young man who later spent five 1 the political mind?
106 Robert E. Lane Patterns of Political Belief

One meaning or function of economizing in political thinking is bookkeeping, stance, the court now says that "right to counsel" is part of "due process," arguing
striking a balance between liked and disliked features of a political system. This has that the concept itself "logically7' includes this protection. Things hang together as
the important consequence of determining whether a person is, in sum, for or against syndromes (spots and fever, for measles), with third-factor causes. Experience puts
the regime, the administration, a certain policy. Thus, people must manage their things together, as with conditioned reflexes or any kind of conditioning. Convention
thinking so as to know where they stand on such large constructs as the following: does this too, as in the sequence of phrases "thank you" and 'you're ,welcomen or
(1) the constitutional order, the regime, the administration, and such various com- the defeated candidate's concession acknowledgment after an electoral defeat is
binations of these as may be referred to as "the establishment"; (2) the net balance apparent.
of costs and benefits flowing from the established order to the individual and the The economizing that links idea-elements to each other operates through
groups with which he identifies, compared to what he thinks he and they ought to (1) summation of good and bad and the consequent bookkeeping; (2) encouraging
have; (3) the net balance of influence the individual has, compared to what he consistency (value compatibility, means-ends compatibility, means-means compati-
thinks he ought to have. bility, theoretical compatibility) ; (3) discovering likeness in relevant properties (thus
Economizing also has a consistency function; that is, the management of the locating things in the same class covered by the same concept) and link+g concepts
parts of the belief system in such a way as to bring together elements that have not or principles to their illustration or exemplars; (4) identifying parts anq wholes-
been brought together before so as to reveal compatibilities and incompatibilities. especially, for our purposes, parts of whole processes, but also parts of systems, or-
These are of four kinds : ( 1) compatibility' of ends-for example, the problems ganisms, syndromes; (5) employing the connections provided by experience, either
presented by simultaneous advocacy of individualism (the autonomy of the indi- conditioned or reflected upon; (6) definition, as may be established by an authori-
vidual) and community (the familial model of common decision making and shared tative interpreter, like the Supreme Court or the supreme lexicographer; and (7)
destiny) ; (2) ends-means compatibility-perceiving the relationship, for example, nlisting conventional arrangements, as in codes of behavior, linguistic usages, familiar
between welfare policy and preferences for "equal pay for equal work"; (3) means- ~xtaposition,common origins. The mind, of course, must make the links to relate
means compatibility-that is, perceiving the way two policies complement or oppose ne idea-element to another; the thinking process may rest on observational acuity,
each other (so that the economizing thinker would discover the difficulties in arguing interpretation of purpose or insight into function, free association (links provided
at the same time for decentralized educational decision making and against revenue by unconscious experience), analogical reasoning, conditioning, andJor inductive and
sharing) ; and ( 4 ) theoretical compatibility-for example, arguing for the overall deductive reasoning. But given these possibIe sources of relationships, it is too much
principle of the inevitability of class warfare and for principles of mediation and to expect that the ordinary man, unused to the traditional judgmental yardsticks of
arbitration in labor disputes. Western political thought, would find these yardsticks the best or most economical
Economizing, in this sense of managing the interrelationship of the parts, way of relating idea-elements.
has a placing or concept-attainment (Bruner, Goodnow, and Austin, 1956) com- But there is another meaning of economy-namely, thrift-that is important.
ponent; that is, the capacity to place the item into a set composed of elements whose What is it that economical political thinking might be thrifty of, might save? I
shared properties are the important ones for the thinking at hand (thereby reducing believe the answer is time and effort. (We will later examine the concept of eco-
the time and effort involved in appraising each item of an interrelated set of items nomical as "efficient7'-meaning economically productive of desired consequences.)
individually). Thus, to know that TVA, Amtrak, and the Post Office are all examples I mentioned stereotypes above as economical of time and effort; so is a conditioned
of government public enterprise permits some economies of thought for those who reflex. We need some criteria to decide what is a good use of thrift and what is a
support and those who oppose government ownership of the means of production. bad use. The criteria must rest first on alternative uses of one's time and effort; in
(For other purposes in the same general domain such concepts as bureaucracy, public this connection, some ideas from a recently published piece on the good citizen
utilities, and service industries are more useful.) The same economizing through (Lane, 1972, pp. 299-31 7) may help to illuminate the problem.
assigning things (principles, events, policies) to concepts-that is, concept attainment For the concert pianist, who has no background in world events, little interest
-applies to such larger matters as strategic armament limitation treaties and the in public affairs, and a great and consuming love of music (and a social contribu-
concept of the balance of power; reliance on unequal property taxes for educational tion to make there), the cost of developing a political philosophy and of relating this
financing and the concept of equality of opportunity; and licensing requirements for to a policy preference is high. For the Washington lawyer, with much interest, daily
"subversives" and the concept of freedom of speech and of association. The econo-. informational inputs, a capacity to employ experience and previously acquired skills
mizing of course, takes place by relating principles to exemplars or genus to species. in relating his idea-elements to each other and to their informing philosophies
Economizing takes place by recognition of the relationship of part to whole. +
(values causes), the cost is low. Thus, the thrift value of economizing on time and
In general, the stages of a process hang together because of their relationship to a set effort is very different for each person, even if the consequences (gains or losses) of
sequence: as cross-examination of witnesses may be related to due process in an a policy-say, a tax measure on identical incomes-are in many ways identical.
adversary procedure, or as hearings may be related to the legislative process in The calculations of cost and gain of political thinking are complex, as may be
American democracy. O r the part-whole relationship may be definitional; for in- :en by the following formula (Lane, 1972, p. 311) :
108 Robert E. Lane Patterns of political Belief 109
f
Cost of modifying the critic of his own judgment regarding the rightness or wrongness of an individual's
one's own situation Cost of any indi- i policy preferences from another point of view, especially from the critic's own view
Cost of following according to an cated effort to I of the public interest. I t is not a totally relativistic standard; that is, it does not
to and understanding estimate of the alter the course
gain in following - the news + impact of an + of history imply that everything is as good as everything else, according to whatever an
the news On a - event on the self individual or a culture may say is preferable.
chain of events In addition to the bookkeeping and consistency functions, then, economizing
Probable gain or loss Probability that the event will in political thinking takes into account the following calculations: ( 1) the costs of
implied by an event x mature so that the indicated gain

i
the time and effort in framing a policy and in participating in other ways; (2) the
in the news or loss will take place probability of making any difference in the world through this expenditure of time
and effort; (3) the opportunity costs of the time and effort expended; and (4) the
But even this set of calculations is more complex than at first appears. To every probability of adopting a correct policy, where "correctness'' depends upon maxi-
probability figure in the formula, one must attach a degree of certainty; that is, a mizing one's own enduring values and need gratifications.
weight indicating the confidence one attaches to the figure. Further, one must think These references to "making a difference in the world" and to "maximizing
in two different ways (as in statistical calculations) about the relative preferred
values" suggest that a political idea is a production good; it is instrumental in achiev-
risks in being wrong: Is it better to take action when one might be wrong, or not ing other values. This is true in two senses. The first has to do with the societal
to take action when one might be right? Further, the costs of gaining the requisite product. From this point of view, the "correctness" of the idea does, indeed, depend
information and calculating probabilities, certainties, and risks are greater than those upon information and social theories regarding ends-means relationships. In the
involved in isolated reflection, for it is important to talk about these events and
second sense, the idea-element is productive of individual benefits in a person's social
policies so as to rehearse one's ideas, receive criticism, anchor them in the judgments milieu: career advancement, popularity and friendship, family harmony or sense of
of others (Lane, 1962, p. 102). Thus, the time-and-effort cost figures run high.
independence from family. For these products, the correctness of the idea is to be
But costs are to be evduated as "high" or "low" not only according to the
judged not by what the world would be like if the policy were implemented but
labor costs that may be calculated but also according to the benefits that flow from
rather by the effects of the expression of ideas on other people. Like the calculations
the "correct'' decision. By what standard, then, might one judge a "correct" policy
expressed in the formula above, the calculations here are complex: Group A is made
preference for a certain m-an? Marxists are willing to say that a working-class man
happy to degree X, but Group B is made unhappy to degree Y; and these scores
who is not a socialist has misjudged his own interests, that he has a "false conscious-
are multiplied in each case by the estimate of the importance of each group's degree
ness"; but'history has not shown that even economically the working class is better
of happiness to the holder of the opinion and summed in the bookkeeping operation
off under socialist or communist .regimes. In any event, the observer's judgment of
described earlier.
the "right" and "wrong" policy .preferences for a given person at a given time, with But, as everyone knows, opinion elements are consumption goods as well as
his plural values and multiple needs, rests upon shaky philctsophical foundations and,
production goods; the mere expression of an opinion without further consequences
I believe, flounders on the difficultiesof interpersonal comparison that have given the
ha;s values: it may reveal to the spokesman his mastery of a subject, his own im-
economists such difficulties in comparing utilities of different persons.
The psychoanalysts have a better posture in determining what opinions and POrtance, his high moral posture. Opinions-express anger, determination, depth of
de!spair. The very expression of opinions is literally self-serving; it may provide the
behavior are "right" for a specific individual. They, seek to find out the complex
sarne gratification that a writer finds in writing for his secret file of unpublished works.
needs and values that give enduring gratification to a patient, and then to guide
Bu t again, the calculations regarding the benefits gained are not simple; for the
him to give up copihg patterns that provide short-term "secondary" or "neurotic"
incjividual is both agent and critic, as the terms guilt and sober second thought imply.
gain at the expense of long-term life satisfactionand to adopt behavioral and thought lle may not be pleased with- what he has said or even what he has thought. Con-
U.
patterns that are expressive of hk idiosyncratic enduring values. In the same way;
sequently, it is a net benefit or gratification that must serve as the criterion. In sum,
one could say that a citizen adopts a correct policy preference when, after that degree therefore, taking the individualistic point of view outlined above, we must estimate
of reflection indicated by the above formula, he expresses his long-term interests and
the correctness of an opinion according to both the societal and the social-production
values, whatever these may be. functions and according to the consumption function of an opinion. It is against
This standard does several things. First, it avoids substituting the observer's these net benefits, weighted and summed, that the costs of an opinion must be assessed.
values for the subject values, the mistake made by the Marxists. Thus, it permits
Only then, it seems to me, does the term rational make sense.
an individual to prefer security over economic gain, popularity or respect of peers
over class interest, freedom of choice over security, without the stigma of "irra-
tionality," which such choices often provoke. Second, although this standard makes
I Core Belief System
If the judgmental dimensions of Western culture are rather poor organizing
use of h e ~ndividual'sown enlightened self-interest as the criterion of right and.wrong principles for political thinking (on grounds of their elusive constraints on particular
for his own preferences (and thus adopts a modified relativism), it does not strip policy issues and their remoteness from the ongoing lives of individuals), and if social
 110 Robert E. Lane Patterns of PoIitical Belief 111

values are only little, if any, better, where shall we turn for such principles? What their operation. The common man is anthropomorphic. The human sets that people
does constrain political thought? The clue to the answer, I think, lies in taking (including historians and social scientists) identify are these: families, cliques and
further the experiential and personalistic point of view outlined above and searching associations, races and other ethnic groups, communities of region or place, nations,
for ways in which,ordinary individuals organize their political thinking. From this occupations, rich and poor and middle strata, religions and secular ideological groups,
point of view, we need to know how a person thinks about himself in society. The interest groups, parties. Stratified, these become castes and status groups, social
'
answer to such a question' is likely to be at the level of inarticulate axioms-in- classes, specialized elites.
articulate, at least, until we ask the right questions. The first of these terms, himself, In framing an identity, a person is likely to place himself among these human
may be conceived as an identity; the second term, society, we will examine as a sets, and at least two general postures toward them are important. The first is the
universe of human sets. relationship of the individual to his set-his mobility in and out of the set and his
To interpret the interaction of these two elements, the individual.employs capacity to dissent; in short, his individualism. The second posture is the relationship
ideas that are the common man's analogs to the philosopher's several domains: among sets: the lines of status, the terms of conflict, the nature of national and
What and who are authoritative? What values and goals are to be sought and worth international community. The democrat believes in the plurality of sets, adjustment
seeking; what needs do these values and goals reflect? What is the individual's moral within them, and fluidity between them.
good; what is an appropriate moral code? What causes effect individual and social Authority. Authority is power plus legitimacy. Legitimacy was accorded to
change; how shall we explain social events? What concepts of time and place are to government by the men of Eastport (Lane, 1962) on three grounds: ( 1) the people
be employed? What is it to "know"; what is true? How shall truth be discovered? (people like me) in the end control governmental policy; (2) the government is
If we know these things about an individual or group, we can predict his or the group interested in us and is generally responsive to our needs; (3) the constitution is a
members' policy preferences better than if we rely on the concepts of Western politi- legitimate source of power, and the rules laid down therein are sanctified, workable,
cal thought. These personal-philosophic concepts are ingredients of constraint; it is and fair. Further, the means of selecting political authorities give us good leaders,
these things that cause one idea-element to be associated with another in a political kept good by their responsiveness to the people.
belief "system." But these views do not give orientation on what to believe when opinion on
Identity. An identity (Erikson, 1956), of course, is the complex answer to a particular policy is divided. Asked whom they would turn to for advice, the people
the simple question "Who am I?" The answers are, in the first place, social; that is, of Eastport were reluctant to name a person or a group or a leader; they wanted
placing the self in a social context of country, ethnicity, occupation, family. In this ;o vest that authority in themselves, whatever the reality of the case might be.
sense they link to the reliance on. group reference that most studies find important in Further, "experts" had to be challenged, for they might be interested parties. What
voting decision and in policy preference. In the second place, they are individual. s authoritative here, in the realm of political knowledge, is the process of argument-
Asked to describe themselves, the men of Eastport (Lane, 1962) mentioned such I U ~not too prolonged, for too much talk is fatiguing. Charisma, the magical leader,
qualities as "imagination," "getting along well with people," "I don't blow my stack s not eagerly sought.
at the least little thing.'' Such concepts of individual identity can be linked to per- Concepts of authority probably do give some constraint to mass belief systems
sonality theory, with its concepts of conflict resolution, need achievement, approval ~ u less
t than might be expected. Knowing whom an individual trusts might help to
motive, aggression, dependency, and so forth. ~redictwhat idea-elements go together; but in their own terms the common men are
If the concept of identity is closer to the person than, say, "liberalism-con- ;elective shoppers, "buying" this idea here and that one there.
servatism," it seems more remote from policy preferences. But the links are there. A Values, Goals, Needs. Values are discussed in the language of philosophy and
person's identity tells him something about his self-interest and, because it includes anthropology, needs in the language of psychology; but they come together, in some
g o u p reference, his group interests as well. Further, it tells him something about his measure, as source and destination, motive and expression, cause and reason. They
proper role in political affairs, what he can and ought to do, what is appropriate for ?re not reciprocal, for although every need seeks a value, values do not always re-
him. In some ways an identity gives information on what one "ought" to get from lect needs-partly because values include both what is desired (needed) and what
government, what one ought to give to government (am I rich or poor?), what rules s desirable (preferable but not necessarily preferred). As mentioned above, general-
are appropriate for "me." By its social reference, an identity establishes fault lines in zed values tend to be weak constraints on political beliefs, but this is not true of
society-indicating friends, allies;opponents, even enemies. These may or may not leeds.
follow class lines; for most Americans they follow ethnic lines. And an identity, through I t has been argued (Maslow, 1954) that needs are' hierachically ordered in
establishing levels and orientations of trust and distrust (often by generalizing from ;uch a fashion that one must satisfy certain of them before one can consider others,
the self), indicates a posture toward self-government and toward governmental relia- and there is some evidence that this is true. Thus, a man must satisfy his physiological
bility and responsiveness. needs first; it is a condition of survival. The political belief systems of persons strug-
I Human Sets. For the individual, society is made up of institutions and peo- gling with these needs (Knutson, 1972a) is short-term, egoistic, concrete; the test
ple; but the institutions, being generally remote, become personalized or at least 'or a policy preference is "What immediate benefits in food, shelter, relief from
identified with human groups who are thought to control, or benefit or suffer from,
112 Robert E. Lane Patterns of Political Belief 113

security are little different. Both of these needs lead to subservience to the strong in hang together. Gardiner, a moderate conservative from a prosperous home, "says"
exchange for food and protection-the patron-client relation the world over. At the (here I interpret and paraphrase) :
third level is the need for affection and belongingness, sometimes expressed as love,
sometimes as community, a form of generalized love. Deprivation of this need leads to Parental support and encouragement have led me to adopt the family
hostility, prejudice, autism; those deprived adopt political beliefs reflecting these ideology and model, and to accept my family's status and the value placed
on status. I accept the status and status values. From them I receive a
views in a generalized mistrust of others. Fourth, the need for esteem, generally
flow of benefits: ascriptive leadership, deference, indulgences, power.
latent as long as the more basic needs are unsatisfied, emerges as a stimulus to political These benefits serve to validate my parental code and values, for they
activity-but also as an inhibition. The search for esteem to satisfy deficiency feelings prove to pay off. Paying off in the present, these values become the basis
in this respect need not give political orientation (left or right, authoritarian or of my future calculations, the basis for political argument and rationaliza-
equalitarian) but gives urgency to the search and indicates the rewards (recognition) tion. Therefore: I am a Republican, a capitalist, a "moderate" conserva-
that must be satisfied. Self-esteem and the expressed esteem of others are separate tive [p. 2791.
but related; both are important. Finally, there is self-actualization-the need that
seems close to our concept of self-fulfillment. This need is sometimes expressed in McDonald, a liberal, makes explicit reference in his self-analysis to the
the phrase "What a man can be, he must be." Like the highest stages of moral analogy between the society and the family; he comes from a liberal professional
reasoning, embodying empathy and internalized conscience, self-actualization em- home. He could be said to argue (as I reconstruct his argument) :
braces others as well as the self. The self-actualized person, whether conservative or
liberal, is humane, open-minded, present and future oriented, undefensive, cosmo- Society rewards and punishes like the family; the behavior and attitudes
politan. which are rewarding in the family are properly and naturally transferred
The theory embraces both an ontogeny, in the modified sense of the psychic to social behavior and attitudes. Control of aggressive and rebellious feel-
evolution of the individual, and a phylogeny, in the sense of stages of history. Analysts ings and more or less compliant behavior toward paternal (good) author-
have not had great success with stages of history, as Comte, Spengler, Sorokin, Marx, ity is rewarding; it produces reinstatement of love and acceptability.
and now Charles Reich (1970) all bear witness. Stages of individual development Generalized social rules and norms are desirable, legitimate, not too con-
have fared better--especially where there is no inevitability attached to the develop- strictive; hence I accept them, and rebellion is unwarranted. Big business
mental theory. The theory of a need hierarchy has progressed beyond speculation, for is bad authority, and hence criticism and challenge of it are not dangerous;
several recent studies (Inglehart, 1971b; Knutson, 1972a; Simpson, 1971) reveal socialist dogmatic authority is too constrictive, hence not acceptable (yet
that a need-deficiency theory does help to explain political belief systems. "something I intellectually need"). The Democratic party allows latitude
From my own investigation of the basis of political thinking in need gratifi- for deviance, hence is acceptable and rewarding. Therefore: I follow
social rules and norms; I am against the power of big business; I am not
cation, examined through the instrument of student ideological self-analyses (Lane,
a socialist; I am a Democrat-and for the welfare state [pp. 281-2831.
1969), it seems quite clear that liberalism and conservatism are themselves shaped
by the uses to which an individual puts his opinions, their serviceability for what he The point to be made here is that conservative and liberal ideologies are mediating
wants to accomplish. Thus, the young men who expressed and revealed a need to be concepts between the individual identity (with, in these cases, a need for continuing
liked and a concern about their likeability tended, more than others, to bid for the
identification with the family) and policy preference.
affection of underdogs with a liberal "tolerance" and to require of the government
Not moral axioms but moral reasoning is the clue to the relationship between
the same kind of propitiation of others that they required of themselves. Those who
morals and political beliefs. As with politics, one has a better understanding of what
were marked by aggressive needs might be either liberal or conservative; but those
a man believes if one knows the way he thinks than if one has a map (or, more
whose aggression was restrained largely by fear of the consequences to themselves
likely and less informatively, an inventory) of his beliefs. In the study of Political
were conservative, while those whose aggression was restrained by concern for the
Thinking and Consciousness discussed above, we found some young men seeking,
victims tended to be liberal (see also Thibaut and Riecken, 1955).
rather painfully, to reveal a high moral stance to the world and to themselves. Those
We are interested in what holds opinions together, making for clusters instead
of isolates. As we have said, the deductive logic from, say, conservatism to a policy
who adopted a primitive moral reasoning based upon external rules turned out to
preference has only weak constraining power, but the logic of experience in gratifying be conservative; those who were guided by internalized consciences--on Kohlberg's
needs may be more powerful. In Political Thinking and Consciousness (1969), I (1964) scale those at stages 5 and 6-were liberal. Furthermore, for the middle-class
inferred some chains of reasoning for several young men who were particularly close person, an impression-management morality (stage 3) tends to produce a liberal
to their families and wished to continue their familiaI identity, continuing roughly orientation, probably because an expressed preference for a social arrangement
in their fathers' footsteps. The constraints represented by these internal arguments, providing some element of gain for groups other than one's own, the deprived and
first for a conservative and then for a liberal, help to show how certain idea-elements suffering, gives a better "impression" than would a preference for social arrangements
114 Robert E. Lane Patterns of Political Belief 115
clearly beneficial to oneself and others like oneself. For the working class, this argu- And social scientists believe that there are no uncaused events, and yet they do not
ment would not hold. give up their own sense of undetermined choice. I t is the old free-will dilemma, to
Yet moral codes do not greatly constrain policy preferences-partly because which no solution is in sight.
moral prescriptions and premises must necessarily conflict; that is, one moral con- Setting for Political Interpretation. The currently popular phrase em-
sideration is selected at the cost of another: efficiency (parsimony) at the cost of ployed to account for the malaise of the third world, "the frustration of rising ex-
generosity, love at the cost of work, civic duty at the cost of familial duty, and so pectations," reflects one way in which a sense of time may be out of phase with the
forth. This inevitable conflict is reflected in the moral apothegms of society: "Blessed maturation of events.' In another sense, as the economists point out, the difference
are the meek" but "Fear God and take your own part"; "It is a fine thing for rich between preferences for money "now" versus money "later" is reflected in the interest
people to be philanthropic" but "Take heed for the morrow" (suggesting a more self- rate; but since many individuals-'may discount the future at a higher rate, they may
interested thrift). Prudence and initiative come into conflict in the familiar opposi- be improvident. A welfare society is designed to help such people through imposing
tion: "Make haste slowly" or "Look before you leap" versus "A stitch in time saves mildly coercive social security tax provisions. But, of course, tl?e main theme in time
nine" and "He who hesitates is lost." In morals as in the law general principles do perspective, reflected in the value orientations employed by F. R. Kluckhohn and
not decide individual cases, for the life of the law (or of morality) is experience. And, Strodtbeck (1961), is the differences among a past orientation (as revealed in the
as Converse (1964) found for most of his sample, general ideologies do not decide thinking of American Indians), a present orientation (Mexican American), and a
policy preferences; experience is more important. I
future orientation (Protestant Americans). These orientations have constraining
Explanations of ,Events. Policy preferences imply the beliefs that some gov- i effects on policy preferences, as the very terms reactionary and progressive imply.
ernmental action will effect some preferred change; thus, they rely on theories of
The constraints upon policy preferences and other political thinking made by
cause and effect. Lying behind the conservative ideology is a theory of the effective-
a sense of "place" are often embedded in concepts of community. What is the rele-
ness of the market in achieving a just distribution of rewards; the liberal (in the
vant community to be taken into account: the village, the region (American
American but not the European sense of the term) believes that governmental action
South?), the nation, the Occident, the Afro-Asian world? The men of the Eastport
can sometimes distribute rewards better, or at least can modify injustices or ineffi- study (Lane, 1962) clearly considered locality important in their private lives (be-
ciencies in the market allocation. Cutting across this explanatory principle is the
cause of family and familiarity), whereas in politics the nation was most important,
principle of "cabalism" (Lane, 1962, pp. 113-130), the belief that small groups of for the media had nationalized the news. In the study of student orientation (Lane,
powerful men or interests, either good or bad but always out of sight, decide things.
1969), on the other hand, localism was minimal; some of the students regarded the
These "cabals" might work through manipulating the market or through manipulat- I American nation as their theater of important events and their criterion of relevance,
ing government, or most probably both. This is a kind of conspiratorial theory of
whereas for others the world order (in which the American nation was one important
cause; it relieves those who fear that no one is running things but frightens those who
unit) was significant. By providing these criteria for attention and relevance, the
prefer to believe that things are pretty much what they seem.
sense of place and community guides policy preferences fulIy as much as any ideologi-
Those who adopt great-men theories of cause may be conservative (the cap-
cal posture toward liberalism or conservatism,
tains of industry are the great men) or liberal (the innovative presidents are the
Epistemology. The main epistemologies go under the names intuition, faith,
great men). But those who believe that some kind of organic process, perhaps bene-
or divine revelation; rationalism, empiricism, or ( a mixture of the last two) scientific
ficent but at least unchangeable, is at work in society-so that natural laws inscrut-
method. All of them lodge somewhere on the abstract-concrete continuum. The
able to mankind are controlling our destinies-will most likely bepconservative.These
emphasis in the preceding discussion-upon experience as constraint versus ideology
"organicists" hold that conscious human intervention, in the form of laws and pro-
as constraint-seems to argue, contrary to much informed opinion, that the concrete
grams, is impotent or, if not that, malignant in its ignorance of consequences. Con-
end of the continuum is more useful for opinion formation than abstraction. I do not
versely, those who have a "scientific" approach to society, believing that through accept that inference, for the reason that abstraction from experience, employing
informed inquiry we can discover the cause-effect relationships between human
experiences comparatively and analytically, is a form of abstract thinking with as
actions and tlieir consequences, may be liberal, for the government can make things
many credits as the deductive process of comparing case with principle and decid-
better, but they may equally be conservative, for it may be that private enterprise is
ing accordingly. Similarly, a failure to compare, contrast, anticipate, or in other ways
a better and more beneficial agency of change.
to rehearse alternatives in the light of experience, reduces the coherence and con-
But any great optimism about the constraining effects of the adoption of
sistency (as described above) of a political belief system. Without the rehearsal of
causal theories of these global character should be discouraged by consideration of
alternatives, order is given to responses by the sequences in which they are presented;
the Marxist, the Christian, and the social scientific paradoxes. The Marxists believe they are dealt with in the terms presented, one at a time. This procedure gives over
that there is a historical inevitability to the dialectical process leading to the socialist I
1 to ccsociety"or "history" the power to give order or coherence to a belief system; the
society, but they also believe that dedicated communists are necessary to achieve their
capacity to anticipate and to consider alternatives takes back this power for individual
goal. The Christians believe that divine providence controls human events, but also 1
I use.
that their own freely chosen actions are central to the guiding of human destiny.
The capacity to abstract the self from the environment, to see the self now
116 Robert E. Lane

and the self later as related but different, to hold in solution the emotional angry
self while one deals with the problems of self and others in less angry terms-such
a capacity gives to information a greater utility, and to reasoning a greater persuasive
power.
These three capacities-relating event or problem to concept, rehearsal of
alternatives, and separation of self from environment-are all capacities toward the
abstract end of an abstract-concrete dimension. They are the products of matura-
tion, education, intelligence; and they are informed by personal experience.
Other dimensions and concepts of knowledge guide the constraining functions
in organizing belief systems. If one is not good at cognitive balancing, one tends to
organize one's views by preference for message source rather than message content.
If one is dogmatic, one organizes beliefs according to their utility in defending the
dogma (and this is the risk implicit in the use of ideologies as constraining devices).
Steeped in a philosophy of science, a man asks for causal chains and evaluative
criteria before he decides-if, indeed, he ever does decide. SOCIALIZATION
Richard G. Niemi

p o l i t i c a l ideas-like the consumption of

cigarettes and hard liquor-do not suddenly begin with one's eighteenth birthday.
The analogy does not stop there. For in their political habits-as in their consump-
I tion habits-young people are influenced by parents and other family members and
1 by the schools. And yet the inculcation of traditional values and behavior is not
I
completely successful; for whether it is in poIitics or in deciding what to smoke, new
ideas and habits often come from the younger members of society. The analogy also
pertains to the history of research on political learning and on learning about drink-
ing and especially about smoking. While both politics and smoking have been around
for a long time (sometimes simultaneously, as in smoke-filled rooms), only in the
last fifteen years or so have serious efforts been made to understand the,respective
learning processes. Understandably, there is much to be learned-witness the naive
thought that stopping broadcasting advertisements of cigarette smoking would curtail
I the habit; but we have come a long way in a decade and a half.

I I thank Roman Hedges for critically reading a draft of this chapter.

~ 117
 118 Richard G. Niemi Political Socialization 119
The present chapter, henceforth restricted to political learning, will survey learning has led to an awareness that children develop political or quasi-political
much of the knowledge gained in recent years. But it will do so with an eye toward notions at a very early age. Thus, the studies we cite below involve youngsters from
understanding the meaning and importance of specific research findings that have from the first or second grade on.
been uncovered. In addition, areas in which further research is necessary will be At the same time that we recognize the consequences of political socializa-
pointed out. tion for political systems, we should also recognize that political learning is important
We will not try to provide a precise definition of the term political socializa- for the developing personality of each individual. Thus, for example, while we might
tion. (For attempted definitions see Easton and Dennis, 1969, p. 7; Sigel, 1965, p. be interested in political efficacy partly for its effect on the political system, we might
2.) Greenstein (196813, p. 551) properly indicates that socialization can be thought be equally concerned with how a person's feeling of competence in dealing with the
of either very narrowly-meaning, chiefly, civics classes in high school--or broadly- political world and society in general helps him meet his own needs and goals in life.
meaning "all political learning" (see also Greenstein, 1970a). To us the main feature (This subject is discussed at length by Smith, 196810.) Similarly, we might wish to
of most socialization studies is that they involve learning at the preadult stage. In-
observe the development of moral judgment and behavior from the point of view of
deed, one contribution of the field of socialization is that it alerted researchers to the
the society (to determine, for instance, how well a population obeys written and un-
notion that political learning has roots deep in childhood. But we shall not confine
written rules and laws) ; but we might also wish to consider a more individualistic
our attention solely to preadults. We will on occasion draw from adult studies; and
point of view (to ask, for instance, why some individuals develop habits that deviate
later on we will have to confront directly the question of the relationship between
from conventional moral patterns and what happens to them as a result).
early learning and adult political life.

Youthful PoliticaI Learning I History

I t is something of a tradition to note that writers ever since Plato have been
It is generally assumed that preadult political learning is important because
concerned with the political socialization of youth. One writer (Connell, 1969) even
what is learned early will have a significant impact on adult political attitudes and
delights in pointing out that Plato was far from the earliest author to concern himself
behavior. It is argued, for example, that one of the difficulties of managing a newly.
with this matter. However, the era of serious academic attention is 'much shorter
formed nation is that individuals bear no allegiance to those particular political
indeed.
boundaries. Instead, citizens of the new nation retain loyalties developed in the past
In the area of genera1 socialization, without a particularly political focus,
and consequently have little interest in or loyalty to the new nation (LeVine, 1963).
studies have been done at least since the turn of this century. Such work has a long
Similarly, when new and different regimes come into power in a nation, the new
leaders often attempt to establish new modes of political thought-sometimes by history and is still actively pursued today. Piaget alone has done research over a forty-
reteaching adults (Fagen, 1970, pp. 318-320) but more often by changing the system year period. More recently, several journals, such as Child Development and the
of educating youth (Azrael, 1965, esp. pp. 257-258). Merrill-Palmer Quarterly, have been devoted solely to this area. The political rele-
Even in established nations, the presence of "subcultures of discontenty' vance-of much of this literature is marginal or at least not clearly spelled out. But it
(Greenberg, 1970c, ch. 5-8) is seen as potentially disruptive of stability and peace. is useful for gaining a wider understanding of child development and for placing
Conversely, preadult attitudes favorable to the existing political system are often political developments in a broader framework of general socialization.
thought to be a major source of system stability (see esp. Easton and-Dennis, 1969). Political socialization as such has some of its intellectual origins in educa-
Particularly in times of unrest and instability, the underlying support for a nation tional research, which also goes back to the turn of the century. In particular, we
which develops early in life goes a long way toward explaining the maintenance of might mention the scholarly study of formal aspects of civic training made in the
political systems. Iate 1920s and early 1930s by Merriam (1931) and others. Another part of the
Although the correctness of this assumption is a matter of utmost importance, intellectual ancestry of political socialization research is found in the studies of
and one we shall return to later, we will dwell on it here only long enough to make national charatter-studies undertaken during World War I1 and after (Inkeles and
one further point. If, as seems obvious, youths do not suddenly develop political Levinson, 1969)-which attempted to specify internation personality differences and
ideas at age eighteen, and if early learned attitudes are of cardinal importance, then to understand their developmental origins. Finally, several scattered studies in the
it behooves us to seek the origins of political learning. But that implies not stopping 1 9 3 0 ~the
~ '194.0~~and the early 1950s concerned political attitudes specifically. Many
with sixteen year olds or fourteen year olds, but studying children as early as they of these are summarized by Hyman (1959). In addition, there were studies of politi-
begin to develop political or politically relevant ideas-whenever that time is. We cally related attitudes, such as children's perceptions of social class and race (Holling-
do know that children as early as first or second grade can articulate ideas about shead, 1949; Stevenson and Stuart, 1958).
authority and about political figures. We also know that children's responses are The continuous and direct study of political socialization began in the middle
sometimes unstable (Vaillancourt, 1970)' which raises the question of just when to late 1950s under the impetus of three more or less simultaneous developments.
meaningful attitudes take shape. In any event, the search for the origins of political First, Hyrnan's book Political Socialization (1959) reviewed much of the scattered
 120 Richard G. Niemi Political Socialization 121

literature that existed up to that point, and indelibly impressed upon political scien- Having now seen the rationale for and the history of studies of preadults, let
I
tists and psychologists that political learning is indeed a challenging and worthy us turn directly to what some of the studies have found.
subject in its own right. The other two sources were studies begun under the direction
of Easton and Hess at the University of Chicago and by Greenstein in New Haven. Children and Political Authority
The rather surprising findings of these two studies about the idealization of children's
political ideas have been a source of fascination and, with the college uprisings of Do children think about political authority figures; if so, what do they think?
the mid-1960s, a source of consternation. Ultimately, most of these findings were The answer to this question is quite thoroughly documented in the Chicago and the
summarized in three works (Greenstein, 1965; Hess and Torney, 1967; Easton and New Haven studies. Both of these studies, along with the large number of pretests
Dennis, 1969). The New Haven and Chicago studies were both based on quasi- conducted in the Chicago study, were conducted in the late 1950s and early 1960s
longitudinal research designs aimed at identifying the developmental path of political and were limited almost entirely to white, urban children. We shall have to return
attitudes over the elementary school years. We will draw freely upon them in talking to the question of just how representative these results are of other times and other
about children's views of authority. individuals. Moreover, both studies relied on survey responses, and it has been sug-
In the mid-1960s another study, with a different research design and em- gested on the basis of in-depth interviewing that survey responses may overestimate
phasis, was undertaken at the University of Michigan. In this study interviews were the extent to which children view authority favorably (Knutson, 1972b).
held with high school seniors and their parents, teachers, and school principals, with The Chicago study obtained questionnaires from second- through eighth-
the purpose of more adequately determining the sources of young people's ideas. A grade children and the New Haven study from children in the fourth through eighth
major report of this study is forthcoming (Jennings and Niemi, 1974). grades. Thus, when we speak of early political learning, we have in mind children
As with any emerging field of inquiry, texts soon appeared (Dawson and as young as seven or so. In fact, it has been suggested that some rudimentary political
Prewitt, 1969; Jaros, 1972), along with several readers (Sigel, 1970; Greenberg, notions are brought to kindergarten by children, but it is difficult to establish this in
1970c; Adler and Harrington, 1970; Orum, 1972; Dennis, 1973), review essays any systematic way because of the problems of verbal communication about rela-
(Sears, 1968), and an extensive bibliographic effort (Dennis, 1971). These sources tively abstract subjects at that age.
provide an overview of what has been done to date. Easton and Dennis (1969, pp. 391-393) provide four terms to describe major
More recently, studies of political socialization have taken at least three new features in the development of children's feelings about political authority: politiciza-
directions. First, there are studies of subgroups of the American population, especially tion, personalization, idealization, and institutionalization.
groups likely to hold attitudes different from those of middle-class American whites. Politicization refers to the fact that young children fairly quickly learn that
Examples include studies of Appalachian youths (Hirsch, 1971), studies of blacks there is an authority above and beyond family and school figures. This politicization
(Lyons, 1970; Rogers and 'Taylor, 1971; Greenberg, 1969, 1970a, 1970b; Liebschutz often comes about in rather simplistic ways-for instance, when a father obeys
and Niemi, 1972) and studies of Mexican Americans (Cornbleth, 1971; Garcia, traffic laws so as not to run afoul of the policeman-but it is no less effective because
1972). Second, the host of studies of socialization in other countries indicate the of that. By the time children are seven or eight years old, then, they have become
degree of generalizability of some of the findings on American youths. Finally, the aware of an external force that demands some support, obedience, and (usually)
few methodological studies undertaken (Vaillancourt, 1970; Niemi, 1973) get at respect. They may understand this external force only sketchily, and terms such as
some of the serious methodological quesbions that must be considered when any new government may be foreign to them until they are older. But awareness of such a
population, but especially a young population, is first studied. power makes possible the development of other political ideas from an early age.
Most of the findings cited below are based on survey research, although Personalization refers to the fact that children first and most easily become
some studies using projective and semiprojective techniques (such, as Adelson and aware of political authorities through individual men-most commonly, the president
O'Neil, 1966; Knutson, 1972b) and content analysis (Litt, 1963) have been used. and the policeman. Collectivities such as Congress and the Supreme Court and other,
(See also Chapter Fifteen of this volume.) Admittedly, survey research can be un- more abstract concepts such as government are apparently less easily understood,
reliable, especially when used with young children. Respondents may give unstable although understanding increases throughout the elementary grades. I t is significant
answers, may check substantive answers when "don't know" would be more appro- that personalization should be found in a country such as the United States, where
individuals are presumably subordinate to the constitutional system. This finding
priate, or may give what they see as socially acceptable responses, Moreover, closed-
suggests that the young child's emphasis on individual personalities is due to cognitive
ended survey items often fail to bring out the way in which children organize their
limitations. To put it simply, it is far easier for the young child to learn about the
rudimentary political thoughts and the way in which their political ideas relate to
president than about a group such as Congress.
nonpolitical (but politically relevant) orientations. Despite these possibilities, few
A finding of particular interest to us older cynics is that idealization also
psychologists, and even fewer political scientists, have actually used other methods,
characterizes children's views. That is, to most children political authority seems
although there is a growing movement in that direction (Lane, 1969; Dennis, Bill-
trustworthy, benevolent, and helpful. 'To a surprising degree children respond that the
ingsley, and Thorson, 1968; Greenstein and Tarrow, 1970; Merelman, 1971a). policeman and especially the president "would always want to help me if I needed
122 Richard G. Niemi. Political Socialization 123
it," that "they almost never make mistakes," that "they know more than anyone." the common viewpoint seems to be that young blacks very quickly learn "what the
Most children agree rather strongly that what goes on in the government is all for score is" and soon develop negative images of political authority. Interestingly, how-
the best. Most agree with the blanket statement that all laws are fair. Such extremely ever, the data on this question are far from one-sided. Jaros (1967), for example,
idealistic responses soon fade away. But if indeed the earliest learned attitudes persist reports almost no differences in views of the black and white children in his sample
to some extent in later life, the child's idealization of political authority does help us in Detroit. Jennings and Niemi (1974, ch. 5), Rogers and Taylor (I971), and
understand the attitude, on the part of numerous adults, of unquestioned support for Kenyon (1969) found almost no differences in political trust between black and white
the government. students. In contrast, a number of studies (Dennis, 1969; Lyons, 1970; Greenberg,
Finally, the development of children's viewpoints is characterized by institu- 1969, 1970a, 1970b; Liebschutz and Niemi, 1972) report fairly consistent differences
tionalization. Young children gradually learn to associate with depersonalized objects in the expected direction. Reviewing these studies, Abramson (1972) made it clear
such as the government and Congress. In fact, they transfer to them some of the that the explanation for these contrasting results lay in a rapid decline in political
qualities previously attributed to the president and the policeman. This, of course, is trust in the late 1960s, a decline especially notable among blacks.
a necessary step in a country in which allegiance is owed to constitutional authority Turning to contrasts between sexes and across social classes, we find some
rather than individual leaders. differences in the expected direction. For example, brighter children become politi-
As children get older, these images change quickly and dramatically. In part cized faster than those who are not so bright. Girls personalize and idealize political
we have already described this development. They come to emphasize institutions authority more than do boys. Lower-status children likewise have more personalized
more than individual personalities. They become aware of deficiencies in individual and more benevolent views of authority. More important, however, the pattern of
persons and inequities in the law. There is a growing differentiation between role and development is again very much the same in every case: an early personalization and
incumbent, so that, for example, the presidency remains highly regarded even if the idealization followed by a decline in those features and a rise in other characteristic
incumbent president is seen as less than wholly praiseworthy. Interestingly, while views, such as cynicism and a belief in the importance of institutions.
teenagers' views of political figures are far more realistic (or less heroic and benev- Just how generalizable are these patterns, then? Are there instances of real
olent) than those of young children, they are still considerably more trusting and less deviation from the development we have seen among most children in the United
cynical than those of adults. Clearly, the process of growing cynicism continues well States today? Several of the key features we have pointed out are probably universal.
into adult life, although at a slower rate than during the elementary and high school Politicization, for example, must occur everywhere. Under some circumstances this
years (Jennings and Niemi, 1968a; 1974, ch. 10). process may unfold more slowly than in the United States (Pammett, 1971) ; in
How generalizable are these patterns? Do they characterize today's children, other instances it may occur very quickly. If children are raised by the ,state, so to
and do they characterize the views of minority-group children and those of other speak, in communal groups organized or controlled by the central authority, they
groups in the society? The best available answer seems to be that the patterns may very quickly learn about the transcendence of political authority. In fact, they
described are highly generalizable, although the level of positive or benevolent views may learn that political authority is, the major outside source of authority-rather
of political authority may often be lower than what was found by the earlier re- than the authority of family, of schools, of secondary groups, of religion, and so on.
searchers. In other words, children today, when compared with older children and An attempt to bring about this kind of socialization process has in fact been made in
adults, still seem to be relatively favorable toward the government and specific the Soviet Union (Azrael, 1965) and probably also in China, with the collectivization
authority figures. At the same time, their views appear to be somewhat less favorable of child rearing and with the centralization of the schools. It has not worked as well
than the views of the children in earlier studies. as political leaders might like it to because of difficulties in breaking down some
Comparisons across time are difficult to make because of variations in the authority relations-especially familial and religious authority; but it certainly has
sample, in questions, and in methods of gathering data. Nonetheless, there are some had some effect.
suggestive results. Tolley (1972) found that President Nixon's actions regarding the Personalization and institutionalization likewise are probably nearly universal,
Vietnam war received considerably less than full support from children. And in a because, as mentioned earlier, it may simply be easier for children to grasp individual
study conducted in the late 1960s, Liebschutz and Niemi (1972) found that white personalities first and abstract conceptions later. We would not go so far as to say
students in Rochester, New York, were considerably less likely than the children in that this pattern can never be reversed, but in the normal course of events it probably
the earlier Chicago study to consider all laws fair. At the same time, these later describes the development of most children.
researchers found the same overall pattern of much higher belief in the fairness of Idealization seems to be the one concept that clearly is not universal. Some
laws at the early grades. Thus, although more directly comparable data are needed, young children, in fact, grow up disliking political authority-evidencing "hostiliza-
changes seem to have occurred in the level but not in the pattern of support for tion" (Easton and Dennis, 1969) rather than idealization. The one outstanding
political authority over the past decade. example of this kind of development was found by Jaros, Hirsch, and Fleron (1968)
Among subgroups in the population, the most obvious group of children who
during research in the Appalachian region of eastern Kentucky. A few figures call
might be expected not to share the developmental pattern described above are urban attention to the sharpness of the contrast between their findings and those in earlier
black children. We are speaking, of course, about young children and not adults, but studies. When asked, for example, how hard the president works compared with most
 124 Richard G. Niemi Political Socialization 125

men, only 3 percent of the respondents in the Chicago study-but 41 percent of the system then established would be subject to the same forces for change. Instability
children in Knox County, Kentucky-felt that he works less hard. Similarly, when of political regimes, as is common in many parts of the world, might be a far-reaching
asked their view of the president as a person, 8 percent of the students in the Chicago result. In contrast, the present socialization system, with its tendency toward early
study, in contrast with 26 percent of the Knox County children, said that he is not idealization, promotes stability in the political system, with the concomitant possibility
a good person. What is perhaps most significant, however, is that the pattern of of stagnation.
development itself is different among the Appalachian children. There seems to be The point we are making is not that stability is more or less desirable than
little change in their responses by age, so that evkn relatively young children are not change, but that the socialization system plays a crucial role in the balance between
impressed by or are even cynical about political authority. stability and change. Alteration of the method by which we socialize children,
The extent to which idealization or hostilization characterizes young children whether accidental or planned, may very well have far-reaching consequences for
in other cultures is a fascinating question. There are only hints of an answer. Hess the future of the political system.
(1963), for example, found that the degree of idealization is strongly related to age We might point out in passing that the socialization of attitudes toward
in the United States and Japan, while feelings about authority figures are virtually authority may also have serious consequences for individual well-being. One might
unrelated to age in Puerto Rico and Chile. His samples, however, were small and wonder, for example, whether an individual is more likely to develop ego strength
were intended only to be suggestive. There have been a number of suggestions that and a feeling of identity if he first learns that the political authorities in his country
French children are relatively cynical about political authority (Wylie, 1964, esp. are extraordinarily good and only later is introduced to some of the complexities of
ch. 10; Roig and Billon-Grand, 1968; Abramson and Inglehart, 1970) ; but, as making political judgments; or whether it is better from this perspective if he is
Greenstein and Tarrow (1969) note, sound empirical evidence is lacking. At the trained from the beginning to have an open mind, to examine questions for himself,
very least, then, we can conclude that idealization is less than universal and that and to make his own decisions about what he considers best.
there are probably significant instances of children who develop a hostile attitude Before leaving the area of children's views of authority, we should call athen-
toward political authority at a very early age. tion to one other type of study-namely, children's awareness of and understanding
We might well ask at this point just what the consequences are of the pattern of national identity and foreign cultures. (A good summary of this work is found
of youthful development that we have surveyed. Since the extremely benign, and in- in Davies, 1968.) I t has been found that children often have an egocentric view of
deed naive, views of early childhood are quickly rcplaced by more realistic and even the international system. A child can understand that children in other countries
cynical views of adulthood, are the early attitudes of any real consequence? In part are "foreigners," but he finds it difficult to grasp that from the point of 'view of
this question is impossible to answer until we resolve the difficult issue of just how individuals in other countries he himself is a foreigner. In addition, certain geo-
much constraint early learning places on later attitudes. Obviously, the childhood graphical notions-for instance, that a city is within a state and that a state is
attitudes are rejected in favor of more realistic ones, but it may well be that the within a country-are difficult to conceptualize (Piaget and Weil, 1951). Although
earliest learned responses are never totally forgotten and, moreover, that adults strict stages of development may not be followed (Jahoda, 1963a, 1963b, 1964),
would be more cynical than they are were it not for the early idealization of authority. certain viewpoints, along with an understanding of abstract concepts, generally do
If what is learned early in life is indeed influential in later years, the implica- not develop before specified ages (Adelson and OWeil, 1966, 1969). Finally, it has
tions of early idealization may be of singular importance. That is, childhood views been documented that quite early in life children may develop unflattering stereo-
of political authority may be an important source of stability in the American political types of other cultures and nations (Lambert and Klineberg, 1967).
system (Easton and Hess, 1962; Greenstein, 1965). This point might be made most
forcefully if we consider the possible consequences of altering the educational system Agents
to purposely instill in children at the earliest possible age a questioning attitude
toward everything, including the foundations of political authority. A completely Considerable effort in the study of political socialization has centered around
objective presentation might conceivably be made of the good and bad points of a the question of where children get their attitudes. Most of the work has focused on
variety of political systems. Individuals could be encouraged to make up their own the family and the school, with proportionately less work involving peer groups, the
minds about which system is best and about the proper means of bringing about that media, and other sources of influence. Likewise we will concentrate here most heavily
kind of system. Such an instructional program probably would bring about a gradual on the family and the school.
weakening of support for the present system and give rise to the probability that Family. Until recently one could easily summarize the academic and the
change would occur. popular stereotype of socialization in the family: "We are all like our parents,
Whether such a program would be good from a normative point of view especially our fathers." Before more firm evidence was gathered, the reasons for this
depends on one's attitude toward change and stability. Recently many young people view seemed substantial enough. First of all, because the family is the earliest sociali-
in particular have argued that our political system is stagnant and that change is zation agent, we can easily slip into thinking that it must therefore be the most
called for. If ch-dnge is considered desirable, perhaps altering the socialization system important one. Second, since the young child is emotionally attached to family mem-
in the direction indicated would help bring it about. But, of course, any subsequent bers, the considerable impact of family viewpoints would seem likely. Third, certain
126 Richard G. Niemi Political Socialization 127

findings about partisanship, at least as they were typically interpreted, supported the Turning to other orientations studied, we find some additional surprises.
notion that children generally adopt parental attitudes (Maccoby, Matthews, and There is no need to go through them all in the detail that we did with partisanship,
Morton, 1954; McClosky and Dahlgren, 1959; Hyrnan, 1959, ch. 4; Campbell and but a few examples can be drawn upon to illustrate significant points. Consider, for
others, 1960, pp. 146-149). Moreover, researchers were hard pressed to uncover example, the student-parent relationship for political cynicism. Cynicism is a general
significant amounts of adolescent rebellion regarding political matters (Nogee and sort of dimension; it is not a response to one particular item or to a particular event
Levin, 1958; Lane, 1959a; Middleton and Putney, 1963). Finally, it seemed obvious but, rather, a generalized orientation toward the political system and toward political
that fathers are more instrumental in passing along political views than mothers, authority as a whole (see Aberbach, 1969, and the references cited therein). Thus,
although this viewpoint was probably due as much to male chauvinism as to empir- even if individuals' reactions to specific events and questions were to differ, we might
ical findings. expect a reasonably strong resemblance between the overall or generalized attitudes
Since partisanship formed the primary empirical basis of the traditional of young people and their parents. It is something of a surprise, therefore, to find
stereotype, we might begin to summarize the empirical evidence that is now available that the correlation for this particular measure is only .12. Similarly, on a general
by taking a close look at this attitude or orientation. (Qn the concept and measure- measure of political interest--quite apart from particular manifestations of this
ment of partisanship.or party identification, see A. Campbell and others, 1960, 1966.) interest, such as reading about politics or participating in politics-the correlation is
Our discussion draws heavily on Jennings and Niemi (1968b; 1974, Part 11)' who only .12.
interviewed a national sample of high school seniors and their parents. Some of the Having seen these results, we miglnt turn instead to particular behaviors or
main results are also supported by an imaginative "sibling" study conducted by Hess attitudes toward narrowly defined issues in the hope that perhaps these would yield
and Torney (1967, pp. 97-99), in which the similarity between siblings was compared higher correlations. This hope is not very well fulfilled, it turns out. On some salient
to that between matched pairs of nonsiblings. The results below are presented largely issues the correlation rises to as much as .3 or so. On certain other issues, however,
in terms of correlations. The use of correlations is intended to make the comparisons it is even lower than the correlations cited for generalized orientations. Indeed, out
as comparable as possible across items differing in numbers of categories (so that of some twenty different orientations measured (Jennings and Niemi, 1974, ch. 6),
percentage comparisons necessarily would vary widely) . there was only one-namely, presidential preference in 1964--for which the correla-
Turning to partisanship, we find a correlation between parents' and students' tion rose above that observed for party identification. In the case of presidential
partisanship not of .8 or .7 or even .6, as we might have expected, but of .47. How preferences, the correlation was .59-still not of a magnitude sufficient to assume com-
might we evaluate this particular correlation? From one point of view it is not too plete correspondence between parental and student preferences.
"bad" a correlation. It is certainly in the expected direction, and it is clearly much Some of the implications of these findings will be taken up later. Two other
greater than zero. There i~ a definite simiIarity between parents and their children. questions need to be confronted at this point. First, we might reasonably ask why the
Partisanship is, to a degree, passed on from generation to generation. correlations are not higher than they are. Second, we might want to know whether
Some further understanding of the meaning of this particular correlation may the student-parent correlations are very high for certain types of students-for in- -
be gained by taking a more detailed look at the underlying relationship. First of all, stance, those who are especially interested in politics. These questions might be
as a simple percentage table would reveal, there are not really very many families in answered in a number of ways, but here we will concentrate on a few of the major
which the parents are Democrats and the children Republicans, or vice versa. In results.
fact, such families amounted to only 7 percent of the total in 1965. Hence, in this One possibility is that the correlations would be higher if we used children's
extreme respect there is considerable continuity in the American political system. On perceptions of parental attitudes rather than parents' own reports about themselves.
the other hand, if we divide the population into Democrats, Independents, and That is, children may think they are more similar to their parents than they really
Republicans, we find that only about 60 percent of the student-parent pairs show are. The results of a whole set of analyses indicate that, for high school students at
identical partisan feelings. This means that if we consider the Independent category- any rate, the student-"perceived parent" correlation is likely to overestimate the true
and movement into and out of this category-more carefully, there is a greater degree student-parent correlation by about .l. Recall, for example, the true student-parent
of partisan fluidity than the previous~comparisonsuggested. Among the Democrats, correlation of .47 for partisanship. The student-"perceived parent" correlation in this
for example, one third of. the students deviated from their parents at least to the case is .58. Hence, as we would expect, the correlations involving perceptions are
degree of becoming Independents. Among the Republicans fully half of the students greater than the true student-parent correlations. However, the increase is not so
departed from parental ways, and among the Independents the figure was about one large as to explain entirely why the actual relationships are no higher than they are.
half. Thus, there is far from a one-to-one correspondence between parents and Another reason that the student-parent correlations are not considerably
students in the matter of partisanship. With these percentages as a background, the higher is that other parental characteristics impinge upon the development of student
.47 correlation seems to make considerable sense. I t indicates only a small number of attitudes. In other words, some family influence is not monitored by the student-
major deviations, representing movement all the way across the partisan distribution, parent correlations. Thus, for example, when a parent who feels that he is a Democrat
but it also indicates a fair amount of shifting back and forth among narrower ranges votes in a particular presidential election for the Republican candidate, his children
of the partisan dimension. are much less likely to adopt his attitude than wkien his vote is consistent with his
128 Richard G. Niemi
I Political Socialization 129
stated partisanship. Ironically, in such instances it is the parent's own behavior that example, have studied the effects on children when the father is gone for long periods
leads students away from parental paths. of time (Tiller, 1957; Gronseth, 195'7; Lynn and Sawrey, 1959; Burton and Whiting,
A third, glaringly important factor in explaining the degree of student-parent 1961; see also references in Walter and Stinnett, 1971, pp. 81-82). Only a few
similarity and dissimilarity is the extent to which the parents themselves agree politi- political scientists have picked up this lead. Pinner (1965) suggests that parental
cally. For most political attributes, the fact of having similar parents raises considef- overprotection leads to political distrust and disaffection. Langton (1969, ch. 2)
ably the probability that youths will follow in their footsteps. For example, whether suggests that children from maternal families are more authoritarian and less politi-
the parents agree with each other seems to be more important than the family's cally interested and efficacious than children from nuclear families. Lane (1969,
degree of interest in politics, the compatibility of family members, demographic ch. 14) has related identification with the family to political feelings of college-age
features such as region, or personal characteristics such as race and sex. students. Clarke (1973) finds that blacks in father-absent families are more cynical
Even when the parents agree, however, student-parent agreement is far from about politics than are those in two-parent families.
complete. Most of the correlations reported by Jennings and Niemi are still below .40
Studies of family relationships have not yet progressed to the point that out-
(the highest being .57 for partisanship and .63 for presidential preference). Even standing and well-documented conclusions can be drawn from them in regard to ,
when combined effects are considered, there are few sizable increases in the magni- politically relevant dimensions. But they clearly define an area in which future study
tude of the correlations reported. Even, for example, when the parents agree with one is needed. They also call attention to the need for being cautious about concluding
another, the parents and the child are both interested in politics, and family members that the family in all its respects has little influence on the child's political develop-
I
get along well with each other, there are still many instances of student-parent dis- ment.
agreement.
I A second qualification to our earlier summary of parent-child similarity is
A complementary point concerning parental influence involves children of the likely fact that there are significant differences across geographical and cultural
parents who disagree with one another. Whether we look at the agreement of these boundaries. Butler and Stokes (1969, ch. 3) found that in many respects the pattern
students with their mothers or with their fathers or with some amalgam of the two
of political socialization is very similar in Great Britain and the United States. They
parents, the typical student-parent correlation is far below what is observed when the
note, however, that in one sense fathers are more important in the socialization
mother and father agree. Thus, while there is a set of homogeneous parents whose process in Britain, because many of the British respondents could not recall their
children are more like them than overall comparisons suggest, a complementary mother's political feelings; however, this inability to recall the mother's partisanship
group of students are frequently unlike~theirparents. This, then, is another factor
is declining among the younger voters. Converse and Dupeux (1962), in a classic
that holds down the overall student-parent agreement rates.
article on France, point out that in that country many individuals do not know even
There remains another question to be considered: the question of the father's
their father's political persuasion. As a result, they contend, the socialization process
influence on the children versus the mother's influence. Recall the stereotyped notion
is severely altered; specifically, partisan identification is considerably curtailed.
that politics is a man's business and that most socialization can therefore be traced
Whether the relationship between socialization and partisanship observed by Con-
to the father. This hypothesis, like so much widely believed but little-tested lore
verse and Dupeux will generalize to other countries remains to be seen, although
about the family, has proved to be resoundingly incorrect. The balance of influence
there is suggestive evidence that it does not fully apply to Great Britain (Tapper and
between the parents appears to be remarkably even and, if anything, is tipped in the
Butler, 1970). In any event, it is clear that various~historical and institutional
mother's favor. Evidence for this statement is found in a comparison of the student-
characteristics alter the development of party identification and of the influence of
mother and student-father correlations in families in which the parents disagree. Of
the family thereon (Converse, 1969; Dennis and McCrone, 1970).
some twenty comparisons made, the student-mother correlation was higher than the
So, too, there are cross-national differences in other characteristics of the
student-father correlation thirteen times, with three being dead heats and four
family. Almond and Verba (1963, ch. 12) observed that participation in family
showing at least minimal advantage for fathers. The differences usually were not very
decision making varies from one country to another. However, in all the countries
large, but the results clearly deny the one-sided influence of the father. The same
that they studied, participation in family decisions is related to what they call sub-
conclusion was reached in a study of college-age children by Thomas (1971). In-
jective competence (political efficacy) . Likewise, Langton (1969) notes that maternal
terestingly, another recent study suggests that in matriarchal Appalachia the influence
families are much more frequent in Jamaica than in the United States, and while
of the mother is considerably stronger than that of the father (Hirsch, 1971, pp. 39-
mother dominance of the nuclear family is found here, it is only a partial analog
41). to complete absence of the father.
Clearly, these conclusions about the evidently much smaller impact of the
School. Trying to assess the role of the school in the political socialization
family than was previously thought deserve further consideration, for they have
process is a most fascinating if perplexing task. With all the resources devoted to the
implications concerning the entire socialization process. Indeed, they are so important
school and with the amount of time that children spend within school walls, we
that we will come back to them after we have looked at other socialization agents.
would like to think that we get some return on our collective investment. With our
Another factor that needs to be considered, especially by those well versed in
devotion to civics courses at certain grade levels, it seems only right that certain
psychological work, is the potential impact of family relationships. Psychologists, for
kinds of attitudes and orientations should be developed or strongly reinforced by this
 130 Richard G. Niemi Political Socialization

effort. And teachers naturally are frustrated if their efforts are not met with some a broader conclusion. Questionnaire results revealed almost no changes between the
development in the attitudes and behavior of their students. freshmen and senior levels of high school in political interest, activities, and a variety
Now in some ways there can be little doubt that the school has a significant of political attitudes. While these data can be countered with some positive findings
impact. In sheer volume of knowledge possessed by an individual, whether it be of change (Hyman, 1959, p. 53; Jennings and Niemi, 1968a), the wide array of
knowledge of political or nonpolitical material, the school almost certainly plays a measures used by Hess and Torney weigh in their favor. Similarly, Merelman
predominant role. And yet when it comes to political ideas and attitudes, it is far (1971b) comes to basically negative conclusions about the role of the school in stu-
less clear just what role the school (grade school, high school, or college) plays. Let dents' civic training.
us begin by reviewing some of the conflicting research results. Some aspects of the informal milieu of the school have also been subject to
Several well-known studies argue and/or show that the school or a particular question. Several studies have found little or no effect of participation in extracur-
course within a school has a significant impact on political attitudes. In most cases, ricular activities, whether at the elementary school level (Hess and Torney, 1967,
however, a more or less serious qualification must be added to the statement of pp. 120-125) or at the high school level (Ziblatt, 1965).
positive effects. Hess and Torney (1969, p. 101), for example, conclude unequivocally Finally, although the results are by no means completely one-sided (Feldman
that '?he public school appears to be the most important and effective instrument of and Newcomb, 1969, pp. 19-23), the various college studies lend themselves to the
political socialization in the United States." Yet their conclusion has been properly general conclusions cited above. Specifically, political science courses apparently do
criticized (Sears, 1968; Niemi, 1969) on the ground that their evidence shows merely not increase political interest among students (Somit and others, 1958), and even
that significant changes take place during the elementary school years and that there presumably new and exciting ways of teaching about politics apparently do no better
is no necesary link between these changes and what actually goes on in the school- in stimulating interest and learning than do more traditional ways (Robinson and
room. others, 1966) .
One might also cite Newcomb's (1943) study of the effects of Bennington Many of the same mixed findings pertain to studies done abroad (see Cole-
College on the students enrolled there. This is an excellent study, which convincingly man, 1965a; Dawson and Prewitt, 1969). In new nations, for example, the schools
shows that the college had a strong and real impact. And yet it is clear that this was allegedly bear an important part of the socialization of national identities (LeVine,
an extreme result, not often to be duplicated; Newcomb's own study shows that 1963; Coleman, 1965a, p. 22). Presumably they are especially effective in countering
several other colleges in the Northeast failed to have the same kind of impact on their the more traditional socialization of the family and can be manipulated to promote
students. the interests of the new national unit. While this argument sounds convincing, a case
I study in Uganda (Prewitt and Okello-Oculi, 1970) suggests that manipulation and
A number of other studies lend additional support for the impact of the school.
Litt (1963) found that civics,classes had a uniform impact on certain political atti- control of the education system does not work as easily in practice as in theory:
tudes and a significant effect on other attitudes under some but not all conditions. "However attractive programmed political education might appear on paper, the
Patrick (1972) reported that an experimental course, entitled American Political political and administrative realities caution us not to expect significant results"
a
Behavior, achieved some of the desired changes. Levin ( 1961) and Langton ( 1967) (p. 620).
found that the partisan or class environment in the school had an effect on children's With all of these conflicting reports and conclusions, what are we to conclude
attitudes, although it should be noted that this was an effect of the informal milieu about the school? One conclusion seems inescapable. The effects of the school are
rather than of the content of the course work. One might also draw on studies of highly variable-depending at least on the quality of the teacher, the class material,
the college, which generally show "decreasingly conservative attitudes toward public the social and political composition of the school and classroom, particular circum-
issues" (Feldman and Newcomb, 1969, p. 326) between the freshmen and senior stances of time and place, and even interactive effects such as the correspondence
years, All of these studies together, along with many more casual observations, suggest between what is taught in the classroom and what is informally taught outside of
that schools often make a significant impact on the development of political attitudes. school. A second conclusion is more programmatic in content. It is suggested rather
Nonetheless, the qualifications that we added to the above findings, in addi- uniformly that extracurricular activities seem to have virtually no effect on students'
tion to a number of more negative results to be mentioned, indicate at a minimum political views, such as their feelings about democracy. While such activities might
i
that the effect of the school is far from uniform. A major study by Langton and well be justfied on other grounds, it is difficult to support them on the basis of their
Jennings (1968) found virtually no impact of civics courses on a cross-section sample supposed contribution to the development of democratic citizens.
of high school seniors, although these courses did have a meaningful impact on the The conclusion about the variability of the school's impact deserves further
black students in the sample. Although the results of this study must be qualified- consideration. My own feeling is that the school has an enormous impact, but
since, for example, the study did not take into account the possible impact of other I precisely what affects each student is so variable that it is difficult.to measure the
kinds of courses-the national scope of the sample, the variety of the effects looked overall impact of any one component part. A particular course, for example, may
for, and the utter absence of effects raise a serious question about the impact of civics have an important effect on two or three students because of their particular interest
courses. in the course, their precise stage of development at the time of the course, or the
One of the pretests reported by Hess and Torney (1967, pp. 8-10) suggests way the class material coincides with parental and peer influence. The same course
 Richard G. Niemi Political Socialization 133
I
may have no effect on the rest of the class. Similarly, a particular teacher, or a most attention of all socialization agents. Several other agents should be mentioned,
particular teaching style, may "turn on" certain students while at best not turning however. There is a considerable literature on these agents, but most of it is not con-
off the rest of the class. Effects may be even more specific than these examples suggest. cerned with their effects on political views of young people.
A particular reading or a portion of a particular reading may have special significance Analysis of peer-group relationships is complicated by the variety of ways of
for one person in the class while for the rest of the class it follows the proverbial path defining and studying them and by the difficulty of carrying out adequate study
in one ear and out the other. I designs. Thus, for example, while sociometric networks of friends have often been
Thus, a particular teacher or a particular set'of lectures or a particular set of I determined, seldom has the next step been undertaken-actually interviewing indi-
readings or a particular class composition each may affect only one or two students 1 viduals and their friends in order to see just how homogeneous the friendship groups
I
in an entire class. Yet considering that during a three- or four-year period of high are and to make inferences about the influence of friends. One such study concluded
school and four years of college a student has a large number of classes, it is fully that, outside the area of partisanship, students are more like their friends than unlike
possible that every single student is affected in a serious way a t least at one or two them and sometimes even resemble their friends more than their parents (Jennings
points during his schooling. The point is that if we measure the effects of individual and Niemi, 1974, ch. 9 ) . Moreover, it is clear that similarity among friends is not
components of schooling, and even of individual classes, we might find very little an artificial product of social background similarities. In addition, the influence of
.effect, even though over the course of four years nearly everyone is affected and sbme friends seems evident from the finding that when friends and parents agree, student-
students deeply moved a number of times. parent similarity is enhanced, while when friends and parents disagree, student-parent
Such a view of the overall impact of the school has at least two consequences. II similarity declines.
The first concerns the way in which we look for school effects in future research. If Other studies (Levin, 1961; Ziblatt, 1965; Langton, 1967) support the notign
I
we continue to look for the impact of a particular kind of teacher, or one particular that informal peer groups are influential during and after adolescence. These studies
class, or one particular method of teaching, and so on, we are likely to continue suggest that the informal milieu is more important than participation in formal group
getting mixed results. OccasionalIy we will come upon a teacher or class or method activities. Studies of nonpolitical subjects (Hollingshead, 1949; Coleman, 1961) sup-
that will have a broad impact on a group of students, but most often we will not. port this conclusion. The studies also indicate that peer influence increases over the
If instead we concentrate on analyzing individual students, trying to see what the adolescent years, although Hirsch (1971, pp. 71-77) has argued that this is not true
effects of all the components of the school were on each of them individually, and in Appalachia because of its matriarchal culture.
only then aggregating across individuals, we might get a good estimate of the overall An interesting cross-cultural perspective on peer influence is provided by two
impact 'of school on young people. We might also be able to specify how many were I
recent studies of child rearing abroad. Bronfenbrenner (1970) studied collectivized
influenced by teachers, for example, even though any one teacher may have influenced child rearing in the Soviet Union. He argues that in this system peers are very
only a few students. An approach stressing the individual would also make fuller use important in teaching and regulating behavior but that at the same time there is
of the gradations in development of children at the same age or grade level. Such more adult direction and leadership than in our own country. Peers are also seen as
variations have received little attention in the past, even though they no doubt help important in Israel's kibbutzim, although again the setting provides adult leadership
account for the variable impact of particular classes and of school curricula generally. and guidance (Bettelheim, 1969). Although these examples may seem quite distant
The second consequence of viewing the schooI in the way suggested is that it from our own culture, a substantial increase in day-care centers in this country would
becomes much more difficult to evaluate curriculum and other school changes since involve some of the same features as described in these two studies. I t will be worth-
it is clear that students will be differentially affected by any proposed changes. Some while for future researchers to pay close attention to the political and social ramifica-
will be positively affected, some negatively, while many will not be affected at all. tions of this kind of more collectivized child rearing. Interestingly, both Bronfen-
Nonetheless, we might in some. ways be in an improved position. We might, for brenner and Bettelheim conclude that there are valuable aspects of raising children
example, have a much improved estimate of what proportion of students was affected in collectivized settings. Whcther day-care centers can and should provide the same
in what way by social studies courses." advantages (without even greater disadvantages) is an important and exciting new
Such a point of view admittedly might reduce or eliminate our ability to area for study.
generalize about the effects of particular components of the school situation. It could In the modern world, where so much of our time both as ~ o u t h sand as adults
likewise serve as an excuse to avoid procedures such as students' ratings of teachers, is taken up by thC mass media, one might expect the media to be a prime socialization
on the grounds that each teacher should be expected to reach only a small proportion agent. The last word on media influence is not yet in. As more and more sophisti-
of his or her students. This is not, of course, what we suggest. But we do feel that cated "packaging", is developed, the potential for massive, unidirectional influence is
viewing the educational process in a more individualized or idiosyncratic fashion I enhanced, Nonetheless, the media so far-instead of changing people's attitudes-
would allow us to estimate more precisely the overall effect of the school. In the seem primarily to reinforce attitudes formed elsewhere (see references in Dawson
long run, it might also help us design better tests of the effects of particular school and Prewitt, 1969, pp. 198-199). Whether this situation will continue is uncertain.
characteristics. While increasingly sophisticated uses of the media may be developed, other factors
Peers, Media, and Events. Families and schools have received by far the may work against massive influence. Ironically, it is sometimes suggested that the
134 Richard G. Niemi Political Socialization 135
I
large-scale development of cable television, with specialized channels for conveying during early childhood. It now seems clear that a significant amount of political
news and public affairs in depth, will erode rather than add to people's political learning continues well into adulthood.
understanding. Presently, when all three major networks broadcast a particularly If the patterns of political development in childhood are quite similar across
newsworthy item, such as the national conventions, people are forced to watch them social groups, however, they are quite different across social groups of adults. Educa-
or to abandon television completely. If at some future date such programs are tion, above all, makes an important difference in the mode of development during
carried only by specialized channels, many people may abandon political viewing in 1
adulthood. To take bbt one example, media usage continues to rise into late adult-
favor of westerns, situation comedies, and old movies. I hood for people with less than an eighth-grade education. The same curve slopes
I

Finally, we would be remiss if we did not point out the significance of political upward more slowly as education increases, so much so that among college-educated
I
events to the socialization process. A commonly cited example is the so-called people there is, if anything, a slight decline after the college years (Converse, 1971;
"Depression generation." The Depression and the images it helped form in people's Jennings and Niemi, 1974, ch. 10).
minds played a dominant role in shaping American political party competition over The greatest amount of changemseems to take place relatively early in adult
the ensuing generation. And perhaps we now have a "Vietnam generation" of life, but it is surprising to us how much change continues into middle adulthood
people who are more alienated, cynical, and uninterested in conventional party and even to retirement age (A. Campbell and others, 1960, p. 162; Easton and
politics than the previous generation. There is' no doubt that a major event affects Dennis, 1969, ch. 14; Jennings and Niemi, 1974). What happens after retirement
the people who live through it--especially the people who are coming of age during is only now being properly studied, with the increased awareness of and interest in
that event (Beck, 1974). One readily observes, for example, the increasing proportion I what happens to older people generally. Whether people graduany disengage from
political life or maintain their interest and awareness is a subject of some debate
of Democrats among people who were in their twenties during the decade of the I
I
1930s (A. Campbell and others, 1960, p. 154). Similarly, differences between politi- (Cummings and Henry, 1961; Glenn and Grimes, 1968; Glenn, 1969; Verba and
cal "generations" are readily observed in contemporary Britain (Butler and Stokes, 1 Nie, 1972).
1969, ch. 11). I It is clear, then, that adult attitudes are not completely determined by child-
What is needed now are1 more precise measurements of just what constitutes hood development. Of what significance, therefore, is early learning? There is, as one
a generation. For example, just how different are the attitudes of youths who came would sqspect, not nearly enough evidence about this point, but the data that do
of age in the early 1960s and the late 1960s? How homogeneous are members of the exist support the notion that early learning does influence later development in
Depression or the Vietnam generation? And especially, how and when do the effects important ways.
I First, however, we must be clear about exactly what the question is that we
of an event lose their impact? For example, to what extent will those who grow up in
are asking. Often the question is phrased "Do early or later experiences have a
the 1970s be part of the Vietnam generation? Presiunably they cannot escape some
greater impact on the attitudes and behavior of adults?" Almond and Verba's (1963,
of the cynicism characteristic of the late 1960s, but the events of those days will
ch. 12) comparison of family and the school versus job experiences suggests this
probably have less impact than on individuals who lived through them. But just how
formulation. The same point seems to be raised by the Prewitt, Eulau, and Zisk
to measure these differences has not yet been determined. Cohort analysis (Evan,
(1966-1967) study of whether early socialization or later political events make the
1959; Crittenden, 1962; Glenn, 1970; N. Cutler, 1970; Klecka, 1971), a relatively
Teater difference in legislators' role orientations. I t is to be expected, I feel, that
new type of study, may provide us with a far better understanding of the extent and
orrelations between present attitudes and those a short time ago (or related present
the precise nature of the impact of events on the development of political attitudes. ttitudes or experiences) should be higher than correlations between present attitudes
Clearly, this is a fruitful area for future study and for the development of new I and attitudes from, say, twenty years ago. If this is what we mean by the attitudes
measures and methods of analysis. of twenty years ago being less important, then this formulation and method of study
would be appropriate. But such a result does not seem to me to mean that attitudes
Early Learning and Adult Political Life learned twenty years ago are unimportant or of no consequence.
We might more appropriately ask how high or low the correlation is between
To this point we have talked almost exclusively about preadult experiences. present attitudes and those of twenty years earlier. Only if we continually find very
Now we must confront directly the question of how relevant these experiences are for low or nearly zero correlations between early and later attitudes can we argue that
adult political attitudes and behavior. There are at least two ways of approaching this socialization has no effect or is unimportant. (Even then we might still want to look
problem. One way is to ask whether there is late as well as early learning or whether for connections between early, basic values and later attitudes.) From this perspective
most political learning1 takes place prior to, say, the adolescent years. I would not want to argue that early attitudes are less important than attitudes or
A few years ago it was reasonable to argue that almost all learning takes place events more closely related to the present time, although they are clearly less adequate
predictors of current attitudes or behavior.
1 We will not deal here with socialization into particular roles, such as that of legis-
lator (Barber, 1965) and a variety of nonpolitical bccupations (see the references and OnIy a few true panel studies exist (studies in which the same respondents
discussion in Brim and Wheeler, 1966). were interviewed at more than one point in time) from which we can determine the
136 Richard G. Niemi Political Socialization 137
similarity of attitudes at widely spaced intervals in time. Bloom (1965) summarizes Niemi, 1974, ch. 12). The most relevant points can be summarized in the form of
a number of such studies. Unfortunately for our purposes, most of these studies deal several general conclusions.
with nonpolitical topics, although one deals with a "conservatism-radicalism" meas- I

I There is a large potential for malleability in young people's views of politics.

ure. Of five measurements over a fourteen- to twenty-year period beginning in early Note that we say there is a large potential for change. Very often, as we argued,
I
adulthood, the correlations were .22, .38, .53, .57, and .60. Obviously these are all there is in fact a significant degree of similarity between attitudes of individuals at
of a sufficient magnitude to support the notion that what is learned by early adult-
hood acts as a significant constraint on subsequent developments. 1 one point in time and a t a much later point in time. Yet many attitudes, even parti-

The most extensive long-term, directly political panel study was carried out I san feelings, are not so firmly fixed that major changes are impossible. Indeed, the
degree of change that has been observed in adults suggests that at least well into
in the early 1960s by Newcomb and his associates (1967). Earlier we noted New- adulthood attitudes are not so firmly entrenched as has often been suggested. But
comb's study of the tremendous affects of Bennington College on its students in the particularly among young people, attitudes can change to a remarkable degree in a
1930s. In the early 1960s Newcomb located most of the women who were in the remarkably short period of time under the proper stimulus.
original sample and reinterviewed them. He then tested to see whether those who A major part of our understanding of this malleability in young people rests
had been converted to a more liberal attitude in college had now reverted to their on the finding that children are often like their parents in the aggregate but not like
precollege views. Such a test is by no means trivial, for it is clear that what was their own parents. In numerous instances it was found that young people as a group
liberalism and conservatism in the 1930s would not have the same connotation today. were similar to parents as a group (or varied in ways wholly consistent with life cycle
And, of course, one could not repeat precisely the same questions that were used in changes, such as having less cynical views). Yet even in these instances, the correla-
the earlier study. Nonetheless, Newcomb shows convincingly that Bennington did have tions between student and parent attitudes were low, suggesting that there were many
a long-term effect on the women who attended it in the 1930s. Interestingly, New- compensating changes; some children were more liberal or more cynical or more
comb found that a surprising number of Bennington graduates remained in liberal cosmopolitan than their parents, while others were more conservative, more trusting,
social circles. This is perhaps another illustration of the effects of their having been or more locally oriented. Hence, even when young children as a whole look very
converted. That is, they may have deliberately chosen liberal friends, hence finding much like the older generation, we cannot conclude that most children are following
support in their adult years for their newly found liberal views. in the footsteps of their own parents, with little possibility of change. T o a greater
The pieces of evidence cited are, of course, far from adequate. In particular, extent than has been realized, children are forming their own political views. Thus,
none of the studies examines the continuity of attitudes from early or middle child- when social and political events capture their interest, young people are sufficiently
hood to later years or over the critical period of late adolescence to early adulthood. free of parental influence that rapid changes in their attitudes are possible.
Fortunately, a study of the later period is now beginning (under the direction of Children are often like their parents, we would argue, simply because of
Jennings and Niemi). Nonetheless, in spite of the need for additional research, the inertia. If there is nothing to push a child in one direction or another, it is easiest
presently available evidence strongly suggests that socialization at one point in time to be like his parents or at least like adults as a whole.. Admittedly, it is difficult to
does have long-term effects. The assumption that early learning is an important predict when events will come along that will move young people. Also just what
factor in the developmental psychology of attitudes does receive support. these events will be is perhaps unpredictable. During the 1950s, for example, "the
bomb" seemingly had the potential for stimulating a radical alteration in young
PoIitical Socialization ahd Social Change people's views, and yet it failed to do so. What we might do is to speak of prob-
abilities of marked attitudinal changes. Presently we cannot, and perhaps we will
Social events and political socialization research combined in the 1960s to I
never be able to, predict the exact occurrence of unstabiliiing events. And yet the
create an intriguing intelIectua1 puzzle. On the one hand, the point was made that present socialization system in the United States arranges for what we regard as a
if young children had benevolent attitudes about political authority during the late I
relatively large potential for this kind of change.
1950s and early 1960s, then why did these same individuals react strongly against This is not to make a value judgment about the present socialization system.
authority in the middle to late 1960s? Similarly, if we adopt the traditional view that Some would prefer more potential for change while others would prefer less. Again,
children are much like their parents, then why was there a rejection of traditional the critical point is to emphasize the part played by the socialization process in
values during the 1960s? On the other hand, if we accept at least tentatively the determining levels of, or perhaps we should say probabilities of, stability versus
point of view that children are not much like their parents, then why is there some-
change.
times little rejection of parental values, as in the 1950s? Or to put it all in a single I t should be obvious that none of what we have said denies the influence of
question, can the process of political socialization explain both continuity and change I families, schools, and other socializing agents. Families, schools, and most other
in the social world? agents do channel children along the same paths as adults; and, as we noted, in the
The overview of the political socialization process that we have provided in I absence of powerful forces children will turn out to be like their parents, at least in
this chapter gives some tentative answers to these questions (see also Jennings and the aggregate. Thus, very often stability will mark the passage from one generation to
138 Richard G. Niemi

another. And even when change and conflict are the order of the day, as in the
middle to late 1960s, the instability would be even greater were it not for family and
school influence.
What we have learned of the political socialization process thus contributes to
our understanding of the likelihood of stability versus change and, more important,
provides us with insight into the mechanisms underlying that balance. Tipping the
balance toward greater change or stability involves ethical and normative questions,
which go well beyond the scope of empirical political research. But such tough
questions will have to be faced as we gain more and more knowledge about how to
control our social as well as our physicaI environment. Political socialization research
should allow us to face these problems with at least some knowledge of the conse-
quences of the process, the rate, and the sources of political learning.
AUTHORITARIAN
PERSONALITY IN
I
CONTEMPORARY
PERSPECTIVE

T h e day after Hitler came to power in

Germany, Max Horkheimer, director of the Institute of Social Research in Frankfurt,
moved from his home in the suburbs to an apartment close to the railway station.
Within a few days he and other members of the institute staff were in Switzerland,
where, against the likelihood of a Nazi takeover, they had set up a branch of the
institute. How did it happen that these people were able to see the implications of
what was happening and to take appropriate steps when most other German
intellectuals and Jews clung to the belief that catastrophe could still be averted?
(Psychoanalysts in Vienna were still carrying on their practices as usual, believing
I
The preparation of this chapter was supported by a grant (K05-MH 12892-0151)
I from the National Institute of Mental Health.
140 Nevitt Sanford Authoritarian Personality in Contemporary Perspective 141.
that nothing serious would happen to them.) The reason was that they had carried accomplished what it did because of its way of combining psychoanalytic theory
out studies of political outlook in Germany and arrived at a very striking finding- and American social psychology;. more precisely, its authors found ways. to test
namely, that whereas a great majority of working-class Germans expressed themselves psychodynamic hypotheses through the use of interviews, attitude scales, projective
as staunchly Social Democratic when asked polling-type questions, a significant num- techniques, and statistical methods.
ber revealed themselves as highly authoritarian when asked questions that encouraged In this chapter I want to summarize the general approach and the major
them to express indirectly some of their underlying attitudes and value orientations findings of T h e Authoritarian Personality, paying particular attention to those por-
(questions that nowadays would be called "projective"). The conclusion of the insti- tions that seek to convey something of the complexity of the personality processes
tute staff was that Hitler probably would come to power and that if he did there forming the authoritarian pattern. This should prepare the ground for my discussion
would be no effective opposition from German labor. of how I see our work today and how my thinking about personality and ideology has
This is probably one of the few cases in recorded history where social scientists changed since 1950. I shall also attempt to evaluate some of the research carried out
were willing to stake everything on the truth of their research findings. But that is by others since our work was published, and suggest what ought to be done now.
not the point here. The point is that much was known about authoritarianism and Our studies did not begin as an inquiry into authoritarianism or as an effort
the other emotional appeals of fascism, much had been written about an "authori- to test any specific hypothesis. They began, in 1943, when Daniel Levinson and I
tarian character structure," before my colleagues and I started the series of investiga- accepted a proffered donation of $500 for a study of anti-Semitism. We thought the
tions that led to our publishing T h e Authoritarian Personality (Adorno, Frenkel- place to begin was with the construction of a scale for measuring this form of
Brunswik, Levinson, and Sanford, 1950). In America, probably the best-known work prejudice. Scores on this instrument could then be correlated with a variety of
on this general subject in the early 1940s was Erich Fromm's (1941) Escape from personality.and sociological factors. As devotees of psychoanalytic theory, we assumed
Freedom. In this book Fromm, a former member of the Institute of Social Research, the wholeness of personality and supposed that anti-Semitism has sources deep
set forth his theory of the "sado-masochistic characterJyin relation to modern totali- within the personality. And we were guided in part by psychoanalytic theory in
tarianism. His basic notion was that ever since the breaking up of the established composing items for the scale.
authority of the Middle Ages, ever since the demise of what Walter Lippmann later We carried out this study with the thought that it had to be complete in and
called "the light and the leading" of the medieval church-state, men of the Western of itself. Soon, however, the renewal of our grant, and further funding at a slightly
world have been ambivalent about the burdens of freedom and dignity, and their higher level, made it possible for us to devise a methodology for the systematic study
often unconscious wish to escape from these burdens renders them in various degrees of anti-Semitism and personality. Else Frenkel-Brunswik and Suzanne Reichard now
susceptible to totalitarian propaganda. The individual in whom this susceptibility is joined Levinson and me, and we all conducted interviews with subjects who had
strong has an emotional need, often unconscious, to submit to authority-in the first obtained extreme scores on the anti-Semitism (A-S) scale. These interviews, which
instance parental figures; at the same time, he has unconscious impulses of aggression ranged over wide areas of ideology and personal background, became the major
toward these same figures, which impulses are often redirected toward outgroups such source of our knowledge of the differences between people scoring high and people
as ethnic minorities. Fromm thought of this psychodynamic structure as sado- scoring low on our scale, and of the personality structure of people who seemed
masochism; but realizing that this term was commonly used in reference to perversion I particularly susceptible to anti-Semitic and fascist propaganda. The interviews were
and neurosis, he used the term authoritarian character to refer to the same kind of
structure in ordinary people.
I supplemented by the Thematic Apperception Test, the Rorschach test, and "pro-
jective questions." It was our intention to conceptualize material from all these
Horkheimer and other staff members of the Institute of Social Research had I
sources in such a way that it could be quantified and made the basis of questionnaires
already written up their study of Gennan workers-in Studien iiber Autoritat und which could be administered to large groups of subjects2 I t was our idea, in other
Familie (1936), a volume containing a chapter by Fromm. Other European psy- words, to bring methods of social psychology into the service of concepts and theories
chologists, most notably Reich (1933) and Erikson (1942), were writing about from the dynamic theory of personality; we thereby hoped to make "depth psycho-
fascism and anti-Semitism from a psychoanalytic point of view. Probably Maslow logical" phenomena more amenable to mass-statistical treatment and to make
(1943a) was the first American psychblogist to join Fromm in discussing authori- quantitative surveys of attitudes and opinions more meaningful psychologically.
tarianism and in insisting that social attitudes serve important personality functions, We did not, however, in our quantitative work, go directly from anti-Semitism
although Stagner (1936) had earlier argued that fascist attitudes should be regarded to personality dynamics. Instead, we asked about the relations between what subjects
as parts of an integrated system of outlook, beliefs, and attitudes. thought about Jews and what they thought about other people and other issues. Were
Thus, the contribution of T h e Authoritarian Personality lay not in the new the trends found in anti-Semitic ideology-its generality, stereotyped imagery, de-
ideas it presented but in its empirical demonstration of the coherence of various
beliefs, attitudes, and values associated with anti-Semitism and fascism and of the admit that the weight of evidence is on the side of the propositions or, at the very least, that
functional role of this ideological system within the individual's persona1ity.l I t they are now reasonable. (See Brown, 1965, p. 525.)
2 Funds for this work became available from the American Jewish Committee through
1 "Demonstrate" may be too strong a word. Not all psychologists who have looked the good officesof Max Horkheimer. Our cooperative arrangement with the committee also
into the matter agree that these propositions have been demonstrated. But they would have to made it possible for us to welcome T. W. Adorno to our group.
 142 Nevitt Sanford Authoritarian Personality in Contemporary Perspective

structive irrationality, sense of threat, concern with power and immorality-also hypotheses concerning the structure of authoritarianism also yielded some evidence
expressed in the individual's thinking about group relations and social issues generally? favoring them-though their confirmation awaited the gathering of further data. For
Accordingly, we developed scales for measuring generalized prejudice, or ethno- example, the interviews like some questionnaire material, showed unmistakably that
centrism (E) as we called it, and political-economic conservatism (PEC). a distinguishing feature of the highly ethnocentric subject is his tendency to glorify
The E scale was made up of items pertaining to (a) Negroes and Negro- his parents. And the interviews also gave evidence of ambivalence in this subject's
white relations; (b) other minorities, including not only ethnic groups but minority relationship with his parents. It was usually not long after the statements of glorifica-
political parties and religious sects, foreigners, "Okies," "zoot-suiters," criminals; tion that a note of complaint or self-pity began to creep into the interview. How
and (c) patriotism-items dealing with America as an ingroup in relation to other might one demonstrate that overt glorification of the parents is functionally related
nations as "inferior" outgroups. This scale of thirty-four items had a split-half relia- to underlying hostility toward them? One way would be to use a projective technique
bility of .91, which we took as an indication of a very high degree of generality in to obtain an independent measure of the latter and see if the two vary together.
ethnocentric ideology. The correlation of E with the anti-Semitism scale was .80, Unfortunately, this is not simple. What are the TAT signs of repressed aggression?
which led us to believe that anti-Semitism should be regarded as primarily an aspect Certainly not the frequency and intensity of aggressive actions by heroes of the
of a broad pattern of ethnocentric ideology. Ethnocentrism was found to be related stories. These seem to be, for the most part, indications of aggression that is accepted
to political conservatism, the correlation of E and PEC being .50. by the ego; it was more pronounced in the low scorers than in the high scorers on
Two patterns of conservatism could be distinguished : a traditional laissez-faire the A-S and E scales. But Betty Aron (1950) did conclude that there is more ego-
conservatism and a "pseudoconservatism,)' in which profession of belief in the tenets alien aggression against the parents in the stories of high scorers, the indications being
of traditional conservatism is combined with a readiness for change of a sort that such things as the frequency with which parent figures are the victims of affliction
would destroy the very institutions with which the subject appears to identify him- or death and the frequency and intensity of aggression against parental figures on
self. There was evidence that the pseudoconservatism items were the more largely the part of characters with whom the storyteller is not identified.
responsible for the correlation between E and conservatism. The concept of pseudo- Such conclusions as these still call for independent validation, and the same
conservatism has been followed up and amplified by Hofstadter (1965). hoIds for the projective questions. The material elicited by this procedure is for the
On the basis of our work with the three scales and the clinical procedures most part on the same level of personality as scale responses. Responses to the open-
noted above, we proceeded to formulate the psychodynamic structure which we ended questions could easily be-and they sometimes were-translated into scale
thought is expressed in ethnocentrism and in various opinions and attitudes associated items. Thus, the projective questions yielded a large amount of material that in-
with it. Since we assumed that action always depends upon the situation the person is dependently confirmed scale findings on the difference between ethnocentric and
in as well as upon personality, we referred to this structure as a potential for fascism, nonethnocentric subjects. But, more than this, the material from the projective
a susceptibility to anti-Semitic propaganda, a readiness to participate in antidemo- questions called for interpretation, for the conceptualization of underlying trends that
cratic social movements. In the remainder of the chapter I call this structure would explain the pattern of overt expression.
ccauthoritarianism,"which is in keeping with current practice. It is worth noting, Two of the projective questions read as follows: (1) "We all have times when
however, that in our work, in our early publications, and in the writing of The we feel below par. What moods or feelings are the most important or disturbing to
Authoritarian Personality we did not use this terminology. In our search for the you?" (2) "There is hardly a person who hasn't said to himself: 'If this keeps up
correlates of anti-Semitism, we wanted to allow for the possibility that they em- I'll go nuts.' What might drive a person nuts?" These two questions, like the six
braced more than, or were constituted of something different from, the "authoritarian others used, brought out numerous differences between high and low scorers on the
character structure" of Fromm ( 1941) and Maslow ( 1943). I t was not until .our E scale. The "lows" are most disturbed by conscious conflict and guilt feelings,
book was finished that we thought of its title, which was intended to indicate our frustrations of love and dependence, consciousness of hostility toward loved objects;
indebtedness to earlier writers and our concern with a pattern of personaliky orga- they suppose that people are "driven nuts" by inner psychological states or by a
nization. dominating environment. The "highs," on the other hand, are most disturbed by
I t must be understood, however, that I am here discussing what nowadays is violations of conventional values by self or others or by a threatening or nonsupport-
called "right-wing authoritarianism." As will be pointed out below, some of OW ing environment; they are also more disturbed by, and state that people are "driven
subjects who obtained extremely low scores on the A-S or the E scale, particularly nutsy' by, what Levinson called "rumblings from belowv-intimations that suppressed
the "rigid" low scores, were observed to have something in common with the extreme passivity, or anxiety, or hostility might break into the open.
high scores. In the years since 1950 a great deal of attention has been focused on
"authoritarianism of the left'' (see, for example, Rakeach, 1960), and much has been Authoritarian Personality Dispositions
learned about certain characteristics that people at the two ends of the political spec- The major personality dispositions which, we concluded, largely make up
trum have in common. The focus here, however, is not on these commonalities but authoritarianism may now be discussed in turn, with attention to their organization
on the distinctiveness and coherence of attitudes, beliefs, and values usually found on and their reIation to anti-Semitism and ethnocentrism. This set of related tendencies
the political right. will then be examined in the context of a more general psychodynamic theory of
The interviews and projective techniques which yielded the major specific personality.
144 Nevitt Sanford Authoritarian Personality in Contemporary Perspective 145

In our conversations with anti-Semitic subjects, we observed that most of their This narrowness of consciousness is cIosely linked to two other tendencies
accusations against Jews were couched in conventionally moralistic terms. This theme noted in highly prejudiced individuals: superstition and stereotypy. Superstition
was also pronounced in the original A-S scale items. These subjects, we believed, indicates a tendency to shift responsibility from within the individual onto outside
were revealing not so much bad experiences with Jews or an adaptation to a general forces beyond his control; these forces then appear to the individual as mystical or
climate of pinion as a need to adhere strictly to conventional, middle-class values, fantastic determinants of his fate. Stereotypy is the tendency to think in rigid, over-
a disposition to feel anxious at the sight of or the thought of any violation of these simplified categories, in unambiguous terms of black and white, particularly in the
values-something that could be attributed to instability in the individual's own realm of psychological or social matters. We hypothesized that some people, even
value system. This disposition we called conventionalism. The term refers not merely those who are otherwise "inteIligentJWmay resort to primitive explanations of human
to conformity with middle-class values but to rigid adherence to such values, to an events at least partly because they cannot allow many of the ideas and observations
overaccent upon them, and to responsiveness to contemporary external social pres- needed for an adequate account to enter into their calculations; because these ideas
sure. are affect-laden and potentially anxiety-producing, they cannot be included in the
Submission to authority, desire for a strong leader, subservience of the conscious scheme of things. The assumption here is, of course, that many of the
individual to the state had for some time been put .forward as important aspects of common phenomena of prejudice are superstitions or stereotypes.
the Nazi creed. It was thus not surprising that these themes were prominent in our As suggested above, when an individual is forced to submit to powers or
interviews with highly prejudiced subjects. Authoritarian submission, as we termed agencies with which he is not fully in sympathy, he is left with a nagging sense of
the hypothetical generalized disposition of personality, was conceived of not as a weakness. Since to admit such weakness is to damage self-respect, he makes every
balanced, realistic respect for valid authority but as an exaggerated, emotional need effort to deny it-sometimes by projecting weakness onto outgroups (according to
to submit. Here, as with conventionalism, the individual is assumed to be oriented the formula "I am not weak; they are") or by using the mechanism of overcompen-
toward external powers or agencies of control rather than under the direction of a sation, whereby he seeks to present to the world an aspect of power and toughness.
conscience of his own. This "power complex" contains elements that are essentially contradictory. The
Authoritarian submission is closely related, conceptually, to authoritarian power-centered individual wants to have power but at the same time is afraid to
aggression. Both attitudes, according to theory, spring from underlying hostility seize it and wield it. He also admires power in others, and is inclined to submit to it,
toward ingroup authorities, originally the parents. The individual strives to keep this but at the same time is afraid of the weakness thus implied. A common solutionifor
hostility in check by overdoing in the direction of respect, obedience, and gratitude ,uch a person is to align himself with power figures, thus gratifying both his need to
toward the ingroup authorities and by displacing the underlying hostility toward lave power and his need to submit. By submitting to power, he can still somehow
these authorities onto outgroups. This is the most essential connection between author- ~articipatein it.
itarian aggression and ethnocentrism. But it appears that the tendency to displace Although authoritarian aggression provides a very broad channel for the
hostility is more general than that seen in the common forms of prejudice; the xpression of underlying hostile impulses, many of our prejudiced subjects seemed to
greatest variety of people and actions are likely to become objects of condemnation. leed still other channels. Possibly because of numerous externally imposed restrictions
Moreover, the kinds of things for which the individual would punish other people upon the satisfaction of their needs, they seemed to harbor a great deal of generalized
are the same as those for which he himself was punished or for which he feels in hostility, which would come into the open whenever it could be justified or ration-
his heart he deserves to be punished, in as much as he would like, unconsciously, to alized. Such rationalized, ego-accepted aggression (not including authoritarian
do these things himself. aggression) we termed destructiveness and cynicism. The cynic, in our terminology,
In addition, our highly prejudiced subjects showed, both in the interviews rationalizes his aggressiveness by attributing a similar aggressiveness to everybody
and in some of the A-S and E scale items with which they heartily agreed, that they else. In his view, it is "human nature" to exploit and to make war on one's neighbors.
disapproved of a free emotional life, of the intellectual or theoretical, of the im- It seemed a fairly safe assumption that minority groups were often the victims of this
practical. They tended to attribute these characteristics to their outgroups. And there undifferentiated aggressiveness.
was theory at hand to explain the relations of these attitudes to prejudice and to the The mechanism just described (and often alluded to in our discussion of
personality trends just discussed. The individual who had had to repress hostility authoritarian aggression and of superstition) is, of course, a form of projection.
against his parents and others who appeared to be strong-and who was thus forced Indeed, projection has a crucial role in the whole theory of prejudice as a means for
into submissiveness, which impaired his self-respect-would naturally be required to keeping the individual's psychological household in some sort of order. The most
maintain a narrow range of consciousness. Self-awareness might threaten his whole essential notion is that impulses which cannot be admitted to the conscious ego tend
scheme of adjustment. He would be afraid of genuine feeling because his emotions to be projected onto minority groups. We conceived of projectivity as a general
might get out of control, afraid of thinking about human phenomena because he feature of the personality, a feature that can be considered independently of the
might think the "wrong" thoughts. The term anti-intraception, borrowed from object onto which the projection is made. It is manifested mainly as preoccupation
Murray (1938), describes such an individual-his general attitude of impatience with "evil forces" in the world, with plots and conspiracies, germs, sexual excesses.
with and opposition to feelings, fantasies, speculations, and other subjective or (For a more detailed discussion of projection, see Chapter Fifteen.)
L'tender-minded''phenomena. Concern with sex seemed to deserve special consideration, since our prejudiced
146 Nevitt Sanford Authoritarian Personality in Contemporary Perspective 147

subjects showed marked inhibitions in this sphere, and moral indignation over the offering a somewhat stereotyped and idealized picture at the beginning of the inter-
sexual behavior of other people. Sexual immorality was one of the many violations view and allowing negative features to make their appearance only when there was
of conventional values which they attributed to minority groups. This ego-alien questioning about details. In contrast, low scorers more often undertook an objective
sexuality is another characteristic of the typical prejudiced person. appraisal, with good and bad features mentioned in their place. High scorers tended
We assumed, then, that these were the major personality dispositions under- to deny that there was any conflict between the parents; the lows usually described
lying overt anti-Semitism and ethnocentrism. We still needed to determine how some conflict in more or less realistic terms. High-scoring men usually described their
these dispositions relate to one another and to the structure of personality as a whole. homes as dominated by the father; low-scoring men more often described homes in
In our theoretical work on these questions, we leaned heavily on the Freudian con- which there was general orientation toward the mother.
cepts of superego, ego, and id. We considered that these features of the personality Discipline in the families of the more authoritarian men and women was
have characteristic modes of functioning in the highly ethnocentric subject. Speci- characterized by relatively harsh application of rules, in accordance with conventional
fically, the superego is strict, rigid, and relatively externalized; the id is strong, values; and this discipline was commonly experienced as threatening or traumatic or
primitive, and ego-alien; the ego is weak and can manage the superego-id conflicts even overwhelming. I n the families of subjects low on authoritarianism, on the other
only by resorting to rather desperate defenses. According to this formulation, the first hand, discipline was more often for the violation of principles, and the parents more
three of .the personality dispositions just considered (conventionalism, authoritarian often made an effort to explain the issues to the child, thus enabling him to assimilate
submission, and authoritarian aggression) all have to do with superego functioning. the discipline.
The strict superego demands-as a means for .keeping unacceptable impulses in In view of the more authoritarian subject's obvious inclination to put as good
check-punishment of "wrongdoers" in the name of those authorities to whom the a face as possible upon his family and his childhood situation, we were inclined to
subject has submitted. Anti-intraception, superstition and stereotypy, and pro- assume that such negative features as appeared in his account were probably to be
jectivity may be regarded as manifestations of a relatively weak ego. Anti-intracep- taken more or less at their face value; that is, to believe that the high authoritarians
tion involves the primitive defense mechanisms of repression and denial. Superstition came, for the most part, from homes in which a rather stern and distant father
shows an inclination to shift responsibility onto the external world, as if the ego were dominated a submissive and long-suffering but morally restrictive mother, and in
giving up its attempts to predict and control, while stereotypy is an attempt to deal .
which discipline was an attempt to apply conventionally approved rules rather than
with complex events by means of oversimplified categories. Projectivity is the con- an effort to further general values in accordance with the perceived needs of the
sistent use of another relatively primitive mechanism of defense. Power and toughness, child. i
another manifestation of ego weakness, involves an overaccent upon the conven- This account of the genesis of authoritarianism in the individual personality
tionalized aspects of the ego (for instance, the emphasis on "will power") ; but this emphasizes early experiences in the family. It may well be asked, "What makes
variable, like destructiveness and cynicism and sex, also expresses with a minimum parents behave in ways that promote authoritarianism in their children?" We cannot
of indirectness the activity of id tendencies. merely say that they are authoritarian themselves, for this would be pushing the
A major hypothesis guiding our investigations into the origins of authoritar- question of the origins of authoritarianism indefinitely into the past. Instead, we must
ianism was that such central structures of personality have their beginnings in ex- consider that family life, within which personality largely develops, is constantly
periences of early childhood. By taking the many differences between the reports on under the influence of various economic, social, and historical processes. We carried
childhood of subjects scoring high and those scoring low on the A-S and E scales, out no studies of such processes, as they might have affected the parents of our
and considering these in the light of contemporary knowledge and theory of per- subjects, but we did offer some speculations; for example, that middle-class parents
sonality development, we were able to put together a plausible account. who have climbed rapidly and feel insecure about their new status are disposed to
I t may be helpful at this point to sketch very briefly the contrasting accounts discipline their children in ways that favor the children's dcvelopment of authori-
of childhood by high and Iow-scoring subjects. High-scoring men more often de- tarian tendencies. That is, people may consistently behave in authoritarian ways, or
scribed their fathers as distant and stern; the low-scorers as relaxed and mild. High- even develop authoritarian trends in their personalities, because of the social environ-
scoring women characteristically saw their fathers as hard-working and serious, while ment of the present as well as the past.
low-scoring women more often perceived their fathers as intellectual and easygoing. If contemporary social situations and processes help determine the behavior
The mothers of high-scoring subjects, both male and female, were more often said of parents toward their children, might they not also help determine the behavior of
to be kind, self-sacrificing, and submissive; the mothers of low-scoring subjects were these parents toward outgroups? Many social scientists have thought so. Indeed, at
more often described as warm, sociable, and understanding. High-scoring men the time we began our work contemporary economic and social factors and processes
tended to accent the mother's moral restrictiveness; low-scoring men, her intellectual were generally considered the major determinants of ethnic prejudice. More inward,
and aesthetic interests. High-scoring women generally described their mothers as more individualistic factors, we thought, were rather neglected; and we were deter-
models of morality, restricting and fearsome; low-scoring women usually were able mined to give them their due. We assumed, however, that in the determination of
to offer more differentiated criticism. ideology, as in the determination of all observable actions, there are situational as
In general, high scorers gave a rather undifferentiated picture of their parents, . well as personality factors at work and that explanation requires the careful weighing
 148 Nevitt Sanford Authoritarian Personality in Contemporary Perspective 149

of the role of each kind of factor. Accordingly, we studied anti-Semitism or ethno- ethnocentrism. We had to leave to future investigators the question of what would
centrism in relation to income and occupation; to membership ,in political, religious, be the relationship between E and intelligence test scores in samples with wider ranges
and other social groups; and, of course, to those favorites of the social scientist, of IQ, and the question of how, ultimately, the interactions of native intelligence
education and intelligence. Our methods, however, were not well calculated to give and psychodynamic processes were to be formulated.
these factors their due, mainly because our 2099 subjects were drawn almost ex- Intelligence, of course, is correlated with number of years of formal education;
clusively from the middle socioeconomic class, and were mainly white, non-Jewish, not surprisingly, therefore, .we found a slight negative correlation between ethno-
and native born. centrism and amount of education. Since relatively few of our subjects had been to
We found no simple relationships between ethnocentrism and socioeconomic school less than eleven years, our samples were biased in the direction of higher levels
factors or social group memberships. There was only a slight tendency for higher of education. There was great variability, however, among subjects of the same edu-
incomes on the part of the subject's father to be associated with lower scores on the cational level: being a highly intelligent college graduate is no guarantee against
E scale. The offspring of Republicans were slightly less ethnocentric than the off- ethnocentrism.
spring of Democrats, the difference being significant at the 5 percent level of confi-
dence, Subjects who had a political preference different from that of their fathers, F Scale
however, were much lower on the E scale than were subjects whose political pref-
erence was the same as their fathers'. This permitted the kind of psychological inter- After the empirical and theoretical work described above had been done, we
pretation we favored: having the same political preference as one's father may be conceived the idea of constructing a scale for measuring potential fascism in the
an indication of submissiveness to ingroup authority, which favors ethnocentrism; personality. We wanted not only a scale for measuring prejudice without mentioning
having a different preference is a sign of critical independence, which favors freedom the names of any ethnic minorities but a means for quantifying the fascist potential
from ethnocentrism. l
I
and estimating its strength in various groups of subjects. Our basic idea was that the
The story was much the same with religious group memberships. The religious personality dispositions expressed in anti-Semitism and ethnocentrism would be
denomination of parents did not prove to be significant for ethnocentrism. When the expressed in various other ways as well. For example, if a subject's tendency to
religious affiliation of the subject was considered in relation to his or her parents, attribute weakness to Jews springs from his own underlying fear of weakness, that
however, interesting results were obtained. Subjects whose parents had the same fear might also express itself in an overaccent upon his own strength and toughness;
religion were higher on E than subjects whose parents had different religions or 1 hence we considered the item "An insult to our honor should always be punished"
differed in the sense that one was -religious and the other not. Also, subjects who. I mainly an expression of the "power and toughness" dimension discussed above.
I
agreed with their mothers in the matter of religion were more ethnocentric than those For every item in the new scale, which we called the F (for prefascism)
who did not agree with their mothers. In sum, for religious subjects mere membership scale, there was a hypothesis (sometimes several hypotheses) about its connection
in a religious group is not as significant for ethnocentrism as the way in which with prejudice. The major source of these hypotheses was, of course, the studies that
religion is accepted or rejected: if its acceptance reflects conventional or authoritarian had gone before, particularly material from the interviews and the Thematic Apper-
submission, we may expect to find some ethnocentrism; if its acceptance reflects the I
I
ception Test; but, at the same time, we made use of the general literature, both
subject's own experience and independent thought, we may expect relative freedom empirical and theoretical, on anti-Semitism and fascism. Once a hypothesis of this
1 kind had been formulated, a preliminary sketch for an item was not far to seek:
from ethnocentrism.
Intelligence and education must be considered as factors in authoritarianism, something an interviewee had said, a phrase from the daily newspaper, a fragment
even though we were not able to study them thoroughly or systematically. Scores on of ordinary conversation. The item was then improved or made maximally useful
one or another of the standard intelligence tests were available for several groups, to our purposes through successive trials with subjects.
totalling 560, of our subjects, the great majority of whom had IQs of 100 or more. Although the items could be classified as mainly expressive of one or another
There was a low but dependable relationship between low intelligence and ethno- of our dimensions of authoritarianism in personality, most of the items were con-
centrism. I t was hard to tell which to accept more, the lowness or the dependability. sidered manifestations of more than one of these dimensions. Thus, in the item "An
I t has often been remarked that prejudice is a form of stupidity in social matters. insult to our honor should always be punished," the idea of punishment, which
But does stupidity cause authoritarianism, or does authoritarianism cause stupidity? would not have been necessary to an expression of overaccent on honor, was included
We believed that suggestibility and lack of discernment would favor acceptance of because it was thought to express authoritarian aggression. Again, the item "Sex
our E items, but we were inclined to see these traits as aspects of the total personality. crimes, such as rape and attacks on children, deserve more than mere imprisonment;
I t seemed hardly likely that subjects could free themselves from such psychodynami- such criminals ought to be publicly whipped" seems a particularly good example of
determined tendencies as rigidity, concreteness and stereotypy of thinking, authoritarian aggression, but we thought of it as a combination of this disposition
narrowness of ego bounds, and awkwardness in the face of psychosocial phenomena and our "defense against sex" dimension. This way of proceeding was in keeping
when they sat down to take intelligence tests. On the other hand, we thought of with our notion that the several personality dispositions we had conceptionalized
incompetence 'as a major source of low self-esteem and aggression and, hence, of were dynamically interrelated, and largely constituted a structure-or "syndrome,"
150 Nevitt Sanford
Authoritarian Personality in Contemporary Perspective 151
as we called it. We wanted to develop a measure of the whole syndrome, and to do
other. The items we had grouped together as mainly expressive of a given personality
so in an economical way. To this end we packed as much meaning as we could into
disposition tended to be intercorrelated (.I1 to .24), but the items of one group
each item, trying to design it so that as many as possible of the facets of authoritar-
correlated with one another no better than they did with numerous items from other
ianism would be expressed. clusters.
We used the above-mentioned theory concerning authoritarianism and the
This last finding, taken together with the fairly high internal consistency of
total personality in the same way. According to theory the superego, the ego, and the
the scale as a whole, seemed sufficient justification for our thinking that the F scale
id can be separated only by abstraction. In actuality, the functioning of any one of
measures one thing-in statistical terms a very general factor, which varies in amounts
these agencies depends at any moment upon the activities of the other two; expressed
from one individual to another. At the same time, however, we thought of this one
attitudes and values are not readily classifiable as manifestations of superego, ego, or
thing as something highly complex-a syndrome, as we preferred to say. A syndrome
id but are to be understood as expressions of the relationships among these agencies.
is a pattern of dynamically related variables; in the ideal case a change in one
Consider the item "He is indeed contemptible who does not feel an undying love,
variable will bring change in all the others. Where a syndrome is fairly common in
gratitude, and respect for his parents." On the surface, this item expresses authori-
a population of individuals, its unitary character will be reflected in intercorrelations
tarian aggression and authoritarian submission and, hence, might be classified as
among its constituent variables.
primarily a superego item. But the theory was that agreement with this extreme
We could not say that all the variables that make up the authoritarian
statement might well mask an underlying hostility toward the parents. In other
syndrome are touched upon in the F scale; nor could we say with certainty which
words, we hypothesized that unconscious hostility toward the parents is a distin-
variables are the truly fundamental ones. It seemed clear, however (as will be docu-
guishing feature of the highly ethnocentric person; our problem was to determine
mented below), that we had a measure that could make some highly significant
how this tendency might give itself away in an attitude scale. One answer was
differentiations among individuals.
through signs of a reaction formation, this mechanism being a common one in the
According to theory (Murray and others, 1938; Sanford and others, 1943,
highly ethnocentric person. Thus, the present item had to do with an interplay of
pp. 20-22), the variables in a syndrome may be quite loosely held together. A variable
superego, ego, and id; an underlying unconscious, ego-alien tendency, coming mainly
from the id, has led to anxiety of punishment (superego), which the ego seeks to ward
off or reduce by transforming the forbidden tendency into its opposite. But this is
~ which in one individual has a place in a given syndrome may in another individual
appear in a different syndrome; and its nature may be somewhat different, depend-

I
ing on the context in which it is found. And, what is particularly important, syn-
not all. This is merely the authoritarian submission expressed in the item. "He is
dromes themselves, though conceived as distinguishable patterns, have no true
indeed contemptible" is an expression of authoritarian aggression. The ego must, so
independence; their nature too will depend upon the still broader context of per-
to speak, be doubly sure that punishment is avoided, and that the original id tendency I
sonality within which they have their being. Thus, individuals may exhibit the same
finds some sort of gratification; hence, it joins forces with the punitive agency and
syndrome in about the same degree and yet differ among themselves in numerous
imputes the "badness" to other people, who may then be freely aggressed against in
significant ways. No syndrome can ever totally embrace a person. Even when authori-
good conscience.
tarianism is pronounced, what emerges in behavior will depend upon what other
It must be noted also that each item of the F scale is in the positive direction;
I syndromes are present. It is not proper, then, to speak of an authoritarian type of
that is, agreement with it is taken as an indication of potential for fascism. Here we
person. But one may speak of types of authoritarianism. Authoritarianism may vary
followed a procedure that had served us well with the anti-Semitism scale (Levinson
I
from one individual to another, not only in quantity but according to which of the
and Sanford, 1944). Setting out to study susceptibility to anti-Semitic propaganda
constituent variables are relatively pronounced, a matter which may depend upon
in the midst of World War 11, we used items that stated an anti-Semitic position and
what other factors are at work in the personality. Some of these variations in authori-
at the same time offered a "pseudodemocratic facade" or a rationalization for agree-
tarianism may be common in large populations.
.
I
ment: "There may be some exceptions but Jews . ." The question was how far I
Some varieties o! high authoritarianism were labeled, by ~ d o r n o ,surface
would a subject permit himself to be lured into agreement with antidemocratic senti-
resentment, conventional, authoritarian, tough guy, crank, and manipulative. Surfa~e
ments. Interviews with subjects who obtained high or low scores on this scale satisfied
resentment refers not so much to any deep-lying tendency in the personality a$ to a
u s that our instrument was effective in identifying individuals who were relatively
state of affairs in which the individual is provoked to prejudiced and authoritarian
anti-Semitic and those who were the opposite. Items on the F and E scales, which
modes of behavior by externally imposed frustrations. The conventional pattern
were composed in the same way as the original anti-Semitism scale, were those that
emphasizes conventional values and determination by external representatives of
we considered fascistic in content (with varying degrees of explicitness) and that
the superego. The authoritarian pattern is similar to F r o m ' s conception of the sado-
subjects might agree with without violating their democratic self-conceptions.
masochistic character. The subject achieves his social adjustment by taking pleasure
The F scale in its final form contained thirty items and had a split-half in obedience and subordination, while remaining ambivalent toward his authorities.
reliability of .90. The average interitem correlation was .13; the average item-total
In the tough guy the accent, as might be expected, is on power and toughness and
scale correlation, .33. The correlational analysis did not indicate that our hypotheti-
on destructiveness and cynicism. The individual is prepared to do almost anything to
cal dimensions of authoritarianism were to any significant degree independent of each
protect himself against what he perceives to be a hostile world. The outstanding
 Nevitt Sanford Authoritarian Personality in Contemporary Perspective 153

feature of the crank is projectivity, with superstition and stereotypy also looming selves with negative terms like "international bankers," "bloodsuckers," "apostates,)'
large. I t is as if the individual has withdrawn into an inner world, where he con- "verminJJ; but when they said "BIood will have to be shed," their audiences knew
centrates upon self-aggrandizement and the protection of his self-conception by full well whose blood. Thus, scales with items expressing hostile attitudes toward
projective formulas. In the manipulative pattern anti-intraception is extreme. There ethnic minorities not only risked being unfair to these minorities but also might not
is a marked deficiency of object cathexis and of emotional ties. In the extreme case get at the full potential for racism and fas~ism.~
people become objects to be handled, administered, manipulated in accordance with The F scale in its final form, however, seems to have served its purpose fairiy
the subject's theoretical or practical schemes. well. Correlations with the E scale in various groups ranged from .62 to .87 and
The patterns that seemed prominent among subjects low in potential fascism averaged .77. This should not, of course, be surprising, since, as indicated above, the
were labeled rigid, protesting, impulsive, easygoing, and genuine liberal. The rigid E scale was thought of as expressing personality dispositions which, taken together,
low scorer appears to have most in common with the overall high pattern. Here the were supposed to be measured by the F scale. Consistent with this finding is the
absence of potential fascism, instead of being based on concrete experience and F-PEG correlation of .57.
integrated within the personality, is derived from some general external ideological The only data we had on our subjects' politically relevant overt behavior, as
pattern. T o quote Adorno and others (1950, p..772), "We encountered a few sub- distinct from their verbal behavior, was their membership in the groups we ap-
jects who had been identified ideologically with some progressive movement, such proached in order to obtain responses to our questionnaires. As indicated above, we
as the struggle for minority rights, for a long time, but with whom such ideas-con- could not tell whether socioeconomic class as such is a determinant of ethnocentrism,
tained features of compulsiveness, even of paranoid obsession, and who, with respect but there was some evidence that membership in certain groups is an expression,of
to many of our variables, especially rigidity and total thinking, could hardly be an authoritarian outlook. Thus, it seemed consistent with our general theory to find
distinguished from some of our high extremes." In the protesting Iow scorer the that inmates of San Quentin prison obtained the highest mean F scale score of all
decisive feature is opposition to whatever appears to be tyranny. The subject is out the groups we studied, while psychiatric clinic patients obtained a relatively low
to protect the weak from the strong; he can perhaps lead or at least be effective in score. This finding seemed, at the least, to express a difference between those who try
revolts, but finds nothing to do once the revolt has met with success. In the impulsive to overcome frustration or resolve conflicts by striking out against the environment
pattern unconventionality is the outstanding theme. The subject is able not only to or blaming others as opposed to those who seek solutions within themselves. In
sympathize with what is different but to be different. The easygoing pattern is the general, the main difference between our low-scoring groups and our high-scoring ones
opposite of the manipulative form- of potential fascism. I t is marked by imagination, was that the former in some way expressed liberal or progressive or humanitarian
capacity for enjoyment, and a sense of humor that is often directed to oneself. The attitudes or thought; the latter did not. Thus, middle-class men recruited from the
genuine liberal is close to the psychoanaljrtic ideal, representing a balance of superego, university and movie communities of Los Angeles, from the PTA, or from the lay-
ego, and id. Perhaps the outstanding features of this pattern are moral courage and men's league of the Presbyterian church were significantly lower on F than were
a sense of personal autonomy. The subject in whom it is highly developed resists any members of a Berkeley service club, while working-class men who attended classes
interference with his personal convictions and beliefs, and he does not want to at the California Labor School or who belonged to a "militantJ' union were lower
interfere with those of others. than other groups of working-class men.

Correlates of F Scale Research Based on F Scale

If one knows personally individuals in whom one or another of the above Almost as soon as The Authoritarian Personality appeared, it became the
patterns is pronounced, has carried out case studies of them, or has seen them in object of vigorous criticism, both for its methodological shortcomings and for its
intensive psychotherapy, it might easily strike.him as foolhardy to suppose that all I neglect of "authoritarianism of the left."' The studies using questionnaires were said
this complexity could somehow be boiled down and expressed in a thirty-item scale. to be flawed most importantly by "acquiescence response setJJ ( a general tendency to
Although members of our research team had had these experiences, we persisted in acquiescence) and by inadequate sampling, while the efforts to quantify interview
our efforts to make "depth psychological processes amenable to mass-statistical treat-
8 Anti-Semitism is still the giveaway of the most dangerous political orientations. It
ment." The effort seemed worthwhile not only because of the light it might shed on
is prtcisely because of their relative invisibility that Jews have projected onto them the most
personality organization and functioning but because American social scientists irrational destructive fantasies. And anti-Semitism is as hard to get at as ever. For example,
probably would not put much stock in our clinical findings unless we were able to Vice-President Agnew's "impudent snob1' speech aroused a great deal of uneasiness and led
render them in quantitative form. to accusations that he was appealing to people's worst motives, but I could find few people
I t seemed to us particularly important to have an index of anti-Semitism and who were struck by how much it was in the style of the fascist agitators of the 1930s.
4 Useful reviews of T h e Authoritarian Personality and of research on authoritarianism,
racism that made no reference to Jews or other minorities. Lowenthal and Guterman
particularly research based on the F scale, have been published by Dicks, 1951; Christie and
(1949), Adorno (1946), and others had shown that the fascist agitators of the Jahoda, 1954; Sanford, 1954, 1956; Christie and Cook, 1958; Bay, 1958; Brown, 1965;
1930s, in America and elsewhere, made no mention of Jews. They contented them- Kirscht and Dillehay, 1967; Greenstein, 1969; and Knutson, 1971.
 154 Nevitt Sanford ~uthoritarianPersonality in Contemporary Perspective 155

and projective material suffered from failure to rule out several kinds of observer response set is a factor in the F scale; therefore, correlations of this instrument with
bias. other scales made up of positive items--for instance, the A-S and E scales and
As noted above, all items of the A-S, E, and F scales are in the positive Rokeach's (1960) dogmatism scale-are somewhat higher than they would other-
direction. I n reviewing this procedure for constructing the F scale, critics not sur- wise be. This state of affairs has not, however, washed out any findings of the original
prisingly asked how we could tell whether high scores on the scale were due to study or created serious doubt about the generality of authoritarianism. Christie,
authoritarianism or to response set. Cohn (1953) was probably the first to suggest Havel, and Seidenberg (1958) have shown that the F scale correlates with variables
that the F scale is in part a measure of acquiescence; but far more attention was in which response bias is not a factor, and, as shown above, the conclusions from
accorded the work of Bass (1955), who composed reversed versions of the F scale T h e Authoritarian Personality do not rest on findings from the use of this instrument
items, administered both the original scale and the reversed scale to the same subjects, but on the convergence of findings from various procedures. Future students of
and obtained a correlation between the two scales of only -.20. Christie, Havel, and authoritarianism or similar personality structures, however, if they choose to use
Seidenberg (1958) argued that Bass's'reversed items were not in fact psychological scales rather than some more revealing and trustworthy instrument, will probably be
opposites of the original items, and proceeded to show that the composition of such well advised to compose them of both positive and negative items.
opposites is an extremely complicated business. As indicated above, each F scale item Concerning samplihg, Hyman and Sheatsley (1954) concluded from their
conveys a complex pattern of connotations. Nevertheless, Christie and his associates masterful critique that the results reported in T h e Authoritarian Personality could
showed that an F scale with equal numbers of authoritarian and equalitarian asser- not be generalized to any known population because of the unrepresentativeness of
tions can be constructed. Numerous other studies, some of them highly ingenious, the people studied. Even today the problem of sampling is far from solved, for
have been directed to finding out how much of the variance on the F scale is due investigators still rely heavily on middle-class subjects, college students in particular.
to acquiescence. Conclusions have ranged from "virtually all" (Peabody, 1961) to Nevertheless, during the years since 1950 the F scale (or shortened or equivalent
"none worth bothering about" (Gage and Chattergee, 1960). Some writers agree versions of it) has been used with huge numbers of subjects of the greatest variety:
with the authors of T h e Authoritarian Personality that acquiescence is an expression I
Jews (Adelson, 1953; Himelhoch, 1950), blacks (Selznick and Steinberg, 1969; Kel-
of "authoritarian submission" and that therefore the use of only F positive items man and Barclay, 1963), Lebanese (Prothro and Melikian, 1953), Germans (Cohn
increases the validity of the scale. Kirscht and Dillehay (1967, p. 25), in their valua- and Carsch, 1954), Japanese (Niyekawa, 1960), teachers in seven Western European
ble survey of research on authoritarianism in personality, sum up the controversy as nations (Levinson, 1958), Catholics (Brown and Brstryn, 1956), political leaders
follows: "Even after fifteen years of research, the influence of acquiescence on scores
from the F scale is difficult to assess. This is due in part to mechanical problems in
isolating acquiescence, but it is also due to the likelihood that the interaction between
i at the local level (Harned, 1961), people with low occupational status and little
education (Martin and Westie, 1959), Washington lobbyists (Milbrath and Klein,
1962), and children (Lyle and Levitt, 1955).
I
acquiescence and authoritarianism is complex." Couch and Keniston (1960) showed F. H. Sanford (1950a) and MacKinnon and Centers (1956) used quota
that the response set to agree ("Yeasaying") is a personality characteristic in its own samples in greater Philadelphia and Los ,4ngeles, respectively, while Janowitz and
right. I t is not, apparently, a component or manifestation of authoritarianism, for Marvick (1953) and Lane (1953, 1955) used national probability samples. Not all
these workers produced evidence that the two personality dispositions are inde- of these studies had to do directly with T h e AuthoP-itarian Personality's basic con-
pendent. Rorer (1965), on the other hand, has pointed to the low correlations among ception of personality and ideology, but none produced results inconsistent with it.
different measures of acquiescence and thus questioned the conception of acquiescence As Roger Brown in his judicious review says, "On the level of covariation, of one
as a unitary variable. variable correlated with another, the findings of T h e Authoritarian Personality seem
To readers who have worried about authoritarianism in the individual and in to me to be quite well established. Anti-Semitism goes with ethnocentrism goes with
society ever since Horkheimer ( 1936), Frornm ( 1941) , Maslow ( 1943), Reich anti-intraception goes with idealization of parents and self goes with a rigid con-
(1933), and others drew attention to the phenomenon, and who have never felt ception of sex roles, etc." (Brown, 1965, p. 525.)
there was any difficulty about recognizing it when they saw it, this controversy about Many studies have served to widen the circle of covariation. For instance,
response bias might seem trivial. T o more than a few psychologists, however, the high scores on the E or the F scale have been shown to be associated with high scores
questions of how to formulate authoritarianism and how to resolve it into its con- on various other attitude scales, such as scales for measuring rigidity (Gough and
stituent elements have seemed important, for these questions have considerable bear- Sanford, 1952), misanthropy (Adelson and Sullivan, 1952), xenophobia (Campbell
ing on our understanding of the relations between authoritarianism and intelligence, and McCandless, 1951), dogmatism (Rokeach, 1954), traditional family ideology
education, socioeconomic background, and cognitive and emotional development. (Levinson and Huffman, 1955), nationalism (Levinson, 1957), and a custodial
Acquiescence, unfortunately, does not seem to be a very promising candidate orientation toward mentally ill patients (Gilbert and Levinson, 1957). These corre-
for the status of component of authoritarianism. There are serious questions about lations afford confirming evidence of the generality of the authoritarian outlook.
what it is and how general it is; moreover, as Couch and Keniston (1960) suggest, Other studies-a great many of them-have used the F scale, or some variant
its underlying determinants in the personality seem to be almost as diverse and com- or rough equiyalent of it, as a prediction of overt behavior. For example, experiments
plicated as those hypothesized for authoritarianism itself. Nevertheless, acquiescence have &own that authoritarianism is associated with mental rigidity, especially under
 Nevitt Sanford Authoritarian Personality in Contemporary Perspective 157

conditions of anxiety (Neuringer, 1964) ; with suspiciousness and untrustworthiness factors seems even more crucial. As Sanford and Comstock and their zssociates
in a two-person game situation (Deutsch, 1960) ; with conformity to group pressure (1971), in their analysis of events such as the My Lai massacre, and Hersey (1968),
(Crutchfield, 1955; Canning and Baker, 1959; Vaughn and White, 1966) ; with in his account of the Algiers Motel incident, have stressed, the actors in these inci-
punitiveness toward people of low status (Roberts and Jessor, 1958) ; with autocracy dents believed that they had some kind of permission to do as they did-partly
in leadership roles and, in followship roles, readiness to be satisfied with appointed because propaganda or cultural norms depicted the victims as beyond the pale; but
leaders (Haythorn and others, 1956aJ 1956b) ; with obedience to an experimenter's in addition, they were all suffering from fatigue, fear, confusion, and momentary
authoritative commands (Elms and Milgram, 1966) ; and with resistance to efforts group pressures. All this is not to say that personality does not enter into the determi-
to change attitudes by educational means (Katz, McClintock, and Sarnoff, 1957). nation of such events. Authoritarianism, in particular, is a concept to explain the
Other studies, focusing on behavior in natural situations, have shown authoritarianism varying degrees of susceptibility in individuals to the kinds of situational pressures
to be associated with totalitarianism in clasroom teachers (McGee, 1954), failure just described.
to be nominated for leadership positions (Hollander, 1953), low ability to adapt to To work out the ways in which personality and the situation interact in single
an unstructured educational setting (Goldberg and Stern, 1952), disinclination to dramatic episodes is, of course, extremely difficult, for the necessary observations are
participate in political affairs (F. H. Sanford, 1950a; Milbrath and Klein, 1962), hard to obtain. One can do better at the analysis of behavior in organizational or
tendency not to vote in elections (Janowitz and Marvick, 1953; Lane, 1955), and institutional settings, where the situational pressures on individuals are more manifest
preference for candidates seen as conservative, regardless of party (Leventhal, Jacobs and durable. Levinson and his associates have probably gone further than anyone else
and Kudirka, 1964). in demonstrating some of the interactions between institutional characteristics and
These findings have strengthened the argument in favor of a central and authoritarianism in personality. Studying staff members in three mental hospitals, they
relatively deep-seated personality structure, which helps to determine behavior in a showed first that authoritarianism ( F scale) goes with a "custodial" as opposed to a
wide variety of situations. However, the findings are in the form of general relation- humanistic orientation toward the mentally ill. They then showed that the hospital$
ships-correlations between variables in populations of people. Individual differences and subsections of them, differed and could be reliably rated with respect to the
are large-in all the studies, and factors other than authoritarianism clearly enter into degree of custodialism expected of their employees. Their major conclusion was that
the determination of all the kinds of behavior studied. As stated in The Authoritarian there is congruence between the policy requirements of a social system such as a
Personality, p. 36: "Specific social attitudes, if adequately measured, will undoubtedly hospital and the modal personality of its members. Authoritarianism in personality
be found to correlate with a variety of external and contemporary factors, and if one helps to determine who will select (and be selected for) a given organizational role
studies only specific attitudes he may easily be led to the belief that this is all there and who will remain in it, while life in the role (with which an individual may
is to it. Consistent trends in the person can be revealed only by subjecting him to identify himself) and in the organization (which may be oppressive or liberating)
a variety of stimuli, or placing him in a variety of different situations, or questioning may increase or decrease authoritarianism in personality (Pine and Levinson, 1957;
him on a wide variety of topics; then, according to the present hypothesis, consistent Gilbert and Levinson, 1957; Greenblatt, Levinson, and Williams, 1957). Levinson
trends (that is, personality) will always be revealed." and his associates offer what appears to be the right approach to the study of institu-
An example of how far situational factors can go in determiniag what might tional racism or sexism or other forms of antidemocratic corporate behavior. The
be regarded as a specific form of "authoritarian behavior" is afforded by Milgam's individual occupants of an institution's diverse roles may be replaced without chang-
(1963, 1965) experiments on what he called obedience. In what he described as a ing that institution, but the study of the personalities of individuals who live or work
learning experiment, Milgram instructed his subjects, university students and towns- there is necessary to a complete understanding of their behavior and of the meaning
people, to' pull switches which they thought would give Hevere and dangerous shocks of corporate actions. Such study is also necessary to an understanding of the impact
to subjects who failed to supply correct answers to test questions. To Milgram's sur- of organizational life upon individuals and of how and why some individuals are
prise, a large fraction of his subjects readily obeyed his instructions. Aut!loritarianism able to resist organizational pressures (Howard and Somers, 1971).
in personality may well have been a factor in determining obedience to the experi- Situational factors (or, one might better say, a range of nonpersonality
menter's words "You have no choiceJ' (when they might have walked out) ;but the factors) help to determine not only overt social behavior, individual and collective,
results of this experiment strongly suggest that ordinary decent people can be in- but also one's responses to any item of an attitude scale. The F scale item "Obedience
duced to carry out socially destructive acts in a situation embodying such features and respect for authority are the most important virtues children should learn," for
as a prestigeful university, a powerfully authoritative experimenter, the presumed exampleJ was acceptable to almost everybody in Prothro and Melikian's (1953)
requirements of science, and ignorance and confusion in the face of complicated sample of Arab students and in F. H. Sanford's (1950a) representative sample of
technological products and arrangements. greater Philadelphia. I t certainly seems fair to say tkat most of these subjects, in
The same line of thought has been used to explain why a great majority of agreeing with this statement, were merely expressing a cultural norm rather than
professors at the University of California at Berkeley finally submitted to the iin- coping with an unresolved conflict of personality. Again, on all occasions when I
position of a'loyalty oath. Situational pressures, economic as well as social, became have discussed the details of the F scale with groups of people who were not psychol-
almost overwhelming (Sanford, 1953). In riots or massacres the role of situational ogists, someone has protested, "But I agree with item X because it is true." No
 Nevitt Sanford Authoritarian Personality in Contemporary Perspective 159

doubt each item of the F scale has an aspect of objective truth for some people, in Africa (Pettigrew, 1958) obtained significantly higher F scale scores than do their
the sense that it reflects more or less accurately the real world in which they live or counterparts in the United States. And, speaking of students, the F scale has been
in the sense that they find it rational or practically adaptive to act as if the statement administered to many large samples of students, in all classes, in various higher d u -
were true. Indeed, it has been argued with some persuasiveness-for enrample, by cational institutions. According to the summary of this work by Feldman and New-
Miller and Riessman ( 1961) and Campbell and others ( 1960)-that "working-class comb (1969), authoritarianism is significantly reduced over time, that is, is lower
authoritarianism" is mainly to be explained on this basis. And again, an "authori- among seniors than among freshmen, in most colleges, a phenomenon that might be
tarian" response to an F type of item might be a matter of simple ignorance. largely explained as adaptation to campus culture. What seems quite clear is that
Hyman and Sheatsley (1954) in their methodological critique aptly point out that culture is a main determinant of the fact that F scale scores differ, in all classes,
the authors of T h e Authoritarian Personality simply were not justified in attaching from one institution to another-being higher, for example, in church-related col-
psychodynamic or even ideological significance to a subject's mentioning George leges than in secular ones, higher in vocational institutions than in those in which the
Washington or General MacArthur as his most-admired person when they had not Iiberal arts are accentuated.
established whether the subject knew the names of my other great men or women. Findings concerning the role of SES are not so clearcut. Lipset (1959)
We were aware of these and various other possible determinants of response marshaled evidence from various studies, including T h e Authoritarian Personality,
to particular scale items, for example, a subject's discussion with his friends shortly to show that authoritarianism is more pronounced in the lower than in the middle
before he filled out our questionnaire, but we relied on the trend of response in the classes. Miller and Riessman ( 1961) took strong exception to Lipset's conclusions,
scale as a whole-behavior in numerous, varied microsituations. (It is unfortunate which, however, were soon to receive support from Lipsitz's (1965) analysis of data
that the studies of national samples mentioned above used very short versions of the from three surveys of opinion. The difficulty here seems to be that different types of
F scale, for this raises questions about whether their measure of authoritarianism authoritarianism have to be considered and that when large samples of lower-class
was accurate and whether they were measuring the kind of authoritarianism we were people are brought into the picture cultural, or subcultural, differences become highly
talking about.) And we relied, for support of our psychodynamic hypotheses, upon important.
the correlations of the F scale with other variables. But Hyman and Sheatsley (1954) Low education apparently is a major factor in authoritarianism and is largely
had an important point when they said that since we had not studied samples in responsible for correlations with low SES as well as with certain culture patterns.
which nonpsychodynamic factors might loom large, it could not be said that our main Selznick and Steinberg (1969) have brought together evidence from various studies
hypotheses was proved by the data we presented. of the relations of education to prejudice and authoritarianism-including their own
This kind of criticism has been the taking-off point for many studies of the -and have advanced more completely and effectively than anybody else what is by
determinants of authoritarianism, mainly studies in which the F scale or some variant all odds the most important criticism of T h e Authoritarian Personality. They start
of it has been the dependent variable. Some of these studies have tested some of our by recognizing that the findings of the original study are largely intact; it is the
hypotheses concerning the origins of authoritarianism in childhood in the setting of theory to explain the findings that has to be called into question. The original theory,
the family. Harris, Cough, and Martin (1950) found that ethnic prejudice in children in their view, is not so much wrong as unnecessary since an alternative theory can
was asbociated with an accent on obedience, strict control, and inculcation of fear better explain the facts we now have. The alternative theory is, roughly, a cognitive
on the part of their parents. Frenkel-Brunswik (1954) reported that prejudice and theory, in the tradition of Hyman and Sheatsley (1954), Rokeach (1960), and
associated personality characteristics in children were found most often in families McClosky and Schaar (1965). The most essential idea is that authoritarianism can
where relationships were "characterized by fearful subservience to the demands of be explained as resulting from a lack of cognitive development, without recourse to
the parents and by an early suppression of impulses not acceptable to adults." Lyle psychodynamic or "ego defensive" theories. As Selznick and Steinberg (1969, p. 141)
and Levitt (1955) found that scores on a children's antidemocratic scale were write, "What is at issue is not the face meaning of F beliefs but the reasons they
correlated with parental punitiveness as measured by a sentence-completion test. are accepted. Does acceptance of F beliefs have psychological sources and intellectual
Baumrind (1967), on the basis of careful observational studies of nursery school consequences, as the original study claimed? Or does acceptance of F beliefs have
children and their parents, distinguished three patterns of parental behavior toward intellectual sources and psychological consequences, as their relation to education
children: authoritative, authoritarian, and permissive. She concluded that the author- strongly implies?" These authors go on to argue that Western society contains at
itative pattern is most favorable to the development of desirable traits in children, least two cultures, the common and the enlightened, and that anti-Semitism and
while the authoritarian pattern tends to give rise to distrust, withdrawal, discontent- other forms of prejudice are simply aspects of the common culture. The question then
traits that may be thought of as forerunuers of authoritarianism. is not why people accept anti-Semitic views but why some do not (p. 169). Selznick
But studies of this kind have been few. Much more interest has centered on and Steinberg are not simple-minded about the ability of just any sort of formal
finding determinants of authoritarianism in culture, socioeconomic class (SES), and education to reduce prejudice (p. 191) ; they say that "at its best, education involves
low educational status. Culture comes very much to the fore when the F scale, a total belief system of a fairly general and abstract nature"; some people are less
translated, is used in different nations. Workers in Germany (Cohn and "receptive" to it than others, and some apparent freedom from prejudice in the
Cars&, 1954) and students in Lebanon (Prothro and Melikian, 1953) and South upper middle class is simply "conformity."
 Nevitt Sanford Authoritarian Personality in Contemporary Perspective 161
A discussion of this thesis requires that we ask what education is. And what tarianism that is "more general." His conception is indeed more general; but it is
is culture, and how is it transmitted? To consider culture first, Kroeber and Parsons also emptier. Anti-Semitism, or the tendency to ascribe all manner of evil to enemies
(1958) define it as "transmitted and created content and patterns of values, ideas, foreign and domestic while thinking only good of ourselves, is expressive of a man's
and other symbolic-meaningful systems." According to a major anthropological deepest passions, most certainly his basic needs and inner conflicts, not just his style
tradition culture is transmitted mainly by the.family rather than by the school, and of thinking. We were right, I believe, when we searched for the appeals of fascism
it becomes incorporated by personality not only through conformity or routine learn- in the 1940s, to study personality comprehensively and in depth, with attention to
ing but through a variety of psychodynamic processes including unconscious identifi- the interaction of cognitive and psychodynamic factors; and I would advocate .the
cation with parents. The individual personality is not a mere carrier of culture; each same course in the study of destructive ideologies today. What matters most in the
I assimilation of a social element is a product of something from the culture and some- study of prejudice and violence is who the victims are, how they are conceived, and
thing already in the person. Patterns of upbringing vary from one culture to another, who sanctions, by what means, the attacks upon them (Sanford and Cornstock,
but within a given culture they have enough generality and enough impact on per- 1971). It makes a big difference, and to a great many people, whether the victims
sonality that we may discuss a modal personality for that culture. Personality is not are hippies at home and Asians abroad, with violence sanctioned by government
I:
so fixed or durable, however, that it may not at any time of life change under the agencies, or policemen and "the establishment," with sanctioning by a small and
iI influence of cultural change. And individuals may, at any time, contribute to culture. intimate group of true believers. To say that the actors in both cases are dogmatic or
I t follows that differences in F scale scores from one culture or subculture to violent, without specifying what they are dogmatic about or violent against, is not
another do not by any means contradict or downgrade the role of personality in enough, and to imply that it is enough is to skirt the real issues. Established authority
authoritarianism. Selznick and Steinberg, however, would leave personality out reserves the right to say what is dogmatic and what kinds of violence are legitimate
'

altogether, relying simply on the learning of cultural norms or elements as an ex- while exempting itself from any such categorization. Government-sponsored studies
planation for authoritarianism. No doubt such learning, or nonlearning, influences of violence, initiated in connection with the political activism which began in 1964,
F scale scores and may even be largely responsible for high scores in the "surface have been directed to violence per se or objectless violence, and, where the individual
resentment" type of authoritarianism and the "rigid" type of low authoritarianism. was concerned, they have been carried out according to the popular assumption that
But any alternative theory of authoritarianism must explain the major finding of the for this "ill" there must be a cure, such as a pill or an operation. These studies carry
original study-the patterning of the component dispositions. Why should anti- the implication that it is out-group violence that is bad, while our violence is, if not
Semitism be associated with anxiety about sex, with self-glorification, with sub~nissive- good, at least necessary, there beicg so much violence to contend with.
ness toward authority? To say that this is "in the culture" or is due to a general lack It is hard to make the argument for the primacy of cognitive style hold water.
of enlightenment (as Selznick and Steinberg do) is not an explanation. Why should Kirscht and Dillehay (1967, p. 132) strongly support R-okeach. But then they go on
the dispositions go together in a culture? I would argue as we did in The Authori- to say, "Certain domains of belief often serve as foci for the close-minded style"
tarian Personality, that they go together because they constitute a dynamic system, in (p. 133). Seeking to explain this statement they fall back on precisely the sort of
an individual or in a social group. In any culture the common emotional impulses formulation in psychodynamic terms that pervades The Authoritarian Personality-
of individuals are shaped through shared experience in the social group, and w-ays that is, that mental rigidity comes largely from anxiety, which comes from the arousal
of controlling these impulses are developed in the individual and in the group, thus of impulses such as sex and aggression, which might be expressed in ways that run
forming and favoring cultural values. In any culture some individuals accept pre- counter to powerful cultural demands.
vailing values and social codes without necessarily needing them for defensive pur- Such formulations as this rode the crest of the wave in psychology during the
poses, but such individuals most quickly adopt the cultural elements they do need 1940s, and The Authoritarian Personality participated in this trend; thus, one can
for such purposes; and any individual may develop methods for coping with the appreciate the concern of some psychologists to give the cognitive a fair chance
problems of life in the same way they were originally developed in his culture. Cul- (Sanford, 1970b). But they don't have to go overboard or risk driving the whole boat
ture and personality continuously interact, in mutually supporting ways. onto the rocks. In the foregoing discussion, unfortunately, it has been necessary to
Although Selznick and Steinberg do not invoke personality theory to help speak of personality and culture and of the psychodynamic and the cognitive as if
explain the origins and working of the common culture, they at least assume indi- they were sharply separated opposites, in competition for the distinction of being
vidual differences in receptivity to enlightenment; and here, as noted above, they the more important. That was not in the spirit of The Authoritarian Personality,
give the intellectual the primary determining role. On this point they have received which, though it gave most of its attention to dynamic theory of personality, also
a great deal of support from psychologists. Rokeach (1954, 1956, 1960) has taken showed awareness of the intricacies of personality-culture and psychodynamic-cogni-
the lead in arguing that the best way to conceive of authoritarianism-in his view tive relationships. Thus today, although I find little in the facts or theories of The
"authoritarianism of the left" as well as authoritarianism of the sort so far discussed Authoritarian Personality that seems fundamentally wrong, I have no difficulty in
in this chapter-is not in terms of psychodynamic processes or even psychological urging that we now stress the intsraction of personality and culture, of psycho-
content but in terms of cognitive style, most importantly the "open and closed mind." dynamic and cognitive processes.
Rokeach states that his main intention has been to arrive at a conception of authori- The last comes directly into focus when we consider education, which I as
162 Nevitt Sanford Authoritarian Personality in Contemporary Perspective 163

well as Selznick and Steinberg consider the essential counter to authoritarianism. knowledge about variability, probability, and so forth. Her instructor found her
But what is education? Selznick and Steinberg-in company, it seems, with most crying in the corridor after a class, and her complaint was "You are ruining every-
adherents of our enlightened culture-seem to assume that education is a matter of thing I ever believed." But I would argue that instruction in statistics would not have
learning content and training intelligence, while I see it as a means for developing been so effective had this young woman not loved her teacher.
the whole personality. When my colleagues and I began to study education at the If this kind of change occurs in college, what, we may ask, about the schools?
college level in 1952, we asked ourselves how authoritarianism might be reduced I have not been able to find any relevant longitudinal studies of junior high and
through education. We soon realized that the reductions over time in E and F scale high school students, but I would hypothesize that some relatively unprejudiced
scores that regularly occurred in a liberal arts college were not solely a matter of children become more prejudiced and some relatively prejudiced children become
adaptation to campus culture, for correlations began to accumulate between degree more liberated as they go through school--depending on the kind of school environ-
and direction of change in scale score and a wide range of factors and processes ment they encounter at different stages of personality development. Adelson (1971)
within the academic environments. Since 1952 my colleagues and I have published has reported a highly significant study of the development of political ideas. He found
many articles and several books treating this subject in detail (D. Brown and others, that early adolescents are highly authoritarian, with significant reduction occurring
1956; Sanford, 1957, 1958, 1959a, 1959b, 1960a, 1960b, 1961, 1962a, 1962bJ 1963, in middle and late adolescence-largely because of cognitive development. For
1964aJ 1964bJ 1966aJ 1966bJ 1967, 1968a, 196815; Katz and others, 1968; Axelrod example, the early adolescent's relative inability to think abstractly or to cope. with
and others, 1969), and the trend in college studies that we started has resulted in the relativity of values is highly favorable to his acceptance of authoritarian stereo-
an outpouring of publications relevant to the fate of authoritarianism in college. types. But the early adolescent's extraordinary bloodthirstiness toward deviants of
(For summaries of this work, see Webster, Freedman, and Heist, 1962; Katz and all kinds must have to do with the problems of these young people in coping with
others, 1968; Chickering, 1969; Feldman and Newcomb, 1969.) Unfortunately, their own impulses and with the authorities around them. Which comes first, the
psychologists, sociologists, and political scientists are prevented by the norms of cognitive failings or the inner conflicts? Is it necessary to ask this question about
academic culture, which virtually prohibits their having anything to do with educa- processes so intimately bound up together? Certainly the problems of adolescents
tion, from reading any of this literature. ( I could find only one reference to it in are the more difficult the less well the individual is able to think, but abstract think-
the dozens of papers, books, and monographs consulted in the course of preparing ing and relativity of values can be taught-though probably not to children who are
this chapter.) I t is not possible adequately to summarize it here. The main point is afraid of their teachers or by teachers who cannot think abstractly themselves. Adel-
that freedom from authoritarianism increases with education-not only because of son has opened up a rich area for future research.
exposure to enlightened culture, not only because of intellectual development, but
because personality itself changes under the impact of well-directed education. What To Do Now
Various features of the college environment work, or can be made to work, to reduce
authoritarianism in personality: governance that helps the student become his own The trouble with research on authoritarianism is that it has bken research
authority, course content that helps him become aware of his own impulses, teaching on The Authoritarian Personality. Instead of following the lead of this work-
that challenges preconceptions and gives practice in criticism, faculty members who adapting its comprehensive, exploratory, empirical approach to the study of other
are models of independent thinkers, and a general climate of freedom and respect for problems, such as the appeals of communism or the new populism-personality psy-
the individuaL5 chologists have shown an obsession with the F scale. This preoccupation is not due
In all this we deal with interactions, often very subtle, between the cognitive just to laziness or a lack of imagination; it has to be understood in the light of the
and the psychodynamic. For example, I know of one young woman whose system of recent history of psychology and the politics of research. Since the early 1950s we
primitive, right-wing authoritarian ideas was changed through a class in psychological have witnessed a gradual downgrading of the concept of the person and the growth
statistics. She could not maintain her rigid mental compartments in the face of of a kind of puritanica1 future orientation, expressed as a concern for building an
eventually useful science on the model of nineteenth-century physics. In addition,
it Several writers have called attention to the irony in the fact that when our study
there has been mounting pressure within-and from above-university departments
of anti-Semitism and fascism was published in 1950, the Cold War was on and national
attention was focused on communism. There is irony also in the fact that at about the timk for more and quicker publications. As a result, and with the full support of govern-
our work on the college experience (its potentiality for liberating students through developing mental funding agencies, a virtual ban has been placed on comprehensive inquiries
broader perspectives, relativity of values, self-insight, the courage and knowledge to challenge into personality, on case studies, on naturalistic and exploratory studies-in short,
authority, and so forth) was beginning to catch on among educators, national attention on virtually everything necessary for the discovery of anything.
became riveted upon students who were already "liberated." Of course, as the Feldman and We should now launch a whole new attack on the problems of personality
Newcomb (1969) survey shows, the students involved in "protest" or "activism" were never and politics, aiming at psychological understanding of the issues and the patterns of
more than a small minority of college students in the country at large. This does not down-
grade the significance of student activism in the 1960s, but it does remind us that the great thought and action observed today. If we proceed in the way the authors of The
educational challenge was, and remains, how to help students develop beyond ~uthoritar- Authoritarian Personality went about their work, the results cannot fail to be exciting
ianism. and important. This fresh approach should take into account the fact that authori-
164 Nevitt Sanford Authoritarian Personality in Contemporary Perspective 165
tarianism of the sort we studied in the 1940s is a psychohistorical conception-it both thinking and feeling were inhibited, and consciousness constricted, because of a
owed much of its content and structure to the period in which it was observed and fear of impulses.
formulated. Indeed, its particularity in this respect accounts toaagreat extent for the N1 this seems a pretty far cry from the anti-intellectualism among middle-
value of the concept. class college students, which has been the subject of much comment in recent years.
Max Horkheimer, in his foreword to our book, referred to the potential fascist There is little interest among these students in the old issue of the things of the mind
we studied as "a new anthropological species," a person in whom the ideas and skills versus the practical, and recourse to authority of some kind as the source of truth
typical of a highly industrialized society were combined with irrational or antirational and knowledge seems to have been replaced by recourse to feeling and intuition.
beliefs. According to this view, we could not hope fully to understand authoritarian Thinking (disciplined thinking, at any rate) and feeling are separated and set in
leadership of the 1930s through studying Oliver Cromwell or Cotton Mather; and, opposition, the former being rejected apparently because of its association with "big
by the same argument, the study of prefascist personalities in the highly industrialized science" or that super-rationality (thinking separated from feeling and value?) &at
society of the 1940s cannot be expected to tell us all we need to know about right- has got us into so much trouble as a nation-uncontrolled technology, Vietnam, and
wing extremism in the postindustrial society of today. so forth. This way of sorting things out favors the expansion, or perhzps the altera-
Indications of change in the content and structure of what we called fascist tion, of consciousness and the direct expression of impulses. This new anti-intel-
potential are not hard to find. For one thing, many users of the F scale have called lectualism may have something in common with the d d : both may be ways of coping
attention to out-of-date items: items that make direct reference to events of the time with a world that becomes increasingly hard to understand or to do anything about.
("It is best to use some prewar authorities in Germany to keep order and prevent But we cannot really tell without carrying out, today, theoretically based clinical
chaos") as well as items that seem to require a different interpretztion today in view studies of the kind reported in The Authoritarian Personality and summarized abdve.
of changed circumstances or changes in the general climate of opinion-items, for The argument here is that personality itself-not just issues and ways of
instance, that make reference to work, science, authority, and sex. regarding them--changes with the times. The prototypic right-wing authoritarian
Again, scale iteins that were once considered expressions of personality now of the 1930s and 1940s now seems a bit quaint; and as we look to the future, it seems
have ideological significance. Recently a student asked me about personality measures that we have to worry less about old-fashioned dictatorships than about domination
that might be predictive of ideological trends among college students. I suggested, by corporate systems that nobody seems able to control. And if the "authoritarian
among others, the social maturity (SM) scale, developed by my colleagues and me family," the one presided over by a father whose sternness and rigidity increased as
in the early 1950s (Webster, Sanford, and Freedman, 1955), which now has a place he lost his grip on affairs in the larger world, is changing or disappearing in pace
in the Omnibus Personality Inventory (Heist, 1960). Our idea had been to produce with the liberation of women and the decline in the birthrate, we can certainly expect
an ideology-free instrument for measuring the F syndrome, by using items from change in the type of personality we saw developing in this sort of family.
established personality inventories such as the MMPI (Hathaway and McKinley, When I say that personality changes with the times, I do not mean that the
1951). Ouflproduct, a scale comprising 123 items, had a reliability (KR-20) of .88 general laws governing its functioning will have to be rewritten or that we must
and correlated .74 with the F scale in large samples of Vassar freshmen. The student abandon our sense of kinship with the ancient Greeks or the neurotic personalities of
was back the next day to say that he could not use this instrument because it was Freud's time. There are universal human dilemmas, such as how to gratify basic
loaded with ideology. An examination of the scale showed me that he was quite right. emotional impulses in ways that do not violate the norms of one's culture, and there
Items that in the 1950s could be regarded as expressive of such generalized personality are, apparently, universal laws governing the ways in which inner conflicts are re-
dispositions as punitiveness, conformity, sex-role preference, or cynicism would lved: for example, the conditions under which one or another defense against a
surely be identified by today's students as middle class or as expressive or Conscious- aoscribed impulse will be chosen. Naturally, we expect little historical change in
ness I or I1 (Reich, 1,970). ese basic functions.
We have to consider, too, that personality dispositions which in the 1940s A given child's ways of coping with inner conflicts bear great similarity to
belonged to the F syndrome today belong to something else, or may indeed be differ- those of many other people in his society, and we arrive readily enough at the notion
ent in meaning or in underlying source. Consider, for example, anti-intellectualism, of a major pattern of culture. We know from anthropologists that patterns of culture
which we saw as an important feature of potential fascism. In what we then called differ from one society to another, and we can conclude from the same kinds of
anti-intellectualism, theory was rejected in favor of "the factsw-the tangible and observations they have made that culture patterns may change from time to time in
objective; intellectual activity, considered as a source of pleasure or meaning, was the same society. As early as the 1930s American psychoanalysts were complaining
rejected in favor of keeping busy with practical affairs or throwing oneself into action that they could not find among candidates for their treatment any "simple hystericsH
and adventure; intellectual activity as a source of insight or as a guide to living was the type Freud described; and they were soon to learn that such patients had
rejected in favor of external authority, religious faith, or the conventional wisdom. come rare in Vienna too. To explain how and why this change took place would
Anti-intellectualism was associated with opposition to feeling and, it seemed to us, ke us too far afield. The point for us here is thdt we have to deal in such instances
lack of capacity for feeling. I t was our view that in the high scorers on the F scale w t h highly significant change-not, to be sure, in human nature itself, but in
 Authoritarian Personality in Contemporary Perspective .
Nevitt Sanford 167

processes that are central to the functioning of personality; specifically, such processes changes in personality itself. Forced by overwhelming pressure to go against con-
are observable in the ways in which sexual and aggressive impulses find guilt-free science in signing a special loyalty oath, individuals suffered a change in the structure
overt expression and in preferred types of defense against proscribed impulses: for of the superego and in the relations of the superego and the ego. What happened
example, repression in 1890, as contrasted with the separation of behavior from within a relatively short span of time then may also happen very gradually, as when
feeling, or acting as if one had no problems today. a dean over the years comes to identify himself with an authoritarian social role. By
Changes of this kind are reflected in instruments like the F scale. This instru- the same token, authoritarianism in personality can be reduced in college, and quite
ment does, to be sure, rest on some assumptions concerning universal, or nearly uni- probably in other kinds of organizations. Adelson (1971) says it is reduced in high
versal, human ways: for example, that parents cannot bring up children without school. He suggests that such reduction is mainly a maturational process, and per-
frustrating them and thereby arousing hostility, and that hostility against parents is haps it is. When high school students have been studied as much as college students
a potent source of anxiety. We must, however, allow for the possibility of historical have, however, we may find that certain school experiences reinforce adolescent
change in respect to who does the punishing, the kinds and severity of punishment, authoritarianism and that other experiences speed up the developmental process.
and the ways in which hostility and anxiety are managed; and we should expect The personality syndromes most useful in understanding and predicting politi-
changes of these kinds to be expressed in political behavior and ideology. cal behavior will surely embrace both cognitive and psychodynamic factors. I t is
Rokeach (1960), in seeking a more general conception of authoritarianism, interesting to note in this connection that the potential for fascism we described in
may have been concerned about the time-boundedness of our syndrome and was 1950, by definition a highly complex phenomenon, has never been satisfactorily re-
seeking a conception of personality characteristics-cognitive style-that would be solved into measurable components. The F scale resists satisfactory factor analysis-
predictive of political extremism in all times as well as in various places. But cog- largely, I think, because each of its items is complex, designed to be expressive of
nitive processes are not immune to historical change; in any case, they are bound up more than one of the theoretical components we conceived of. I suggested in 1954
with psychodynamic processes that most certainly change over time. The political that "in place of further analysis of the present F scale, one perform the [factor]
behavior that we really worry about is usually particular, Iike the fascism and the analysis with an instrument three or four times as long. . . . A greatly lengthened
communism of the 1930s and 1940s or our Manichean foreign policy in the Cold F scale could easily be composed of items we discarded or items suggested by our later
War period and the Vietnam policy of the Johnson-Nixon administrations. The clinical findings, and this is to say nothing of the large number of items written by
rightist authoritarian we studied in tfie 1940s participated in and was shaped by, other workers in the course of constructing scales that correlated with F" (Sanford,
even as he helped to shape, the life of his times; and we should expect the same of 1954, p. 24). Harold Webster, Mervin Freedman, and I had this very much in mind
political man today. when we set about developing the 123-item "nonideological" social maturity scale
Although a man's personality and his politics are all of one piece, the two can mentioned earlier (Webster, Sanford, and Freedman, 1955). The simple and straight-
be separated, by abstraction, for purposes of study. We must bring to bear upon con- forward items of this scale fell readily into classes that made sense, but we never got
temporary political behavior and thought all the knowledge we have of personality around to a formal factor analysis.
and its development. Political scientists have to find a "human basis for the polity" O'Neil and Levinson (1954) had much the same idea that we did. They
(Knutson, 1972a), and the best candidate, surely, is personality theory. We can build performed a factor analysis using eight items from the F scale, eight items pertaining
on the kind of theory of human functioning that went into The Authoritarian Per- to enthnocentrism, six to religious conventionalism, and ten to traditional family
sonality and on that work's demonstration of the clustering of attitudes and opinions, ideology. Four factors were extracted: religious conventionalism, authoritarian sub-
of the intimate relations between cognitive and psychodynamic process, and of the mission, masculine strength facade, and moralistic control. The last three of these
importance of irrational factors in the determination of political outlook. Even more factors correspond to hypothetical dispositions considered in the construction of the
important, we can build on its way of studying personality and ideology-not the original F scale. Factor analysis of a greatly lengthened scale might provide more
specifics, which have their shortcomings, but the genera1 approach. And we can build empirical support for our original formulations, or it might give a different and more
on some developments in personality research that have occurred since 1950. adequate picture of the dimensions of authoritarianism,
How shall we look at personality today? I should say that dispositional struc- Robert Holt (1972) is addressing himself to this same problem. He wants to
tures of personality, such as rightist authoritarianism, are less closely related to "dimensionalize, in an area that has been dominated by typological concepts." And
behavior, less fixed or durable, and owe less to early childhood experiences than the he takes the same view of the F scale that I have been trying to put forward here:
authors of The Authoritarian Personality supposed. Particularly important, from the "The method by which the scale was constructed, relying on the fact that the various
point of view of personality theory, is the notion-barely touched upon in The types of items did in fact covary, froze into an instrument and into an operational
Authoritariart Personality-that personality structures, even "deep" or "central" ones, definition of a concept a particular configuration of attitudes that existed at a partic-
are sustained by the social system in which the individual lives and can change when ular time and place in history, partly for extrinsic reasons." His further work will,
that system changes. Study of the loyalty oath controversy at the University of Cali- of course, take into account the likelihood that the prevailing pattern of authoritarian
fornia at Berkeley (Sanford, 1953) showed that situational pressures can be of attitudes has shifted. I t will also take into account recent progress in personality
crucial importance in determining behavior and may, when sustained, bring about research. About the time The Authoritarian Personality was published, Jane Loe-
168 Nevitt Sanford Authoritarian Personality in Contemporary Perspective 169
vinger suggested to its authors that the authoritarianism we studied was best con- it works. Duckles, Comstock, and others are now developing a more elaborate scale
ceived as a stage of ego development. The Authoritarian Personality had a great deal based on the Frornm-Maccoby theory. I should like to see the instrument for studying
to say about ego weakness and ego strength, finding many apparent manifestations destructiveness (Sanford and Comstock, 1971) supplemented by instruments for
of the former among the high scorers on the F scale, many apparent manifestations of characterizing and measuring self-conception, unconscious sex identity, and uncon-
the latter among the low scorers. Loevinger proceeded to view the whole matter in a scious fantasy, and then used in the study of various kinds of political behavior and
developmental perspective. On the basis of data gathered by means of a sentence- outlook, including images of and attitudes toward ethnic minorities, females, and
completion test, she defined six stages of ego development: (1) the presocial and males.
symbiotic, (2) the impulsive, (3) the conforming, (4) the conscientious, (5) the What are the political problems and patterns that most invite study today? I
autonomous, (6) the integrated. If one reads her Measuring Ego Developnzent am willing to leave this largely in the hands of political scientists, though I would
(1970j in conjunction with The Autho~itarianPersonality, it is easy to come to the suggest that they look again at the types of highs and lows described in The Authori-
conclusion that the first three stages are more characteristic of high scorers, the last tarian Personality. The differences between surface resentment and other types have
three of low scorers on the F scale. This study, it seems to me, is a nice example of been explored to great advantage by D. Katz and his associates (Sarnoff and Katz,
the way to make progress in personality research: to proceed from a vague and global 1954; Katz, McClintock, and Sarnoff, 1956, 1957; Katz, 1960), and by Pettigrew
notion to a set of differentiated and measurable variables. Anyone about to study (1958) and Smith (1965), all of whom have made a distinction between "cognitive"
personality and politics today would do well to include among his instruments for and "ego-defensive" authoritarianism. The "rigid low'' calls attention to what people
measuring personality the Loevinger sentence-completiontest. scoring at the two extremes on the F scale have in common. "Left-wing authoritar-
But, of course, in such a study we should not wish to limit ourselves to possible ianism" has so far been poorly conceived and poorly investigated. I have insisted else-
components of the 1950 version of'right-wing authoritarianism. Jeanne Knutson, in where (Sanford, 1956; Sanford and Comstock, 1971) that the F scale and Rokeach's
undertaking her comprehensive study of personality in relation to political participa- (1960) dogmatism scale are flimsy vehicles for carrying the enormous weight of left-
tion, leadership, and socialization, looked far and wide for relevant personality varia- wing ideologies. The role of personality in such ideologies still awaits thorough study,
bles before settling on the following scales : security-insecurity ( Maslow, 1941-1 942), as do other patterns suggested by our types of high and low authoritarianism. In
anomia (Srole, 1956), manifest anxiety (Taylor, 1953), status concern (Silberstein studying such patterns I would proceed in much the same way that we did in the
and Seeman, 1959), esteem (Milbrath and Klein, 1962), faith in people (Rosenberg, studies leading to The Authoritarian Personality; and I would be guided, today, by
19541, dogmatism (Rokeach, 1960), intolerance of ambiguity (Budner, 1962), some principles of "action research."
threat orientation (Martin and Westie, 1959), the F scale, and her own measures of Let us start with naturalistic observations of, and interviews with, likely sub-
the needs in Maslow's hierarchy (Knutson, 1972). She began this work in 1967. A jects, covering a wide range of possibly significant factors. Let us then observe what
few years later she very probably would have included Loevinger's sentence-comple- goes together and form hypotheses about these coherences, check these hypotheses in
tion test (Loevinger and Wessler, 1970) and the Kohlberg (1971) measures of moral more directed interviews and with the use of projective techniques, and carry out
development. some case studies to reveal- something of the organization of relevant processes in
There is, of course, a lot of intercorrelation among these measures, and the individuals. Then we can define some variables with precision and develop measures
question of what are the basic factors must be left open. A correlational analysis of of them, which can be put into appropriate research designs.
all the items of these scales and Robert Holt's proposed dimensionalization of every- The subjects of this kind of research must also be its clients; that is the first
thing that goes into contemporary right- and left-wing authoritarianism might very principle of action research (Sanford, 1969, 1970~).Suppose we want to know about
well yield a similar set of factors. personality patterns in black consciousness, the new left, or the women's liberation
The scales and other measures mentioned here, with the exception of the F movement. We would not proceed by asking questions whose implications were un-
scale, seem heavily weighted on the side of what the psychoanalysts call ego psy- known to our subjects or by inducing them to take part in experiments whose pur-
chology. The personality psychologists may be missing a bet. The F scale, for all its poses were concealed. We would instead ask, in our own way, questions that they
faults, at least considered that man is a whole, that he is alive and filled with passion. want answers to, and that have a relevance to their purposes, .and they would be
Maybe the troubles of the F scale scared the psychologists away from trying to get at the first to receive a report of our findings. We, as scientists, put the pursuit of truth
man, so conceived, by means of objective and quantitative methods. As far as I know, at the top of our value hierarchy, and nobody will object to that. But since what we
the only scale developed in recent years that is based on psychodynamic theory is do in the name of this value has practical consequences, we have to decide whether
the Fromm-Maccoby scale for measuring "love of life" versus opposite tendencies we are content merely to be neutral (which means that our "facts" become available,
(Fromm and Maccoby, 1971). This instrument has already been called, by a leading almost exclusively, to those who already have everything) or whether we are com-
psychologist (Smith, 197lb), worse than the F scale in its methodology, and I agree; mitted to building a more humane and more rationally organized society. If the latter,
but it has demonstrated an extraordinary capacity to predict political behavior- few of our prospective subjects, or subject-clients, will refuse to work with us.
for example, in separating McGovern from Nixon supporters in August 1972 (Com- The other day I was talking with a student about his study of different white
stock and Duckles, 1972). Apparently the only thing it has going for it is the fact that image patterns of black people in relation to more general political orientations. He
170 Nevitt ~anjford

asked, "How do I get reactionary subjects?" That is the question-the question for
research. If he can find out how to persuade reactionary subjects to take part, on the
basis of their own knowledgeable choosing, in a psychological inquiry, he will know a
great deal about how to reduce the paranoid style in American politics. Since that
is what we really want to know, why not go about finding out directly, by approaching
subjects in the spirit of action research?
"Reactionary subjects" can be approached directly. They are, after all, law-
abiding citizens who openly favor quality education, scientific inquiry, and decency
in interpersonal relationships. The first thing to do is to quit stereotyping them (as I
have perhaps just done) and then to see if one can keep one's head while engaging
them in a discussion of politics. The stereotyping is a serious matter. I t follows almost
inevitably from our delineation of types of personality and politics. We must force
ANOMIE,
ourselves to remember that this "reactionary," this "high scorer," who makes us a
little anxious, is still very much of an individual and not all that different from our-
selves. Like everybody else, he can further develop his personality; and to help him
and others do that is the ultimate aim of all our work with him.
ALIENATION,AND
POLITICAL BEHAVIOR
J . Milton ringer

]It is scarcely surprising, in a period of

almost continuous conflict within and between nations, that students of politics should
devote a major share of their attention to the sources of disorder, of disaffection, of
political negativism. This attention matches a widespread concern among other
observers of the contemporary scene.
The prevailing images of our culture are images of disintegration,
decay, and despair; our highest art involves the fragmentation and dis-
tortion of traditional realities; our best drama depicts suffering, misunder-
standing, and breakdown; our worthiest novels are narratives of loneliness,
searching, and unfulfillment; even our best music is, by earlier standards,
dissonant, discordant, and inhuman. Judged by the values of past genera-
tions, our culture seems obsessed with breakdown, splintering, disintegra-
tion, and destruction. Ours is an age not of synthesis but of analysis, not
of constructive hopes but of awful destructive potentials, not of commit-
ment but of alienation [Keniston, 1965, p. 4; see also Sykes, 19641.
 HANDBOOK
Jeanne N.Knutson OF
General Editor
POLS[T]CGAL
PSYCHOLOGY

Jossey-Bass Publishers
San Francisco Washington London 1973
170 Nevitt sadford

asked, "How do I get reactionary subjects?'That is the question-the question for

research. If he can find out how to persuade reactionary subjects to take part, on the
basis of their own knowledgeable choosing, in a psychological inquiry, he will know a
great deal about how to reduce the paranoid style in American politics. Since that
is what we really want to know, why not go about finding out directly, by approaching
subjects in the spirit of action research?
"Reactionary subjects" can be approached directly. They are, after all, law-
abiding citizens who openly favor quality education, scientific inquiry, and decency
in interpersonal relationships. The first thing to do is to quit stereotyping them (as I
have perhaps just done) and then to see if one can keep one's head while engaging
them in a discussion of politics. The stereotyping is a serious matter. It follows almost
inevitably from our delineation of types of personality and politics. We must force
ANOMIE,
ourselves to remember that this "reactionary," this "high scorer," who makes us a
little anxious, is still very much of an individual and not all that different from our-
selves. Like everybody else, he can further develop his personality; and to help him
ALItENATION,AND
and others do that is the ultimate aim of all our work with him.

BOLITI[GAL BEHAVIOR
J . Milton Tinger

][t is scarcely surprising, in a period of

almost continuous conflict within and between nations, that students of politics should
devote a major share of their attention to the sources of disorder, of disaffection, of
political negativism. This attention matches a widespread concern among other
observers of the contemporary scene.
The prevailing images of our culture are images of disintegration,
decay, and despair; our highest art involves the fragmentation and dis-
tortion of traditional realities; our best drama depicts suffering, misunder-
standing, and breakdown; our worthiest novels are narratives of loneliness,
searching, and unfulfillment; even our best music is, by earlier standards,
dissonant, discordant, and inhuman. Judged by the values of past genera-
tions, our culture seems obsessed with breakdown, splintering, disintegra-
tion, and destruction. Ours is an age not of synthesis but of analysis, not
of constructive hopes but of awful destructive potentials, not of commit-
ment but of alienation [Keniston, 1965, p. 4; see also Sykes, 19641.
 J. Milton Yinger I Anomie, Alienation, and Political Behavior
173
I
In such a context, which is by no means new in human experience, certain revived by Durkheim in The Diuision of Labor (1893) and, much more fully, in
terms acquire unusual salience. Critical words are revived or new ones are invented Suicide (1897). There is some ambiguity in Durkheirn's usages, but the predomi-
that seem to help us deal, both emotionally and intellectually, with the flood of nant meaning is clear: Anomie is a state of society in which norms have lost their
disturbing experiences and observations. If we are not careful, a few "big" concepts power to regulate behavior. A literal translation of the Greek source might read "no
become the code words of an intellectual tradition and are used as explanations rather laws" ( a nomous). I t is deregulation. There is, in Parsons' phrase, a lack of struc-
than as labels for complex situations that require explanation. Our vocabularies tured complementarity in human interaction. Thus defined, anomie is a property of
become involved in the desire to simplify our enormous problems-whether of action the cultural system; it is not a quality of individuals. I t is a near equivalent of
or research. Use of key words becomes the mark of the sensitive person, a demonstra- anarchy, if government is the referent, where no group or political system is granted
tion that he is au courant, if not avant garde. Under such conditions, these words take sufficient legitimacy to govern.
on a great variety of meanings, as they spread from their original base to one related Although Durkheim used anomie most commonly to refer to a quality of
phenomenon after another. In the end they may refer to vast problem areas; they groups with low levels of agreement on norms and values, he occasionally used it to
may be more indicative of a mood of the times and of an intellectual tradition than refer to lack of integration in functions. An industrial crisis, for example, would be an
of a specified referent. indication of anomie insofar as the various activities of a specialized work force did
Anomie and alienation are among the code words of contemporary intel- not combine in ways to achieve mutually accepted goals. If the two meanings are
lectuals. If we do not use them with care-subjecting them to linguistic and method- combined, we would say that a society is anomic when it exliibits a low degree of
ological analysis-we may, unwittingly, illustrate the truth of Walter Kaufman's consensus and a low degree of integration. Merton (1968) has suggested, however,
appraisal: "As it becomes more and more impossible to keep up with developments that a special kind of dissensus, in a context of low integration, characterizes anomic
in different fields, most people feel a growing need for bargain words that cost little societies. He notes that industrial societies particularly have encouraged their mem-
or no study and can be used in a great variety of contexts with an air of expertise" bers to strive for newly accessible goals; at the same time, less attention has been paid
(in Schacht, 1971, p. xlk). This risk is the greater because our key terms have both to teaching respect for, or furnishing opportunities for, culturally approved means.
value-laden and objective components; they are critical appraisals of society and of Anomie, in Merton's sense, exists when agreement on goals is not matched by agree-
individual tendencies at the same time that they are attempts to describe. John ment on appropriate means to those goals, or when many persons are caught in
Horton (1964) may exaggerate when he writes that anomie is "basically a utopian circumstances where approved means do not yield expected satisfactions (pp. 185-
concept of the political right:' focused on the need for social control, while alienation 248). In Merton's words, "An extreme cultural emphasis on the goal of success
is "a utopian concept of the radical left,'' focused on domination; but he makes clear attenuates conformity to institutionally prescribed methods of moving toward this
the need for keeping alert to the critical and evaluative connotations of the words as . .
goal. . Norms are robbed of their power to regulate behavior" (p. 223).
we explore their descriptive and analytic uses. My own preference is to consider this a particularly important kind of anomie
Despite the great frequency with which alienation and anomie are used in in industrial, and probably industrializing, societies, but not to use it as a definition
research and in critical writing, there have been relatively few attempts to subject of anomie. It is perhaps too culture-bound to be useful as a general description of the
the concepts themselves to analysis. (For exceptions to this statement see Keniston, phenomenon of deregulation, of the absence of pattern in the mutual expectations
1965; Merton, 1968; Schacht, 1971; Seeman, 1959.) In his article on Durkheim in for social action.
the International Encyclopedia of the Social Sciences (1968), Talcott Parsons states Anything approaching an objective reading of the extent of anomie has proved
that "Anomie has become one of the small number of truly central concepts of difficult to achieve for two reasons. First of all, the implicit critical and evaluative
contemporary social science." Yet there is no article on anomie in this encyclopedia element in the concept allows each of us to load our own perspectives onto the reading
or in its 1931 predecessor. Parsons refers to the concept only briefly, as do a few process. Second, deregulation as a quality of social systems is intrinsically difficult
others writing for the Encyclopedia. Then are articles on alienation in both encyclo- to measurd. Each of these requires brief comment.
pedias, but the earlier one refers only to its economic-legal meanings and the latter Problems in Definition. Almost ,all commentary on anomie (see, for in-
mainly to its historical uses. stance, DeGrazia, 1948b) assumes that it is an unfortunate quality; but such an
We are left, then, with highly significant words of broad and open-ended assumption implies a comparison with some other social arrangement that is prefer-
meanings. Some of our attention will have to be given to questions of usage and able. When specific values are stated, anomie can take on a .desirable connotation.
definition as we seek to assess the importance of anomie and alienation for politics. "There was probably a high level of anomie in the American colonies in 1775-a
necessary aspect of the disengagement from the British' Empire. Perhaps increase in
Anomie
I anomie is an inevitable part of the process whereby old tyrannies are broken and new
values given a field for growth" (Yinger, 1965, pp. 189-190). Even so, anomie is
undoubtedly associated with chaos and human suffering under some conditions. What
Although anomie (or "anomy," in an English spelling) was used at least as
early as'.the' sixteenth century to refer to a condition of disorder, doubt, and uncer- we lack, however, are clear indications of the consequences of given levels of anomie,
tainty (Merton, in Clinard, 1964, p. 226), it nearly disappeared from view until experienced by specified populations or subpopulations, under particular circum-
 J. Milton Yinger I Anomie, Alienation, and Political Behavior

stances. In the absence of such indications, values and fears play a large part in our social distinction may well signify low anomie; the lack of such memberships indicates,
assessments of the extent and the effects of anomie. Violence, discord, and deviation at least under some conditions, high anomie.
are highly visible and readily seen as clear indexes of anomie. In a somewhat nostalgic Various forms of collective behavior may be good indexes of anomie. I t may
Gemeinschaft-Gesellschaft tradition, we may tend to compare contemporary societies be high, for example, in societies where religious sects or cults are proliferating, where
with partially imaginary models of small, smooth-running, integrated rural societies, revitalization movements express efforts to recapture a lost sense of cultural identity,
where all members agreed on goals and means and were well socialized to their where third-party or other "splinter" political groups indicate a reduced sense of
positions. ! allegiance to the traditional or established political processes,
Measures of individual attitudes and values-if they are clearly recognized
In my, judgment,
- - the extent of anomie in a society is a crucial variable for the
student of politics and of other social processes. I t will become a powerful concept, as indexes of anomie, not manifestations of it-are likely to be among the simplest
however, only if we can learn to measure it reliably and validly, overcoming our ideo- ways to assess the extent of deregulation. Levels of alienation, as we shall see, are
logical readings of the evidence. The tendency today may be to overlook signs of often used in this way. I n contrast, the extent of tolerance, as a value or expressed
consensus and integration, to forget the invention of various techniques of mediation in behavior, might indicate the degree to which members'of a society share a cultural
and arbitration, to disregard the emergence of new norms out of conflicts that weak- perspective. A good index, of course, changes in close correlation with the variable
ened the old norms. Since many periods in history characterized by sharp value it indexes-as a column of mercury rises or falls in a glass tube when temperature
disagreements and conflict have been followed by periods of greater consensus and changes. Psychological and social indexes are seldom so valid; they must be used in
integration, there are clearly norm-building as well as norm-destroying processes in conjunction with other measures.
human societies (Becker, 1960). Were this not the case, we would have to think in More direct measures would go Beyond the aggregation of individual attitudes
terms of an analog of the second law of thermodynamics, with social systems inevi- or other tendencies to study the extent to which a group of interactors live in a
tably running down from some*presumed aboriginal state of integration. We will be similar cultural world. This procedure has been followed by Jessor and his associates
wise to think of existing levels of anomie and the trends as qualities of society to be (1968) in their study of a tri-ethnic community. A sample of the community's mem-
studied, not assumed. References to the reciprocal quality, integration, can be of great bers were asked to indicate whether they thought that certain activities were pre-
value in identifying anomic situations (Freedman and others, 1956, pp. 170-203). scribed, permitted, or proscribed. The modal answers and the variances were
Problems in Measurement. There are strong tendencies, as we shall see, to calculated. The higher the proportion of responses in the modal category and the
reduce anomie to a psychological term. (We will examine such uses in the next smaller the variance, the lower the level of anomie. By this process Jessor and his
section, since they are conceptually closer to alienation than to anomie as originally associates were able both to define the concept operationally and to treat it as a
defined.) Even among those who have kept a strictly structural meaning, however, variable, with a range of possible scale values (see especially pp. 244-258). I t may
measures of levels of anomie have been only poorly developed. Research on the effects be well to repeat that this is a group measure; it represents a loss of pattern in mutual
of anomie on politics (or of politics on the level of anomie) will remain rudimentary expectations for social a:tion. Where the modal category is small and the variance
until more adequate measures have been designed. Doubtless some combination of around it large, anomie is experienced even by those in the very center of the mode
direct measures, indexes, and individual items aggregated into group measures can as well as by those who are distant from it.
produce a valuable scale of anomie. If Merton's (1968) somewhat narrower definition of anomie is used, measure-
Durkheim used international and economic crises and family instability as ment requires, as he notes (p. 229), the combination of several sets of data: informa-
indicators of anomie. Crime rates and other forms of "deviation" have been used tion on exposure to cultural goals and to norms that regulate efforts to achieve those
as indexes. (One should not be surprised that anomie thus measured is associated goals; acceptance, within various segments of a population, of those goals and norms;
with deviation. And any inference of causation, running in either direction, is un- the extent to which the goals are accessible-the jobs, schools, and other supports in
justified when such a procedure is followed.) Data on the size and cost of "social the opportunity structure; and the extent of discrepancy between accepted goals and
control" structures-police forces, for example-may produce a useful index of their accessibility.
anomie. Records of the availability and use of mediation and arbitration procedures These references to problems of measurement may help to explain why
are valuable. They indicate arrangements for preventing value disagreements from andmie has been poorly developed as a research concept. Most studies use a few
spiraling into more serious forms of conflict. They could be read alongside records individual characteristics or attitudes as an index of anomie, or even let the distinction
I
of group conflict. between a group and an individual property slip away entirely. An adequate social
Anomie may be more characteristic of societies with heterogeneous popula-
tions; it therefore could be "measured" by data on cultural origins; ethnic identity; 1 psychology of political behavior is blocked by this process, for what we need are
clear analytic distinctioqs among elements, a clear designation of individual and
religious, racial, and lingual diversity; and other dembgraphic information. A related group levels, and then studies of the results of their various combinations. This pro-
dimension is the group structure of a population. Cross-cutting memberships that I cedure is followed in the study of many social scientific problems. We have no diffi-
increase communication, contact, and shared perspectives across important lines of culty distinguishing between the group fact of a wealthy society and a particular
I
 Anomie, Alienation, and Political Behavior 177
176 J. Milton Yinger
who know only one way to spell a word), which are currently used more often to
individual's wealth. It is clear that to be poor in a wealthy society is a different
refer to a property of individuals than of groups. We are not likely to be much better
experience from being poor in a society characterized by poverty. The same distinc-
off if we shift to German and compare the overlapping meanings of HegelysEntaus-
tion, though often disregarded, is essential in studies of anomie and alienation. In
serung renunciation, surrender), Marx's Entfremdung (estrangement, alienation),
most cases it is more diffic~iltto measure a structural property than an individual
and Weber's Entrauberung (disenchantment, discovery of the senseless quality of the
property. More data are required; problems of reliability are greater. Any one or two
measures are likely to yield an incomplete picture of the social structure. Perhaps a world).
combination of several of the sets of data I have suggested above, however, can begin In the face of such complexity and confusion, perhaps the wisest thing for the
writer-or at least the reader-is to turn quickly to the next chapter. But there is
to produce an adequate measure of anomie. Such a measure remains, several decades
after Durkheim, a primary research need. something about the word alienation that wili not let go of the modern mind. I t
It is important not to confuse anomie with possible responses to it. Perhaps draws us into a huge and indefinite but highly significant aspect of the human situation.
the most frequently cited part of Merton's work on anomie is his analysis of types of Whether as a critical and polemical term or as an analytic construct, we cannot set
responses. Upon experiencing the disjunction between means and ends which Merton it aside. It would simply reappear with a different spelling. Although it seems im-
defined as anomie, a person may act in a number of different ways. His response possible now to get a reasonably precise definition of alienation that would be widely
depends upon the severity of the experience, his other inclinations, his social location, accepted, we can at leasf sketch its meanings as a "sensitizing" concept. The tasks of
and other variables. Merton ( 1968, pp. 193-2 11) lists five possibilities: A person may research, then, are to take some segment of the concept, to design ways to measure
continue to accept the culturally approved means and ends as best he can; he may that segment, to study its relationships to other segments, and to examine the ways in
"innovate" in means (for instance, he may steal to try to achieve a culturally urged which the various forms of alienation are expressed in particular contexts. Only
standard of living) ;he may act ritualistically by accepting the means, even if the ends gradually will we discover which of the many individual tendencies that various
are thus denied; he may retreat into apathy or neurosis, trying to solve the contra- authors have chosen to identify as fonns of alienation actually "belong together," on
diction by inaction; or he may rebel, in an effort to create a new pattern of means the class if not the species level (in the sense that the bat, the whale, and the lion
and ends without the contradictions. Although this paradigm has great relevance for belong together as mammals).
politics, it has not been systematically applied to political behavior. It has been most Much of the contemporary interest in alienation stems from Marx. His em-
intensively used to examine the varieties of delinquent behavior, a topic we cannot phasis was on the work situation but spreads easily to politics, religion, and other
explore here (see Clinard, 1964; Cloward and Ohlin, 1960; Cohen, 1965) ;but Me& elements of the "establishment." "What constitutes the alienation of labor? First, that
ton's scheme is valuable because it requires a clear distinction between anomie and the work is external to the worker, that it is not part of his nature; and that, con-
behavior and thus encourages us to search for individual factors-alienation among sequently, he does not fulfill himself in his work but denies himself, has a feeling of
others-that must be brought into the system of explanation. misery rather than well-being, does not develop freely his mental and physical energies
but is physically exhausted and mentally debased. The worker, therefore, feels himself
Alienation at home only during his leisure time, whereas at work he feels homeless7' (Marx,
1844, pp. 124-1251. .
In a few hours' time, one can identify hundreds of articles and books that use Fromm, another major source of attention to alienation, focuses somewhat
the term alienation. Its roots go back hundreds of years in theological work, at least more on the individual experience than the external force, but his definition is not
to Hegel in philosophy and to Marx in political and economic thought. I t is promi- inconsistent with Marx's: "By alienation is meant a mode of experience in which the
nent in existentialist writing, both religious and secular, in some forms of psycho- person experiences himself as an alien. He has become, one might say, estranged from
analysis, in social science, in literature, and in many polemical and critical studies of himself. He does not experience himself as the center of his world, as the creator of
modern society. Perhaps its currency in the United States today can be shown by his own acts-but his acts and their consequences have become his masters, whom
noting that three of the five persons giving nominating speeches at the 1972 Demo- he obeys, or whom he may even worship" (Fromm, 1955, p. 120).
cratic party convention referred to alienation. There is clearly a polemical quality in the use of alienation by Marx and
Since the boundaries set by this word, broad as they are, are indistinct, we Fromm. Alienation is bad, and it is largely the product of bad institutions. In trying
must also note the great frequ&ncywith which other terms are used to refer to similar to asess its value as a concept for political psychology, however, we perhaps ought to
phenomena. There are various degrees of overlap in the meanings of alienation and remind ourselves that most of us applaud certain alienations. We can evaluate them
the meanings attached to pessimism, powerlessness, cynicism, "ressentiment" (as only in terms of their influence on the attainment of stated goals; "if most men are
discussed by Max Scheler), despair, feelings of external (in contrast with internal)? mere caricatures of what they might be, it is quite possible that in order to become
control, existential anxiety, authoritarianism, lack of efficacy, distrust, credibility gap,
more humane they must first become more estranged. This notion finds support not
apathy, necrophilia, meaninglessness, normlessness, self-estrangement, and isolation.
only in Hegel but also in some of the world's great religions" (Kaufman, in Schacht,
To make our task even more complicated, we must add the words anomia, anomy,
1971, pp. liv-lv).
and anomie (like Mark Twain, social scientists have no sympathy fdr ignorant people
 J. Milton Yinger Anomie, Alienation, and Political Behavior 179
Alienation, as I shall use the term, implies a time dimension: one experiences moves into new and more complicated personal relationships, he may attempt to
the loss of a relationship or value. One cannot be estranged without having been handle his "exile" by interpreting his condition as the human condition. Having
attached. The term is useful only if that from which one is alienated is specified. The projected one's own feelings of anguish onto "society," one cannot afford to permit
politically alienated feel estranged from political structures and processes that for- any attachment, any trust of others, because this might expose the true location of his
merly were accepted as valuable means to desired goals. This is a rather severe limi- difficulties. Keniston ( 1965, pp. 53-54, 61 ) describes the situation well :
tation of the concept, but I think it is essential. If the feeling of powerlessness is an
aspect of alienation, we must distinguish it from the experience of powerlessness as In all young men and women the advent of adulthood releases immense
an endemic fact of life. The causes and consequences of alienative powerlessness and new energies and potentials, which in most are centrally involved in es-
endemic powerlessness are quite different. tablishing new intimacies with the opposite sex. This new learning is
I t is important to ask whether alienation is a diffuse condition, unrelated to seldom smooth; but when itlis severely blocked by unresolved needs and
new structures and goals, or whether it implies a kind of zero-sum situation, in which frustrations from the past, it takes but a slight catalyst-and often no
alienation from one group or set of standards implies integration with others. Most catalyst at all-to transform these energies into rage, scorn, and aggres-
commonly, alienation is thought of only in the first or total sense. Alienated persons sion, often symbolically directed against those who have stood in the way
are said to be without goals, negativistic, distrustful. Keniston (1965) found that the of full adulthood. . .. The alienated prefer to maintain that the world is
extremely alienated young men whom he studied wholly distrusted any affirmation. a dark and gloomy place rather than to say simply that they are pessimis-
"Theirs is an ideology of opposition, and the world offers so many targets for their tic. Psychologically, it is important that they see the world, the state of the
universe, as causing and justifying their own pessimism.
repudiation that they have little energy left for the development of affirmative values"
(p. 69). This we might call acute alienation. We are dealing, however, with a varia- The structural and the interpersonal sources of alienation are not, of course,
ble, not an attribute. Few people are cut off from all group attachments and goals. *
mutually exclusive, and many authors (Barakat, 1969; Keniston, 1965) emphasize
One who is alienated from an established church may join a sect--or a golf club. The their interaction. On an analytic level, nevertheless, the distinction is important, for
politically alienated may join protest movements. Some of the alienated young see the balance of sources in particular cases may critically influence the strength and
themselves as sharing a contraculture or achieving integration through "Consciousness direction of behavior-including political behavior-that expresses the alienation.
111." In short, measures of distrust, lack of efficacy, and the like, should always refer The great complexity and diversity of phenomena that have been labeled
to particular structures and values. Efforts should be made to discover the strength alienation have prompted a number of scholars to describe its several "dimensions."
and nature of the attachments that alienated persons do have. If there are none, this Such an effort is often based on the assumption that some unity underlies the several
is a special case,that requires separate analysis of its causes and effects. forms. It is not enough, of course, that they happen to be called by the same name
Alienation, then, is the experienced loss of a relationship and of a sense of to affirm this unity. A number of factor analytic, correlational, and scaling studies
fiarticipation and control, with reference to prevailing social structures. It may or have investigated the question empirically, with helpful if not definitive results (see
may not be associated with new attachments designed to counter the loss. Self-aliena- Middleton, 1963; Neal and Rettig, 1963, 1967; Pearlin, 1962; Simmons, 1966;
tion is not a different phenomenon but is an internal view of the same loss; for Streuning and Richardson, 1965). In my judgment, however, we are still at the stage
central to self-definition are the attachments to the interpersonal and institutional marked by an intuitive feeling that certain individual tendencies are somehow re-
world that surrounds each person. lated: they come from the same causes, they reinforce one another in many instances,
Sources and Dimensions. It would take us too far from the study of the and they are manifestations of the sense of exile that some people experience.
political implications of alienation to review the literature dealing with it as a Whatever the outcome of efforts to find an underlying unity, it is clear that
general concept, but three questions require brief comment: What are various the term alienation is now used to refer to a wide variety of different tendencies. In
explanations of the origin or sources of alienation? What are its several dimensions, a seminal paper, Melvin Sccman (1959, pp. 784-790) has described five logically
,or forms? And how may these be measured? distinct meanings in the literature on alienation: Power1.essness is "the expectancy or
Two traditions, which we can somewhat inexactly call the Mamian and the probability held by the individual that his own behavior cannot determine the
Freudian, furnish many of the explanations of the sources of alienation. In the first, occurrence of the outcomes, or reinforcements, he seeks." When experiencing mean-
an oppressive economic or political structure produces the sense of estrangement. inglessness, "the individual is unclear as to what he ought to believe." Normlessness
Many recent studies document the relationship, if not the exact causal connection is "a high expectation that unapproved behaviors are required to achieve given
(Blauner, 1964; Lipsitz, 1964; but see also Seeman, 1971) . Using a somewhat broader goals." Individuals are isolated "who assign low reward value to goals or beliefs that
sociological frame of reference, many authors see alienation as a result of frustrations are typically highly valued in a given society." And self-estrangement "means to be
produced by the social strucutre or as a consequence of its anomic qualities (Parsons, something less than one might ideally be if the circumstances in society were other-
1950, pp. 256-267; Etzioni, 1968; Fromm, 1955) . Those in the Freudian tradition
wise."
emphasize the interpersonal, and particularly the childhood experiences as sources These terms have been used frequently in later research and commentary;
of alienation. If a child feels severely cut off from necessary emotional supports as he but unfortunately, precisely shared meanings have not evolved. Powerlessness, for
 180 , J. Milton Yinger Anomie, Alienation, and Political Behavior 181

i
1 example, which is perhaps the critical variable for the,student of politics (or at least Before we comment on the scale that Srole developed, it may be helpful if we
the most researched), is defined in various ways. Political alienation, to Thompson examine possible confusion resulting from such word transpositions. This shift
and Horton, is the reaction to perceived inability to influence events-not, as defined occurred, in all probability, because of the frequently observed, if not well-measured,
by Seeman, as perceived inability itself (Horton and Thompson, 1962; Thompson correlation between anomie and alienation. Anomic settings produce alienation; and/
i
and Horton, 1960). Despite such differences, the identification of several aspects of or alienated individuals increase the level of anomie. To some degree, one can be
II alienation is a step toward understanding the complexity of the phenomena encom- used as an index of the other. The closeness of this relationship, however, makes the
I1
passed by the term. drawing of careful analytic distinctions all the more necessary. Some of the most
i important questions arise in connection with nonalienative behavior in highly anomic

1'
ij Measurement. Another necessary step is the development of reliable and
I valid measures for each aspect, or a scale of some combination of aspects. We can only situations and with alienation among those who live in relatively eunomic settings,
illustrate the work being done along this line. Dozens of measures have now been characterized by strong normative agreement.
developed, through the use of intensive case studies (Keniston, 1965) ;content analysis In fields where one term is used as an index of another, and where measure-
ii
(Taviss, 1969) ; and especially questionnaires, in which a number of respondents are ment is, in any case, imprecise and conceptualization shaky, there is a risk that the
index will be confused with the variable it is used to measure. Since anomie as a

1~
asked to express their degree of agreement or disagreement with statements believed
to indicate some element of alienation (Dean, 1961; Finifter, 1970; Mizruchi, 1964; structural fact is difficult to measure, we run the risk of losing sight of it when we
Seeman, 1959; Srole, 1956). (For a discussion of psychobiography as a method, see use the same word for a psychological variable. To be sure, Srole used a different
Chapter Eleven in this volume.) Some studies simply record the answers to a series spelling; but in a large share of the papers that use his scale or discuss it, the spelling
!Il reverts to anomie or anomy. In what turns out, after the first paragraph, to be a
i!; of questions. The Harris poll, for instance, notes that, in 1972, 50 percent of a na-
valuable paper: McClosky and Schaar (1965, p. 14) write: "Almost all work on
Ill tional sample agreed with the statement "People running the country don't really

1 care what happens to people like me"; in 1968, 36 percent had agreed and in 1966, anomy to date has employed 'sociological' explanations to the virtual exclusion 'of
28 percent (Plain Dealer, June 19, 1972, p. 18-B). Other questions in the poll were all others. This paper claims that the standard sociological theory of anomy has
serious conceptual weaknesses and cannot satisfactorily account for many of the

1
"The rich get richer and the poor get poorer," "What I think doesn't count very
much," "I feel left out of things going on around me," "People who have the power relevant facts." This is a curious statement, lamenting that a sociological term has
are out to take advantage of me." When answers to these questions are combined into been used sociologically. In point of fact, it seldom has been. Reduction to individual
an index, 47 percent of the 1972 national sample were recorded as alienated, compared measures; often without explicit intentions of aggregating into a group measure, is
with 40 percent in 1971. It is interesting that among supporters of President Nixon far more common. Thinking of an analogous problem in the physical sciences, I
the percentage was 36; among Senator McGovern's it was 53. doubt that one is likely to find in a journal of organic chemistry a paper by a phy-
i sicist lamenting that chemists have been studyingthe molecular properties of various
Several authors have designed scales and tested them for unidimensionality.
combinations 3 C, H, and 0 without at the same time studying their individual
I~ Middleton (1963) found that answers to questions dealing with several of the mean-
ings of alienation isolated by Seeman fell into a Guttman scale, with a coefficient of
reproducibility of .90. Those who agreed with the norrr~lessnessstatements, for
atomic properties. Nor will we find such a journal using the same word for carbon
and alcohol (or oxygen and sugar, for the teetotalers). Of course, sociological theorists
example, were likely also to agree with statements referring to work ejtrangment, cannot account for many of the facts of behavior by studying anomie alone, although
I social estrangement, powerlessness, and meaninglessness. Since the number of items some sociologists are likely, following Durkheim, to forget the analytic quality of the
I
I used in preparing the scale was small and the sample was drawn from one city only, discipline. Psychological factors must be brought into the interpretation. At this stage,
~ though, I wish that lexicography were more highly developed among social scientists.

~
we cannot be certain of the general usefulness of this scale; but it represents a valuable
I step toward more adequate measurement. I t might lead to the use of alienation as a general "focusing" concept, referring to
As noted, a number of writers (Davol and Reimans, 1959; Lowe and Daman- ' feelings of estrangement within individuals, and to the use of anomie to refer to a
kos, 1968; McClos$ and Schaar, 1965; Nelson, 1968; R. A. Wilson, 1971) have coadition of deregulation, of loss of pattern in member interactions, within groups.
I employed anomie as a rough synonym for alienation or, more precisely, have taken a This wish stems from no passion for order (in this somewhat disorderly fellow! but
term that originally referred to a property of groups and used it to refer to an from a passion for conceptual clarity that wiIl help us go about our business of under-
individual, psychological property. It seems appropriate to discuss the use of anomie standing human behavior.
! To return to Srole's (1956) scale of alienation (anomia), we may be wise to
(or anomla or anomy) as an individual quality here, in a section dealing with
i alienation, in order to distinguish such use from anomie defined as a quality of list the questions used, for they have been employed in many studies. The scale is
I groups. This redefinition of Durkheim's term happened first, perhaps, in somewhat composed of only five statements, and respondents are asked to express their agree-

I incidental references by Robert MacIver and David Riesman; but the usage became
widespread only after the presentation, and later publication, of a paper by Leo
Srole (1956). Ifi his terminology, anomie (changed to anomia in the published ver-
ment or disagreement with each statement: "There's little use writing to public

'Put see the comments of Srole (1965) and Nettler (1965), with the response from
McClosky and Schaar.
sion) refers to an individual's "self-to-others alienation."
182 J. Milton Yinger Anomie, Alienation, and Political Behavior 183
officials because often they aren't really interested in the problems of the average
shall mention only one more, a scale designed to measure political alienation. Draw-
man." "Nowadays a person has to live pretty much for today and let tomorrow take
ing on the work of Seeman, but modifying his list of types of alienation, Finifter
care of itself." "In spite of what some people say, the lot of the average man is getting
(1970, pp. 390-391) posits four forms of alienation toward the polity: powerlessness
worse, not better." "It's hardly fair to bring children into the world with the way ("an individual's feeling that he cannot affect the actions of the government") ;
things look for the future." "These days a person doesn't know whom he can count political meaninglessness ("political decisions are perceived as being unpredictable") ;
perceived political normlessness ("the individual's perception that the norms or rules
I t is not clear just what this scale measures, and Srole included no tests of
intended to govern political relations have broken down") ; and political isolation
validity. I t has not fared well under methodological examination (see, for instance, ("rejection of political norms and goals that are widely held and shared by other
Carr, 1971; Meier and Bell, 1959; Miller and Butler, 1966; Proctor, 1971) . Because members of a society"). Using the data from the national probability sample em-
the items are simple and clear, however, and because many researchers have agreed
ployed by Almond and Verba for T h e Civic Culture (1963), Finifter found twenty-
with Srole's estimate that they seem to have "face validity," the scale has been widely
six questions related to political alienation. When factor-analyzed, twenty-one of
employed. The relationships of "anomia" to social class (Bell, 1957; Mizruchi, 1960; these questions fell into two distinct clusters-a powerlessness dimension and a per-
Nelson, 1968; Simpson and Miller, 1963) to prejudice (Angell, 1962; Hamblin,
ceived normlessness diqension.
1962; Lutterman and Middleton, 1970; Roberts and Rokeach, 1956; Srole, 1956),
Finifter's work is a useful construction of a specifically political alienation
and, as we shall note in a later section, to politics have been extensively studied.
scale. Our predictive power is increased by thus designating a precise referent for
Another widely used scale that is in the alienation domain, although the term
respondents' alienative attitudes, rather than dealing with alienation in general. One
is not explicitly employed, has been developed by Rotter (1966). He uses a twenty-
of the critical problems for research is to discover the conditions under which various
item scale (six of the statements are "filler" items) to measure what he conceptualizes
forms of alienation converge and when they diverge. When such information is
as a generalized expectancy for internal or external control. This expectancy is
combined with measures of the separate modes of alienation, we have a way of avoid-
judged important because reward and reinforcement, in the psychological sense, are
ing a picture of alienation as a uniform and simple tendency (Seeman, 1972). Martin
influenced by the degree to which a person perceives them as "contingent on his own
and Bengston (1971) set five basic institutions (political, economic, educational, reli-
behavior or independent of it." Respondents are asked to choose between two state- gious, and familial) against the five modes of alienation designated by Seeman.
ments in each of twenty-nine sets. For example: "Many of the unhappy things in Using this twenty-five-cell matrix, one is able to compare, for example, feelings of
people's lives are partly due to bad luck" or "People's misfortunes result from the political with educational powerlessness, or perceptions of normlessness across the
mistakes they make." Four of the pairs of statements deal with politics. For example: five institutions. For all three of the age groups into which Martin and Bengston
"The average citizen can have an influence in government decisions" or "This world divided their respondents, political alienation (accumulating the five modes) was
is run by the few people in power, and there is not much the little guy can do about highest; economic alienation was second. Of the five modes, powerlessness was highest
it." and meaninglessness second, again for all three age groups. Such findings might not
The concept of "internal-external controly' is mentioned here because, like be repeated with different samples, using different measures; but the procedure is a
scales of ."anemia," it measures one aspect of alienation as I have defined it. Its valuable way to speed the "mapping" of alienation. I t helps us avoid the assumption
meaning is close to that of powerlessness. Rotter himself notes: "The concept of that the tendency is uniform and pervasive-an easy assumption among those for
alienation . . . does seem related at a group level to the variable of internal-external whom alienation is more a polemical than a descriptive term. I t seems impossible-
control. The alienated individual feels unable to control his own destiny. He is a and probably unwise, even if possible-to remove the critical element from the con-
small cog in a big machine and at the mercy of forces too strong or too vague to con- cept. What we need is a parallel in use in which alienation is measured in sufficiently
trol" (p. 3 ) . The internal-external scale, or variations of it, has been used in studies objective terms that its use in research can be strengthened.
of responses to discrimination (Coleman, 1964b, 1966a; Gurin and others, 1969) ; in Only with improved measures will we be able to discover how much projection
studies of learning (Bullough, 1967; Gurin, 1970; Seeman, 1963, 1967a) ; and, as there is in the persistent theme of alienation. If a person does not know who he
we shall note later, in studies of politics. is, where he belongs, what he believes, he may try to deal with the resulting discom-
Space limitations prevent discussion of the many other scales of alienation. We fort by declaring that alienation is widespread. If such declarations of alienation
occur, then other persons-perhaps less alienated themselves, but hearing from
In a more recent version, four items have been added to the scale; but these have many articulate people that modern man is alienated-seek to explain the presumed
not yet been widely used in research. Since the new questions are also all stated in the same
direction (with agreement indicating anomia), the scale has not avoided problems of response widespread alienation. They believe that they ought to believe it is there, since many
set. The four additional items are: "Most people really don't care what happens to the next fervent voices have told them so. (And among the ambiguous signs of human
fellow." "Next to health, money is the most important thing in life." "You sometimes can't behavior, perceptions can be selective to produce the necessary result.)
help wondering whether anything is worthwhile." "To make money there are no right and I do not know that this is true but am simply emphasizing the importance of
wrong ways anymore, only easy and hard ways." (See Survey Research Center, Measures of measurements that persons from different perspectives can agree upon. There may
Social Psychological Attitudes, 1969, p. 175.)
be a general underestimate of alienation and anomie in the contemporary world-a
 184 J, Milton Yinger

tendency to see order where it is lacking, of feigning meaning and coherence where
they are in fact weak. Or, even if some oversensitive" persons exaggerate today, they
may be alerting us to nascent, latent, but powerful trends leading to more alienation
and anomie. Such warnings may help to set reverse processes in motion. In either
case, it can be argued that one of the key fu~lctionsof the intellectual-and the artist
-is to anticipate serious difficulties and help to set self-negating prophecies in
motion. We need also to know, however, when the opposite occurs, when overesti-
mation promotes self-fulfilling prophecies. I t seems likely that the most effective
policies will be based on accurate appraisals.

A Field-Theoretical View

Despite the large number of valuable studies of alienation and anomie, we

have still, it seems fair to say, some distance to go before we achieve clarity of
definition and precision of measurement. We are even further from research that
combines the two in order to examine behavioral outcomes under a range of condi-
tions. Alienation is not behavior but a tendency. Even those who are strongly
alienated do not lack other predispositions that can be activated in some contexts.
Anomie is not a uniform fact but varies in its implications for behavior, depending
upon the tendencies-including the levels of alienation--of those who experience it.
It is not enough, of course, that we combine only these two variables into a system
of explanation, for there are many contextual facts besides anomie and many
individual facts besides alienation. In a formal research design that is still too simple
but may illustrate the necessary multivariate approach, we might combine four
variables, each dichotomized. As Figure 1 shows, there are sixteen possible outcomes
(of individual political behavior or of religious response, for example) from such a
system. Knowing that a person is high or low in alienation is only a start toward
understanding or predicting his behavior. A highly alienated, upper class person in a
wealthy society, for example, might be apathetic toward politics; a similar person in
a poor society might be deeply involved in a counter-revolutionary movement.
There are few problems with the methodology of such an approach. Regres-
sion analyses, multiple-classification analyses, and the like, can be applied to much
more intricate designs. What we lack are adequate data. "There is not a single em-
pirical investigation of anomie and deviant behavior," Merton wrote a few years
ago, "that has succeeded in mouti~lga research design that syst~maticully,rather
than impressionistically and qualitatively, includes simultaneous analysis of col- ,
lectivity, subgroup, and individual attributes in relation to deviant behavior" (in
Clinard, 1964, p. 240). That statement is not literally true today. For illustrations of
studies that combine individual and group measures, see Jessor and others (1968)
on deviation; Warner and Defleur (1969) on discrimination; Rushing (1971) on
class differences; Finifter (1970) and Kirby (1971) on politics. There is, however,
still a shortage of systematic social-psychologicalstudies that examine the independent
and interactive influences of anomic settings and alienative tendencies.
For our purposes in this chapter, it may be adequate to consider alienation
and anomie as independent variables. Our attention is directed toward their separate
and combined influences on politics. In fact, of course, their sources also require
explanation. Among these sources are various influences that flow from the conse-
 186 J. Milton Yinger Anomie, Alienation, and Political Behavior 187

quences of anomie and alienation. The effects of anomie and alienation become entitled "Political Effects of Alienation and Anomie," but that topic cannot be
causes in a feedback process. Thus, in an analysis of an evolving system, we might sharply separated from its counterpart, "alienative and anornic effects of politics."
.designate rapid social change (especially when it is experienced at different rates), Some of our attention therefore will be directed toward research in which political
growing social heterogeneity, persistent war, and the rapid expansion of wants as institutions and processes are treated as the independent variables.
some of the sources of anomie. Alienation may spring in part from these causes (just There is much more empirical work dealing with alienation than with anomie
as it may contribute to them) but also from stressful socialization. Shifting focus, we defined as a structural fact. I t is sometimes possible, however, to infer measures of or
see anomie and alienation as the independent variables. They lead to various out- references to anomie from analyses of alienation. A study of powerlessness, for
comes, depending upon various intervening conditions. These outcomes, in turn-be example, may give indications of the lack of opportunities to achieve culturally
they opposition, apathy, withdrawal, or other '~responsesY'-feed back into the system approved goals, which is anomie in the Merton sense. A study of disenchantment
out of which they came. One of the most important and difficult questions relates to and the withdrawal of legitimacy, a second major form of political alienation, may
the conditions under which the behavioral outcomes heighten or reduce the aliena- note the general normlessness of the social environment, that is, its anornic qualities
tion and anomie from which, in part, they flowed. This field can be sketched as in the Durkheim sense. We will call attention to these references or inferences when
shown in Figure 2, with the 2 on the return arrows indicating that "consequences" they occur, even if the explicit subject of a study is political alienation.
can either increase or decrease their own "causes." (A political or religious movement Among the many questions that might be examined with reference to the
springing from an anornic setting among alienated persons msiy "cure" their aliena- political effects of alienation and anomie, several seem of special importance: In
tion by reducing the normlessness of the environment, or it may divide a society what ways are political effects related to religious or other possible effects of aliena-
further and compound the frustrations of its members.) tion and anomie? How do alienation and anomie affect political participation? How
do studies of mass society contribute to our understanding of the interactions of
politics, anomie, and alienation? To what degree can political deviation, to the left
+ or to the right, be accounted for by alienation and anomie? Each of these questions
deserves extended examination. Our hope will be, in the limited space available, to
Withdrawal state the issues in a way that will help to clarify their discussion and promote further
study.
Religion or Politics for the Alienated? With a little change of language, the
following description could be applied to many different kinds of movements at
+Opposition many different times in history.

Once upon a time of great turmoil, people came together sharing their
outrage over the evils and corruption of the world. Together they read
the works of a prophet of a century before, discoursed upon them and
+ added their own commentaries. Rejecting the venal luxury of the exalted,
these elect, seeing the true light, embraced a life of poverty. The present
age must be coming to an end so they dedicated themselves to ending it,
thus ushering in the coming Last Age in which the pure would be united
Studies of the relationship of politics to alienation and anomie seldom can
in brotherly communion and the present rulers chastised, punished, and,
be put into such a system, since most of them deal explicitly, and appropriately, if necessary, eliminated. The teachings spread. Multitudes were fired with
with some small part of the field. As long as we recognize the level of abstraction the new faith. Established powers pronounced against the dangerous new
on which they work, we can heighten our understanding of the psychological and teachings and persecuted the heretics. The believers, in turn, heightened
sociological sources of political behavior. However, if we try to see them in the their fury against the authorities and sharpened their determination for
context of this larger field-theoretical statement, perhaps we can avoid "the fallacy apocalyptic change. They must actively bring about the New Order
of misplaced concreteness" that Whitehead warned aginst. I t is rather easy to forget [Endleman, 1972, p. 31.
that one has abstracted from a complex system.
This description appears in a review of a book dealing with contemporary crises in
Political Effects of Alienation and Anomie American universities, but it refers to thirteenth- and fourteenth-century sectarian
movements in Europe. In the prevailing interpretation (see Yinger, 1970, ch. 14-15),
If the preceding discussion has not sharply delimited the meanings of anomie religious sects are seen as expressions of alienation, as protests against societies ex-
and alienation, it may at least have alerted us to the cautions necessary in applying perienced as frustrating and anornic, although "loss of faith" and "deprivation" are
these concepts to the analysis of politics. We turn now to that task. This section is more likely to be the terms used. "The last root of all sectarianism lies in the aliena-
188 J. Milton Yinger I Anomie, Alienation, and Political Behavior 189

tion of some g o u p from the inclusive society within which it has to carry on its life" into political forms or political movements into religious sects. Different aspects of
(Stark, 1967, vol. 2, p. 5). alienation are made operative in the two situations. And religious and political move-
If alienation can be among the sources of either a political or a religious ments are often quite different in their implications for society and for their own
movement, additional evidence must be brought in to determine when one or the. members.
other will occur. To be sure, a movement may be a mixture. Whether Gne is referring Despite these differences, however, there are also important similarities. The
to the seventeenth-century English Levelers, for example, or to the post-World Wax protest movements emphasize the normative breakdown of the social order; they
I1 Soka Gakkai movement in Japan, or various aspects of the current Black Power seek to increase the sense of alienation of their members from that social order. And
movement, both religious and political elements are apparent. In some contexts, how- 1
then they promise the way-the way to a new arrangement of things here or here-
ever, they are competitive, or at least mutually exclusive. H. Richard Niebuhr after and to a source of meaning and power.
attributed the relative absence of sectarian protests in the West during the last Political Participation. It is sometimes assumed, perhaps too quickly, that
century to the availability of radical political movements. I t is significant that in estrangement from political structures, personnel, and policies, or a high level of
recent years there have been radical sectarian protests, particularly among American normlessness among the members of a society leads to apathy, political withdrawal,
blacks, who feel relatively powerless politically. There has also been some conservative I
and a low level of political participation. Yet it is also observed that the politically
religious revival among those alienated from the trends of contemporary life. In a I
estranged may be swept up in enthusiasm for a political movement. Also, in some
recent study, as yet unpublished, I found that university students in Japan, Korea, societies at least, periods of political crisis (one element of which is low normative
Thailand, New Zealand, and Australia consider religious activity more effective than agreement on political matters, or anomie) produce extraoidinarily high political
political activity in dealing with man's most serious problems. In the sample of 850, participation. Here, as elsewhere, the need is for a careful specification of the condi-
I
51 percent and 46 percent responded favorably to the two questions dealing with tions under which these two different effects occur.
political efficacy; 65 percent and 79 percent responded favorably to the two questions The results reflect, in part, different meanings of the term political alienation.
I
dealing with religious efficacy. I Olsen (1969), for example, distinguishes between political incapability and discon-
I
Such items leave us with the question: When will the alienated turn to tentment. Lane (1962) sees a syndrome involving three attitudes: I am the object,
I
religious, when to political, and when to other kinds of activity (or inactivity) ? \Ye not the subject, of political life; the government is not run' in my interest-it is not
can give only a general answer to the question, based on three kinds of information. my government; I do not approve of the way decisions are made-rules of the game
Political action is more likely when there are extensive channels of communication are unfair. Finifter (1970) hypothesizes four dimensions of politicaI alienation:
to political groups, when the values of the surrounding culture and subcultu~ are powerlessness, meaninglessrless (political decisions are perceived as being unpre-
I dictable), normlessness (perception that the rules intended to govern political rela-
expressed strongly in political terms, and when the individual involved has acquired I
1
tendencies that sensitize him to political cues. Religious structures, cultures, and tions have broken down), and political isolation (rejection of the political norms
tendencies, of course, promote different modes of expressing one's alienation. and goals widely shared in the society). Factor analysis of questions related to politi-
References to religion as a response to anomie and alienation, perhaps as an cal alienation, however, revealed only two distinct dimensions: political powerless-
alternative to a political response, are less common than discussions of alienztion ness and perceived political normiessness. This is close to Olsen's distinction; it is also
from religion (Pin, 1962) and religion as a-source of alienation, generally based on close to Lane's distinctions if the last two attitudes he lists are combined.
the contention that the religious establishment contributes to the support of unjust However we finally subdivide the concept of political alienation, we need to
social arrangements (May, 1949). undoubted!^ these conditions are common, recognize its complexity in two senses: its own internal structure and its relationships
particularly in societies undergoing rapid social change. Qoe consequence is likely to other forms of alienation. We cannot assume, for example, that there is a close
to be increased support for "apocalyptic" political movements that promise to cure association between feelings of political powerlessness and other fornls of powerless-
the society of normlessness and to lift the burden of the feelings of alienation. "In , ness. There is doubtless a close connection in many cases; but the relationhips require
France, for example, ecological studies which contrast degree of religious practice specification. Some persons who have strong feelings of efficacy in one setting-
with communist strength show that the communists are most successful in regions owners of a small bnsiness, for example-may experience feelings of pov~erlessness
in which the 'anticlerical' wave had previously suppressed much of the traditional in politics precisely because they cannot match the degree of control there that they
fidelity to Catholicism (Lipset, 1Y63aJp. 530). have in their occupations.
I t is important to know, of course, what conditions lead to a political move- Although it furi~ishcsvaluable guidelines for the study of the relationship
ment that is, in part, an expression of alienation from an established order, including between alienation and political participation, we cannot review the extensive litera-
the religious order; and what conditions lead to a religious movement that is, in I ture dealing with more general questions of participation. This much, however, needs
part, an e,xpression of alienation from the secular order. Moreover, one expression to be said : The earlier affirmations of a high negative correlation between alienation
of alienation may develop into another. A critical problem for future research is to and participation have been extensively qualified by the more recent research. When
specify more fully the conditions that transform religious manifestations of protest the relationship is specified by type of organization, by level of mobility orientation
190 J. Milton Yinger , Anomie, Alienation, and Political Behavior 191

of the respondents, by education, by ownership versus managership, and other varia- between powerlesmess and participation, with many other variables being controlled;
bles, the simple corrrelation is reduced and sometimes disappears (Cutler, 1969; but there was no correlation between normlessness and participation. This second
Neal and Seeman, 1964; Nelson, 1968) . form of political alienation was most powerfully associated with lack of faith in
A primary task of studies dealing with political participation is to design people, a variable that explained only a small part of the variation in political
reliable and valid measures. Although no one scale has emerged as the definitive one, powerlessness.
there is substantial agreement on the items to be included. Olsen (1969), for exam- I n the face of such findings regarding the various forms of political aliena-
ple, has measures of media exposure, frequency of political discussion, voting, and tion, it is clear that we shall have to use any single-factor indexes with care. They
political involvement (reading party literature, giving money, doing volunteer work) . may obscure important relationships.
Erbe (1964) has combined many of the same items into a scale, using modifications A number of studies have specified the political participation-political aliena-
of a procedure developed by Julian Woodward and Elmo Roper: " ( 1) frequency tion relationship in various ways. Using a probability sample of Berkeley residents,
of informal political discussions ('frequently,' two points; 'occasionally,' one point) ; Templeton (1966) found that alienated persons, as measured by a modification of
(2) role in informal political discussions (atempts to dominate, two points; equal Srole's scale, tended to withdraw from national politics but not from Iocal politics.
role, one point) ; ( 3 ) attempts to influence public officials (on two or more issues, When channels for the expression of discontent are readily available, as is more likely
two points; on one issue only, one point) ; (4) personal political campaigning (two to be the case on the local level, the alienated, on the basis of this evidence, take
points if any i n last four years) ; (5) financial contributions (one point if 'yes') ; (6) part. On occasions, of course, such channels are available on the national level.
number of times voted in last four years (six or more, three points; three to five times, Approval of George Wallace has often been interpreted in this way, although seldom
two points; once or twice, one point)'' ~[p.2021. with substantial empirical support.
Further work is needed on such scales of political participation to discover if Because Templeton did not distinguish between powerlessness and normless-
all the appropriate data have been included, weighted in a reasonable way, measured ness as forms of political alienation, we cannot say how emphasis on one rather than
accurately (not simply dependent upon memory), and otherwise refined. They repre- the other might affect the direction of political participation in his sample. A number
sent a substantial foundation, nevertheless, for empirical work to build upon. of other useful studies of community politics (for example, Horton and Thompson,
What, then, are some of the findings relating levels of political participation 1962; McDill and Ridley, 1962) also combine several questions into one alienation
to various forms of alienation? Using a sample from three small midwestern towns, measure, leaving open the question whether or not feelings of political powerlessness
Erbe (1964) found a simple correlation between alienation, measured by Dean's affect political behavior differently from feelings of distrust or normlessness,
(1961) scale, and low political participation. The relationship disappeared, however, In a study of the associations between political knowledge, organizational
in a partial correlation employing socioeconomic status and organizational involve- memberships, and political alienation in Sweden, Seeman (1966) limits his discussion
ment. It was the association of alienation with these latter variables that produced the to one type of political alienation, powerlessness. Through a number of controls, he
original simple correlation. Using his two measures of political alienation-incapa- found that those with strong feelings of political powerlessness were less likely to
bility and discontentment-Olsen (1969) found in another midwestern city a small have knowledge of political affairsS3They were also less likely to be interested in the
but significant negative correlation between political participation and his alienation discussion of politics or in keeping up with international affairs. (See also Levens,
measures. In one out of eight possible comparisons, the relationship was not sig- 1968.) Although it is difficult to establish the time order of these relationships,
nificant: those high in discontent were no less likely to vote. This finding should be Seeman argues cogently for the interpretation that powerlessness (low expectation
put alongside the fact that, when education is controlled, the business and professional for control) inhibits the learning of political information and interest in political
respondents had the highest rates of discontentment. Their alienation does not dis- affairs. It would be interesting to know whether there are cyclic reinforcements to
courage participation in the form of voting. It is important to emphasize, however, such patterns, from powerlessness to low levels of knowledge back to powerlessness.
that they were characterized by discontentment, not by powerIessness, emphasizing It would also. be valuable to know whether such cycles might be reversed. Some
the importance of this distinction between types of political alienation. Aberbach evidence for a reversal is found in a study (Gottersfeld and Dozier, 1966) of sixty-
(1969) found that a related distinction-between lack of trust and feelings of in- two persons being trained as organizers for a community-action program in East
efficacy-helped to interpret the 1964 election. Political trust was negatively asso- Harlem. When they were measured by Rotter's internal-external scale, those who
ciated with the Goldwater vote. Interpersonal trust was not, nor was political or were experienced as organizers felt less powerless than newcomers to the program.
personal efficacy. There is no reason to think that alienation need always be the independent variable.
The importance of the distinction between types of politicaI alienation is Mass Society and Politics. Questions of anomie and alienation reverberate
clearly shown by Finifter's (1970) analysis of a national probability sample. The through the discussions of mass society. Although these terms themselves are not
political participation index was formed out of answers to six questions. The political always used, descriptions of normlessness and estrangement are common, with impli-
powerlessness index was derived from answers to eleven questions, which strongly cations that their intensity has created a fundamentally new kind of society, gesell-
intercorrelated in factor analysis. The perceived ~oliticalnormlessness index was
derived from the answers to ten questions. There was a high negative correlation Recently, Seeman (1972) has reported similar findings for France.
 J. Milton Yinger Anomie, Alienation, and Political Behavior 193

schaftlich in the extreme. Louis Wirl-hys(1948, pp. 2-3) description of a generation the atomized [anomic?] society, and subjectively the alienated population. There-
ago is relatively more objective than most. Mass societies, he wrote, are characterized fore, mass society is a system in which there is high availability of a population for
by great numbers, widely dispersed, with heterogeneity of interests, values, and power. mobilization by elites" (pp. 32-33).
Since the members of these societies, though exposed to the same media, know few A critical quality of mass societies, according to most interpretations, is tl-~e
of their fellow recipients or the producers of the contents, there is continuous ex- relative absence of intermediate groups that can serve as buffers between the indi-
perience of anonymity. The mass is relatively without leadership or a program of vidual and the total society with its powerful governmental structure. A pluralistic
action. Tnere we few common customs; hence, the population is open to suggestion, society, in Kornhauser's terms, has a structured heterogeneity that differentiates a
to currents of collective behavior. Individuals are substantially unattached; their society but does not divide it. Dividing lines do not reinforce one another; therefore,
memberships are largely for special interests, where part of self is expressed, in con- they serve as channels of allegiance to the total society rather than as a substitute
trast with the wholeness of experience in the primary groups of more communal allegiance. But war, depression, rapid urbanization, and other shocks to the social
societies. Durkheim would not have been uncomfortable with Wirth's description order destroy such intermediate groups. As a result, according to this line of a r p -
as an account of anomie. ment, several resources brought to society by such groups are lost. They satisfy some
More recently, psychological qualities believed to be characteristic of those needs and hence lower the reliance on politics; they serve as channels for communi-
who experience mass societies have been given more attention. Mass man is often cation to the decision makers and hence lessen the need for or likelihood of direct
regarded as a poor participant in a democratic society: "The psychological type intervention; they serve as checks on one another; they reduce the alienation of
characteristic of mass society provides little support for liberal democratic institutions. individuals by giving them meaningful attachments and some feeling of control over
The mass man clearly is available for mobilization by mass movements, since he lacks their lives; and their leaders often help support the larger system into which their
a strong set of internalized standards and substitutes standards of the mass. There- own authority is bound.
fore, in the absence of an acceptable self-image, the individual seeks to overcome the We have noted that the evidence for such effects is anything but decisive.
anxiety accompanying self-alienation by apathy or activism" (Kornhauser, 1959, p. Some groups may be the expressions of alienation without affecting its intensity;
112). others heighten it by giving it a vehicle for expressions; many are associated with
We cannot undertake to trace here the lineage of these ideas, back to the lower alienation. We are still in need of differentiated research that will help us dis-
ancients, the church fathers, or even to Ttinnies and Durkheim. The contemporary cover how particular groups influence persons of given levels of alienation in specific
reader will recognize them as attempts to account for the Escape from Freedom or contexts.
The Lonely Crowd. There is some nostalgia in the descriptions of modern urban life; Political Deviation. One of the best ways to study how anomie and aliena-
and ideology sometimes interferes with observation. "Everyone is against atomism tion influence politics in contemporary societies is to study political deviation. There
and for 'organic living.' But if we substitute, with good logic, the term 'totaly for are risks involved in such an examination. It may seem to support the assumption
'organic,' and 'individualistic' for 'atomistic,' the whole argument looks quite dif- that mass societies are uniformly riddled with anomie and alienation and peculiarly
ferent" (Bell, 1960, p. 27). It would be foolish, of course, to disregard the differences vulnerable to antidemocratic political movements. In fact, however, some of the
in human experience produced by the immensely expanded network of communica- , forces influencing politics in such societies (tendencies toward equalitarianism; the
tion, the greatly increased mobility, industrialization, urbanization, and the other universalistic standards of bureaucracies; the production of a deeper national, as
critical elements of modern life. The cultural "blueprints for action" and the char- contrasted with a local or class, culture and experience) lend support to democrzcy
acter structures more or less shaped to work with those blueprints are severely ' (Gusfield, 1962; Manti, 1970). Another risk is that studies of deviation tend to focus
strained, just as they have been in earlier periods of drastic transition-from food on extreme deviation. Until we learn to measure anomie and alienation more pre-
gathering to settled agriculture, for example. The task is to examine the contemporary cisely, only the more extreme varieties show up on our recording instruments. We
situation, in our case in terms' of its politics, without comparing it with an imaginary , will therefore miss the possible reformist sensibilities and activities of the slightly
past. alienated, and we will overlook the flexibility and freedom that may require a small
Interpretations of mass society, as Kornhauser (1959) points out, are of two dose of anomie. This is another version of the ancient problem of freedom and
major types. Most nineteenth-century and some more recent commentators lament order.
the loss of standards, the destruction of cultural elites, the loss of liberty in an effort Despite the risks, the student of alienation and anomie has much to gain by
to attain equality. Mankind has moved into an "era of crowds" characterized by the studying poiitical deviation, both of the left and the right. Before turning to that
of the unqualified." Opposed to such aristocratic interpretations are the task, we need to note the current difficulties in defining deviation. T i e defining
interpretations of democratic critics of mass society. They characterize it in these process itself is being sharply examined and challenged. Many scholars and activists
terms: "(a) a growing atomization (loss of community) ; (b) widespread readiness declare that those with the power to label (whether the labeling refers to criminality,
to embrace new ideologies (quest for community) ; (c) totalitarianism (total domi- mental illness, level of intelligence, or political legitimacy) rather than consensual
nation by pseudocommunity). In this universe of discourse, mass society is a con- standards determine who is deviant, This current debate is itself a sign of anomie.
dition in which elite domination replaces democratic rule. Mass society is objectively
194; J. Milton Yinger Anomie, Alienation, and Political Behavior 195

Were standards fully shared, there would be little difficulty in knowing when they or revolution-designed to destroy the social order, not to communicate with it; and
were being violated. still others define them as protests, as a more or less legitimate form of communica-
DiPalma and McClosky (1970) resolved the issue of definition by designating tion, a desperate shout trying to get above the noise pollution of our civilization to
as deviant those who disagree with a statement that is accepted by 70 percent of a demand change.
population under study. By summing answers to a series of questions, one can design I t is clearly of great political importance how a protest movement is defined
a scale of deviation. Their questions dealt mainly with political values and tradi- by various segments of the public. According to 'Turner (1969), "Five theoretical
tional rights, not specific political issues. For example (the deviant response is given perspectives can be used to predict when the protest interpretation will be made:
in parentheses) ; "Our freedom depends on the private enterprise system" (dis- (1) events must be credible as protests; (2) an optimal balance is required between
agree), "We need a strong central government to handle modern economic problems appeal and threat; ( 3 ) protest interpretation is often an aspect of conciliation to
efficiently" (disagree), "To bring about great changes for the benefit of mankind avoid full-scale conflict; (4) protest interpretation can be an invitation to form a
often requires cruelty and even r~thlessness'~ (agree) . DiPalma and McClosky found coalition; and (5) protest interpretations can be a phase of bargaining by authorities"
a steady increase in rates of alienation and anomy (an individual measure in their (p. 815; see also Olsen, 1968).
usage) as the rate of deviation increased. This was true of both a Minnesota state Political Deviation to the Left. With these problems of definition in mind,
and a national sample. There were, it should be noted, a number of interaction we can profitably examine some of the many evidences that political deviation stems,
effects that qualify this general relationship. Among the college-educated sample, in part, from higher than normal levels of alienation and anomie. (Once again, this
alienation and anomy scores vary much less drastically between conformers and is not a valuative statement. I would personally lament some and applaud other
deviants than they do among noncollege respondents. Among the latter, noncon- forms of political deviation.) Radical deviationists are those who want to push ahead
formity was associated with particularly high alienation and anomy scores. From the to a hoped-for new world. They have never "had it good"; or, seeing some objective
data a t hand we cannot tell which came first, the alienation or the deviation, or improvement now, their aspirations have been drastically raised, and they want to
whether they were mutually supportive. And there is no clustering of the deviation continue the direction of change. Reactionary deviationists want to go back to a
answers around specific political movements and activities; hence, the importance of previous (real or imagined) world in which they and those with whom they identify
the relationship for political behavior can be determined only by further study. were better off; or where their present high levels of income and influence seem
DiPalma and McClosky used a purely objective definition of deviation: dis- less likely to be threatened by "upstarts." These statements are insufficient definitions,
agreement with a majority (70 percent) view. In a time of deregulation, however, however. I n a democratic polity, far right and far left are also defined in terms of
the majority view, the traditional answer, and the standard speech pattern are sub- their unwillingness to grant legitimacy to the existing political processes. Extreme
jected to criticism; the customary labeling process is exposed. I once heard an rightists see the society in which people like themselves prospered (or so they be-
American scholar give a learned lecture on the Far East. During a question period, lieved) disappearing, and they see no way within the system to restore it. Extreme
an Asian demolished the speech with a gentle inquiry: "Far? Far from what?" leftists see their hope for a new society frustrated, and they see no way within the
Today, we hear the same question with regard to many issues, although the inquiries system to attain it. Seeing themselves as perpetual minorities, they cannot accept
are often more strident. "Welfare mothers," who in the past have simply been as- temporary electoral defeat; they promote polarizing strategies designed to change
signed (or denied) help, now enter the political arena to fight for their checks and the system.
to try to increase them as a matter of right. How far are they from "legitimate" These political outcomes, whether of the left or the right, are products of
political activity? O n this anomic issue, our society does not know whether to class interaction among the objective situation, particular individuals' places in the system,
them with chiselers or as minor-league colleagues of the five cotton farmers who, in and their tendencies to interpret their circumstances in a given way. Research seldom
1970, were paid over one million dollars each for not planting cotton. The majority, specifies these ingredients; hence, we are left with probability statements. Studies of
I would guess, see them as deviationists-taking a position beyond legitimate political , communists and ex-communists in noncommunist societies speak of "loss of faith,"
debate. A few label the cotton farmers that way also; but since the farmers were neurotic tendencies, feelings of emotional and social rejection, which overlap to
dealing with a law passed by Congress and signed by the president, it is more diffi- some degree the concept of alienation (Almond, 1954; Cantril, 1958; Ernst and
cult to make the label stick. Loth, 1952). Although they do not generally demonstrate that these are more char-
In a time of stress and change, deviation thus becomes politicized, with various acteristic of their subjects than 03 the general population, this seems a reasonable
minorities demanding the right to resolution of issues through political processes. And deduction from the evidence presented. Whether or not these people live in more
political debate becomes "deviationized," with the protagonists and antagonists anomic settings, they experience their social worlds as anornic and respond by seeking
trying to brand their opponents as beyond the political pale (Horowitz and Liebo- a new normative system, a faith that will help them overcome their sense of mean-
witz, 1968). This process is not yet well understood but is beginning to get careful inglessness and powerlessness (Grossman, 1949) .
study, particularly with reference to protest activities. Even a cursory study of com- More recent studies of leftist political deviation have dealt with protest move-
mentary on riots, for example, will reveal that some observers interpret them simply ments among students, or young people generally, and among Negroes. Remembering
as a series of individual criminal acts; others regard them as elements in a rebellion the problems associated with defining both deviation and protest, we can profitably
 J. Milton I'inger Anomie, Alienation, and Political ~ e h a & r 197

examine a few of these studies. Questionnaires given to a sample of students at the members of the radical Students for a Democratic Society with members of the
Berkeley and Los Angeles campuses of the University of Califdrnia revealed only conservative Young Americans for Freedom, Westby and Braungart (1970) found
low and insignificant correlations between "personal powerlesstiessy' and participation significantly higher levels of alienation-as measured by responses to questions deal-
in a protest demonstration (Kirby, 1971). Agreement with the "social powerlessness" ing with attitudes toward parents, school, and religious leaders-among the former.
items, however, was significantly correlated with participation in protests. (Among These measures are rather weak, however, and we are left by such studies with the
the statements used were "The average citizen can have an influence on the way task of sorting out the varieties of political response that can flow from various levels
the government is run" and "Real decision-making power in this country is in the and types of alienation. Using her two dimensions of political alienation, Finifter
hands of a few men w& make up a power elite.") Kirby argues that this relationship (1970) suggests that when both are low, conformative participation is a common
is in part the product of a counterculture which recruits support frorn alienated end response; when powerlessness is high but perceived normlessness low, apathy is the
nonalienated alike. Activism is not associated with expressed feelings of loneliness modal response; with the opposite combination, one finds a reform orientation-
or personal powerlessness but with sharing the views of this counterculture. He protest groups working within the system; when both are high, there is likely to be
emphasizes, in our terms, the anomic setting-the availability of sharply competing extreme disengagement-separatist and revolutionary movements, complete with-
normative systems-as a source of protests. drawal. This is a useful way to structure the problem; but the fourth category leaves
In their widely cited books, Roszak (1969) and Reich (1970) g'ive a some- undetermined the conditions under which either withdrawal or revolutionary protest
what different interpretation. Their descriptions of technological rigidity-of dis- occurs. Additional structural and psychological variables need to be introduced into
order, corruption, and loss of self-are severe. The protests they see among the the system to differentiate between these two contrasting responses.
young, however, are basically apolitical. I n Roszak's view, "the alienated young are Additional variables are used in some of the research dealing with prptests
giving shape to something that looks like the saving vision our endangered civiliza- movements among America's black population. Various forms of alienation are
tion requires" (p. 1) ; but the vision is an emphasis on psychic and cultural rebuild- among the forces influencing these protests, with the six designated by Fendrich and
ing-indeed a counterculture-not on politics. This is also Reich's view: "There is Axelson ( 1971, p. 25 1) being a representative list. Black political alienation, as they
a revolution coming. I t will not be like revolutions of the past. It will originate with describe it, has these dimensions: distrust of white authorities, whom they regard
the individual and with culture, and it will change the political structure only as as incompetent and biased; rejection of a public philosophy that pays only lip ser-
its final act3' (p. 2). vice to justice, equality, and opportunity; a positive black identification-a set of
These are, as I read them, essentially "religious" books, describing one road attitudes that foster a separate social psychological identity; support for leaders and
organizations that reject traditional reform and advocate drastic change; favorable
to salvation and proclaiming that alienated young people have started down that
sentiments for political and social separation; and support for the use of violence.
road. Just as the Quakers pulled away from the intensely aggressive political and
Although this is a useful list, it requires some modification. Except perhaps for the
economic interests of the Levelers in seventeenth-century England, some current
first two dimensions, the reference is to responses to alienation and the environment
h~manistic( I would not call them secular) sect movements emphasize withdrawal
in which it is experienced, rather than an account of various aspects of alienation. It
and noncooperation with a harsh world, to the end not that political power can be
leaves out powerlessness, one of the most thoroughly studied dimensions.
seized but that politics will be transformed. Roszak cites the founder of the Quakers, In examining the relationship of alienation to protest movements among
in fact, in describing his view of how best to deal with the crises of modern life: blacks, we must not forget that alienation is probably much more widespread than
participation in protest activities. I t may be a necessary cause of protest, although it
The process of weaning men away from the technocracy can never be clearly is not a sufficientcause. Blacks may resort to protest because they are relatively
carried through by way of a grim, hard-bitten, and self-congratulatory
powerless and lack political resources (Wilson, 1961), or protest may appeal to a
militancy, which at best belongs to tasks of ad hoc resistance. Beyond the
tactics of resistance, but shaping them at all times, there must be a stance wider "third party," thus creating a political resource in the face of powerlessness
of life which seeks not simply to muster power against the misdeeds of (Lipsky, 1968) ; but many who experience powerlessness have not been involved in
society, but to transform the very sense men have of reality. This may protests. Conditions that promote extensive communication of shared feelings of
mean that, like George Fox, one must often be prepared not to act, but alienation and encourage the formation of groups contribute to protest activities
to 'stand still in the light,' confident that only such a stillness possesses (McPhail, 1971). They have been carried on, it should also be emphasized, in a
the eloquence to draw men away from lives we must believe they in- strongly anomic environment. Rioting and other forms of protest are sharp chal-
wardly loathe, but which misplaced pride will goad them to defend under lenges to the legitimacy of existing economic and political structures at a time when
aggressive pressure to the very death-their death and ours [p. 2671. opposition to traditional educational procedures, national war policies, the civil
movement, and other aspects of national life have made normlessness highly
No doubt this is one response being made to politics by the alienated. Along- visible.
side it we must note the continuation of protests and radical opposition. Comparing MTe have few direct measures of the anomic situation in which protest activ-
198 J. Milton Yinger

ities occur. Thinking of anomie in the Merton sense, however (a gap between the
culturally approved goals and the structurally furnished means), we do have a good
index of it in measures of perceived relative deprivation. Real gains in income, for
T Anomie, Alienation, and Political Behavior 199

lence. The two measures are doubtless correlated, so that firm statistical deductions
are not justified.
The importance of multivariate interpretations is shown in a number of other
example, can increase discontentment if they are experienced as psychological losses, studies. Gordon (1972) found that participation in the Newark riot was most likely
because they do not represent any closing of the gap separating one from a comparison to occur among those who were low in political trust but high in political efficacy-
group. Using an instrument designed by Cantril, Crawford and Naditch (1970) not among the more fully alienated, those low in both. Ransford (1968) combined
divided a sample of Detroit-area residents into those who saw a low discrepancy three variables to produce a valuable interpretation. Interviews with a sample of
between their present situation and "the best possible life" and those who saw a 312 Negro male heads of households secured information on social contact with
high discrepancy. The latter consistently scored significantly higher on attitudinal whites, as well as on feelings of powerlessness (the internal-external scale) and racial
questions dealing with militancy. For example, 54 percent thought that riots help dissatisfaction. Although he treats isolation as a measure of alienation, it might well
the Negro cause, compared with 28 percent of those who perceived low discrepancy. be regarded as an anomic element in the structural sense, since it involves facts
Unfortunately, various indicators of external structural discrepancies have regarding interaction across the race line, not perceptions of separation. All three
not been added to measures of perceived discrepancy between goals and means. In- variables are significantly associated with answers to the question concerning willing-
dividual perceptions do not determine the structure within which behavior occurs, ness to use violence to secure Negro rights. There are a number of interaction effects
the supportive or challenging behavior of others, or the levels of satisfaction and cost among the variables. Perhaps the most interesting is that among those who are
attached to given decisions. Psychological research tends too easily to overlook the isolated from contacts with whites, powerlessness and racial dissatisfaction are
distinction made by H. A. Murray and others between an alfiha and a beta press: particularly strongly associated with violence proneness. Isolation is less decisive in
"The former are those elements in the objective environment, as seen or inferred by its influence among those who are low in powerlessness and dissatisfaction. Such
the trained observer, that can affect behavior. They need not necessarily be easily multivariate analysis may help to reconcile the somewhat contradictory findings
observed; in fact, perceived factors are often rationalizations. The beta press is concerning the alienative qualities of segregated neighborhoods (Bullough, 1967;
made up of the forces acting upon an individual or group as perceived by them" Wilson, 1971).
(Murray, 1938, pp. 126,290). The combined effects of the three variables measured by Ransford are
Lacking direct measures of anomie, of discrepancy between goals and means, decisive: Respondents low in contact, high in powerlessness, and high in dissatis-
Crawford and Naditch (1970) used their measure of perceived discrepancy in a faction contrast sharply with those who are low in alienation by these indicators.
helpful way. They cross-tabulated it against the Rotter internal-external control Sixty-five percent of the former, but only 12 percent of the latter, are willing to use
measure. Using data gathered by Ransford shortly after the Watts riot, they found violence.
that only 12.6 percent of their respondents characterized by internal control and Of the many other studies that explore the relationship of various alienation
low perceived means-ends discrepancy expressed a willingness to use violence to measures to protest activities among blacks, we will comment on only one more.
obtain Negro rights. In contrast, 57.1 percent of those characterized by external Reporting on a Detroit-area sample, and using a five-point scale of political trust,
control and high means-ends discrepancy indicated a willingness to use violence. Of Aberbach and Walker (1970) indicate that 54 percent of those lowest in political
those who were cccross-pressured" (external-low or internal-high), 25.5 percent ex- trust (and only 17 percent of those highest in trust) answer "yes'y or "maybe" to the
pressed a willingness to use violence. question "Can you imagine a situation in which you would riot?" This relationship
The importance of this combination for politics is shown in further cross- is particularly strong when cross-tabulated by reported experiences of discrimination.
tabulations with reference to other questions. Crawford and Naditch (1970) used a Because we are dealing with extremely complex relationships, the patterns are
large sample drawn for the United States Civil Rights Commission from Negro men by no means cleqr. We are beginning to sort out, however, the kinds of alienation
and women in metropolitan areas of the North and West. Among other questions, and the levels of anomie most closely associated with political---or politically rele-
subjects were asked: "Have you ever taken part in a civil rights demonstration?" vant-protest movements among American Negroes. The great need in future re-
Those who were high in internal control (the highest third) and high in perception search is to achieve a more systematic combination of structural and psychological
of means-ends discrepancy (again, the scale was divided into thirds) were most variables. When the whole field of forces is the unit of analysis, interaction effects
likely (28.2 percent) to answer yes. Those high in external control and low in means- among individual and group factors will receive the attention they deserve.
ends discrepancy were least likely (3.8 percent), with the two "cross-pressured" With reference to the general topic of the effects of anomie and alienation
groups falling in between ( 15.5 percent and 17.1 percent). on radicalism, we should emphasize that the relationship is not determinative. To
In comparing responses to these two politically relevant questions, one might
some degree, radicalism is simply the expression of cultural training. Matza (1961)
speculate that an increase in the sense of internal control, as a measure of low
argues persuasively that student radicalism, far from being contraculturaI, is an
alienation, increases the likelihood of participation in a civil rights demonstration.
expression-perhaps an exaggerated expression-of values and beliefs rooted in the
On the other hand, an increase in feelings of means-ends discrepancy, as an index
dominant culture. Populism, visions of a transformed society, and evangelistic fervor
of anomie, raises the likelihood that a person will express a willingness to use vio-
 J. Milton Yinger

are scarcely inventions of rebellious youth. Organizational effects must also be taken
into account: Unionized unemployed Negro workers are more militantly class-
T Anomie, Alienation, and Poilitical Behavior

selected, the historical period, the society, and other contextual factors is high. This
helps to account for the inconsistencies; it also requires that one speak tentatively.
conscious than the unorganized (Leggett, 1968, ch. 5). In addition to such cultural McClosky's (1958) well-known work on this topic is more a study of cons,prva-
factors, deprivation influences the degree of radicalism. Although the idea that tism as a political philosophy than of right-wing political movements. After several
deprivation by itself leads to radicalism seems inadequate, certainly many forms of checks for reliability and validity, he produced a twelve-item conservatism scale,
relative deprivation heighten the sense of injustice. Use of the concept of relative which contains many of the themes in a wide range of material describing the
deprivation modifies the Marxian interpretation considerably, and it helps to conservative image of life. Sample statements are "Duties are more important than
account for the fact that the strongest protests and the most severe revolutionary rights," "No matter what the people think, a few people will always run things
trends occur zmong those who have experienced a period of steady gains but who see anyway," "People are getting soft and weak from so much coddiing and babying."
those gains against even more rapidly rising hopes and envisioned opportunities Extreme conservatives (those who scored between 7 and 9 on his twelve-point scale)
(Davies, 1962; Edwards, 1927; Simpson and Yinger, 1972, pp. 715-719; and Chap- tended to be in the high ranges of his "anomie"'. (an individual measure) scale.
ter Nine below). This argument, however, brings us back toward Merton's theory Fifty-nine percent fell in the top third, compared with 30 percent of the moderate
of anomie. It is easier for a society to lift goals and hopes than it is to revise systems conservatives, 16 percent of the moderate liberals, and 4 percent of the liberals (0-2
of opportunity. Riesmah once spoke of the difficulties involved in trying to "ration on the comervatism scale). The same pattern appears in connection with the measure
dreams." One might say that many contemporary societies are highly anomic because of alienation, with the following percentages falling in the top third of the scale as
they do not distribute dreams as unequally as they distribute income and power. one moves from extreme conservative to liberal: 45, 27, 20, 11,) McClosky is careful
Political Deviation to the Right. There are many similarities of environment to note the necessary qualifications in interpreting these findings. The sample is drawn
and of alienation between those who deviate to the political left and to the political wholly from Minnesota; there is inadequate use of various controls that might help
right. Both see a rapidly changing world where guidelines to action have become to separate types of conservatives; the relationships may be simply correlational, dis-
problematic; both tend to feel estranged from that world. There are subtle differences, guising the underlying causal influences of additional variables; and, in particular,
however, in the nature of that estrangement. One might say that those on the far his "extreme conservatives" are not identical, granted his scale of measurement, with
right see a glass that is half empty, and they fear that it soon will be entirely empty "right-wing authoritarians," although he does not describe precisely how they differ.
unless present trends can be stopped and reversed. Those on the far left see a glass That modes of measurement are crucial is shown by comparison of the results
tliat is half full, and they fear that it never will be entirely full unless present obstacles of studies that use different sca!es. Schoenberger (1968$, for example, identifies con-
to change can be removed. The reality component of these feelings may be fairly servatives as those belonging to the New York Conservative party. On most per-
small, but it is seldom lacking from the equation. A political movement that call sonality measures they do not differ significantly from members of the Republican
emphasize the shared discomfort of the left and the right may, for a time, combine party. Some of these measures-for insrance, authoritarianism and misanthropy-
the left and right in expressions of populist discontent. When the reality of the situa- overlap the domain of alienation, although the term is not directly used. Schoenberger
tion is examined, however, the shared component proves to be too weak, and the believes that his findings contradict those of McClosky; but until similar measuring
movement breaks apart. processes are used, we must reserve our judgment. I t seems likely that membership
Certain structural conditions and cultural training lead more to the rightists'
in a large, well-organized party is a quite different phenomenon from selecting the
fear of loss than to the leftists' frustrated hopes for continued change. These condi- conservative response to a variety of questions. With such widely differing measuring
tions are found in greater than average amount among those groups who have re- instruments being used in the two studies, it is particularly important to control for
ceived fundamentalist training, whether Protestant or Catholic; those clinging to other variables (education, income, residence, and the like) before comparisons are
traditional white supremacy; those in small businesses who feel the constant threat
made.
of loss of freedom of action; those whose money is "new" and not entirely "justified" ' In a study of a Northwest sample, Rohter (1970) distinguished between the
by cultural standards; those trained to old-fashioned military notions, rather than
rightists and nonrightists on the basis of group memberships (for instance, the John
to long-range strategic planning intricately connected with diplomatic issues. When Birch Society and the Liberty Amendment Committee), content analysis of letters to
several such influences converge on a person, the kind of life he sees as essential is
the editors m several newspapers, and various attitude measures. His four-question
so severely threatened no amount of activity within the system seems capable of scale of alienation does not distinguish among various themes but seems to tap the
reversing them. When he is moved to action, it is characteristically to try to stop sense of powerlessness most fully. There is a significant relationship between the
something (to repeal, to impeach, to get out), not to accomplish something new. degree of "radical rightness" and the level of alienation. Nine percent of those
Empirical support for these propositions is extensive but not entirely con-
lowest on the radical-right scale are: hie5 in alienation, compared with 40 percent of
sistent. We can examine only a small part of the research on the far right that uses those high on the radical-right scale. Of those low in alienation, only 8 percent are
the concepts of anomie and alienation (Abcarian and Stanage, 1965; Bell, 1963;
high on "rightnessm; 36 percent are low on "rightness." Rightists and nonrightists
Hofstadter, 1965; Sokol, 1968). In this research, sensitivity to the questions used
also differed significantly on a three-item sense of powerlessness scale. None of these
(most of this research is based on interview. or questionnaire data), the samples
relationships is decisive. Clearly many other factors are involved, as Rohter empha-
202 J. Milton Yinger'

sizes. But the study brings further supporting evidence to the proposition that sight-
wing political beliefs and activities express deep-seated feelings of alienation. We
need simply to emphasize again the need for distinguishing between the sense of
alienation that comes from a belief that things have changed for the worse and aliena-
tion based on a belief that things have failed to change sufficiently rapidly or in
the correct direction to prevent deterioration.
A full exploration of the association between anomie, alienation, and the
radical right would require a comparative study. A comparison of American right-
wing politics with, for example, "baasskap" in South Africa, Action Francaise, ultra-
nationalism in Japan, and Nazism would help to reveal the various mixtures of
individual psychological factors and the structural-historical factors that can lead
to extremism on the right. Equally valuable would be a study of anti-Semitism; for
PATTERNS OF
as a political movement-if not as an individual, culturally taught prejudice-it has
been a vehicle for the expression of negativism, of opposition to the emergence of
the modern world, of resentment over the directions of change, and of deep-seated
estrangement in many lands. We cannot explore the vast literature on anti-Semitism
here (see Simpson and Yinger, 1972, ch. 9-10) but recommend it as invaluable to
the student of the political right.
Daniel Katz
Conclusion

Having used the space available for this chapter, we must summarize its
argument briefly. Our understanding of political behavior is enhanced by the in-
troduction of measures of anomie and alienation. These phenomena, although highly
interactive, require separate measurement and analytic treatment. Each is a complex
variable, with its boundaries not yet well established; hence, many more years of
effort will be needed to design reliable and valid scales. In particular, there is a
great need for the addition of behavioral to verbal measures of alienation and of
direct to indirect measures (aggregations and indexes) of anomie. Because the two %e first bridges between psychology and
phenomena are so close empirically, there is particular need to study their interactions political science were built around the relationships of personality variables to poli-
and their separate influences. A high association between alienation and some form tical events and often around the psychopathological aspects of personality. The
of political behavior, for example, might be interpreted as a causal relationship if great pioneer in the bridging of the two disciplines, Harold Lasswell (1930, 1948),
the possible prior influence of anomie on both had not also been examined. In a has convincingly described the role of personality factors in political behavior, with
more likely relationship, the reciprocal influences need to be explored. Finally, it is special attention to compensatory and defensive syndromes. Moreover, he has sug-
important not to emphasize the impact of alienation and anomie to the exclusion gested that the theoretical approximation of the political type of leader would be the
of more old-fashioned factors. Cultural influences, class location, party memberhips, one in whom power is the primary goal-specifically, one high in authoritarianism.
-occupational subcultures, and many other variables affect political behavior. Although This early linking of psychology and politics deserves closer investigation.
these interact with alienation and anomie under some conditions, they are analytically Its major weakness is that it ties psychological processes to the irrational in
and often empirically separate from them. Even among the politically "deviant," political behavior. If people and their leaders behave as one expects them to, it
where the evidehce for alienative and anomic factors is strongest, many other in- implies that a psychological explanation seems unnecessary. If, however, outstanding
fluences are at work. Development of the kind of social-psychological interpretation leaders meet with reverses or if poorly regarded leaders meet with success, or if their
expressed in this chapter will best be promoted by keeping it in continuous relation- behavior seems to follow an irrational model, then psychology is invoked to provide
ship with the larger theoretical system of which it is a part. answers. If a leader like Hitler behaves like a psychopath and if people, instead of *

incarcerating him, enthusiastically support his craziness, then psychologists and, even
better, psychiatrists need to be called in to explain the cruel nonsense. Even less
204 Daniel Katz Patterns of Leadership 205
dramatic examples can be cited to show the attempts to use clinical psychology for more than a single action, or two or three such actions. That is, there must be some
the political leader who does not meet our ideal expectatioha. There is the instance degree of constancy or predictability in the leader's influence over others. Finally,
of Woodrow Wilson's ill-fated attempts to persuade "the small group of willful men a leader is someone who influences an entire group-not just one other individual.
representing no opinion but their own" to accept his warlike policies and later the Can political leadership be distinguished from other forms of leadership-
League of Nations. Wilson's actions have inspired volumes, which attribute to him an for example, from business or educational or religious leadership? One can, of course,
intellectual coldness and contempt toward his colleagues, supposedly growing out of regard the differentiating characteristic of political leadership as its occurrence in
a childhood tyrannized by a sarcastic father (George and George, 1956). certain settings, such as a political party or a government office. This definition does
How, we do not object so much to the attempts at psychoanalyzing key figures not furnish a generic difference with respect to the social psychological processes
at a distance as to the emphasis upon the defensive mechanisms of a single character involved, and in fact such differences are difficult to identify.
in the drama. With perhaps unbecoming ingratitude the social psychologist asks The major difference between political and other forms of leadership is the
about the Iess dramatic incidents: Are these not also motivated actions? Why limit target against which influence is exerted. Two targets are political in nature. One is
psychological study to the irrational and unexpected? The answer that there is no the reformulation or change in group goals, as in the reform or overthrow of existing
problem in the latter case is not a good answer. Sometimes men act rationally and systems or, conversely, the mobilization of forces to resist system change. The other
sometin~esirrationally. What. factors are at work in one case and not another? And is the allocation of resources and rewards which may or may not involve system
the boundary between the rational and irrational is not always that easy to draw. change. Politics, then, has to do with decisions about policies and about resource
A psychological approach to political science is not a study of psychopathic allocation. Political leadership is concerned with affecting such decisions and is of
personalities in politics or even of more general pei.sonality factors related to pcilitical course zound in other settings than the political party or the government. Two areas
decision making. Rather, a psychological approach is a different level of analysis, of leadership would, then, be excluded from our conception of political leadership.
which studies the behavior of people in politicnl settings, whether the behavior is One would be the application of existing rules by bureaucrats; here, no policy change
situationally induced or whether it derives from childhood socialization practices. It and no shifts in the distribution of resources among people are involved. The mem-
is the microapproach to social sciences-dealing with the perceptions, cognitions, bers of Congress, in passing tax bills, are engaging in a political process, but the
expectations, and motivation of people. In contrast, the more traditional political internal revenue officers who carry out the provisions of such legislation are not
science approach, at the macrolevel, is concerned with collective outcomes and their acting politically if they follow their prescribed roles. They may have power over
relation to institutions. The political culturologist is just not interested in psycho- people in enforcing the rules, but they are not exercising political leadership. If they
logical explanations, whether clinical or social for he is dealing with the broad sweep were to go beyond their duties and utilize their position to favor one group over
of patterns of events, which for him have their own logic. At the microlevel, the another, then they would become political. In some forms of eastern bureaucracy it
political behaviorist, however, frequently can join forces with the social psychologist is not uncommon for bureaucrats all along the line, from top to bottom echelons,
in a common research enterprise. They enter such undertakings not as representa- to exploit their positions and become political figures.
tives of unique disciplines but as behavioral scientists. The recent sourcebook in The second area that lies outside political leadership is what French and
political psychology edited by Greenstein and Lerner (1971) attests to the interdisci- Raven (1959) call expert power-influence exerted in collective task accomplishment
plinary character of this field, with papers from political scientists, historians, sociolo- because of superior knowledge. We do not speak of the technical expertise of the
gists, social psychologists, and clinical psychologists. engineer as political, even though it is influential. He may increase group output, but
Though such joint ventures in political psychology have made great progress he does not have anything to say about how this increased product is to be distributed.
in recent years, they have been largely confined to electoral behavior, political sociali- He is nonpolitical. He is not concerned with changing the social system (the power
zation, and citizen involvement in political affairs. The study of political parties and relationships between people) but is oriented toward the task and toward people as
leadership within the party and in governmental structures is not very far advanced objects for task accomplishment.
as an interdisciplinary science. The chairman of a meeting, in following parliameiltary procedures, can
assume the role of expert or the role of political leader. If the chairman is there
General Leadership and Political Leadership because he is a knowledgeable parliamentarian and because he plays his role im-
partially, political leadership is not involved. If, however, he uses his chairmanship
Leadership is the process by which one individual consistently exerts more to recognize his friends and to gavel down his opponents-in other words, if he
influence than others in the carrying out of group functions, If all members are becomes a partisan-he is exerting political influence..In the first case he was facilitat-
equipotential and substitutable in determining group outcomes, the group will con- ing the outcome of the group process and giving heavy priority to the ruIes in run-
tinfie to function in some fashion even if it loses a particular member. If all members ning the meeting. In the second instance he was intervening in the ongoing activity
are equipotential in function, then there is no need for the concept of leadership. AS and giving priority to outcomes which he favored.
McFarland, (1969, p. 155) puts it: "The leader is the one who makes things happen We are saying, then, that following the rules of the game in general does not
that would not happen otherwise." Moreover, a leader influences others through require political leadership even though the follower of the rules is in a role that
206 ~ a d i eKatz
l Patterns of Leadership

carries authority. The rules may be inequitable, but it is their formulation and ex- States (Blondel, 1963; Milbrath, 1965). Even the militant left and minority groups
tension or modification, not their execution, that is political. are interested in seeing more of their young people trained as lawyers.
In excluding from political leadership the administrative skills for carrying The lawyer role has some interesting differences compared with the role of
out policies and the technical expertise for task accomplishment, we are recognizing the ideologist or policy formulator. The man who influences people through his
that not all system functioning is political in nature. Parsons (1960) distinguishes conceptualization of goals is not as avowedly partisan as the lawyer pressing the
between the technical, the managerial, and the institutional subsystems of social claims of his client. The ideol'ogist moves toward a broader rationale. Though he
structures. The technical (or production) is concerned with the task, whether pro- seeks reforms or revolutionary change, he usually speaks in general terms so that
ducing automobiles or teaching a foreign language; the managerial, with decision conversion to the cause is possible. The ideologist, then, tries primarily to appeal to
making about and control of people; the institutional, with relationships to other people's values and to-their basic trust in the sincerity of his own beliefs, whereas the
social systems. The production subsystem is thus not basically political in nature. This partisan pleader is expected to manipulate the situation to the advantage of his
is an old but neglected distinction in social science. Oppenheimer, in his classic followers and to seek compromises for their benefit.
volume The State (1914), distinguished between two means of acquiring wealth: Not all ideological leadership is political in nature. To the extent that the
the economic, by producing it; and the political, by getting possession of the title to leader formulates policies which energize and direct some group to achieve its ob-
it. Weber (1922), less interested in class conflict, formulated his bureaucratic model jectives in competition or in conflict with other groups, he is a political leader. If,
without regard to its political functioning and saw its roles filled by those chosen for on the other hand, his ideology embraces all of mankind, so that there is no outgroup
their competence for their tasks. His was an engineering approach to social organiza- or enemy, his leadership is more religious than poIitical. Thus, Mahatma Gandhi was
. a religious leader in his humanitarian philosophy, his egalitarianism, and his ethic of
tion. Though the theoretical distinction between political and nonpolitical influence
has long been recognized, there are practical difficulties in its application, since psy- nonviolence. On occasion, when he departed from his own broad ethical principles
chological distinctions do not always follow formal organizational patterns. The in practical situatio~sto a more limited tactic of winning concessions from the British
engineer can exploit his technical expertise to influence company policy in nontechni- rulers, he moved in the direction of political leadership.
cal matters. As long as the system operates mechanically, on the basis of an accepted A second characteristic of political leadership is that the representative of
reward structure, political processes are minimally involved. In practice, there is partisan interests becomes involved in a constant process of social exchange between
always a mix of the political and the nonpolitical; but we have been suggesting areas his followers and other competing groups. He obtains concessions and in turn per-
and conditions in which the dominant feature of the mix may be more political than suades his people to yield some ground. In this interaction he often becomes an
nonpolitical. interested party in his own right. He needs power of his own to operate effectively
One domain of particular interest with respect to this distinction is that of as opportunities arise, and part of his reward is the possession of power. His followers
decision making in organizations. In democratic societies participation in decision grant him some freedom to act for them, and in return he secures benefits for them.
making is seen as a rewarding process in itself, in which people should share. Many Hence, a triple set df social exchanges develops: between groups, between the leader
people thus may regard anyone above them not merely as an administrator following and his followers, and between leaders. In the leader-follower relationship, the leader
rules or a technician employing expertise but as a power figure. When subordinates obtains favors and exacts concessions for his following, and they in turn are expected
challenge the legitimate authority of superiors, they are attempting a change in the to support him as leader. In this continuing exchange process, the leader faces the
structure of the situation. As they join the issue, the situation becomes political. challenge not of "What have you done for me?" but "What have you done for me
What differences are there in the social psychology of political leadership ' lately?" Many of the bosses of political machines in American cities had their posi-
compared to nonpolitical leader~hipif we accept the above criteria concerning tar- tions undermined when federal patronage and federal programs supplanted the local
gets of influence? These criteria, it will be recalled, have to do with exerting influence spoils system.
in two directions: to affect the allocation of rewards and to change the existing social Because of his insecure position in this social exchange relationship, the leader
structure or to prevent such change. For one thing, the political leader is either a seeks to buttress his own position by extending his power beyond the immediate situa-
partisan representative of his group or an ideologist or formulator of policy (politician tion. He needs to build up some reservoir of obligations to himself, both from those
or statesman). As partisan representative he seeks prestige, privilege, or power for below him and those above him in the social structure. He seeks to increase his con-
some group in an attempt to advance its cause or protect its interests or some com- trol of rewards and sanctions and to make many people beholden to him. Hence, he
bination of both. He may internalize the values of his client group, or he may repre- devotes time and effort to acquiring power for the future as well as for the present
sent it without personal convictions, as does the lawyer representing a client. This in to carry on his group functions effectively. Power seeking, however, can readily
fact may be one of the less obvious reasons why political leaders in all Western become an end in itself. The empire building of the political leader is often construed
democracies are drawn so heavily from the ranks of the legal profession. Lawyers out- as indicative of the seeking of power in and for itself. Basically, power seeking is
number the members of any other profession in the parliaments of European democ- built into the political process, although the political leader may not have this as his
racies, in the American Congress, and among the governors of states in the United only personal motive. The leader who finds a concern with power personally un-
208 Daniel Katz Patterns of Leadership

congenial, however, will find politics a difficult and rough road. Many leaders do position is like the thesis of W. F. Ogburn (1922), who showed that many great
show a mixed pattern of motivation, with power seeking as only one of the drives inventions and discoveries are hit upon by more than one person at approximately
involved. the same time. Historically, however, these two assumptions (about the structure of
The manipulative politician, who intervenes as the wheeler and dealer in the the field and the presence of a number of people of approximately the same abil-
political process, fits the pattern just described. But how about the consensus seeker, ities) are frequently not realized. Just as every football team is not three-deep in star
who attempts to adjudicate competitive claims? The political compromiser, however, quarterbacks, so a group or a nation may at times lack the great men who could
does not play the role of impartial arbitrator as in a labor-management dispute. He does help to determine its destiny. Part of leadership ability, moreover, is the vision to see
not stand above the scene or outside the scene, but is part of it; he is an active when and where there are gaps or weaknesses in the social structure to permit move-
protagonist, if not of some of the parties, then of the solution he favors. Moreover, he ment and redirection.
seeks acquiescence not by bringing antagonists into communication with one another Sidney Hook (1943) has carried the argument further in a ap-
but by keeping them apart. In third-party intervention to help resolve conflicts, the proach to heroic leadership. He hypothesizes not only that the decisive outcome can
role of the third party is to restore communication between the conflicted parties, be attributed to a particular individual but also that no other person could have
find common goals, and have the parties reach some solution through a full under- functioned in similar fashion. Without Lenin, for example, the czarist regime would
standing of their own motives and those of their antagonists. The political leader, have fallen, but the successful Bolshevik revolution would not have occurred. Hook's
however, plays a more active role, and the competing parties direct their pressures method in determining heroic leaders is to use a probabilistic reconstruction of
toward him to exact concessions in their own favor and at the expense of their history, to examine past situations and look at probable outcomes without given key
rivals. figures.
The great-man school, in concentrating upon the personality of the leader,
Controversy Concerning the Great-Man School of History does not seem to recognize that a leader's personal characteristics and values must
fit the needs and aspirations of his following. Outstanding leaders often lose their
If we descr;be leadership as the differential influence exerted by a particular supporters and drop out of sight-not because they have changed but because the
person, is not the logical extension of this view an acceptance of the great-man pattern of wants and desires of their foI~owershas. Winston Churchill, the great war
school of history? According to this doctrine, history is synonymous with the biog- leader of Britain, whose eloquence, wit, and courage made him an almost legendary
raphies of great men. The Reformation was the story of Luther, Calvin, and Zwingli; figure, was rejected by the English electorate in the first election after World War 11.
the French revolution, of Voltaire, Robespierre, Danton, and Marat; the Russian We can still accept the definition of leadership as differential influence with-
revolution, of Lenin, Trotsky, and Stalin; fascism, of Mussolini and Hitler. The out joining the great-man school of history. Personality can at times be the critical
school of cultural determinism rejects this great-man approach to social explanation factor, but individuals may exert influence for a number of other reasons. They may
as a misplaced personality cult. Leaders, say the cultural determinists, are produced have special expertise about the issues in question, they may have command of
by historical forces and are constrained in their roles by ongoing social processes. special resources, they may be in positions of importance. If we think of acts of
Thus, Presidents Eisenhower, Kennedy, and Johnson, though different personalities leadership, moreover, rather than a personality mystique, we will be in a better
and subscribing to different value systems, followed the same foreign policy in Eastern - position to relate acts of leadership to situational forces. This is not to deny that there
Asia. Richard Nixon, the ardent supporter of free enterprise, instituted peacetime are occasions when charismatic leadership assumes critical importance. But this is a
controls over prices and wages. It is the situation, not the actor, that determines the matter of recognizing that for the followership the leader possesses quasi-magic qual-
outcome. ities. The task of the scientist is not to attribute these qualities to the leader but to
Both approaches suffer from their emphasis upon one aspect of the leadership examine what it is about his behavior or his personality and what it is about his
process. The cultural determinists are preoccupied with the situation, and the great- followers that produces these attitudes of devotion.
man proponents are preoccupied with the characteristics of the leader. The fact is
that leadership by definition is a relational process, involving both the leader and Leadership Orientation and Behavior
the people led. J. F. Brown (1936) perhaps has presented the clearest exposition of
this relational character of leadership. In two propositions-(1) the leader must have If there is one clear and compelling finding in studies of leadership, it is the
membership in his group and (2) the leader must come from a region of high discoirery of two fundamental types of leadership orientation and behavior, the task-
potential in the social field-Brown calls attention to the characteristics of the leader oriented and the social-emotionally oriented. In laboratory studies of small groups
as they relate to the social field (that is, he must be high in values, temperamental without formal leaders, Bales (1958) reported that leaders did emerge. Some indi-
qualities, and intellectual abilities esteemed by the group). The culturaI determinists viduals gave direction to the group through their suggestions about the task itself,
assume that social fields are so tightly structured that there is no room for movement others through their supportiveness of their fellows. Few leaders combined both
or maneuvering by the leader, and they also assume that social conditions have patterns, though there were some who did.
created a number of people of equipotential for assuming leadership positions. This Many writers have called attention to these two major dimensions in group
210 Daniel Katz Patterns of Leadership 211

functioning. One of the first was Barnard (1938), who distinguished between the with definite power structures. The structure generally assumes the form of a pyramid
eflectiveness of a group in getting a job done and its eficiency in providing a return and confines those at the lowest level to very little participation in decision making. In
to its members in personal satisfactions. On the basis of her empirical studies everyday life we are fully aware of the limited rule of the precinct leader compared
Jennings (1943) described two types of groups: in one type, relations among mem- to the governor or president, but in the social science literature on leadership we
bers are based primarily upon their working together on a common task; in the tend to ignore these critical institutional settings. Table 1 shows the development of the
other, intermember associations are the main reasons for the group's existence. two main leadership patterns (task-oriented and socioemotionally oriented) at differ-
Jennings recognized that most groups are not exclusively of either type but represent ent levels of the organizational hierarchy. At the lower levels in social and political
some mixture of these patterns. Deutsch (1949a) has similarly written about task structure (for instance, the precinct captain or ward boss in the political party or the
functions and group functions, and Homans (1950) about internal systems (con- first line supervisor in industry), the area of freedom for decision making and leader-
cerned with person-oriented reaction) and external systems (concerned with the ship is narrowly circumscribed. The task is given, and so too are most of the pro-
survival of the group in its environment). Though Cattell (1951) takes a different cedures for carrying it out. The leader at this level can develop technical compe-
approach to the study of groups, in his distinction between maintenance synergy and tence, which we would regard as nonpolitical even though he operates within a
effective synergy he is in agreement with other theorists; for maintenance synergy is political party. Political leadership enters, however, as the precinct captain or first
the energy expended to keep the group in being, and effective synergy is the total line supervisor encourages his people to contribute more to the group effort than they
energy used to solve or complete a task. otherwise would. The skill of the leader lies in assessing what constitutes positive
The extensive research of Fiedler (1967) does raise the question of whether reinforcement for the different members of his group. In summary, the leaders at
leaders themselves have to carry the operation of both task and supportive functions the lowest echelons in social systems exert influence in their skillful use of existing
to ensure effective performance. He has found that group performance can be nega- structure. They are not in a position to elaborate or modify system requirements.
tively related to the socioemotional supp~rtivenessof the group leader, for instance, At intermediate levels in the organizational structure, the scope for the exer-
the leader who maintains psychological distance between himself and his men has a tion of influence expands greatly. Task orientation can now go beyond the completion
more productive group than the man who is personally close to his subordinates. In of a given job and can take the form of extending and developing the organizational
most of Fiedler's field studies, however, the supportive function was operative in structure itself. T o the mastery of technical know-how is added the special skill of
other ways, either through an informal leader or the assistant to the leader or the fact initiative and of innovation. A specific example of such supplementing of structure
that there already was a high level of positive intragroup relations. His contingency occurred in the Democratic party in the state of Michigan on the part of a county
model examines four sets of variables: (1) the structured versus unstructured char- chairman. Her county was one of four comprising a congressional election district.
acter of the task, (2) the power position of the leader, (3) the quality of leader- The county itself was well organized and the county organizations geared nicely into
member relations, and (4) the psychological distance of the leader. Where leader- the state system. But there was practically no enduring organizational structure for
member relations are already good and the supportive function taken care of, then the congressional district in question. The county chairman proceeded to contact her
the leader who is personally close to his men can detract from group output. Where fellow county chairmen in the same district ta set up an appropriate board, which
leader-member relations are poor, then personal closeness is positively correlated could be tied both to the counties at the lower level and to the state and the nation
with performance, provided that either the task is structured or the power position of a higher levels. In business organizations a department head may similarly develop
the leader strong. Fiedler's work, though not in political settings, is the most systema- additional structures to piece out gaps in the existing system.
tic research on leadership and group effectiveness in its measurement and control of When intermediate levels of management, whether in political or other organi-
a number of important variables. zations, introduce and implement changes of this character, we would consider it an
The task and supportive patterns are not commonly found in the same person, act of political leadership. For the implementation requires more than calling on old
since concern for objective task accomplishment can preclude concern for affective mles and describing the need for the change; it requires promises, persuasion, and
interpersonal relations, and vice versa. Furthermore, it is difficult for the task master bargaining. Most organizational blueprints neglect this reality of organizational life.
to be perceived as a warm supportive person by those whose work may receive critical The model as originally devised does not automatically keep spinning on the basis of
attention. This distinction can become exaggerated and take on additional dimensions prescriptions of procedures and invoking of rules. Changes are always necessary in
when organizations deliberately add leaders who are yea sayers to balance the nay the formal structure, even if they are only extensions or the filling out of old struc-
sayers or the reverse. tures. Such changes, to be accepted and to be effective, call for political leadership,
which basically follows a bargaining or negotiating principle. Failure to realize the
Position in Organizational Hierarchy psychological requirements for change has been a frequent cause for the breakdown
of otherwise well-conceived administrative reforms.
We beIieve that the distinction between task and supportive functions is Even when intermediate levels attempt little change in existing structure, they
fundamental in leadership patterns-both in small groups and in large organizational have considerable freedom in developing the social-emotional function of adjudica-
structures, The most common settings for the exercise of influence are social systems tion and the engineering of consent. Claims and demands from the subsystems they
Patterns of Leadership 213

manage have to be met in some fashion that will ensure the loyalty and effective
support of the various groups involved. This is not the role of the judge in assessing
guilt but of the compromiser who wants to keep the system functioning.
In the political system the problem of maintaining loyalty is even more pro-
nounced in that the consent of the governed is a critical variable. I n fact, politics has
been defined as the art of compromise. The engineering of consent is based upon
compromise, and herein lies the dilemma of political leadership. Compromise is well
suited to labor-management disputes over economic issues, Each side gives a little,
and the conflict is settled. But moral and ideological disputes do not provide the
same easy resolution. One cannot give a little on a moral principle. The baby is still
illegitimate even though a tiny infant. The art of the politician, then, is to select for
negotiation and bargaining those issues that do not involve moral principles or to
redefine moral issues so as to deal with certain specifics that are not precedent setting.
For example, black pressure for quotas to ensure more black representation runs
counter to egalitarian principles of treating human beings without regard to race.
The leader meets the issue not by conceding the principle of quotas but by accepting
it as a temporary expedient in the interests of redressing past injustice.
The managers or politicians at intermediate levels also have as a major
function the socioemotional problem of integrating primary- and secondary-group
relations-that is, of encouraging interpersonal relations that will be supportive of
the formal structure. Formal patterns of role relationship are impersonal and by
themselves produce deprivation of social reinforcement; they also reduce spontaneity
of expression and confine the individual to prescribed and often fragmented forms
of activity. Hence, they need to be modified to permit more freedom for social inter-
action and greater scope for individual effort. The lowest echelons lack the authority
to develop such modifications, but at the intermediate level the effective leader can
maintain organizational objectives and yet encourage modifications that will permit
more socia1 interaction and greater group cohesiveness.
One procedure for the officer at intermediate levels is to develop a two-way
orientation in the system, so that he can be representative of those below him and
yet accepted by his superiors. If he relates only in an upward direction, he may be
favored by those above him, but he will not be supportive of his own followers and
hence will lack their support. If he relates only to those below him, he may be known
as a good guy but not as an effective leader, since he lacks the support at upper levels
to accomplish things for his followers. In ah old study in a public utility, Pelz (1951)
found that the foremen who were valued by their men were those who could be
&ective in going to bat for them up the line. The members of work groups were not
as favorably disposed toward the foreman who was friendly and democratic in manner
as by the foreman who could back up his friendly stance through effective action in
their behalf.
When we reach the top echelons in a social sptern, the area of freedom for
the exercise of influence is greater than at any ocher level. This is almost a redundant
st~tementif the hierarchical organization is completely authoritarian in its formal
structure. But in a democratically based system the authority structure can be
changed by the electorate. Policies can be validated or rejected. Leaders can be re-
elected or turned out of office. The authority structure then is not really pyramidal in
form, since the large electorzte is at times the top of the structure; the system there-
214 Daniel Katz Patterns df Leadership 215

fore is more accurately represented as a pyramid within a circle. Hence, top leaders can be less significant than intellectual attributes when we deal with policy formula-
often feel that they are under many constraints from the subgroups comprising their tion and implementation. There is some controversy about the extent to which
public, and thus their margin of decision making is greatly reduced. One university Franklin Roosevelt sought power as against ideological goals. The discussion is
president, after taking office, stated frankly that he could not hold his position very interesting but may be of minor moment. What is major is Roosevelt's deep under-
long unless,he carefully observed the limits beyond which his behavior would alienate standing of historical trends and his keen appreciation of internal and external system
any one of four groups: (1) his board of regents; (2) his faculty; (3) the student forces.
body; and (4) the people of the state, especially their representatives in the state Roosevelt's masterful uses of self-directed networks of intelligence and of self-
legislature. Since the four subgroups placed different constraints upon him, he felt determined timing of decision making are well described by Arthur Schlesinger
that his area for decision making was restricted to a narrow band of the total spec- (1959) and Richard Neustadt (1960). Upon assuming office the president of the
trum. United States can be overwhelmed by the huge, complex bureaucratic structure; and
Nevertheless, in general there is more opportunity for policy formulation and if he does not rise above it, he can become little more than a confused office boy.
its implementation at higher than at lower levels. At times the field of forces may be Roosevelt seized upon two basic dimensions of organizational operations, to which
tightly structured, and the responsible officials may have little room for movement. he could not allow himself to become a captive. One was the information system,
But those below them who seem to have more freedom in reaching decisions are not which necessarily filtered and reduced enormous inputs into usable capsules but
making decisions as far-reaching in scope. Moreover, the perception of constraints by filtered them according to the needs of the bureaucratic structure and its personnel.
the official may underestimate his real freedom. President Johnson probably thought Roosevelt persistently checked information from official sources with as many private,
he had no alternative to continuing the war in Vietnam-a policy supported by two informal, and other channels as possible, utilizing his wide acquaintance with people
previous administrations. It was almost axiomatic that a national leader in office from many walks of life as well as his wife's discerning reports from her variety of
should not take actions that could be attacked by his opponents as destructive of contacts. The second trap Roosevelt avoided was that of bureaucratically imposed
national honor, national prestige, and national interest. But there had been changes deadlines for decisions, If the president accepts the time schedule imposed upon him,
in nationalistic forces, and Johnson and most of the Democratic leaders had missed he is under such constant pressure that the decisions are hardly his own. The un-
the significance of these changes. On the other hand, President Nixon, in spite of his relenting time pressure is what many high-ranking officials report as their most trying
lack of charismatic leadership, did not allow misperceptions of the situation to deter experience in government, since they do not have an opportunity to give adequate
him from embarking upon bold new policy courses toward China and Russia. consideration to major problems. Roosevelt saw this clearly and became the "master
The two dimensions of leadership again can be seen at this level of leadership, of the self-created deadline," to use Neustadt's (1960) expression. Neustadt goes on
but they assume rather different forms. Task-oriented leadership now becomes the to say:
conceptualization of collective goals and the formulation of policy. This can in its
filllest extent mean the initiation of new structures. Franklin Roosevelt formulated Not only did he keep his organizations overlapping and divide
new domestic policies and new international policies which implied structural change. authority among them, but he also tended to put men of clashing temper-
Ile followed through, moreover, to implement new policies-not only by naming new aments, outlooks, ideas in charge of them. Competitive personalities
personnel but by setting up new governmental agencies and creating new govern- mixed with competing jurisdictions was Roosevelt's formula for putting
mental posts. Some leaders are content with espousing ideological changes, others pressure on himself, for making his subordinates push up to him the
with modifications in formal structure; but the outstanding leader attempts both. choices they could not take for themselves. I t also made them advertise
their punches; their quarrels provided him not only heat but information.
The qualities necessary for leadership at the top levels of complex organiza-
Administrative competition gave him two rewards. He got the choices
tions are heavily conceptual and intellectual. The political system, mediating the and due notice, both.
demands of other systems, calls for cognitive skills of a high order, as many business As a result he also got that treasure for a president, time to defer
leaders have discovered in their assumption of government office. Mann (1964), on decision. By and large, his built-in competitions forced the choices to him
the basis of his studies of industrial organizations and governmental agencies, main- early, or at least made him aware that they were coming. He, not others,
tains that at upper levels in the structure the conceptual abilities of the manager far ' then disposed of time to seek and -to apply his own perspective" [pp. 157-
outweigh his technical expertise and his skills in human relations. Katz and Kahn 1581.
(1966, p. 313) have used the concept of system perspective to refer to the ability of
the leader "to see, conceptualize, appraise, predict, and understand the demands and The socioemotional pattern at top levels can be one of charisma. At lower
opportunities posed to the organization by its environment." The necessary frame of levels the socioemotional leader is sympathetic and supportive of his immediate asso-
reference goes beyond the problems of the day to the complex internal functioning ciates and in fact is seen as one of the group. In some respects, the charismatic
of the system in relation to the historical direction of external events. Open-system leader is close enough to the group to permit identification with him, but he is also
theorizing is relatively new in behavioral science, but the outstanding political leaders perceived as having a magic about him that makes him a superior figure. I t is
have long been system theorists a t a practical level. Motivational aspects of leadership precisely this combination of membership character and high potential (noted by
 216 Daniel Katz
Patterns of Leadership 217
Brown, 1936) that enables people to attach themselves to his pers~nalityand then
secure is not the end of such hatred. It appears in attacks against minority groups,
soar to accomplishment beyond their everyday expectations. A comparison of Truman
to whom the undesirable characteristics of their attackers have been attributed,
and Roosevelt is interesting in this connection. Both men espoused the came of the
especially under conditions of social support.
common people vigorously and built up links of identification with them. But with
A similar account of charisma, with some zidded complexity concerning social
Truman the process did not go .much further. The common man was satisfied that
identity, is described by McFarland (1969) in his interpretation of biblical history.
in Truman he had someone like himself who would fight for his interests. In Roose-
For the Hebrews wandering in the wilderness Moses provided a new social and
velt, however, he had a leader whose power to achieve great things was almost un-
personal image. McFarland writes: "Thus in times of value strain, a charismatic
limited.
hero niay appear whose psychological processes are paralleled by his public actions,
To move the concept of charisma closer to operational measures, two criteria
perhaps in a widely appealing resolution of a personality identity crisis that provides
can be utilized. The first criterion is the degree of emotional arousal among the
the critical decisions and values for a new social identity, thereby leading to social
followers, as in the reception accorded the Kennedys compared with other Democratic
change through the establishment of new social structures infused with new ideology"
leaders. Charismatic leaders are reacted to emotionally by both their adherents and
(p. 175). This account applies as much to Hitler as to Moses, in that the Germans
their opponents. The wild enthusiasm of the one group is matched by the deep
were under a value strain and Hitler provided them with a new identity as members
hatred of the other. The operational test for the charismatic political figure might be
of a superior race with appropriate ideology and with new social structures to provide
a well-defined U-curve of affecrivity toward him, with very few people assuiiing
properly for their new status.
intermediate positions. The second criterion is the wide scope and great efficacy of
Charisma involving the symbolic solution of internal conflict is much more
the leader in the perceptions of his followers. The portrait is global and not dis-
common among the leaders of social movements than among the officials of estab-
criminating. Specific weaknesses are neglected in this great figure. Weber (1922), to
whom we owe this concept, talked of this second characteristic-namely, the magic lished structures. The latter are too tied to existing practices to furnish convincing
aura that people attribute to their hero--as the basis of his definition. ideological appeal for the dawn of a new day when the meek shall take over the
earth.
What leads to such high emotion and such exaggerated beliefs about a leader?
A second form of defensive charisma is less complex. It involves no new
On the negative side, one of the first conditions is some degree of distance. Day-to-
day intimacy destroys illusions, as in the valet-nonhero relationship or in romantic ideology but is a continuation of dependence upon the father figure or, more
love affairs. This is one reason why there is littIe charisma at lower levels in the accurately, an identification with the aggressor. The two conditions for such identi-
system. The man or woman above you whom you see every day may have strengths, fication (Sarnoff, 1962) are the possession of overwhelming power by the aggressor
but he also has obvious weaknesses and no charisma. But the leader whom you see and the inability of the person to escape the exercise of such power. Hence, totali-
on great occasions and under special circumstances is sufficiently remote to arousg tarian regimes are the most likely structures to breed such charisma. They do this
charisma. by concentrating power at the top of the structure and by making it dimcult if not
Much more than psychological distance is necessary for charisma, of course. impossible for people to leave the system. A literal example of the latter device was
Three types of interperscnal relations can be distinguished. First, the leader may the building of the Berlin wall. Identification with the aggressor also is likely where
have charisma because he symbolizes the followers' wishful solutions of internal home and school training have been authoritarian; an emotional dependence then
conflicts. Instead of acquiring insight into their deep-lying motives, people seek some is readily perpetuated in adult life toward leaders in a position of authority.
release from their conflicts by projecting their fears, aggression, and aspirations onto A third type of charismatic process, which is not the product of internal
some social objects which allow a symbolic solution. The charismatic leader provides conflict, should be noted. People can magnify the power of their leader because they
the symbolic solution, both through his personality and his program. He does not perceive in him some attributes that can advance their interests. They may not be
analyze the complex causes or true history of people's frustrations but echoes their searching for a new identity, and they may not be driven by internal conflicts. Fully
feelings about the issue and justifies their acting out their impulses against the enemy. aware of the goals they seek, they may be emotionally excited about an able leader
The advantages of the solution are psychological, not rational. There is temporary who might help them toward their objectives. They are not seeking a symbolic solu-
release of tension. Since the symbolic s~lutiondoes not necessitate any digging into tion, 'but they have an emotional attachment and a wish-thinking exaggeration of
causes, the individual can eat his cake and have it too. Since the solution is social the leader's abilities. Eugene McCarthy, for many of his followers, was a charismatic
and not idiosyncratic, people have the support bf some of their fellows. Part of their leader of the first type in symbolizing a wishful solution of internal conflicts; Adlai
conflict generally involves the superego, and some degree of social ,support helps Stevenson was a charismatic leader of the third type.
to counteract superego forces. Hitler had charisma Tor the German people, who- Pye (1961) uses Erikson's (1958) description of identity crisis to explain
with their lower-class and middle-class origins and with their deprivations after the role of charismatic leadership in the political development of emerging nations.
World War I-found in his arrogant rhetoric the power they thirsted for. The psy- Old customs lose their meaning, and old strucrures which bound people together
chology here is very much that of compensatory defense mechanisms as described by crumble. People grope for a new collective identity. The leader who has achieved
Adorno and his colleagues (1950). The repression of childhood hostility by the in- his own personal integration and can articulate it for others becomes a charismatic
figure. His ideological appeal is not merely for agreement on political forms and
 218 Daniel Katz Patterns of Leadership 219

measures but represents a shared orientation on a deeper level-that of personal context; if they are well organized in a role structure the leadership relationship
values. takes on a different form. We would suggest four major dimensions of social settings
These forms of charisma should be related to the types of frustration and in which leadership occurs :
deprivation which leave many people dissatisfied with things as they are. The one 1. Degrse of structure or role-determined behavior of people in relation to
type is a general frustration: people lump together their many personal, family, one another. At the one extreme would be a highly organized role system like the
occupational, and social grievances, with little discrimination or analysis of what the army or like a hospital. At the other would be aggregates of people comprising a
problems are or what the solutions could be. This frustration is more anomic than public or potential public with some common characteristic such as age or occupa-
revolutionary and is accompanied by general feelings of distrust and unwillingness tion. A social movement would occupy an intermediate position in that it is no
to cooperate with one's fellows. The second type of frustration is more of a thwarting longer an unorganized aggregate but a mobilization of people around known ob-
of specific needs; for instance, workers wanting better and more secure jobs. There jectives and known leaders; nevertheless, its common values are still not precisely
is some analysis of the problem and some knowledge of alternative solutions. There defined by implementing norms and its role system is on the primitive side.
is distrust of those on the other side but a willingness to cooperate with those ex- 2. Primary and secondary relationships. A related but not identical dimension
periencing a similar fate. The general frustration can lead to counterrevolution and is the degree to which primary relationships support or oppose the secondary rela-
to our first type of charismatic leader. The populist forces behind George Wallace tionships. The aggregate of individuals may become aware of one another and
could well reflect the generalized distrust of the powers that be. The more particu- interact with one another in local settings, and thus the aggregate can be mobilized
larized distrust may have been important in the nomination of George McGovern. in a public opinion process. This could occur without primary-group involvement,
Here people with specific problems were moving toward specific solutions as articu- but most studies indicate that people move toward acceptance of ideas from the re-
lated by their leader. To lump together the antiestablishment forces as a populist mote leader through the mediation of local influentials with whom they are in
revolt, and thus account for the strength of the following for these two figures, is contact. Similarly, the social movement without much formal structure requires the
to miss important differences in political leadership. support of primary-group participation. Without mass meetings, demonstrations, and
The concept of charismatic leadership has received so much attention in informal sessions, social reinforcement would be minimal. In fact, the vitality of the
recent years that the foreign term charisma appears in the public prints. This movement owes a lot to the opportunities for primary-group involvement.
attention is probably related to the value conflicts in our society, to present as well 3. Relationship of group or unit t o other systems. Some groups and organiza-
as anticipated shocks, and to some of the palpable malfunctionings of old institu- tions are relatively independent of other structures; some are intimately associated.
tional practices. Conventional bureaucratic leadership promises more of the same, Political systems by definition are integrators of other systems and hence show a great
and one would predict the emergence of charismatic leaders in the years ahead. deal of interdependence with them. In addition to the degree of interdependence is
Both policy-making and charismatic leadership have greater opportunities for the character of the interdependent relation. It can be one of mutual interaction or
expression at the top echelons of a social system, and they can complement one social exchange; or it can mean a superordinate-subordinate relationship, in which
another in the great .leader. The new conceptualization of goals needs some char- an organization is a subsystem highly dependent upon a more powerful structure. It
ismatic pattern to secure acceptance. Woodrow Wilson initiated new structures and can be one in which there is e~onomiccooperation but ideological conflict, or vice
formulated new ideology for international relations but lacked the charisma within versa. Intersystem relationship:, have not been systematically identified let alone
his own country to secure support for the League of Nations. But abroad he did studied, so that we can furnish only a few examples of their bearing upon processes
have such charisma, and his doctrine of self-determination for small nations changed of political leadership. One reason they come to the fore in considering political
the map of Europe. leadership is that the political figure frequently has to relate to his own subsystem,
We have referred in passing to the social movement as a more ready place to other organizations, and to some aggregate of publics.
for defensive charisma to appear than the established social structure. In our dis- One implication of looking at the contextual setting in terms of interorganiza-
cussion thus far we have given little attention to other variables than position in tional relationship is the identification of a boundary role as opposed to internal
the hierarchical structure of the leader. Let us turn, then, to a more adequate roles. The leader occupying a boundary position in which he must relate to other
account of social settings before commenting on some of their special characteristics structures will carry out different functions than the leader embedded in the system
in relation to leadership practices. with little contact with external groups. It has been said of Lyndon Johnson that his
experience and knowledge as a political figure was confined to the domestic scene.
Social Settings Hence, when he assumed the Presidency, he was not well quipped for his boundary
role of dealing with other nations.
If we see leadership as a relationship between the leader and his followers and 4. The mix of democratic versus authoritarian institutions in the system
carry through the logic of this position, then we are bound to examine the patterning (varying from the dominantly direct democracy of the kibbutz to the centralized
of the followers and the ways in which leaders and followers communicate and inter- bureaucracy of a totalitarian society). The strength of the kibbutz leader lies in his
act. If the followers are some aggregate of unorganized individuals, we have one powers of persuasion and skills in interpersonal relationships; the strength of the
220 Daniel Katz Patterns of Leadership 221
totalitarian leader, in his use of coercive power in the maintenance of dictatorial ing to known rules while they themselves are free to make up their own rules as they
control. go along.
In addition to this greater freedom, the leader of a social movement has the
Social Movements and Revolution advantage of the vitality of a new cause, a new forumlation of goals. The ideas of
the established order seem trite, and people are habituated to the old appeals to the
The leader or official in a structured institutional setting has many built-in point of virtual indifference. People may move back and forth between a longing
advantages over the leader of a relatively unorganized aggregate of people or even for security and a desire for novelty. When the old way of life meets needs and
of a social movement. 'The organized structure already has some cohesion based upon expectations, people's wishm for novelty are overshadowed by their wishes for security.
interdependent roles or common norms or conmion values, or some combination of When the old pattern no longer meets their needs, they are heavily attracted both by
these three elements. For the unorganized aggregate or semiaggregate setting, a the promises of the emergent leader and by the novelty of his proposal. For certain
sustained basis for integration has still to be developed. In the organized structure the key groups-namely, for artists and writers-novelty has an even greater appeal,
lea.der has formal legitimacy for his position of authority. In the developing social since these groups seek new ways of perceiving life and new conceptions of man in
movement the legitimacy rests upor. the development of an ideology. With established relation to his fellows (Lipset, 1960). Hence, as Brinton (1938) has pointed out,
legitimacy go sanctions against noncooperation with the leader-sanctions not avail- every revolution is preceded by the desertion of intellectuals from the establishment
able in the aggregate setting. to the radicals.
I n the organized structure the leader is in a strategic position of advantage Ideology and charisma are more important for the social movement than
because he has better access to mass media, greater resources and staff, and many fbr the leadership of the ongoing organization. In the ongoing institution, ideological
dependent subgroups. Most leaders in established systems have engaged in a social- leadership can be significant, but it is more a clarification and extension of existing
exchange process over time and have people who are obligated to them as well as goals than the formulation of new doctrine. Social movements, lacking as they do
people to whom they are obligated. Other officers and members of the organization the consistent supportive inputs of established organizations, depend much more upon
thus have an invcstment in their leaders. That is why it is so difficult for newcomers the appeal of the great cause and the charisma of the new Messiah. T o sustain a
to move into the system and assume direction of it in short order. Michels (1915), social movement requires either the continuity of the same charismatic Ieader or a
in his study of political parties, formulated his iron law of oligarchy because he saw well-developed ideology embodying a program of social change geared to specific
these tendencies toward entrenched power as overriding in strength even in formal causes of discontent. Traditional Marxists have realized this in their programmatic
democratic systems. He argued that officials are bound to be more knowledgeable approach, which provides both a social philosophy and a practical strategy and set
about problems than the rank and file; they have more experience in organizational of tactics for social actions. Thus, the movement can suffer some reverses, but there
procedure and can give more time to determining policies and maintaining power remains a consistent body of positive doctrine to give stability to the cause. The new
than can the rank and file. left has not learned this lesson, and its quasi-anarchism provides no stabilizing force.
Outside the established structures, leaders have to start almost from scratch I t has its appeal for overthrowing the older system but no program for replacing it,
to build a following, to achieve unity within it, to maintain it over time, and to give since it is antisystemic in its approach. Hence, like crisis management it operates by
it direction. Critical situations arise which create fairly widespread dissatisfaction the seat of its pants. I t may be strong in one emergency and weak in the next, and
with existing institutions-economic crises, an unsuccessful war, some obvious mal- it fails to build power as have Marxian revolutionary movements in the past. More-
functioning of the system. The emerging leader can base his appeal upon these over, with no clearly formulated goals, its followers can even be co-opted into the
common frustrations and also can find some support from a general dissatisfaction system.
with old institutions. There are also the slightly disturbed, the borderline paranoiacs, The classic account of revolutions and their changing leadership is that of
the cranks, and the misfits who may rally around a new cause. The older radicals Crane Brinton (1938). He describes common characteristics of the French, American,
themselves are afraid of this lunatic fringe and attempt to build their movement English, and Russian revolutions, with the French revolution as the basic prototype.
around a dependable core of those with genuine ideological commitment. Preceding the revolution there is a division of society in which the underprivileged
Noninstitutional leaders have certain distinct advantages. They have more develop power. I t is not a period of severe depression and repression but, on the
freedom of movement in two senses: they lack the many commitmects to existifig contrary, a period of increasing prosperity and rising expectations. The intellectuals
groups and power figures; and their legitimacy resides in a new ideology, which they desert to the revolutionary cause. The old regime meets the rising expectations very
can interpret as they will within very broad boundaries and which has not become poorly and is in fact characterized by a decadence of leac?ership. I t shows no under-
clarified through many specific tests of application. The established leaders, because standing of the depth, extent, or nature of popular discontent. It is indecisive and
they do not want to radicalize the discontented, are often reluctant to use their power vacillating in its use of power. Thus, when the revolt breaks out, the old regime is
to squelch the opposition. Hence, their challengers can justify any action on the part overwhelmed by the tempo of events and offers concessions that would have been
of their own followers and yet hold the established leaders to the restrictions of law adequate a short time before but are no longer adequate. I t essentially plays yester-
and order. They have the advantage of making the authorities play the game accord- day's ballgame.
222 Daniel Katz Patterns of Leadership

With the actual revolutionary outbreak, moderate leaders come to the fore, 1 Killian (1962) has described this process in his discussion of desegregation leadership.
the Kerenskys and the Mirabeaus. They link the new with the old and attempt real As the civil rights movement gathered momentum, old accommodating leaders were
reforms in the structure. Again the timing of events is against them, since by now often replaced by revolutionary leaders. When formal leaders were not replaced, they
the people are demanding more complete change. More radical leaders, whose moved to some degree in a radical direction but were not even perceived by the
promises are consistent with current needs, emerge. A double feedback cycle is set in people in the community as speaking for the blacks.
motion, with competitive leaders reinforcing the demands of their following and in In one Florida city a study (Killian and Smith, 1960) revealed a complete
turn having to move to more extreme positions to maintain their leadership: Moderate turnover in leadership. In 1957, during a bus boycott, panels of white and black
leaders cannot show sufficient action or progress to maintain their position and are leaders-including some who might have been involved in the boycott and some
displaced by the extremists. who had occupied positions of liaison in the past between the white and black com-
The extremists attempt to achieve the revolutionary changes explicit or im- munities-were interviewed. The emergent leaders in the bus controversy were of
plicit in the ideology of the revolutionary movement. The easy change is the liquida- the protest type, with not a single holdover from the accommodating leadership of
tion of old rulers and d d oppressors. The Russian revolution was successful in the past.
removing and dispossessing the old elite. The French revolution was not as thorough Sooner or later-since most people do not sustain their drives toward remote
in spite of its reign of terror. The more difficult change is the building of new social goals-there is some falling off among followers in their support of extreme leaders.
institutions to replace the old. The extremists are better at liquidating their enemies These leaders can misperceive the situation because the vocal people may still talk
than at rebuilding society, They have to meet the aspirations of the people for a aggressively but may be a small minority of their potential backing. The problem
better way of life. As this becomes difficult, their hold over their followers becomes becomes clear when specific action programs calling for mobilization and demonstra-
shaky and a "Thermidor reaction," a swing to the right, sets in. Thus, in France a tion receive less support than anticipated. The militant leaders may then assess the
new conservative leadership arose, with Napoleon as the charismatic figure.
Brinton's anatomy of revolution may not apply to all successful rebellions as
precisely as the analogy to the anatomy of living structures suggests. For example, in
~ situation and move toward a more moderate position. Thus, in the early 1970s some
militant black leaders-who felt they were getting out of touch with their own
communities and some of the older organized groups, such as the church-took a
the Russian revolution (although Stalin did replace Trotsky and opted for building
I somewhat less militant position.
communism in one country rather than for world revolution), the Thermidor The Brinton thesis of decadent elites, their replacement by the moderates, the
reaction hardly paralleled the conservative turn in the French revolution when the swing to extremism, and finally the Thermidor reaction is suggestive of equilibrium
rising bourgeoisie took over. Nevertheless, the Brinton framework provides a useful theory. The old leadership, out of tune with social change, represents an unstable
tool, which with modifications and additions can be applied to revolutionary social equilibrium. The extreme move to the left produces change but also the threat of too
movements. I t calls attention to a number of factors in political leadership to which much change and insecurity. Stability is restored with the swing toward the center.
reference has already been made. 1 Hence, according to equilibrium theory, in spite of some change the social field has
First of all, leadership is clearly a relation between followers with needs and tendencies toward a steady state. However, the fluctuations in social process and in
wishes and leaders who conceptualize, symbolize, satisfy, or promise satisfaction of leadership effects do not really balance out one another. Even after the Thermidor
these needs. Brinton shows that leaders can be quickly displaced when their old tech- reaction, the social pattern cannot be said to have returned to its former state.
niques are inappropriate to rapid changes. What is not fully developed in Brinton's Also explicit in the Brinton analysis is the importance of psychological depri-
account is the basis of common needs in the economic and political structure of the vation as against material deprivation. Revolutionary movements gather steam under
society. We need to identify the lines of divisiveness in a society, the groups with conditions of economic improvement. Once people experience some improvement in
common sources of frustration. The lines may be economic, according to ownership their own lot, they raise their sights for further improvement, as in Lewinian re-
of property; they may be religious; they may be ethnic and nationalistic; or, as in search on level of aspiration. Only, then, when this increasing expectation encounters
French Canada or Northern Ireland, they may be a combination of sources. A group sharp disappointment do people become frustrated and rebellious. This is the thesis
of people with a common set of grievances can become a significant social force if of Davies ( 1962), who combines conceptions of de Tocqueville and Marx in pointing
able leaders come to the fore. We cannot divorce a study of political leadership from out that economic setbacks have revolutionary implications when preceded by
a study of smial structure. According to Trotsky (1932), there was no full-fledged periods of rising aspirations.
Thermidor reaction in Russia because czarist Russia had been moving from feudalism Another necessary condition for rebellion is the realization that one changes
to large-scale capitalism, with few small businesses, and hence had no sizable small things for the better through getting together with those experiencing a similar fate.
bourgeoisie. Turner and Killian (1957, p. 31) emphasize this aspect of revolutionary action and
Brinton's dynamic conception of social process, in which he sees acceleration hold that in addition to conditions of frustration "there must also be a belief in better
of a movement through feedback, is congruent with modern system theory. Leaders .
conditions which can be brought about through collective action. . . As conditions
in certain periods must either become more militant or be replaced by those who are improving for a group . . . the members .. . develop an image of an even
are. Their very militancy, however, only reinforces the extremism of their followers. better state of affairs, as their early gains give them hope."
224 Daniel Katz Patterns of Leadership 225

Similarly, Gurr (1970), in his elaboration of the frustration-aggression hypo- especially to economic investment. Motor transport and technology may change
thesis to account for rebellion and revolution, sees relative deprivation as basic. Spanish society radically in the next decade.
Relative deprivation for him is the discrepancy between value expectatiolzs, or goal Finally, in a society of any size, the totalitarian leader must delegate some
objects to which people feel they are justly entitled, and value capabilities, or positions power to his subordinates. The greatest threat to his rule may come less from social
which people see themselves as capable of attaining, The leader of the social move- movements within his nation than from a palace revolution-for instance, frorn one
ment, or revolution, would play his role in part by making clear to people the dis- of his lieutenants exploiting some of his delegated power. Stalin's long period of
crepancies between their expectations and capabilities and would suggest ways of dictatorship, involving many contradictory actions on his part, can be understood as
joint action to realize their capabilities. his personal interpretation of external and internal threats. In the beginning Soviet
Russia faced external threat and the need for internal restructuring to provide a
Totalitarian Versus Democratic Structures viable economy. Stalin's policies were directed at these difficulties, but his perception
of Russia's needs was readily identified with his own need for personal power.
To the anarchist and to the romanticist, all organizations may look alike in Consequently, he turned his attention from the preservation of the revolution to the
that they supposedly stifle the human spirit. But organized structures, like cities, can liquidation of fellow Bolsheviks who might threaten his own domination. Thus,
differ; and one key to their understanding is the source and exercise of legitimate systemic forces set the stage for Stalin's dictatorial role, but his reign of terror was
power. The source of power can be at the tcp of a hierarchical structure, with some less a response to system needs than a function of his own paranoid personality (Nove,
delegation of its exercise down the line. O r its source can be in the people and its 1964; Tucker, 1965) .
exercise delegated to elected officials. In large democratic societies, there can be Totalitarian leadership, then, moves toward making its control as thorough
some mix of these patterns, since the administrative bureaucracy is not only hierarchi- and tight as possible. Its target of preventing rebellion and counterrevolution expands
cal but also remote from the control of the electorate. 'To make officials more re- to the persecution of anticipated personal competitors for power. Stalin is a textbook
sponsive to the needs of the people is a top-priority issue for large modern demo- case of this pattern, as is Hitler. A dramatic exception can be cited in Marshall Tito
cratic nations. Nonetheless, the social structure in which leadership is exercised in of Yugoslavia, under whose regime the nation shifted from a centralized bureaucracy
good part determines the nature of that leadership. In democratic structures the to something of a decentralized system, and the borders of the country became in-
leader is concerned vrith the engineering of consent, with a social-exchange process creasingly open both for natives and foreigners.
among groups and between himself and his followers, with negotiation and compr9- This exception calls attention to two objective factors so far neglected, as
mise, and even with an integration of conflicting elements in the society. No matter well as to Tito's systemic perspective and charisma. In the first place, we have not
at what level he operates, he knows he is answerable sooner or later both to his mentioned the dependence or independence of a system as a limiting condition on
general electorate and to specific subgroups. In contrast, the totalitarian leader is the totalitarian leader. National states in the modern world are not self-sufficient;
not concerned with a social-exchange process in which he maintains a favorable thus, Spain, in spite of its geographical position and history of exclusiveness, could
balance but with preventing the exchange process from becoming operative,. His orien- not achieve the status of a closed society. Tito had the choice in 1948 of becoming
tation is not to meet the needs of people but to prevent those needs from becoming a Russian vassal or moving toward the West. By choosing to open channels to the
articulated and channeled into collective expression and action. West, he increased the power of Yugoslavia as a third force in world politics.
The totalitarian leader has at least three critical continuing problems. First, In the second p!ace, we have been discussing an ideal of totalitarian society,
all institutions must remain under the control of a single authority system. Not where there are no power groups outside the official political party and where the
only is there a single political party, but the schcols, the press, the church, and even official party leadership dominates all phases of life. But even when there are 110
organized sports and recreational groups should answer directly to the political state. formal structures to support rhe interests of various factions, power still may build
Any organized group with a fair degree of autonomy can become a means of mobil- up in military cliques, among technical experts and professionals, or among func-
izing people against the regime. Soviet Russia's treatment of Jews is less a matter of tionaries. In Yugoslavia there was another potentia1 in affecting nationa! decision
old Russian anti-Semitism and more a matter of totalitarian policy, which finds it making-namely, the ethnic identification of the component subgroups, such as the
dangerous to admit any pluralism into the system. Serbs, the Croatians, the Macedonians, and the Slovenes. These groups had a strong
A second problem is that contact and cornmunicatior, with the outside world consciousness of kind (their own kind) and little affection for some of the other
must be heavily restricted. Internal totalitarian control gives potential rebels no ethnic groupings. The centralized bureaucracy of the federal system had not
chance to organize, but they can get a foothold in other countries and keep alive destroyed these divisions and in :act had too many other diacuities, such as its
some opposition. More important, communication with the outside world can con- eastern inheritance, to function well. By decentralizing down to the level of com-
tribute to economic and social change and can provide new standards and make munity and work place, Tito reduced the strength of the ethnic groupings. Lesser
problems for dictatcrs. Totalitarian regimes need a closed society. The Spanish leaders would have lacked the penetrating insight of Tito into the system forces,
dictatorship mzy over time be undermined through opening its doors to tourists and both within his nation and within the world community. He had breadth of per-
226 Daniel Katz Patterns of Leadership 227
spective and an appreciation of trends of events over time. Moreover, he had the of interests by political leaders requires some compromise and even allows for some
decisiveness of mind to take the appropriate actions at the appropriate time. integration. In Almond's model, democratic leadership involves the social-exchange
Finally, Tito enjoyed a charisma based upon an unusual combination of paradigm of Blau ( 1964) and suggests a dominantly rational decisiorl-making process
factors. He had been the strong, courageous military genius who had led the common by the leaders.
people against the invading Italian fascists and German Nazis. He helped create the There are problems, however, in the use of the concept of social exchange in
new nation. He symbolized the people's cause in rebelling against their own nobility, trading favors and obligations in the political arena. Not only is too rational a
landlords, and plutocrats. He was the George Washington, the Abraham Lincoln, decision-making process suggested, but we know little about the identity of all the
the Napoleon of Yugoslavia, and still alive and vigorous. He could confront powerful agents in the transaction. We do know from careful empirical study that such na-
Russia and rally the Yugoslavs against the threat of foreign domination. Sidney tional political leaders as members of Congress do not receive votes from their
Hook, in making his case for the heroic leader, could well take Tito as one of its electorate in return for the policies they support. Stokes and Miller (1962), in a
clearest examples. thorough study of congressmen and their constituencies, found that the electorate
I n countries that historically have lacked strong voluntary or semivoluntary knew little about their congressmen and their actions in Washington. In fact, over
associations of people, the totalitarian leader can more easily control organized in- half the voters did not even know the names of their representatives. At the presi-
ternal opposition. The colonels in Greece were able to bring off their coup and to dential level there may be more accountability, and Gamson (1968) has used the
remain in power because Greece had weak labor unions, poorly organized political social-exchange model to suggest that a president can build up political credit through
parties, frail professional organizations, and in general no strong secondary associa- past behavior but then may exhaust it, as Lyndon Johnson did in his policies toward
tions. Few, if any, organized structures stood between the individual and the state. Vietnam. It is interesting, however, that in 1956 President Eisenhower was not
The colonels, then, needed no real popular backing when they seized power, nor did evaluated on the basis of his policies or actions as chief executive; his great popularity
they need it to remain in power. Individuals, no matter their number, are powerless was a personal popularity, based upon his being perceived as a nice man, a man of
without organization to summate and integrate their efforts. Totalitarian leaders can integrity, sincerity, and warmth-a good family man. This was the well-documented
maintain themselves for long periods without meeting the needs of their people for finding of the Michigan researchers (A. Campbell and others, 1960) who reported
a better way of life. that favorable references to Eisenhower "already strongly personal in 1952 became
The role of secondary associations, of voluntary and semivoluntary associations overwhelmingly so in 1956" (p. 56).
in mediating between elites and masses, is central in Kornhauser's (1959) thesis of The exchange process between elected officials and their constituencies
mass society. He holds that two sets of factors predispose a society toward totalitarian probably occurs in much more limited contexts than even politicians believe when
movements. One is the autonomy of the elite, the access to it through free elections they so carefully attempt to get mileage out of issues. I t does occur for small segments
and equal opportunity in the educational system. The other is the availability of the of the population, special interest groups who are more knowledgeable about what
masses to manipulation by the elite. I n a society where there are few and weak is going on in Washington. Numerically they do not count, but they may be in a
organizations intervening between the ,state and the citizens, the masses are readily position to influence others. I t also occurs when a problem-such as the Vietnam war
available to persuasion and direction by elite leaders, Mass society, in Kornhauser's or the busing of children-becomes critically salient for many people. On the whole,
theory, represents such a co~ditionof ready manipulability of people; in addition, however, the exchange process is much more common between leaders than between
in mass society elite positions are accessible. Such conditions can give rise to a mass leader and electorate.
movement led by demagogues who can convert the society into a totalitarian regime The complexities of leadership in a democracy are well summarized by
by establishing themselves as rulers and then freezing the elite structure. Greece Neustadt (1960, p. 179) in his study of Franklin Roosevelt, Harry Truman, and
would be an example of such a series of events. Dwight Eisenhower: "Effective influence for the man in the White House stems
Demagogic leadership not only fails dismally to keep its promises to the masses from three related sources: first are the bargaining advantages inherent in the job
but clearly never intended to keep them. The leader of the mass movement who with which to persuade other men that what he wants of them is what their own
promises a more equitable society while planning to take power himself is clearly in responsibilities require them to do. Second are the expectations of those other men
this category; but many leaders in a democratic society also are often guilty of some regarding his ability and will to use various advantages they think he has. Third are
degree of deception. The many subgroups with diverse and even conflicting demands those men's estimates of how his public views him and of how their publics may
make it difficult for the candidate, who needs the support of all of them, to present view them if they do what he wants."
a frank program. However, when there are strong voluntary associations of various
occupational, interest, and factional groups, there is a check upon the sellout of Motivation and Personality of Leaders
political leaders. They have to keep the faith to remain in the game. Almond (1963)
has approached this problem from a somewhat different angle in showing that the Though different characteristics of leaders may appear significant in different
function of political parties and their leaders in a democi-atic system is to articlllate contexts, are there some general motivational attributes that distinguish leaders from
and aggregate the interests of the various subgroups in the population. In this way followers? The distinguishing aspect could be the nature of the motivational pattern
the interests of the people get translated into governmental action. The aggregation or the general level of motivational arousal-leaders might have different motives
228 Daniel Katz Patterns of Leadership 229

or they might be more driven than others no matter what motive possessed them. At lower and intermediate levels in social structure, in our view, power is
In discussing motives, some writers tend to equate activities and motives. Just as not necessarily the dominant value for most leaders, but it does assume more im-
bankers are assumed to have an affinity for financial affairs, politicians are said to portance at higher levels. There is some empirical support for this hypothesis. A
be motivated by a love for power. This in fact was an early thesis of Harold Lasswell study of precinct and higher-level leaders in the Detroit area found such differences
(1948). And one can cite many instances of political leadership that support this (Eldersveld, 1964). About 25 percent of those in the higher echelons (compared to
thesis. Neustadt (1960) though an admirer of Franklin Ro~sevelt,describes his life only 1 percent of the precinct leaders) reported that economic and political gains
as a romance with power, with the White House as his natural home and the Ameri- were their sources of satisfaction from their party activity. Over half of the precinct
can people as his extended family. Biographers of many outstanding political leaders leaders (compared to about one fourth of the higher-level leaders) emphasized social
present similar portraits. Hargrove (1966) does suggest, in his account of six presi- satisfactions such as working with congenial and like-minded people and friends.
dents, that only the presidents of action (Wilson, Theodore Roosevelt, and Franklin Unfortunately, the very top levels of the party structure were not represented, so the
Roosevelt) were power driven, while the presidents of restraint (Taft, Hoover, and evidence is incomplete. A study of precinct leaders in a Norwegian community gave
Eisenhower) were not so motivated. Again, this seems more of an equation of be- very similar results (Valen and Katz, 1964). At this level in the party structure,
havior and motivation than a systematic psychological analysis. And there are in- leaders seem to obtain their substantial gratifications not from satisfying some power
stances of leaders whose other values are much more dominant than their power drive but from being involved with like-minded colleagues in meaningful and in-
drives-who do not choose to run for reelection or who risk defeat because of their teresting activities. More direct questions about career goals in a study of legislators
stand on issues. And those, like Lasswell, who favor a power hypothesis modify and in New Jersey, Ohio, California, and Tennessee showed an. even more inflated
qualify their position to include other motives. Even Wolfenstein (1969), who argues emphasis upon altruistic-contributive as against selfish-exploitative objectives (Eulau
that "leaders crave, relish, and have confidence in their own power and authority" and others, 1951) .
(page 33), also speaks of the need for "a feeling of the rectitude or legitimacy for We are concluding, then, that some generalizations can be made about thd
themselves and the cause they serve" (page 13). motives of political leaders but that these generalizations have to specify social settings
Greenstein (1969) carefully analyzes the relationshjp between personality and have to be statements of a probabilistic character. When we examine the ques-
tion of motivational arousal, there can be no question that political leaders are more
factors and political outcomes in showing the three sets of linkages involved: (1)
Deeper levels of personality are linked to political beliefs, but the association is im- internally driven than their followers. They work longer, they endure greater hard-
perfect in that the political belief is not a simple reflection of a deeper need. ( 2 ) ships, and they seem inexhaustible in their physical energy. Some of this drive,
Beliefs in turn are linked to behavicr, but again the reiationship is not one to one. of course, may be not a matter of motivation but a constitutional factor of sheer
( 3 ) The behavior of individuals aggregates in ways that are often not additive. Thus, physical energy and endurance. Long before American politics had reached its
we cannot predict directly from a depth personality factor to a political decision. But present strenuous character, and even before he himself had reached the White
Greenstein urges that instead of dismissing personality study we follow along the House, Woodrow Wilson (1908, pp. '79-80) wrote: "Men of ordinary physique and
complex set of linkages to provide adequate information about the whole process. discretion cannot be presidents and live, if the strain be not somehow relieved. We
Most social behavior, it is clear, has more than a single determinant, and shall be obliged aIways to be picking our chief magistrates from among wise and
the reward structure of organizations and social movements provides other incentives prudent athletes-a small class." The political Ieader must have not only an extra-
for leaders than the possession of power. In ongoing systems there are many sources ordinary physical constitution but also a high degree of internalized motivation. T o
of reward for leaders: higher income, prestige, affiliation and interaction with one's react to the immediate situation, to be a counter-puncher, is not enough.
fellows, a sense of accolnplishment from tasks completed. Barber (1965) found that I n addition to his physical endurance, the political leader survives the demands
state legislators showed wide variation in motivation and behavior. He was able to upon him because these demands feed his own patterns of motivation. The man
reduce the diversity of their activity to four patterns-each combining temperamental eagerly seeking election may be reinforced by having to make one speech after
another to different groups. An important motive in democratic political systems,
traits, personality style, motives, and skills.
Social movements, especially in their early periods, do not afford a variety of then, seems to be the need for affiliation, the love of social contacts: The need for
incentive of established structures for their activists. Two forms of motivation are affection "probably accounts for a major part of the membership of political clubs
likely to cone to the fore: power and ideologica! goals--either power-hungry indi- and interest groups and, to a lesser extent, the civil service and legislatures" (Knut-
viduals or fanatics devoted to a cause or some combination of the two patterns. son, 1972a, p. 46).Psychobiographic accounts have made a good deal of the need
Sometimes a single leader can be possessed by both motives and can handle the con- for affection in helping to account for the behavior of such important leaders as
tradiction by cloaking his own goals as the justified aspirations of his people. Harold Wilson (George and George, 1956), Trotsky (Wolfenstein, 1969), and Stevens
Lasswell (1930), in his, pioneer work in personality and politics, made this process (Brodie, 1966). One study noteworthy for its direct attack upon the problem by
central to political leadership. According to Lasswell, when private motives are measuring the affiliation needs of local officials and ward chairmen and nonpoliticians
displaced onto public objects and rationalized in terms of public interest, the resu!t (businessmen) reported higher need affiliation scores for the political leaders in the
more noncompetitive jobs (Browning and Jacob, 1964) . The same investigation
is political man.
 Daniel Katz

found no significant motivational differences between politicians and nonpoliticians

I Patterns of Leadership

active-positive type emphasizes a rational mastery of problems and therefore may

231

in power and achievement motivation. The level of position was more important i n , fail to take into account the irrational in politics. Active-negative people are com-
giving high scores on these needs than the political distinction versus the business pulsive characters suffering from a perfectionistic conscience. Passive-positive people
distinction. The need for affection is closely related to the trait of sociability, and are low in self-esteem and feel generally unloved; although they have some positive
most studies of sociability (for instance, Milbrath, 1965) report very significant corre- feelings toward their none-too-great involvement, these feelings are superficial and
lations between participation in politics and the need for social contacts. This rela- their capacity for enjoyment fragile. Passive-negative people are in politics only be-
tionship persists even when socioeconomic status is heId constant. Thus, "sociability cause they feel it is their duty to be there; they lack flexibility and tend to withdraw
should be called a necessary but not a sufficient condition for entering politics. Many or leave the field.
sociable persons do not become active. The reverse is not true, however; a n m - The following presidents are regarded as falling into these four character
sociable person has a barrier to participation in socially interactive political behavior" types: active-positive-Thomas Jefferson, Franklin Roosevelt, Harry Truman, John
(Milbrath, 1965, p. 75). Kennedy; active-negative-John Adams, Woodrow Wilson, Herbert Hoover, Lyndon
The higher drive level of political leaders relative to followers may be true Johnson, Richard Nixon; passive-positive-James Madison, Howard Taft, Warren
of outstanding performers compared to the rank and file in all walks of life: surgeons, Harding; passive-negative-George Washington, Calvin Coolidge, Dwight Eisen-
engineers, scientists, teachers. Other things being equal, the highly motivated person hower.
should exceed the poorly motivated in any field of endeavor. I t remains for empiri- The argument, then, runs that such a chdracter typology has predictive power
cal studies to test the degree of internally sustained drive in political leaders compared for decision making and effective policy formulation of given presidents. Woodrow
to other groups in the population. We have the took for such studies of leaders in Wilson's failure to secure Senate acceptance of the League of Nations can be related
depth interviewing, projective measures, and sampling methods; but so far no one to his compulsive righteousness, which did not permit compromising or negotiating
has combined these techniques in systematic research. Sampling is often the greatest with his Republican enemies. Hoover's inability to cope with the catastrophic De-
weakness, so that we cannot generalize beyond the few cases generally selected to pression during his administration is similarly accounted for by his stubbornness, self-
meet the conveniences of the researcher. In other words, there are no basic methodo- righteousness, and moralism. Finally, the tragic decisions on Vietnam in the Johnson
logical roadblocks to studies of differential value and motivational patterns of leaders administration were the result of Johnson's rigidity and his belief that the war was a .
in various levels in the political structure and in other organizations as well. matter of national honor and criticism of our involvement in it immoral.
Earlier we did not restrict political leadership to those in formal political Barber's typology may well stimulate theoreticians and researchers in the field
roles (party position or public office) since the same psychological processes of exert- because it is so incomplete and contains so many unproved assertions. In the first
ing influence can be found in other organizational settings. Studies that compare place, the theoretical framework is fragmentary, although some of his insights could
formally designated politicians with the rest of the population are thus not definitive lead to the development of a theoretical model. Why his two dimensions, activity-
in identifying the characteristics of political leadership. In the United States, industry passivity and positive-negative affect, should generate the properties attributed to
or the military may offer as attractive opportunities for the politically oriented as do his four character types is far from clear. These properties derive neither from any
the political parties and government. Nonetheless, the comparison of political partici- set of logical propositions nor from careful account of the psychodynamic processes
pants with nonparticipants should give a crude measure of relevant characteristics. assumed.
When positive differences are found, they merit careful consideration. Negative In the second place, there are no well-designed criteria for assigning presidents
differences tell us little. There should also be differences between elected and ap- to the four categories. Wilson, Hoover, Johnson, and Nixon are placed in the same
pointed public leaders, in that the former have to seek validation of themselves and ~ositive-negative category, characterized by a struggle to achieve power though
their policies in different fashion than the latter. Hollander and his colleagues have hampered by a perfectionistic conscience, The impediment of a perfectionistic con-
begun an interesting research program on such differences, and their findings furnish science somehow does not seem to be on target as the major character syndrome of
hypotheses for field research (Hollander and Julian, 19'10). Their research, for Johnson or Nixon or even Hoover. Since the two dimensions of activity and affect
example, suggests that immediately after election, when their credits are high, leaders are supposedly responsible for placing people in theje categories, we again need
are more confident of themselves and willing to deviate from the group; whereas specific criteria for the content analysis of documents, speeches, and records for
later, if their credits are on the wane, they are particularly sensitive to group opinion. measuring degree of activity and amount of positive and negative affect. Moreover,
Barber (1972a) has made a bold attempt to combine level of motivational these measures need to be taken over time and related to dependent and independent
arousal and type of motive in his analysis of presidential character. He selects activity- variables. Johnson, for example, may have enjoyed his exercise of power on the do-
passivity (the amount of energy invested in the presidency) as one dimension and mestic scene, and a negative evaluation of his activity on the international scene may
positive-negative feelings about one's own activity as the second dimension. The have been the result of his failure in that realm rather than the cause of it.
of the two dimensions provides fodr character types: active-positive, I n the third place, though the subtitle of the book is predicting performances
active-negative, passive-positive, and passive-negative. These four character types in the White House, this is post hoc explanation. Because the author was dealing
possess qualities that further differentiate them from one another. For example, the with historical materials, he did not predict future behavior for most of the presidents
232 Daniel Katz Patterns of Leadership 233

considered. He did, however, venture the prediction that Nixon, possessing the in- We have also attempted to relate patterns of leadership to three of these
flexibility oi the active-negative personality, would adhere rigidly to a failing line of dimensions. A pattern represents not a single motive but a characteristic set of actions
policy. In fact, however, Nixon has been the most pragmatic of all recent presidents resulting from motives, personality style, and cognitive skills. It is our conviction that
and has pursued a liberal foreign policy in more open relations with China and progress in this field will come not from attempts to 1ir.k a single motive or a single
Russia and adopted the Democratic line on wage and price controls. On other personality syndrome to leadership achievement, but from a moi-e thorough search
domestic issues he has taken a stand close to the position of the majority of the for consistent leadership patterns as they relate to general but identifiable character-
electorate. istics of the social environment. The personality-trait approach to leadership of a
I t is possible to make predictions when dealing solely with historical materials, past generation of psychologists proved barren. It was replaced by an emphasis upon
where knowledge is lacking about patterns of relationship. One can predict that if personality syndromes or types. This in itself is not enough in that there is a neglect
given variables are operative, then other specified variables will also be present. But of intellectual skills and other aspects of leadership behavior as well as the'relation
both sets of variables need to be specified and criteria for their measurement devE!l- to social settings. When our research and theorizing become more sophisticated, we
oped. For example, Child, Storm, and Veroff (1958) theorized that there would be may abandon the notion of leadership as a linear relation and describe it as a
a relationship in primitive societies between achievement motivation and economic circular process in which there are cycles of events mutually affecting one another.
enterprise. They used achievement themes in folk tales as a measure of motivation With the development of methodology in the social sciences and its increasing
and the presence of entrepreneurs as a measure o: economic activity and found the app!ication to practical and theoretical problems, it is difficu!t to account for the slow
predicted relationship. The Barber thesis, however, lacks such operationalization of research progress in the field of political leadership. The importance of content
its variables and does not carry us much beyond the traditional intuitive approach. analysis, of measurement, of specification of variables, of research design should no
Nonetheless, it may spur othqr workers to formulate hypotheses and develop measures longer be a matter of dispute in the behavioral sciences. Moreover, a number of
for their testing. investigators have demonstrated with considerable ingenuity how research methods
In most treatrrients of pbwer drive and political leaders, there is scant recogni- can be applied to historical materials. Part of the answer may be that we are in too
tion of changing patterns of motivation as societies move from preindustrial through much of a hurry and want complete answers even if they turn out to be fictitious.
industrial to postindustrial stage$., Knutson (1972a) relates such societal shifts to Because of the phenomenal success of natural science in recent years we forget the
Maslow's motive hierarchy (1954) and political participation and leadership in the laborious, careful years of effort that precede seemingly spectacular outcomes. Rut
future. She calls attention to Aronoff's, findings of intergenerational differences in there is no reason for believing that the problems of social science are so simple com-
psychic need satisfaction in a West Indian community as evidence of motivational pared to natural science that we can jump the steps of careful observation, of the
changes which can well affect the political scene. Her penetrating analysis suggests development of measures, of sampling design, of hypothesis testing, of theoretical
that with satisfaction of lower need levels the citizenry can become more self-actualiz- development, and of analysis procedures.
ing and that leadership in a truly democratic society may be based not in compensa-
tory power needs but in self-actualizing tendencies. She optimistically suggests that
some 60 percent of Congress and party officials are minimally self-actualized and uses
V. 0. KeyYs(1966) description of political influentials as support for her position.
Key noted the consensus among the influentials that public opinion should prevail
and that democratic processes should be observed.

Interpretive Summary

I n this account of political leadership we have attempted to achieve some

balance between personality and situational approaches to political influence. It has
been widely recognized that both terms of the relation known as leadership--both
the behavior of the key actors and the social setting-need to be described and
studied. Nevertheless, little has been done to set forth the characteristics of social
settings and their relationships to leadership patterns. We have called attention to
five major dimensions of the relevant social environments : ( 1) degree of formal role
structure; (2) relative autonomy or dependence of system in which leadership is
exercised; (3) the mix of primary and secondary relationships; (4) the totalitarian
versus the democratic character of the authority structure; (5) position of the leader
in the hierarchical structure.
 Aggression, Violence, Revolution, and War 235

restated. They invariably imply or assert the moral notion of mankind's innate
wickedness or the scientific notion of his innate tendencies to violence.
These predictions are based on knowledge that no one really possesses about
the roots of violence in the nature of men. I t seems quite reasonable to suppose that
all men have in common certain basic tendencies to behave in predictable, ordered
ways. I t does not seem quite reasonable to suppose that these tendencies invariably
include an innate desire to do violence. Rather, as I will later argue, it seems more
reasonable to suppose that violence is a response to the frustration of desires that
indeed are innate in all humans.
We cannot make definitive assertions about the certainty or even the high
AGGWES91[ON, probability of total annihilation, but we can diminish the muddle of thought about
violence and its causes. T o do this, we have to consider factors originating in men's
nature (their organisms) and in their environment. But since the commonest asser-

VIOLENCE, tions about violence imply or assert things about men's nature, those factors will here
be most carefully considered that originate within men, in their central control
systems-that is, in their nervous and endocrine systems. These two parts of men's
anatomies will be discussed first psychologically and then physiologically.
REVOLUTION, Some Definitions

AND WAR Much of the emotional wordage about violent behavior results from the
failure of people to make clear what they are talking about. So let me state rather
precisely, at the outset, how I use the terms aggression, violence, revolution, and war.
Aggression is a tendency to engage in hostile and intentionally destructive acts
-purposive acts whose consequence is injury to other living objects (usually people)
James Qowning Davies or damage to things that pertain to other living objects.=
Violence is the characteristic of acts that are successfulIy injurious or damag-

Intent and purpose are here regarded as synonyms and defined very elementally.
I They include acts ranging from a clam feeding itself to a bear scratching itself to human
beings doing either of these things and in addition making love or war. Purpose is not here
regarded teleologically in the philosophical sense but only as goal-orientation. Purpose is
regarded as emerging out of the activation of wants within the organism and as indicating
a tendency to act in a way that can satisfy those wants.
Most interaction between persons and between persons and objects is here deemed to
be intentional, purposive. Whether intent is conscious is a matter I do not intend here to
discuss. If a clam uses its muscle to open its shell to get food, I am willing to infer intent
I n the first Christian century a prophet without raising the issue of conscious intent. Many of men's actions are such that one can
named John wrote what he said was the revelation, to him from Christ, of the end suppose a person knew what he was doing but was unable to control his action. Whether
of the world. I t was imminent and it was horrible, and all the evil people were going a person is able to control his action-that is, to decide whether to do a purposive (goal-
oriented) act is another question to which neurology may sometime provide some answers.
to be destroyed. I n the twentieth Christian century both evangelical and scientific .
Too many mentally ill people knowingly commit violent acts for us to say that these (often
prophets have been predicting, for decades, the same thing, The end is imminent consciously) intended acts are controlled.
(since the end of World War I1 the clock on the cover of T h e Bulletin of the Atomic The factors of intent and conscious control nevkrtheless have to b; considered to
Scientists has variously read five or eight or twelve or seven or two minutes to avoid assuming that human beings have either total control or no control over their actions.
twelve), and everything on earth will be destroyed. The fact that the end has not If the distinction is not made, every traffic fatality would send every surviving driver to
yet come-at least by the time the reader reads this-is altogether ignored. The prison. The legal distinction between wiIlful and accidental homicide remains both necessary
and ill-defined. And the legal concept of willful negligence is a semantic paradox: negli-
ancient and modern reasons for the inevitability of Armageddon are stated and gence implies that one failed to be reasonably aware of the consequences of his action;
willful implies intentional action. (For an earlier discussion of these definitional problems,
see Davies, 1970.)
236 James Chowning Davies Aggression, Violence, Revolution, and War

ing. It is one consequence of an aggressive tendency. Sometimes aggressive tendencies agitation surrounding the Reform Acts of 1832 in Britain, whose consequences were
produce violent-that is, destructive-acts. Sometimes, when sublimated, aggressive comparable in substance to the Jeffersonian and Jacksonian epochs in America.
tendencies produce acts that are constructive, creative, or nurturant. Sometimes, Examples of rebellions, of revolutions that aborted, are these : the Peasants' Revolt in
when one is overwhelmed by threat of injury or damage to one's possessions, he may Germany in 1525, DorrJs Rebellion in Rhode Island in 1842, the Civil War in
suppress aggressive tendencies, and thus they may not lead-at least immediately- America in 1861-1865, Antonio Conselheiro's rebellion in Brazil. in 1896-1 897, and
to violent acts. the Boxer Rebellion in China in 1900.
To avoid pedantic repetition in the discussion that follows, the phrase War is definable as a concerted, coordinated effort by two large societies to
aggressive acts sometimes appears as shorthand for "acts that are the result of oppose each other in mortal combat. The aggressor in war may be defined as the
aggressive tendencies." But it is important tc separate the inferred tendency from the warring party that crosses the frontier of the other, but this is a less central issue than
observable act because not all aggressive tendencies do end in violent acts and not the intention of both parties to the conflict to injure and kill each other as much as
all inferences about aggressive tendencies are accurate. is necessary to win-or at least not to lose-the conflict. Conflict, the generic term,
One more ambiguity about aggression requires clarification: the easy verbal refers to the total class of antagonistic interpersonal acts (both nonviolent and
habit of not distinguishing between the broad category of assertive acts and the sub- violent), ranging from a v violent" argument between brothers to a war between two
category of those assertive acts which are distinguished by the intent to injure or nations.
destroy. In this chapter our problem is to seek explanations for aggression, violence,
Assertiveness is a tendency to interact voluntarily, forcefully with animate or revolution, and war-not all interaction, all change, and all conflict. The discussion
inaninlate objects, which together compose the environment. This tendency leads to therefore sadly lacks the drama of talking about the war between the sexes, the revo-
acts that often, though not always, involve observable contact with persons or things. lution in fashion, the violence of starvation, and about violent argument between two
The contact need not be physical: it may range from waving a hand or shaking a people. But what this chapter loses in the raising of adrenalin levels, perhaps it gains
fist from a distance to shaking hands and saying "hello" to a person next to you--.or, in clarity.
alternatively, hitting him. All these actions are here deemed assertive. Those whose
intent is hostile and destructive are deemed aggressive. Some Rather General Explanations
Opposite to assertiveness is a tendency to be indifferent to the environment-
either withdrawing from it without affect or letting it act on one without approving The ancient urge to understand ultimate causes is so strong that most people
or disapproving, liking or disliking. During anesthesia, a person being surgically sooner or later find it impossible to wait for the truly final statement of ultimate
treated on an operating table is indeed observably indifferent to the surgeon's non- I
causes. They produce such a statement-or adhere to one that someone else has
violent actions. But a person, perhaps a serf or a slave, who is spat upon by a gentle- produced. I t is not only evangelists and ideologists who have stated ultimate causes
man may appear to be indifferent or passive without actually being so. He may later, and responsibilities. Learned men, a subcategory of people, have also produced such
to his own and the gentleman's surprise, set fire to the gentleman's house, an ob- statements. Scientists, including psychologists, have also produced such statements.
viously purposive, violent act stemming from an aggressive tendency that follows It should therefore not logically be a surprise, but probably is a surprise psycho-
being spat upon and appearing to be indifferent. We later consider what it is within logically to note the distinguished and/or famous company to which the syllogism
men that spitting on them frustrates. applies.
Revolution is here defined as a substantial change in the power structure of Freud, Einstein, and May on Aggression. At the age of seventy-four, Freud
'
a society-a change accompanied by substantial violence. If the change is substantial (1930) expressed his notion of a dualism of warring instincts. These instincts, in
but is not accompanied by much violence, it seems more appropriate-particularly Freud's thought, are forces of which men are only rarely conscious.
in an analysis relating revolution to aggression-to call it substantial or even radical One of these instincts is Eros, the force that compels people to procreate
change but not revolution. If the change is not substantial but there is much violence
in the society, it is turbulence but not revolution-and certainly should be seen by
rulers as a very ominous sign that if they do not allow or welcome a change in the
1
sexual y and, when sublimated, to create in other than sexual ways. It is the instinct
that s eks life, that provides (or is, Freud sometimes said) the energy for the con-
tinuation of life and the development of culture. The other is the death instinct, the
power structure, a revolution will very likely take place. A rebellion is here defined fodce that seeks to destroy life, to return all living matter to the inorganic state. We
as an attempted but abortive revolution-that is, one that seeks to change the power
structure by violence but fails to do SO.
Examples of revolutions-according to this definition-are the French,
1
can call it the psychological parallel of the Second Law of Thermodynamics, the law
stating he tendency of all energy states to achieve homogeneity and therefore uni-
form abd minimal activity.
Russian, and Chinese revolutions and also the Nazi revolution. Examples of sub- Overseeing this conflict between life-energy and death, in Freud's system, is
stantial changes in the power structure that have occurred without substantial the ego, the rational set of forces that occasionally succeeds but usually fails to con-
violence are the Jeffersonian and Jacksonian epochs in American history, both of trdl the conflict between Eros and the death instinct. The role of the ego was at
which substantially broadened the power base of the American polity; and the one time compared by Freud to a rider on a rather wild horse; but whether the rider
 James Chowning Davies Aggression, Violence, Revolution, and War 239
seeks to protect the horse from death or to drive it to death is not to my knowledge at least decades if not centuries ahead of the evidence. It was more evangelical than
clearly discussed by Freud. The ego gets most of its energy from Eros and plays a scientific: It was "a sort of mythology."
sometimes constructive role in the struggle that ultimately ends in destruction-the May (1972) presents a synthesis of pychological ideas on the origins of
triumph of death. The kinship of Freud's systematization to the simpler and philo- aggression that is more broad and reasoned than was possible in Freud's time. There
sophically immortal dichotomy between emotion (manifestations of instincts) and have been more wars and violence to reflect on since Freud wrote. May defines
reason (the ego) is clear. aggression as ''a moving out, a thrust toward the person or the thing seen as the
From this dualism of Eros and the death instinct, Freud developed what to adversary" and the goal is "to cause . . . a restructuring of power" (p. 148). He
him was a necessary sequitur: aggression is innate in men; one of their two basic places what I have called assertiveness and aggression into the same broad category
instincts is to kill others and finally themselves, A part of them wants to cancel out and then distinguishes between constructive and destructive aggressios. His analysis
all the tension that life amounts to, all its frustrations, and to erase life-their own of its causes emphasizes the various demands which can produce a thrust toward the
and that of other-forever. adversary when they are frustrated. The desire to establish self-esteem, to count for
In a public exchange of letters with Albert Einstein, Freud stated explicitly something both as an identity and as an active force in society, and to actualize one-
what was not quite clear in his earlier writings: Man is instinctively aggressive. In self-all are demands which make one move against others. His analysis is clear and
this exchange of letters, Einstein raised the question: "Why can men be so easily specific, and it directly attacks, in his own well-defined terms, the stubborn adversary
aroused by their leaders to wage war?" He then answered it by saying: "because man of our ignorance of our constructive and destructive actions in a turbulent, violent
has within him a lust for hatred and destruction" (Einstein and Freud, 1932, p. 18). epoch.
To this Freud (pp. 40-41) replied: "I entirely agree with you. I believe in the Lorenz and Ardrey on Aggression and Sin. We can here consider only two
.
existence of this instinct. . . We [psychoanalysts] assume that human instincts are of the most celebrated of the recent writers who have undertaken to probe into the
of two kinds: those that conserve and unify, which we call 'erotic' . . or else . hidden causes of men's open destructiveness. These are two ethologists: Konrad
'sexual'; and, secondly, the instincts to destroy and kill, . . . the aggressive or de- Lorenz, the Austro-German student of animal behavior under (largely) natural
structive instincts." And he added: "The death instinct becomes an impulse to circumstances, and Robert Ardrey, the forceful American synthesizer of the work of
destruction when, with the aid of certain organs, it directs its action outwards, other ethologists.
.
against external objects" (p. 45) With a notable-and characteristic-self-conscious- Lorenz says that aggression is an innate tendency. It is the product of evolu-
ness, Freud said: "All this may give you the impression that our theories amount to tion, of the natural selection process whereby those genetic mutations that are viable
a species of mythology and a gloomy one at that! But does not every natural science under given environmental circumstances tend to flourish. That is, aggression is
c <phylogenetically programmed," "autochthonous," "endogenous," and "spontaneous."
lead ultimately to this-a sort of mythology? Is it otherwise today with your physical
science?" (pp. 46-47). Lorenz does not always distinguish between acts that involve quick and close contact
Part of Freud's solution for the conflict between the life and death instincts but lack the intent to injure or destroy and those acts that are destructive. But he
sounds much like Plato's reliance on guardians: "That men are divided into leaders does argue that men, like many other vertebrates, engage in destructive action toward
and the led is but another manifestation of their inborn and irremediable inequality. other persons and things. That this is so is too manifest to question. That the causes
The second class constitutes the vast majority; they need a high command to make for such universally observable acts are phylogenetically determined and endogenous
decisions for them, to which decisions they usually bow without demur. In this is not too obvious to question.
context we would point out that men should be at greater pains than heretofore to Scattered clues in O n Aggression show Lorenz's uncertainty about whether
form a superior class of independent thinkers, unamenable to intimidation and fer- aggression is an initial drive or a response to threat. At one point (1967, p. 104) he
vent in the quest of truth, whose function it would be to guide the masses dependent mentions what he calls "the allegedly irresistible drives of hunger, fear, or sex." At
on their lead" (Einstein and Freud, pp. 49-50) .' another (1967, p. 94) he mentions what he calls '< 'the big foury-hunger, love,
It now seems self-evident that Freud was on solid ground in exploring the fight, and flight." The relationship between these drives and aggression, at least in
unconscious determinants of human behavior and in asserting that sex impulses, or O n Aggression, remains unclear. It is possible, within his system, to suppose that the
life-energy itself, are major determinants of a variety of actions-some \of them frustration of these drives produces the universally observable manifestations of
relating to war and to aggression; that is, to the tendency of people to injure and aggression, but Lorenz does not systematically consider the relationship between each
destroy. I t is not at all self-evident, however, that Freud's clinical observations and of the drives or between them and the aggressive tendency. He is by no means op-
his observations of World War I were adequate to support such a confident assertion -posed to the concept of instincts and indeed argues at one point that they are not
of the innateness of aggression. His theory (and that of his successors) was (and is) as such dangerous. He says (1967, pp. 228-230) that the danger of annihilation
I arises not-from men's instincts but from their ability to think conceptually, ab-
2 Roazen (1968,'chs.4 and 5) discusses Freudian and post-Freudian psylhoanalytic stractly, without understanding the roots of human behavior. This theme recurs, as
ideas how society and government not only inhibit but also facilitate, both control and we see, in Arthur Koestler's analysis.
liberate, men from what Freud called their innate aggression (and other instincts). Larenz does not quite leave us with such an inconclusive conclusion. After
240 James Chowning Davies Aggression, Violence, Revolution, and War 241

presenting a massive, exciting array of data and an uncertain set of drives, he then century see it as being in the hands of man, uncontrolled by either natural laws or
proceeds optimistically to exhort. His prescription on the verbal level includes these supernatural forces. Ardrey also fits the mood of profound worry that man is not
phrases, used conjointly and more or less interchangeably: "rational morality," naturally good but naturally evil. He jokingly confesses that he has his own small
"reasoning morality," "rational responsibility," "responsible morality" (Lorenz, 1967, sins and roars at the sinfulness of man. The fit is precise: men are generally sinful, I
pp. 232, 245, 286). In short, an appeal to right reason. The basis for "good" behavior, am a little sinful, and if we all don't straighten up and do right-if we don't act
in Lorenz's view, lies in the nature of men. Any categorical imperative, then, is not contrary to our violent nature and in accordance with our innate territorial demands
really Kant's imperative (arising spontaneously from somewhere outside the nature -then the worId is doomed and in some vague way it's our fault, or more exactly,
of men) but man's. He must act in accordance with his nature. Man should know the fault of the other fellow. (For a recent critique of Lorenz and Ardrey, see Alland,
himself, including both his constructive (affiliative) and destructive (agonistic) 1972.)
instincts. Man must realize his little-tapped potential for affiliative identification Regarding the conclusions of Lorenz and Ardrey, we can appropriately raise
(love) with an expanding circle of human beings and his enormous capacity for the question that Freud raised with Einstein: Does not every natural science ulti-
disguising in the langauge of virtue the act of destruction. mately lead to a sort of mythology? Two facts remain indisputable: first, no one yet
Robert Ardrey (1961, 1966) has also put forth a well-known theory. It is that . knows as much as such writers claim to know about human nature and, second,
"the passion for territory is inborn, its borders learned" (1966, p. 266). Ardrey does humanity has thus far miserably botched the job of committing suicide, of truly fd-
not equivocate; he does not qualify; he does not say that social territory is simply filling its hypothetical death wish.
one kind of territory. He insists that people (and other vertebrates) have an innate If the generality of mankind and its leaders had taken seriously enough during
passion for space-land, water, and air. Territorial passion, in fact, is stronger than the last twenty centuries the evangelical and "scientific" assertions about innate
sex in holding people together, the attachment between males and females is based tendencies to violence, we would long since have committed total collective destruc-
more on territory than on sex, and the rules governing the use of territory are "more tion, human racial suicide. It is perhaps fortunate, that illiteracy is still high enough
permanent than sexual opportunity" (1966, p. 100). This passion for territory, in I that literate arguments postulating violence as an inborn drive are not more widely
Ardrey's view, seems to come ahead of anything else, except possibly food. read. It is surely unfortunate that along with illiteracy go the profound frustrations
What this passion for territory is itself based on is not quite clear. It is not that turn many poor people to violence, once they have the chance. In any case,
based on amity. In fact, amity is the product of enmity, emerging when two or more Hitler succeeded in both genocide and suicide to a frightful degree; the rest of us
people turn their enmity onto external objects as they join in defense of their chil- have for the most part failed. Why have we not expressed ourselves naturally, in one
'dren, their marriage, their home (1966, pp. 271, 273). Ardrey does mention a set glorious, irradiated, cosmic few seconds of deliberate or accidental total destruction?
of basic drives-for security, stimulation, and identity (1967, pp. 170, 333)-and all Just possibly because it is not our nature to do so.
these are satisfied by territory. Identity may be linked to solidarity, which may be Koestler on Selflessness. The words of ethologists who speak with tongues
linked to action, but, as we have noted, amity develops as one turns his enmity which individually are pointed but collectively forked are confusing. An even newer
away from those with whom he shares his territory and becomes aggressive against thesis, contradictory to the aggressiveness argument of Lorenz and Ardrey, is rather
outsiders who attack it. Territory is the categorical imperative. Everything else confounding, And for its supporting evidence it shifts from the observation of animal
appears to be secondary-not just to his analysis but to his belief about the funda- and human behavior in real-life situations to some basic theory and research in psy-
mental wellsprings of human interaction. The relationship between the territorial chology and neurophysiology.
and other lesser passions remains unclear. Only the struggle for territory goes on, Arthur Koestler (1967, 1969) does not trace man's aggressiveness back to an
innate tendency to destroy just for the fun of it. Instead he supposes it to be derived
and from it springs war.
from the urge to dedicate oneself to a cause, a leader, an ideology, and then to
What does become clear is Ardrey's intense moral tone. We must recognize,
proceed to sacrifice one's fellow men if possible and oneself if necessary to that cause,
by whatever terms, that we are all endowed with original sin. What we lack and need -
leader, or ideology. As evidence for this "urge," he cites such occasionally massive
is "a biological morality." The imperative mora! criterion should be the welfare of
popular movements as crusades, wars, and revolutions. And then his explanation
the population, which overrides the welfare of the individual. Biological morality
becomes neuroph~siological:it derives from his belief in the inability of the forebrain,
restrains the individual to the ultimate benefit of the species. I t is the survival of
the cortex, the reasoning part of the brain, to control the midbrain (the hypothal-
the species that matters, not the accidental fact of individual lives. Whatever helps
amus and related structures linked to the forebrain), Indeed, Koestler argues, the
man survive, then, is good. Whatever does not is bad. The virtue is in survival.
emotional part of the brain, which has been called the limbic system, crucially con-
Because of man's innate vice, there must be strong controls, in the interest of sur-
trols and enslaves the reasoning part, the forebrain. The limbic system is "the
vival of the species Homo sapiens. The criteria for controls (other than survival) ghost in the machine." Translated into philosophical language, it is the old argument
and for controlling the controllers he does not consider. that emotion outweighs reason.
Ardrey's popularity, like Lorenz's, is in small part a consequence of his Koestler sees the problem which threatens man's doom by the bomb as an
remarkable writing ability. The large part is that, like Lorenz, he fits a mood of imbalance between what he calls the integrative and the self-assertive tendencies.
deep anxiety about the fate of the world when so many people in our turbulent (In ancient China, it might have been called Yang and Yin.) Selfless dedication to
242 James Chowning Davies Aggression, Violence, Revolution, and War 243

cause, leader, and icieology reflects an excess of the integrative tendency. And this of his nature, psychologically considered. He refers repeatedly to the thesis con-
excess, of course, is the consequence of the failure of the forebrain of man to control tained in the charter of UNESCO that it is not in external cii.cumstances but in the
his limbic system. We return later in this chapter to a discussion of the physiology of minds of men that the origins of violent social conflict are to be found. But he does-
violence. however uncertainly-conceive war to be the outcome of an interaction process
between innate demands and an environment that fails to meet these demands. And
The reductionism with which writers from Freud to Koestler have treated in all his studies he never abandoned his etiological pluralism or forgot the need to
aggression suggests another reductionism: that the main barrier to our understand- analyze environmental forces. And he altogether rejected the reductionist tendencies
ing of aggression lies in the tendency to dichotomize. When explanations and pre- of other students of the origins of conflict.
scriptions are posited in terms of unreason against reason, emotion against reason, One need no longer believe, I would argue, that the aggressive tendency and
limbic system against cortex, hate against love, nature against nurture, man's natural violent acts are the direct, unmediated product of an innate tendency or that they
virtue in a vicious society or his natural vice in a virtuous society-if these pairs are the simple product of environment. One can assert that innate tendencies, dis-
are posed in one form or another, then analysis becomes futile and endless. coverable by psychological and neurophysiological research, bear the same indirect,
Quincy Wright on War. I t is refreshing to note that a man who has con- mediated relationship to environmental circumstances that genes do to the total
centrated his lifelong intellectual endeavors on the study of war has avoided such organism. And one can assert that aggressiveness develops when innate tendencies
reductionism. Quincy Wright ( 1964, p. 108) comes up with four categories of cause are frustrated. The environment by itself cannot produce the organism without the
of war: technological, legal-rational, sociological, and psychological. These causes, in genes; the genes cannot construct an organism-not even an aggressive or territorial
his view, somehow emerge from the interaction process between nature and nurture- organism--out of themselves.
although Wright generally tends toward a neurophysiological, ultimately biological However, it is not the environment but the genetic structure that produces
orientation: "Cultures are but abstractions of common psychological elements in form out of the disordered environment. Correlatively, it is the human central control
aggregates of human beings" (p. 112). He talks about "general tension level" as an systems (the brain and the endocrines) which integrate behaviors that are paradoxi-
abstraction which relates to war as an outcome of individual's tensions and as a cally both constructive and destructive, integrative and disintegrative. I t is reduction-
product of various frustrations. These frustrations, or "apprehensions," relate back ism to say that the central control systems do this all without enculturation, without
to a system of "primitive drives." training and even conditioning. It is also reductionism to argue, as environmental
The main primitive drives are food, sex, dominance, self-preservation, home or sociological monists have argued, that whatever structure and order there is in
territory, activity, independence, and society (pp. 26-27, 320). These are somehow human behavior is altogether the product of environmental inputs. Wright did not
related to processes involved in pursuing satisfaction of other primitive "mechanisms" very systematically integrate into his writings-perhaps could not do so in his time-
like identification, rationalization, repression, displacement, projection, and the scape- either psychological or physiological research. He nevertheless avoided the reduction-
goat, And these things transform "natural human affections, annoyances, arnbiva- ism of most social scientists, who see at the roots of violence only the environment-
lences, and frustrations into group hostilities" (pp. 320-321). The listing of primitive and the social environment at that. And he shunned the simplism of the Social
drives is not quite clear, nor is their relationship to other phenomena. Indeed, at Darwinists and their preoccupied descendants, who can see only man's nature-and
one point Wright even includes among man's "biological instincts" "aggressiveness a destructive nature at that.
and sadism" or "the love of aggression and dominance" (p. 267). The polemics of ethologists have, in an anxious era, focused our attention on
In Wright's view, war is "inevitable in a jungle world; peace is an artificial aggression but have not increased our knowledge of its causes. Wright's open and
construction" (p. 105). War is "natural" in the sense that it is a response to "in- nonpolemical work did not focus a broad reading public's attention on aggression,
ternally generated interests and motivations"; peace is "artificial" in the sense that but-unlike the polemics of ethologists-did increase our knowledge of it.
"its maintenance depends on a general desire to maintain it, on a correct image of
the world as one whole, and on the guidance of political decisions and actions by Sophisticated Psychological Explanations
sound . . . knowledge" (Wright, 1968, p. 466). He does not make clear whether
this general desire is a wholly artificial product of culture or partly a product of An early major assault on the intractable problem of explaining aggression
the primitive social drive. was undertaken in 1939 by a group of Yale psychologists. In their classic pioneering
Wright hopes that this sound knowledge ("psychological, sociological, political, work, Frustration and Aggression, Dollard and his associates stated a set of theoretical
economic, and technological") will enlighten "national hubris" with "rational relationships which now appear oversimple but remain a logically and psychologically
themis." But rational themis in his system must be a part of men's natural endow- elegant proposition of great, though not ultimate, explanatory value. We will first
ment along with primitive drives. At some points he seems to regard it as the en- state their basic proposition and then indicate-how some people have modified it.
dowment of intellectuals only. He does not, however, consider the paradox that man's The theory is stated on the first page of the book: "Aggression is always a
(limited) ability to gather, assess, and use such knowledge is a derivation of man's consequence,of frustration," and "the existence of frustration always leads to some
possessing a highly developed forebrain and that this forebrain is part of man's form of aggression." The authors thus posit a direct one-to-one relationship between
nature, physiologically considered, just as much as primitive, basic drives are part frustration and aggression: if cause A, then effect B; whenever effect B, then cause
244 James Chowning Davies Aggression, Violence, Revolution, and War 245

A. They do not quite say that aggression is the only consequence of frustration but Berkowitz markedly clarified and quantified the relationships between various
that aggression is always a consequence of frustration. In other words, frustration will parts of the aggression process. He pointed out the relationship between the strength
inevitably produce aggression, although it may also produce other reactions. of the forces that are frustrated, the strength of the frustrating force, and the result-
The tightness of the logic contrasts markedly with the sloppy thought of those ant frustrated behavior. He examined how individuals weigh the alternatives of
who had struggied with the causal problem before-and who have struggled with weak or strong response, and of response that is directed toward the source of the
it since. But the Yale group's theory of frustration and aggression is not fully de- frustration or deflected away from it. And he was more specific than the Yale group
veloped. They are quite willing to include within aggressive responses to frustration about what the drives ("instigators" or "goal responses") are, including both physical
the kinds of aggression that are turned inward: self-hate, self-punishment, obse- and self-esteem needs. He pointed out that aggressive acts may also satisfy an ac-
quiousness toward threatening persons. But they do not consider that people may quired motive, may reduce internal conflict. They may also restore self-esteems3
make a nonaggressive response, including the response that Freud and others have More significant than any such accretions to our understanding of aggression
labeled positive identification with the aggressor-not obsequiousness toward the is Berkowitz's emphatic assertion that frustration does not always produce aggression,
aggressor but identification with him. A common example of such identification is the even in sublimated or displaced or internalized form. He mentions not only the with-
anti-Semitic Jew. Another is the "Uncle Tom," the black who identifies with whites drawal alternative to aggression-as when a person is unable to fight back against
and rejects blacks. Rather obviously, identification with the aggressor involves the that which frustrates him or fears worse consequences by aggressing than by not
hostile elements of both self-rejection and rejection of one's group; but it stretches aggressing. He also mentions something that Freud emphasized in his Civilization and
the categorical statements of the Yale group to say that identification with the Its Discontents, the alternative to aggression that may variously be called sublii~a-
aggressor, arising from frustration, is aggression and only aggression. tion or constructjve action.
The Yale group allows in principle for this kind of reaction by talking about I A person may, because of frustration (and several other factors, including his
factors that inhibit the expression of aggression, most notably the fear that expressing particular talents and training), engage in creative acts which only in the most
it will be met with punishment of some sort. They do allow for a reaction which is tenuous sense can be deemed aggressive against other persons or objects or against
turned against (displaced, in Freudian language) some object other than the source self. Such would be the case of Beethoven, who was at times forced into practicing
of the frustration, for a reaction in which the object may be self or other persons or on musical instruments by his mercurial and often drunken father and then became
things. A frustrated person may physically or mentally beat his wife after the boss in a composer. Presumably Beethoven's prime purpose in composing was not to get back
the factory or office has frustrated him. This action lies within the Yale group's range at his father, but perhaps some of the intensity of Beethoven's efforts is attributable
of aggressive responses to frustration, as would a response of self-punishment. I to a childhood-established pattern of work to avoid punishment by father. Such also
But their analysis leaves out two things of great significance. One is the or- 1 would be the case of Lincoln, who said that his father taught him how to work but
ganic aspects of the frustration; that is, what internally causes the person to be never taught him to like it, and who did not go to see his father when he was dying.
frustrated. The other is a full consideration of alternative responses to frustration. Despite his hostile feelings toward his father, Lincoln sublimated his frustrations by
The Yale group mention almost nothing about what it is, internally, that is activity that made him a success.
frustrated. That is, they talk very little about specific demands of the organism The situation that we have arrived at with the help of the above psychologists
which, when blocked, produce frustration. They use the term instigation but-with may be summarized thus: (1) Frustration tends to produce aggression. (2) Aggres-
the shyness of stimulus-response psychologists-talk little about what instigations sion is a common but not inevitable consequence of frustration. ( 3 ) The concepts of
are, except as they may be inferred from overt behavior. To the extent that they talk
frustration and aggression, to the extent that they are related to each other and
about organic demands that are environmentallyfrustrated, they say that a child who
neglect the internal causes of frustration, probably facus too exclusively on the
wants an ice cream cone and does not get one is frustrated, rather than that a child
immediate and concrete rather than the more ultimate and fundamental causes of
with an activated hunger drive will be frustrated if he is denied food. They tend to
violence. (4) Frustration may be one of the factors producing action whose effect is
say that someone disappointed in love will become aggressive rather than that there
not to injure or destroy but to create.
is an innate need for love which, when denied, will produce aggressive action. Need
in the language of the Yale group becomes the ambiguous phrase "goal response."
They thus avoid saying that there are any innate needs and sidestep the tricky Explanations by Social Theorists
problem of purpose and intent. The focus of attention remains relatively shallow
and close to the surface of observable acts. Major social theorists, from Marx onward, who have concerned themselves
A major leap forward in the reciprocal interaction between theory and re- with the causes of revolution lend increasing support to an interaction theory of
search was made in 1962 by a University of Wisconsin psychologist, Leonard Berko- violence-the interplay of fo~ceswithin the organism and the environment. Marx
witz. He clearly defined aggression as "behavior whose goal is injury of some object,"
although he later reintroduced an unnecessary ambiguity in talking about "instru- a The rise in self-esteem that may follow aggressive action was put in poignant
mentalJ' aggression, the kind whose "goal [is] other than doing injury" (Berkowitz, political context by Frantz Fanon (1968). What Sorel (1908) called "the sentiment of
1962, pp. 1, 31). glory," which is fulfilled by combat, seems similar to what is here called self-esteem.
246 James Chowning Davib Aggression, Violence, Revolution, and War 247
(1849) talked about the gradual degradation of the masses as a major step in the I
sequence leading to revolution and the development of class consciousness as a 03 Expected need satisfaction
necessary step between degradation and revolution. Tocqueville (1856) said almost
the opposite: I t is when circumstances are improving that men turn to revolution.
Actual need satisfaction An intolerable gap
Both writers concentrated heavily on the socioeconomic environment. They implied / I between whot people
much but said relatively little directly about intraindividual dynamics. / I
I want and whot they get
Writing in the twentieth century, when psychological knowledge had begun
to develop, the sociologist Pitirim Sorokin (1925, p. 33) did mention human im-
pulses: individual self-preservation, alimentary, sexual, group-defense, and self-
expression. Violent social outbreak is likely when "the 'cramped' hereditary impulse A tolerable g a p between
what people want and
begins to put pressure on the 'brakey-that is, on habit."
what they get
Lyford Edwards (1927), deriving his list of basic human "wishes" from W. I.
Thomas (1923), said that when these wishes for new experience, security, recogni-
tion, and response are frustrated, people feel balked in their expectations and revolu-
tion becomes likely. George S. Pettee (1938) developed further Sorokin's idea of the
"cramp" that faces people whose expectations are frustrated and emphasized the
need to measure quantitative changes over time in order to analyze violent change
systematically.
These separate tributaries have, more or less systematically, jointly con-
tributed to the establishment of a broadening river of theory and research on the
causes of aggression, violence, revolution, and war. What is most exciting is that few
workers on the subject of violence and revolution any longer have the brashness to the frustration of basic needs, a frustration induced by the sudden reversal in gratifi-
assume that they can ignore the central congeries of forces that operate within the cations. My list of basic needs was derived mainly from the need hierarchy of
central nervous and endocrine systems of the human organism as these interact with Abraham Maslow: deprivation of "adequate" amounts of "food, equality, or liberty"
forces originating in the environment. What was the good sense and precocious among people who "differ in their degree of objective, tangible welfare and status"
judgment of Quincy Wright is now relatively universally accepted by all those (Davies, 1962, p. 607). (For a more fully developed statement of the broad political
interested in theory of public violence. I t is less appreciated by many of those who relevance of Maslow's need hierarchy, see Davies, 1963, ch. 1-2; and Knutson,
have done research in the problem, as we shall see. 1972a.)
Somewhat later (Davies, 1969), I broadened and tried to make more explicit
this theory that violence occurs when a period of rising expectations and satisfactions
Recent Theory
is followed by an intolerably growing gap between the two. That is, the theory was
made more explicit by emphasizing that violence becomes increasingly likely when
What Ted Robert Gurr (1970, p. 6 ) called "the renaissance of systematic
any kind of basic need which has come to be routinely gratified suddenly becomes
theoretical and empirical work on political violence by political scientists" began in
deprived. For purposes of joining in political violence of the concerted sort that
the 1960s. The "renaissance," implicitly or explicitly, was the product of a broad
produces riots, rebellions, and revolutions, then, the frustrated aristocrat who is
range of prior writing, perhaps most centrally derived from the work of the historian
denied access to his career aspirations (perhaps by military service in a war he
Crane Brinton, whose work in turn was developed from a variety of sources, includ-
deems supportive not of good principles but of an elite that governs the economy,
ing Tocqueville, Edwards, and Pettee. The first case of the rebirth of this systematic
the army, and the polity) can readily join with his usual class enemies, with working-
study, on historical, sociological, and psychological grounds, was an article of mine
men distraught by inflation, with upwardly mobile people whose education is in-
entitled "Toward a Theory of Revolution" (Davies, 1962). I presented the idea of
terrupted by depression or war or the like. The conjoining factor is the common
a J-curve of rising expectations and gratifications followed by a period during which
state of mind: the profound frustration that develops when the environment in the
there is a short, sharp reversal in gratifications (see Figure 1). In this brief period
form of culture, society, and government denies the opportunity for basic needs to
the gap widens between what people want and what they get, and the probability
get fulfilled.
of revolution increases greatly. The sociological basis for the development of the
In an ingenious combination of the time-sequence theory of Crane Brinton
theory lies in the conflict that develops in a society, when major segments of the
(1938) and of some alienation theory, David Schwartz (1971) has stated a set of
society see themselves as acutely suffering from this sudden "subjective" deprivation,
sequential relationships between the individual and the political system as he and
typically occurring during an economic downturn. The psychological basis lies in
his values become alienated from it; he forms revolutionary organizations, carries
248 James Chowning Davies Aggression, Violence, Revolution, and War 249

out a revolution, and then becomes changed himselE as his alienation is rep!aced by perhaps the earliest, new pioneers are Ivo and Rosalind Feierabend and their asso-
the establishment of his values. ciates (1966, 1969, 1972), who combined the frustration-aggression hypothesis of the
A new integration of psychologically based theory of violence has been pre- Yale group with later developments of theory on the gap between expectations and
sented by Gurr (1970). He establishes as his basic proposition the concept of relative gratifications. From Lipset (1959) they borrowed statistical measures of development
deprivation. Karl Marx had said in his Wage Labour and Capital (1849) that (including gross national product; caloric intake, telephones, physicians, newspapers,
"although the enjoyments of the workers have risen, the social satisfaction that they and radios per capita; literacy, and urbanization). From Lerner (1958) they bor-
give has fallen in comparison with the state of development of society in general. Our rowed a threefold classification of traditional, transitional, and modern societies. And
.
desires and pleasures spring from society. . . Because they are of 2 social nature, they developed their own indicators of political instability on a seven-point scale
they are of a relative nature'' (p. 91). Building on this and more specifically psy- ranging from 0 (change following a regularly scheduled election) to 6 (civil war or
chological foundations, Gurr emphasizes the inadequacy of the frustration-aggression mass executions). The product of their work has been a massive correlation of
nexus and argues that revolution occurs when substantial segments of a society feel degrees of stability with Lipset's measures and Lerner's stages of development. With
relatively deprived. They feel so when they perceive a discrepancy between their statistical and classificatory precision, the Feierabends have established that nations
"value expectations" and the "value capabilities" of their environment. Put more in transition are the least stable and that modern nations are generally more stable
simply, people feel deprived when there is a gap between what they want and what than traditional ones. They have not yet been able to deal adequately with the more
they can get. difficultresearch problem of integrating their data with psychological theory and with
A major difference between the J-curve and this theory is the strong emphasis patterns of relative deprivation.
of the former on the crucial dimension of time--of comparisons between expectations The political scientists doing empirical research in violence and revolution
and gratifications as they change over time. Another difference is that Gum empha- have already shown an ability to analyze empirical data and, more often than not,
sizes the Marxian kind of comparison not just over time but as between the differ- a lack of psychological knowledge. Tanter and Midlarsky have conjoined various
ing rewards of various social groups. Another difference is that the J-curve theory indicators to the J-curve (Tanter and Midlarsky, 1967; Midlarsky and Tanter,
argues that there must be not just a gap but one resulting from a downturn in grati- 1967) ; and Tanter (1969) has conjoined various indicators to a classification of
fications. Although theoretically the gap may result in other ways, as when both sources of conflict between and within nations. In an ingenious analysis on a statisti-
expectations and gratifications rise but expectations increase more rapidly than cal and theoretical base, Russett (1964) has related the land-tenure system in various
gratifications, it is hard to find a case of rebellion or revolution that does not involve nations to their political stability. Bwy (1968) has combined psychologically based
a downturn in gratifications. Part of the downturn before the 1789 French revolution theory with extensive statistical data to produce a thorough synthesis of knowledge
was a set of economic reversals. Part of the downturn before the Black Rebellion of about Latin American political turbulence. In what promises to be a very long-range
the 1960s in America was the use of dogs and firehoses by law enforcement officers project, comparable in duration to Quincy Wright's extended and extensive research,
in Birmingham, Alabama, ,in April 1963-after black people had come to expect Singer (1963, 1968, 1972) has established the research base for long-term trend
that the days of wanton violence perpetrated against them (as in lynchings) had analyses of the antecedents of war. His research to date, however, shows a reluctance
passed (Davies, 1969, pp. 720-723, 727). to work with any kind of motivational analysis. Singer does not seem to deal ade-
Gurr's theory is founded also on a set of basic value expectations (welfare quately (if at all) with a relationship deemed essential by Wright-the relationship
values and deference values) derived from Harold Lasswell and Abraham Kaplan
between war and the wants of men. If Singer should turn to the manageable task of
(1950) and related to Maslow's (1943b) hierarchy of basic needs (the physical,
measuring these wants by various indirect indicators (as the Feierabends, for in-
security, affectional, self-esteem, and self-actualization needs). But Gurr's intega-
stance, have done), then his work should be a great contribution.
tion of Lasswell-Kaplan and Maslow so changes the latter's need hierarchy that it
The interviewing technique used in survey research has begun to be fruitful
becomes a randomly ordered list. Nevertheless, Gurr's emphasis on relative depriva-
for studies of political discontent. Lloyd Free (19'71)-following the seminal work
tion is indeed a significant forward step in synthesizing theory about public violence.
of Hadley Cantril's Pattern of Human Concerns (1965), which deals less specifically
Among other side benefits, his explanation for why men rebel provides a thorough
and scholarly summary and systematization of the work of others. His empirical with political violence-employed an ingenious and promising mode of interviewing.
evidence that real-life revolutions occur in the absence of the J-curve kind of time- I t deals less with overt political violence and attitudes toward violence and more
sequential relative deprivation is less persuasive. with its roots in social and individual discontent. Free successfully interviewed seg-
ments of the population in five countries, to find out whether they expect things to
Quantitative Research get better or worse for them in the years to come. He used a ten-point vertical "self-
anchoring scale," wherby the people interviewed could point to where they are now
The troubled decade of the 1960s has indeed, as Gurr indicated, seen a and where they will be in the future-higher or lower on the scale. Applying the
renaissance of concern, not just for theory but also for research. Among the earliest, technique to Brazil, Nigeria, Cuba, the Dominican Republic, and the United States,
250 James Chowning Davies I Aggression, VioIence, Revolution, and War

Free found that his data correctly predicted political violence in three of the five characteristics of those who commit destructive acts and the common circumstances
nations. in which potentially everyone commits violent acts. This calls for an extension of
the beginnings undertaken by such social scientists as Sorokin and Wright.
Toward Fundamental and General Analyses Some Integrating Principles. A basic premise of my analysis of everything I
I have here discussed is that all behavior is a product of ( a function of) the inter-
The study of violence (both political and nonpolitical) is subject to the same action of the organism and the environment. B = f ( 0 E ) .4 If students of violence
I
problem as that faced by the legendary drunk who looked for his lost keys under the start and end in the environment, they will consider only one of the two fundamental
street light because that was where the light was brighter. The study of violence by variables on the right side of the equation. If, as ethologists seem to be doing, they
social scientists is being undertaken where the data are visible. The data that are start in the organism and end with an analysis of only the most visible evidence of
examined are gross data about such trends as population growth, urbanization, occu- organic determinants-in which environment plays a maximally evident role, which
pational mobility, immigration, and gross national product. The implications of such they ignore-they will only replace the forthright, clear, and holistic conclusions of
data are by no means yet exhausted, but the tendency to look for the keys where the environmental determinists like Marx, Pavlov, and Skinner with a newly muddled
light is brightest is very often undertaken unsystematically, without any theory to set of errors.
make the search systematic. Meanwhile, there is an ever increasing and little-ex- Many social theorists and researchers depreciate both the existence and the
ploited opportunity to analyze violence in plain psychological daylight. activation within the organism of its basic demands as they develop from infancy to
When crudely conceived empirical research is conjoined with sensationalism, maturity. This leaves them free to make any generalizations they wish about the in-
the work tends to become ever more quixotic and its durability ever more ephemeral. herent violence of men. If they deny or if they fail to see or recognize the central
When a social scientist, laden with computerized data, asserts conclusive evidence to relevance of basic demands, then indeed violence, in their thought, becomes not a
prove that ghettos cause violence, one is entitled to wonder what caused and causes means of satisfying such demands but something that is engaged in for its own sake
violence in middle-class suburbs and on middle-class campuses. When a social scien- and is therefore indeed innate in just the same sense that the demands for water,
tist, inclined to even bolder generalization, says that overcrowding causes violence, food, and-I would afgue-dignity are innate.
one is entitled to query why such heavily compacted cities as Tokyo and Rio de Let me summarize some basic hypotheses. Violence is always a response to
Janeiro, instead of blowing up, continue to grow into megalpolises. Or to query why frustration, but violence is only a subset of a still-undefinable total set of responses
violent crimes in recent years are reportedly at least four times as numerous in New to frustration. Violence as a response is produced when certain innate needs or
York as in Tokyo. O r to query why there was such mammoth and concerted violence demands are deeply frustrated. If these needs are met by a nurturant environment,
by the hordes of Genghis Khan and Alexander the Great. Were their troops fugitives the organism usually is "healthy." If they are not, it never fully develops. I t withers
from the ghettos of Peking or Athens or Cairo? and dies after a brief time of flowering, or it turns to violence to gain need satis-
If a serious cause of violence is the symbolic stuff on television, why does it faction. I t never realizes the potential that is innate within it. The human potential
produce violent tendencies among such a tiny fraction of those who watch the cops is reduced to that of primates, lower vertebrates, or crustaceans.
beat the robbers or vice versa? More to the point, did those who watched the real Let me state the relevant motivational hypothesis-a modification of Maslow
violence of the Vietnam war, as reported symbolically on television broadcasts, show (1943b). There are four substantive needs, in descending order of prepotency: the
an increase or a decrease in their desire for American continuation of the war? physical, social-affectional, self-esteem, and self-actualization needs. In the day-to-day
In short, if ghettos are eliminated, big cities dispersed, and all television show- social interaction of every individual, any one need may dominate his attention, but
ing of violence removed, how much fundamental assurance will either social scientists, whenever any lower need (like the physical need for food or rest) becomes active, it
political leaders, or general publics have that they have attended to the fundamental will overpower the activity of the centraI control system that has been responding to
reasons for violence? Perhaps little more assurance than a person who jumps into any higher need (like the self-esteem of self-actualization need). A person may be
a rowboat has about eliminating the causes of floods and storms. With good luck activated to revolt by any one or a combination of need deprivations. The poor man
and more effective use of the innate potential for long-range memory, he may survive is most likely to be activated by physical deprivation-which if it is very severe will
to build dams and levees. Or survive to explore the headwaters of frustration and not deactivate him politically. The person of high status and skill is more likely to be
just stare at the flash floods of overt violence. With bad luck, he remains an evange- activated to revolt by deprivation of his self-actualization need to pursue an occupa-
list in scientific disguise (or the victim of one), and an evangelist (or victim) who tion suitable to his talents and training. But if he is actually hungry, the high-status
prefers shouting in the darkness because he fears to find in the daylight that his great "For a fulIer explication of this premise, see Davies (1963, pp. 1-6). In the present
discovery is only a rivulet. I
chapter I substitute E (environment) for S (situation) used in my earlier version. A similar
If indeed-as is too obviously the case-violence, both by private citizens and prior discussion of this premise is to be found in Lasswell and Kaplan (1950, pp. 4-6).
and by public officials, has occurred in a wide range of circumstances over both I They use the terms acts, nonenuironmental determinants (or predispositions), and environ-
human history and prehistory, it is then indeed appropriate to consider the common ment.
252 James Chowning Davies Aggression, Violence, Revolution, and War 253
person may be activated to revolt by physical deprivation as well as career depriva- and revolutionary leaders as George Washington, Thomas Jefferson, Andrew Jackson,
tion. and Dwight Eisenhower were such as to make their aggressive response to perceived
In addition to these four substantive needs, I postulate three closely inter- political threat, from within or without the country, a reasonable or normal one.
related irnplemental needs: security, knowledge, and power. Each person has some Similarly, though under quite different circumstances, the traumatic experiences of
innate tendency to gain minimal satisfaction of each of these impleinental needs, but Vladimir Ulyanov in his adolescence (the death of his father, his brother, and his
he pursues their satisfaction in the course of seeking means of satisfying one or more sister for reasons attributable: to the tsarist government) are necessary parts of an
of the substantive needs. That is, a hungry and unemployed workingman may exert explanation that regards Lenin's intensively hostile and vicariously violent actions
power by breaking into a bakery. A (substantively) career-frustrated university against the regime to be normal reactions. In the same general category of normal
graduate may seek an (implemental) explanation for his plight by reading and reactors to profoundly frustrating environments may be included such leaders as
adopting an ideology that argues the origins of misery within a particular social Trotsky, Danton, Cromwell, Castro, and Mao Tse-tung.
system. All such normal reactions to profoundly frustrating environments-whether
The basic phenomenon of violence covers a wide range of overt acts, whose the reactors be leaders or ordinary citizens-are, on the dimension of time, relatively
related antecedent events we may call causes. This range covers cases in which there infrequent, for at least two reasons. One is that among normal human beings there
is a great difference between the relative contributions of organic and environmental is a wide range of tolerance for frustrating circumstances. Seen from present-day
causes. Here are four examples of this range, indicating the varying proportions of perspective, it is nearly incredible that slaves in the Middle East and in North and
organic and environmental cause: (1) Some outbreaks of violence may be caused by South America tolerated or even endured their wretched environment. I t is likewise
brain abnormality-resulting from such things as abnormal cells in the amygdala or quite incredible that serfs in Western Europe and particularly in Eastern Europe
a tumor surrounding and destroying part of the hypothalamus; in such cases, the tolerated serfdom until as late as 1861.
abnormality is great enough to produce violent action following only the slightest Another reason is that, in the perpetually continuing dialogue between de-
environmental stimulus. ( 2 ) Other violent outbursts may be statistically rare overt mands of the organism and gratification of these demands by the environment, there
responses to very traumatic early childhood experience-as when one of three siblings is often (perhaps usually) enough flexibility in the customs, laws, and power structure
becomes a killer, all being raised in a family where the father regularly beat the of societies to provide minimal satisfaction of the demands made at any one point
mother and in a neighborhood where violence was common. (3) In individuals with in time. It is when the expectations of people to live more than an animal !ife of
no evident brain abnormality and no evident childhood trauma, violence may result primitive survival and consociation are aroused and when these basic human expecta-
from intense situational incitement to violence, like that facing soldiers in combat. tions are frustrated that revolutions and wars occur.
(4) Finally, in a person with normal brain development and normal childhood These new demands and expectations arise in the minds of people more or
experience, violence may be a response to chronic circumstances that are profoundly less unconscionsly. When a revolutionary leader advocates that the oppressed rise
frustrating but not like the violent matrix in which a soldier or a ghetto inhabitant against their oppressors, he is only verbalizing a response for which the predisposition
acts violently. has been long developing within him because of his own frustrated expectations. The
These four categories are designed to discriminate between re!ative degrees new demands develop first and are articulated later. A person typically experiences
of contribution of the organism and the environment to the overt violent behavior. hunger pangs before he says he is hungry. A person first experiences sexually mature
In each of the four categories there is an interaction between organic and environ- love longings in adolescence, after a long sequence of organic developments that
mental forces. For example, in the first category, a person suffering from brain tumor start in the pituitary. Only after the organic changes and the longings have generated
or other.abnormality acts violently on some cue, however slight, from the environment. within his body will he turn his attention to a girl and say "I love you."
Perhaps it is the mere presence of a wall or of a person. The organism's role is At this point it wolild be possible to veer off into some kind of full-fledged
dominant. In the last category, which includes most of those who get involved in (and premature) statement of revolution and war on a psychological base. This
war or revolution, the environment frustrates demands of the organism that are would surely slake the thirst that the intolerance of ambiguity about violence amounts
altogether normal. To reiterate, from the first to the fourth category the oganism and to and might conceivably be enlightening. But it is more appropriate first to de-
the environment both are necessary parts of the interaction process. In the first case, scribe psychologically the sequence of organic and environmental events that ends in
the organism is overreacting and the environment is not "objectively" frustrating. violence and then to mention some of the research that is the basis for describing
Cases in the second and third categories represent a mix of less-to-more normal or- such a sequence.
ganism and less-to-more frustrating environments. In the fourth category, the en- The first step in the sequence of events leading to violent action is the activa-
vironment is extremely frustrating. tion of some innately rooted demand within the organism. The resulting state of
Political violence is here regarded as falling into the fourth category, for most tension calls for release by or in the environment, which contains the means of satisfy-
of those who participate in it. Most (but not all) military and revolutionary leaders, ing the demand. When a person is thirsty, the environment can provide water. When
and most of those among the general public who join in such violent action, have a person is tired, the environment can provide rest. When he feels lonesome, the
normal organic behavioral tendencies. The circumstances of such American military (human) environment can provide him company.
254 James Chowning Davies Aggression, Violence, Revolution, and War 255

The second step in the sequence is the frustration of one or more demands controlled massive bombing, then responsibility for effecting or blocking change-
of the organism. 'This can take such forms as a failure to provide water (frustrating in Europe in the 1940s and Asia in the 1960s and 1970s--can be avoided.
a person's thirst) or quiet, dark, and comfortable surroundings (frustrating a per- Physiological Research. There is not adequate research to confirm or dis-
son's demand for rest). confirm the four-step sequence from activation of a basic need, to its frustration, to
The third (preaction) step in the sequence is the mental process of seeking the mental process of searching and deciding, and then doing it. But there are
and deciding how to overcome the frustration which the environment places between some seemingly fragmentary pieces of research which such a statement of sequence
the individual's demands and their gratification. This mental process can take various helps to make coherent.
forms: effort to find a path around the blockage; effort to find a substitute object I One line of experiments with laboratory rats has shown what happens when
for the one initially demanded (as in sublimation) ; effort to deny or suppress the they are faced with frustration of the need to be free from bodily harm (specifically,
demand (as in displacement) ; or effort to destroy the blockage. electric shock). In these experiments (Payne, 1968; Berkowitz, 1968) rats are placed,
The fourth step is the one in which the iadiuidual acts. alone or in pairs, in cages with an electrified grid floor. When a rat is alone, it tries
With respect to the elemental demand for water: a path around the blockage to escape from the situation and becomes intensely restless but does not attack the
might be for a thirsty man to decide to look for a well not controlled by the man waIls of the cage or itself.'When two rats are placed in the cage and shocked, each
who refuses to give him water; a search for a substitute might become a search for -will attack the other. This experiment may be regarded as a classically elemental
wine; denial or suppression of the demand might take the form of fantasizing instance of what Freud called displacement. When a substitute figure-some kind of
bubbling streams; an effort to destroy the blockage might be the decision to over- doll-is placed in the cage, an animal undergoing shock will attack it, though not
power the man denying water. so readily as it will another rat. In whatever variation, the experiments show the
During the third step-seeking and deciding-tension builds up and becomes sequence of arousal of the physical need to be free of pain from the environment, the
available for extreme physical effort. If the demand is actually met, the tension is frustration, the search, and the violent act. Another experiment (Lamprecht and
released and the person relaxes. But any mental effort or any subsequent physical others, 1972) found that rats immobilized two hours a day for twenty-eight days
effort that falls short of actually fulfilling the demand will at most release the tension fought more than a control group of rats did. The report also indicated various
momentarily. Whether it be fantasizing babbling brooks or listening to comparable chemical changes that took place following the immobilization, which is one form of
political orators, whether it be deciding to kill (or, in the action step, actually killing) physical deprivation.
the man who owns the well or the ruler who controls the government, such effort In another and now famous line of research, Harlow and his associates (1958,
in itself is of only momentary, mediate, and not ultimate consummatory effect. That 1962, 1970) examined the effects of environmental deprivation of the nonphysical
is, thought-and even action, whether nonviolent or violent-which does not actually need for affection among monkeys. In these experiments, neonate monkeys were
produce satisfaction of the basic demand will not release the tension that builds up denied interaction with all other monkeys, including mother and age mates. Under
when the demand became active. This point is a very crucial one for the theses such conditions, the researchers found, even the regular provision of milk from a
developed i n this chapter. fake mother contrived from wire, hardware cloth, and terry cloth does not provide
Following the third step of deciding how to overcome the frustration comes the measure of emotional warmth necessary for normal growth. Monkeys so raised
the fourth, the action. In the action step, the effort shifts from mental to physical. are not capable of normal play, normal sex, or any other normal social interaction.
The individual makes a move, more or less blindly, which has as its purpose con- When the females that are so raised and thus are uninterested in sex are involuntarily
summation of the basic goal, satisfaction of the basic demand. When it is a matter impregnated and produce offspring, their treatment of their own offspring is abnor-
of water, the action usually is not very blind. When it is a political matter, having mal and usually brutal. What is more immediately relevant to understanding violence
to do with rearranging the power structure of the society, the action is usually almost
is that such abnormally raised monkeys tend to oscillate between extreme withdrawal
blind. The actors who demand political change have a very clear idea of what they
and fierce rage reaction to those approaching the cage. Sometimes in a period of
do not want and a very unclear idea of what they want. Their opponents, the estab-
rage they bite their own arms, tearing the flesh away from the bone.
lished ones, are sure they want things (notably power relationships) to stay as they
I
A similar phenomenon was observed by Rene Spitz (1949) in his work with
are and are not sure why things should stay as they are.
children raised in a foundling home, where they had minimal contact with nurses and
Both the unestablished who want to become established and the presently
none with mothers. These children, like some of Harlow's monkeys, also often showed
established have inefficient notions of how either to effect or to block change. Fantasy
a shift in mood from withdrawal to rage. The sequence following separation of a
is not the exclusive treasure of either of these groups. Violence-oriented thoughts
and images and even violent acts may be (but are not necessarily) fantasies, escapes, child a t about age two from its mother is described by Bowlby (1969, pp. 27-28)
processes that have little to do with effecting or blocking change. One example is as (enraged) protest, despair, and then detachment.
the victory over tsardom that killing Nicholas I1 produced in Russia and the con- The deprivation in the Harlow, Spitz, and Bowlby studies is not physical in
tinued victory of tsardom that Stalin's rule amounted to. Another is the popular the usual sense, though body contact is part of it. The frustration of the nonphysical
assumption that if the ground troops can be replaced with the use of carefully need for affection, or what Bowlby calls attachment, is so crucial as to produce,
256 James Chowning Davies Aggression, Violence, Revolution, and War 257

sooner or later, a rage reaction. Rage and violence may thus be products of either behavior. In some of these cases surgical lesion (by electrolysis) of part of the
physical or nonphysical deprivation. Violence comes from frustration, but frustra- amygdala stopped the violent behavior (Mark and Ervin, 1970).
tion comes from many kinds of deprivation, Other research has been done on cases in which little or no environmental
The strong influence of social ties in inducing violence is also indicated posi- stimulus provoked the violent response. Rabies, the disease whose symptoms include
tively in another experiment. In it there was no radical loss of all social ties but a tendency to attack violently, shows the effect not of ablation or electrical stimula-
merely the threat of the mild social loss that disapproval amounts to. In a contrived tion but of a. virus. The rqbies virus attacks a particular part of the hippocampus,
situation Milgram (1963, 1965) found that subjects threatened with the forceful which is part of the limbic system. Olds(1965) summarizes research in which a
command of the authority figure in the form of the experimenter were willing to rage response was produced by nonviral, nonelectrical, nonmechanical stimulation.
administer what they were falsely told was a near-lethal shock to helpless victims, Potassium chloride was placed on the surface of the hippocampus through a tiny
stooges of the experimenter, who loudly protested the shock. The experimental glass tube (a cannula). When two rats so treated were placed together, they adopted
situation was not a real-life one, but there is ample evidence from combat behavior the posture of attack.
to indicate that people threatened with the loss of the approval of their crucially In a long and still-developing line of research with primates and other high
important military associates are capable of "voluntaryJJkilling of helpless civilians. vertebrates, Delgado (1969) at the Yale Medical School has made experiments
The N8rves. Traditional psychological research has been done with intact with electrodes permanentIy implanted in various parts of the brains of monkeys and
animals and people, leaving only reasonable inference as to what actually goes on of a fierce bull, the electrode being remotely activated. And he has been able to stop
within the "black boxJ' of the central nervous system when some need of the organism aggressive behaviors of dominant monkeys so effectively that monkeys which were on
is frustrated. However, under various names like psychobiology and physiological the top of a social hierarchy became docile and moved down the hierarchy. When
psychology new research and new findings are appearing that already have shed some a switch was then placed in the cage, the monkeys lower in the hierarchy learned to
light on the sequence of neural events. In a book with the appropriate title The switch off the aggressiveness of the top monkey.
Physiology of Hostility, Moyer (1971) has gathered systematicalIy the findings of Other research has provided some kind of mirror image of the effects of
many others and himself. The direct political applicability of such research is yet various kinds of interference that has elicited or inhibited violent responses. In the
to come. But the research does demonstrate that violent behavior is the result of a now famous research in which electrodes were implanted in the "pleasure centersJJ
stimulus applied from the environment to the organism. The neural and endocrine of the midbrain, Olds (1956, 1958; Olds and Olds, 1962) produced the opposite
control systems produce violence, but only upon external stimulus. of aggressive reponses. While not directly relevant to understanding the physiology
In a striking experiment, two German researchers (Holst and Saint Paul, of violence, Olds' research shows the high specificity of various parts of the midbrain
1962) showed what happens to chickens when an electrode is inserted into a particu- as they respond to external stimuli.
lar part of the brain and then a tiny electric current is passed through. Chickens This growing body of neurological research demonstrates that there are many
that knew and were friendly toward the experimenters became suddenly extremely specific parts of the brain which, when stimulated, produce violent behavior. It also
hostile, seeking to fly at and peck their faces. This research follows a long line of demonstrates that such behavior occurs only when there is internal lesion (as in a
investigations into the effects of cutting away, damaging (by cauterization or other- tumor) or external stimulation. No research demonstrates the spontaneous occur-
wise), or merely stimulating various parts of the brain: Walter Cannon and associates rence of violence.
found that removing all of the brain in front of the hypothalamus (the main part of The Endocrines. ,411 this neurological research says nothing about the other
the midbrain) in cats produced animals that were almost spontaneously fierce (as in part of the central control system-the endocrine glands, from the pituitary on down
step 1 of the four-step sequence) but unable to direct an attack at the perceived -that induces or inhibits violent behavior by secretion of tiny amounts of hormones
source of threat. Other investigators found that merely cutting a nerve tract between into the bloodstream. The three most significant endocrine glands affecting violent
the hypothalamus and brain areas in front of it (that is, towaid the forebrain, the behavior are the bean-sized pituitary (also called the hypophysis), the adrenal, and
cortex; "behind" would mean toward the brain stem and spinal cord) produced the sex glands. The pituitary, responding to signals from the brain as it develops,
fierce animals which directed their rage specifically at the source of the stimulation. secretes various hormones at various times to both the adrenals and the sex glands.
The amygdala, a prominent part of the brain structures lying between the The sex glands, on command from the pituitary, develop and in turn secrete into
the bloodstream such hormones as the androgens and estrogens in quantities that
midbrain and the forebrain, contains some substructures that increase rage and
vary by sex. More androgens than estrogens are secreted by male sex glands; more
others that produce placidity. That is, whetl parts of the amygdala are electrically
estrogens than androgens are secreted by female sex glands. Androgen levels have
stimulated, the animal becomes enraged; when other parts are stimulated, it becomes
been found to correlate positively with aggressive, including violent, tendencies.
placid. Removal of the entire amygdala has produced placidity in some animals.
Despite the efforts of some people by verbal means to make males and females
Electrical stimulation at various points of the amygdala has produced rage and
identical in all ways, there remains a difference in degree in aggressive and violent
placidity in human beings hospitalized with a history of epileptic seizures and violent
tendencies between men and women. Part of the difference is attributable to endo-
 258 James Chowning Davies I Aggression, Violence, Revolution, and War 259
crine secretions that are not very responsive to verbal commands or ideologies appraise stimuli so carefully. So they respond more immediately an,d skeletally-and
(Guyton, 1966, ch. 56, 72). less cerebrally.
Perhaps the major endocrine function involved in producing our end product,
violent behavior, is performed by two distinct but similar hormones. The hypothal- A Resume of the Roots of Violence
amus, operating not through the pituitary but through the sympathetic nervous
system, triggers not only the sympathetic nervous system itself but also the inner part None of the research or theorizing on violence has yet provided anything like
(the medulla) of the adrenals which secrete two hormones whose role in the body's an adequate basis for definitive statements about the causes of revolution and war-
response to threat is crucial. These two hormones are adrenalin (epinephrine) and whether on a historical, sociological, economic, psychological, or physiological basis.
noradrenalin (norepinephrine). The nerve endings of the sympathetic nervous system I n defense of even the most pessimistic ethologists (like Ardrey) and psychologists
also secrete adrenalin and noradrenalin, but in tiny amounts whose effects last only (like Freud), it can be said that their pessimism may have the sobering effect of
a few seconds; the amounts secreted by the adrenal medulla are larger, and their diminishing unenlightened optimism. Nevertheless-in my opinion at least-men's
effects last for several minutes. violence is not the direct product of an innate violent tendency. Violence is a response
Some of the effects of adrenalin and noradrenalin on body tissues are the to the extreme frustration of one or more innate substantive needs that range from
same; some are different. Both hormones are critical factors in behavior that may the physical, affectional, dignity, to the self-actualization needs and a response to
become violent in the total psychosomatic effort to protect the body from threats
I the extreme frustration of one or more of the instrumental needs-security, knowl-
perceived (or misperceived) by the visual, auditory, and other sensory parts of the
forebrain. Noradrenalin causes constriction of the blood vessels in the skeletal and
visceral muscles; it causes acceleration of the heartbeat; it causes relaxation of the
1
I
edge, and power-which provide the organic part of seeking satisfaction, in the
environment, of the substantive needs.
It is in the analytical area between innate tendencies and overt behaviors that
intestines; and it releases sugar into the bloodstream. Adrenalin does all of these both theory and research are most uncertain. Environmentalists, including stimulus-
things but has less effect on the constriction of blood vessels in skeletal muscle tissue, response psychologists, find man to be almost entirely the product of the shaping
produces more acceleration of the heartbeat, and much more effect in increasing the given him by his environment. Hereditarians, including those whose view of genetic
metabolic rate of the whole body (including the brain), which thereby becomes determinants is based altogether on their field observations of overt behaviors (visible
more active and excitable. One simplified way of describing the difference in effect of acts), find man to be altogether the product of his genes and maintain a studied
these two hormones is to say that noradrenalin tends more to prepare the individual aloofness from microbiology and from neurophysiology. Their interest in the con-
for physical action in response to threat, whereas adrenalin tends more to produce tents of the black box does not extend to opening it.
an internally aggressive response, one form of which may be to attack oneself. What is, nevertheless, more frustrating even than the tunnel vision of both
The proportions of adrenalin and noradrenalin secreted by the adrenalin environmentalists and hereditarians is the reluctance of physiologists to consider
medulla vary in different species. In man, the secretion is about 75 percent adrenalin man as having any motivations except the simplest physical and emotional ones.
to 25 percent noradrenalin; in lions, somewhat more noradrenalin secretion than Physiological research has typically had to do with the physical demands of the
adrenalin; and in whales and chickens, about 80 percent noradrenalin to 20 percent organism for water, food, sex, affection, and freedom from pain. Work is particularly
adrenalin. In behavioral terms this comparison suggests-but by no means demon- needed on the nonphysical demands of the organism if we are to find out why men
strates-that man's endocrine system, more than that of other vertebrates, disposes turn to violent, aggressive political action in war and revolution. I t is absurd to
him to respond to threat by partially internalized and deliberative activity. Other assert that such violent behavior does not involve specific responses to specific stimuli
species are more likely to respond immediately to threat by skeletal muscle activity,
including physical fight or flight behavior.
These are some quite tentative inferences from research in neurology and
I
I
that threaten people when they lose their sense of dignity. 'There is of course no
dignity gene, no dignity part of the limbic system or of the cortex; but wars have
been fought and people have fought to the death for reasons that have little to do
endocrinology. Perhaps the major distinction between Homo sapiens and other
with survival and a great deal to do with liberty, equality, fraternity, and the dignity
species is the complex of neural and endocrine factors which incline him to deliberate
of man.
and control his responses to the frustration of his basic needs. In man, that is, step We know that the limbic system, the "emotional brain," has some kind of
3 (the seeking and deciding step) is much more significant in the sequence from intermediary role between the forebrain and the brainstem. If there is any one re-
the internal activation of needs to the ultimate response in overt action. The compli- search area that remains largely unexplored, it is the neural links, the nerve-fiber
cated circuitry of the cortex, the forebrain, and the endocrines has a greater role in links, between the limbic system and the cortex. When these and a few other major
. man than in other species in deciding how to respond to stimuli, including those unknown systems are explored, we will then know better why a degraded serf will
perceived as gratifiers of what people want and those seen as frustrators of what burn down the mansion of his lord, why a black man who has never been beaten by
they want. Vertebrates lower than man-including primates and other high verte-
a white man will turn against him for reasons that appear to have little to do with
brates-do not have the amount of cortex that gives them time to sort out and the physical well-being of the black man. Meanwhile, if we continue to believe that
260 James Chowning Davies

the impulse to violence is a basic drive, we might as we11 discontinue our analysis
of the roots of aggression, violence, revolution, and war because then nothing short
of total collective suicide can really solve the problem of violence. If the impulse
to violence is derived from frustration of basic drives, then analysis that seeks to
satisfy basic drives of people in all circumstances becomes appropriate, and the
diminution of violence becomes feasible.

FRAMEWORKS IN

Herbert C. Kelman
A y e d H. Bloom

x e study of international politics typi-

cally has taken the nation-state as its unit of analysis. I t is difficult to challenge the
view that, in the current international system, nation-states are the central and most
powerful actors. The assumption that they are the sole actors, however, can be and
has been challenged (cf. Wolfers, 1959; singer, 1961; Alger, 1963) . One could gain
considerable leverage on the analysis of international political processes by focusing
on a variety of other actors, including individuals. The relevance of individual actors
becomes apparent when we recall that it is individual human beings who make the
 262 Herbert C. Kelman, Alfred H. Bloom Assumptive Frameworks in International Politics 263

decisions and carry out the actions that shape the course of political events; that it processes of foreign policy making or international interaction. The theoretical formu-
is individual human beings who perceive and misperceive, who give and withhold lations tended to view war and peace from the perspective of individual motivation-
support, who compete and cooperate, who kill and die; that it is the fulfillment of often emphasizing aggressive motives and psychopathological manifestations-without
the needs and interests of individuals across the world that constitutes the ultimate sufficient recognition that the role of individual motives can be understood only in
task of international politics. Clearly, the inputs of these individual human beings- the context of the societal and intersocietal processes that culminate in war or peace.
their perceptions, interpretations, assumptions, and reactions-help to shape political In contrast to the earlier work, recent psychological studies are more likely to
outcomes. These inputs represent the locus of analysis of the social psychologist and start at the level of international relations itself, rather than extrapolating from
the source of his potential contribution to the study of international politics. individual or interpersonal behavior. To be sure, some research or writing in this
area still can be criticized for its tendency to "overpsychologize" political phenomena.
Social-Psychological Study of International Relations There is, in fact, no consensus among current scholars about the role to be assigned
to aggressive motives in the causation of war, or about the extent to which one can
Since the mid-1950s, there has been an increasing volume of systematic re- properly generalize from interpersonal (or intergroup) conflict to international con-
search on psychological aspects of international relations. A variety of empirical flict, or about the significance of central decision makers' personality characteristics
studies, theoretical formulations, and policy analyses, focusing-directly or indirectly in the foreign policy process. Nevertheless, it is probably fair to say that the mid-1950s
--on international conflict and other issues in international politics, have made use mark a qualitative as well as a quantitative change in the study of psychological
of psychological (particularly social-psychological) 'concepts and methods. These activ- aspects of international relations. Since that time, conceptualization and measurement
ities are part of the larger evolution of two new and highly overlapping intellectual of psychological inputs into international relations have increasingly started out
traditions-the peace research movement and the "behavioral" study of international from an analysis of international political processes (and domestic foreign policy
relations-which have drawn on theoretical orientations (often, though not exclu- processes) at their own level. Research in the area has increasingly taken place in
sively, in the form of mathematical models), on procedures of data collection, and an international relations context or utilized the archival residues of foreign policy
on techniques of measurement and data analysis developed in the various social decision making and diplomatic interactions. Writings have shown increasing theoreti-
sciences. Research in these traditions, which has become a rather vigorous interna- cal and methodological sophistication, with greater awareness of the complexities
tional enterprise in the past decade, represents the convergence of at least two major one encounters in moving across different levels of analysis. And, most important, two
interests: an interest among an increasing number of international relations specialists groups of specialists have emerged and interacted closely with one another: students
in developing their field as an empirical social science discipline; and an interest of international relations, with a political science background, who are thoroughly
among an increasing number of psychologists, sociologists, economists, anthropologists, grounded in social-psychological concepts and methods; and social psychologists (as
and other social scientists in applying their skills and knowledge to problems of war well as students of other disciplines outside of political science) who have systemati-
and peace. cally educated themselves in the field of international relations.
The interest of psychologists in problems of war and peace and other aspects We shall not attempt a comprehensive review of social-psychological research
of international relations certainly antedates these more recent developments (see on international relations, but a brief listing of the types of studies that can be sub-
Kelrnan, 1968b). Beginning at least in the early 1930s, a series of empirical studies sumed under this rubric may be helpful at this point. T o illustrate these different
explored attitudes toward war and peace, toward aggression, toward international types of research, we have deliberately selected recent publications (after 1965).
relations, and toward one's own and other nations. (For reviews of the earlier work, Reviews of and references to many of the earlier publications can be found in Kelman
see Klineberg, 1950; Pear, 1950.) There were also various theoretical examinations (1965). We shall divide the studies into two major categories: studies focusing on
of the causes of war, within either a psychoanalytic or a general psychological frame- attitudes relevant to international relations held by various groups of individuals; and
work (for instance, Durbin and Bowlby, 1939; May, 1943; Tolman, 1942). What- studies focusing on psychological and social-interactional factors in international
ever the merits of individual findings or hypotheses produced by these efforts, politics and foreign policy decision making. As a rough approximation, one might
,

however, they did not add up to a systematic analysis of the psychological aspects of describe the first category as studies in which psychological factors constitute the
international politics. The empirical research was carried out largely within the dependent variables and the second as studies in which they serve as the independent
context of the general study of social attitudes, with little explicit reference to the variables.
In speaking of attitudes, we have in mind a whole range of dispositional
This chapter is a product of a research program on social influence and commitmknt variables, variously referred to as attitudes, beliefs (see Chapters Three and Four in
to social systems, supported by U.S. Public Health Service Research Grant No. MH17669-04 this volume), values, images, perceptions, expectations. Thus, in the most general
from the National Institute of Mental Health. The chapter draws on several earlier publica- terms, studies in this category are concerned with the ways h which individuals
tions by the senior author (see Kelman, 1965, 1968a, 1969, 1970, 1971, and 1972). Work relate themselves to their own nation and other nations, to the international system
on the chapter was completed during his tenure as Visiting Fellow at the Battelle Seattle as.a whole, to problems of foreign policy, and to the broader issues of war and peace.
Research Center.
264 Herbert C. Kelrnan, Alfred H. Bloom Assumptive Frameworks in International Politics 265

What is the structure of attitudes and images in these domains, what are their Attitudes related to international affairs held by various groups oE individuals
determinants, and what are the conditions of their change? Three subcategories of are a legitimate research interest in their own right. How relevant such research is
studies can be distinguished: to an analysis of international political processes depends on one's conceptualization
1. Attitudes toward foreign policy and international aBairs. The focus here of foreign policy making and international politics. Most analysts would probably
has been on general orientations toward foreign and military policy and toward war agree that attitudinal factors play some role in the process and would therefore
and peace, as well as on opinions on specific foreign policy issues, such as the Vietnam regard research into them to be of at least tangential relevance. For example, even
war. Often studies in this genre have centered on the dimensions of belligerent versus those who assign minimal importance to public opinion in foreign policy making
conciliatory and nationalist versus internationalist orientations. A review of many of would no doubt agree that the moods, expectations, and reactions of the public are
these studies can be found in Eckhardt and Lentz (1967). Among the determinants part of the context within which the decision makers operate. Similarly, even those
of general foreign policy orientations, research has explored both personality variables who look to geopolitical, economic, or organizational factors to account for the course
(see, for instance, McClosky, 1967) and demographic variables (ree Galtung, 1967). of interaction between two nations would probably concede that mutual attitudes of
A recent study by Modigliani (1972a), using public opinion surveys conducted during trust or distrust and perceptions of the other party's intent are likely to have some
the Korean war, tries to tease out the structure of public opinion toward the war bearing on the outcome. The political relevance of any particular attitude study
and to explore the relationship between policy preferences and social class. A number depends on its closeness to the actual context of foreign policy making and inter-
of recent studies (Verba and others, 1967; Brickman, Shaver, and Archibald, 1969; national politics. Thus, the relevance of a piece of research is more readily apparent
Granberg, 1969) have explored opinions toward the Vietnam war in the American if, as in Modigliani's (1972) study, it examines public attitudes in terms of the nature
population and in various special subpopulations. and sources of support that decision makers can expect for various foreign policy
2. National and international loyalties. The focus here has been on the rela- moves; or if, as in Kerr's (1973) survey, the sample consists of participants in an
tionship of the individual to the national system-the sources, the nature, and the international political enterprise; or if, as in Druckman's (1968) simulation, images
exclusivity of his commitment to it. Several analyses of nationalism from a social- are explored as they emerge from and feed back into political interactions. These
psychological perspective have recently been presented (Fishman, 1972; KeIman, studies, in fact, are closely related to the second category of research that we have
1969; Tajfel, 1969). Empirical studies have also explored the meanings of national identified: research on psychological and social-interactional factors that enter into
identity (Herman, 1970; Schwartz, 1967), the different ways in which individuals the processes of international politics and foreign policy malting.
may be integrated in the national system (DeLamater, Katz, and Kelman, 1969), This second category of research focuses directly on international political
and the relationships between loyalty to the national system and loyalty to subnational- processes. I t is concerned with such political outcomes as the type of foreign policy
units (Klineberg and Zavalloni, 1969). Another potentially important focus for pursued by a given nation, the nature of the decision reached on a specific foreign
research is the relationship between loyalty to the national system and loyalty to policy issue, the reduction or expansion of conflict, the means used for pursuing or
transnational or supranational units such as regional organizations, international or- containing conflict, the development of peaceful and cooperative reIationships, and
ganizations, or the global community. Such studies might throw light on the condi- the formation of alliances, of regional organizations, and of transnational institutions.
tions for the development of wider loyalties and the coexistence of multiple loyalties. These outcomes are explored in terms of some of the psychological or social-inter-
A recent study by Kerr (1973) represents a contribution to research on this problem. action processes that may help to shape them: for example, the dispositions and role
It is also unique in that it is based on interviews with an elite sample, centrally behaviors of various relevant actors; the leadership and influence patterns of deci-
involved in at least one aspect of international relations, and that it explores changes sion-making units, znd the internal and external pressures under which they function;
in their attitudes as a function of their direct participation in international political the mutual expectations and perceptions of the interacting nations; the psychological
processes. atmosphere within which the interactions are carried out; and the threats, induce-
3. Images of other nations. There is a long tradition of research in this area, ments, and other bargaining strategies and tactics that are utilized. In these studies,
focusing particularly on ethnocentric attitudes and associated positive stereotypes of then, social-psychological factors generally serve as independent and mediating
one's own nation and negative stereotypes of other (especially "enemy") nations. variables to account for political outcomes. This statement is somewhat oversimpli-
Recent studies have been concerned with multidimensional scaling of such images fied, however, in that analysis often focuses on social-psychological factors as part of
(Robinson and Hefner, 1'967; Wish, Deutsch, and Biener, 1970), with the develop- an ongoing, chainlilie process:Thus, for example, research on the role of mutual
ment of images of foreign nations among children (Lambert and Klineberg, 1967), perceptions in international conflict is concerned not only with the effects of percep-
and with the impact of international contacts on the image: of own and other nations tion on action but also with the way in which action (one's own and the other side's)
(Kelman and Ezekiel, 1970). Of special interest are studies that trace the develop- generates and modifies perceptions, which in turn influence subsequent actions.
ment and change of mutual images in the course of international political inter- We shall again list the research in this area in three subcategories:
actions. Druckman (1968) explored this process in the context of the Inter-Nation 1. Public opinion in the foreign policy process. The focus here is on the role
Simulation, a laboratory simulation of the international system developed by Guetz- that the moods, expectations, perceptions, and sentiments of the public, or of various
kow and his collaborators (1963). special publics, play in the foreign policy process. The shape of public opinion deter-
 Herbert C. Kelman, AIfred H. Bloom Assumptive Frameworks in International Politics 267

mines the population's readiness for various national actions, its interpretation of and provided the data for a study by Holsti, Brody, and North (1969), using content
reaction to international events, its degree of support for various foreign policy analysis of documents authored by key United States and Soviet decision makers
moves-and thus the freedom and constraint within which decision makers operate. during a critical ten-day period. Their analysis focuses on the relationship between
I t may be best to refer to this category of research more broadly as "domestic sources decision makers' perceptions (particularly of the other side's intentions) and their
of foreign policy," to use the title of a book edited by Rosenau (1967). The various actions. They compare the Cuban missile crisis to the events leading up to WorId
War I, on which their earlier research had focused, pinpointing some of the per-
chapters in his volume deal not only with the orientations of the citizenry toward in-
ternational affairs but also with the group and organizational processes through which
these orientations are articulated and with the ways in which they are brought to
i ceptual factors that may account for the very different outcomes. A study by Robin-
son, Hermann, and Hermann (1969) illustrates the use of simulation and gaming
bear on decision makers. The study by Verba and his associates (1967) on public techniques in research on foreign policy decision making. The study investigated the
opinion toward the Vietnam war illustrates the use of public opinion data to assess effects of crisis conditions on the amount and type of search behavior in which deci-
the population's readiness for certain foreign policy directions. By probing more sion makers engaged. A very different type of research in this general area focuses on
deeply than the typical public opinion poll, these authors were able to challenge individual decision makers or on decision-making units, rather than on specific
the assumptions, apparently held by many decision makers, about the kinds of policies decision occasions. Such research might explore the personality styles or the assump-
that the public would or would not support. White (1968), using a variety of sources tions and perceptions of key decision makers, or the operative goals and decision
(including some public opinion data), analyzed the perceptions of the Vietnam con- processes of organizational units with foreign policy responsibilities. Glad's (1969)
flict by various groups in the United States and in Vietnam-including their per- psychobiographic study of Senators Borah and Fulbright provides one illustration of .
ceptions of their own and the other side's attributes, goals, and prospects. He con- this type of research (see also Chapter Eleven in this volume).
cludes that the perceptions of the different groups are, to varying degrees, distorted
and traces the psychological and social processes that generate such misperceptions,
which in turn account-at least in part-for the onset and continuation of a war
that "nobody wanted." One final development that should be mentioned here is the
I 3. Interaction processes in international conpict. Social-psychological analysis
is particularly relevant to the study of interactions between nations or their repre-
sentatives. Many kinds of interaction can be subsumed under this rubric; for example,
international communication in its various forms, normal diplomatic contacts, or
use of aggregate data to infer psychological states prevalent within a population. collaboration within the framework of international organizations. Major emphasis
, in social-psychological research, however, has been placed on the handling of con-
Feierabend and Feierabend (1966), in a cross-national study, developed an index
of systemic frustration (discrepancy between social wants and social satisfactions) flictual relationships--on the ways in which conflicts are pursued, managed, and
from a number of national indicators, and related it to the incidence of domestic resolved. Interest has focused not only on observing the microprocesses involved but
political violence. Such aggregate indices of public moods could be related, in similar also on $discoveringthe conditions conducive to various outcomes; for example, the
fashion, to the incidence and nature of international conflict. conditions under which conflict is expanded or contained, under which it takes
2. Decision-making processes in foreign policy. Research in this category violent or nonviolent forms, under which it is terminated because one party has
focuses on the foreign policy decision makers and decision-making units within a prevailed at the expense of the other or because a cooperative solution has been
given country and explores the psychological and social interaction processes that worked out. A large experimental literature has developed in recent years, based
culminate in state action. A number of writers have used intensive case studies of on the use of the Prisoner's Dilemma (see Rapoport and Chammah, 1965) and
major foreign policy decisions as the raw material for analysis. Thus, Paige (1968) similar bargaining games for the study of conflict. These studies are considerably
presents a detailed analysis of the United States' decision to enter into the Korean removed from the level of international conflict, but they often try to incorporate
war, utilizing the Snyder, Bruck, and Sapin (1962) framework for the study of some of. the crucial variables involved in the interaction between nations. (For a
foreign policy decision making. Joseph deRivera (1968), working with the Korean recent discussion of the logic of such gaming research, see Rapoport, 1970; see also
decision and with several other foreign policy actions, emphasizes the role of the Chapter Fourteen in this volume.) The various chapters in a volume edited by
decision maker's constructioniof reality-which in turn is influenced by personality, Swingle (1970b) bring together many experimental findings on such determinants
group, and organizational variables--on the choices he makes. Janis (1971, 1972), in of the course and outcome of bargaining and negotiation as the type of bargaining
analyzing several critical decisions, traces the effects of group pressures toward strategy selected by the different players, the type of influence attempt (including
uniformity and ingroup loyalty within a small decision-making unit on the cognitive threats and promises) used, the personality characteristics of participants, and the
processes of the participants and hence the quality of the decisions reached. Allison situational context of the interaction. Games have also been used to test the effective-
(1971), in a detailed account of the Cuban missile crisis, contrasts the classical model ness of specific strategies, as in the Pilisuk and Skolnick (1968) study, which was
of the decision maker as rational actor with two alternative models, viewing deci- designed to assess Osgood's (1962) proposals for deescalation. Closer to the level
sions, respectively, as "outputs of large organizations functioning according to regular of international conflict are such laboratory procedures as the Inter-Nation Sirnula-
patterns of behavior" and as "a resultant of various bargaining games among players tion (Guetzkow and others, 1963), in which participants play the roles of national
in the national government" shaped by "the perceptions, motivations, positions, decision makers in an environment programmed to represent major parameters of
power, and maneuvers" of the various players (p. 6 ) . The Cuban missile crisis also the international system. The possible uses of this procedure are illustrated in a study
 Herbert C. Kelman, Alfred H. Bloom Assumptive Frameworks in International Politics 269

by Raser and Crow (1969), which explored the effects of a deterrence strategy based sions. In examining the sources and nature of individual contributions to interna-
on the capacity to delay response on the process and outcome of interaction in the tional political functioning, therefore, we are merely focusing,on a particular nexus
simulated world. A recent paper by Guetzkow (1969) reviews the various uses of in this larger process, in the hope that such an analysis may lead to creative reconcep-
simulation in international relations research. Another approach to the study of tualizations of some old issues in the theory and practice of international relations.
interaction processes in international conflict is represented by some recent "action One can distinguish three principal groups whose individual inputs may even-
research" programs, bringing together representatives of conflicting national groups tually influence outcomes on the international level: decision makers, national leaders,
for face-to-face communication in a controlled environment (Burton, 1969; Doob, and wider publics.
1970; see also Kelman, 1972). The procedure represents a special form of third- National decision-making elites and the elites of international organizations
party intervention (see Young, 1967), with social scientists acting as the inter- clearly enjoy some degree of influence on the choice of policy alternatives and thus
mediaries. I t provides unusual opportunities for observing the dynamics of conflict ultimately on outcomes in the national and international systems. The extent of direct
and exploring some of the conditions for its resolution (see Fisher, 1972). Finally, personal influence of any one member of these elites, however, would depend on
an entirely different approach to the study of international conflict is taken by Garn- such factors as his position in the hierarchy of his national government or inter-
son and Modigliani (1971) in their work on the Cold War. On the basis of a content national organization and the degree of latitude he is permitted, by virtue of his
analysis of newspaper accounts, they developed a chronology of the interactions position, in decision making within the particular area under his jurisdiction. Added
between the two parties to the Cold War for the period from 1946 to 1963, coded in to these factors are the implicit limits imposed on his choice of alternatives by the
terms of several descriptive dimensions. These data were used to test competing expectations of his fellow decision makers, of the larger organization for which he
theories of the Cold War, each of which yields different predictions about the way works, and of the public at large, to whom his organization is at least in theory
each side ought to be reacting to various actions by its antagonist. responsible.
Before concluding this brief overview of social-psychological research on inter- Second, one can generally identify a powerful element outside of governmental
national relations, we call attention to the multiplicity of methods that it employs. structures, capable of exerting substantial influence on the outcomes of nationally
The sources of data range all the way from laboratory experiments to national popu- and internationally relevant decisions. Its influence is in part due to its control of
lation statistics. Between these poles we find simulation studies, sample surveys, inter- such means as financial or editorial support that the decision maker needs to retain
view or questionnaire studies with selected population~,psychobiographies or per- his position and/or successfully execute his policies in the short or long run. Much
sonality assessments, group observations, organizational analyses, ethnographic field of its power, however, stems from its relationship to public opinion. Rosenau (1963,
studies, action research programs, and content analyses of historical documents or p. 6) refers to this source of influence when he speaks of national leaders as opinion
newspapers. This methodological diversity gives witness to the inherently interdisci- makers-that is, "those members of the society who occupy positions which enable
plinary character of this area of research. them to transmit, with some regularity, opinions about foreign policy issues to un-
known persons." By virtue of their positions, these leaders can impede or facilitate
Sources and Nature of IndividuaI Contributions achievement of consensus. They perform, in Rosenau's terms, a "veto-support"
function; decision makers are constrained by their opposition and turn to them for
The varieties of social-psychological research that we have outlined aid in the help in the mobilization of popular support.
study of international politics by focusing on individual actors and the interaction Public opinion, both domestic and international, represents a third source of
between them. We see this effort not as an alternative approach to international individual inputs into the international political system. The moods of the general
relations but as a contributory one. The goal, in our view, is to develop not a com- public and their broad orientations toward national and international affairs are an
plete and self-contained social-psychological theory of international relations but a essential part of the climate in which national policy makers operate. In part, these
general theory of political behavior at the system level in which social-psychological moods and orientations within the population exert a directive influence on the
factors play a part, once the points in the process at which they are applicable have process of policy formulation, in the sense that they impeI decision makers toward
been properly identified. To focus attention on the individual actor does not imply perceptions and actions that reflect public sentiments. Often decision makers are
that such an analysis can provide a total picture of international political processes. not only influenced by these pervasive moods but actually share them with the rest of
Certainly, every participant in both the national and international systems, at what- the population. In fact, these moods may originate in the very elites from which the
ever level he makes his contribution, is heavily influenced by the political, ecnomic, and decision makers are recruited and then spread among the rest of the population, so
social realities that surround him, and acts within the constraints and procedures of that it becomes difficult to specify who is influencing whom. Moods and orientations
institutional structures. Moreover, once an individual actor has developed a point within the population also exert a constraining influence on the policy-formulation
of view, made a decision, or taken a politically relevant action, he has contributed process. Even though the decision maker may have a great deal of latitude on any
only one, of the many and varied inputs that shape every political outcome. There is given foreign policy issue, he usually has to operate within certain broad limits set
no substitute for analysis of the structural mechanisms and systemic processes that by the population.
select among the myriad of individual inputs and mold them into final policy deci- & this regard, the existence of widespread political apathy and malleability
270 Herbert C. Kelman, Alfred H. Bloom Assumptive Frameworks in International Politics 271
in a population does not imply that public opinion is irrelevant to the policy process. acterize important decision makers and opinion leaders may have a direct impact on
Rather, a climate of apathy represents a special state of public opinion-a situation the decision process. (Hermann and Hermann, 1967, provide some suggestive evi-
in which the national leadership can assume that it is free to formulate policy within dence on this point in an interesting simulation of the outbreak of World War I.)
the confines of very broad limits set by public opinion without the risk of losing Dispositions that are widely shared within a society or at least among its major elites
popular support. President'Nixon's appeal to the "silent majority" in support of his may play an indirect, yet often pivotal, role in the process.
Vietnam policy is an interesting example of an attempt by a national leadership to The impact of the idiosyncratic dispositions of prominent decision makers
actually use a state of apathy (or of presumed apathy) as a resource in the political and opinion leaders depends on the relevance of a given personality factor to deci-
arena. Furthermore, indications are that even a silent majority, to whatever extent sion-making behavior; on the centrality of the particular decision maker's position
it does or can exist, will impose implicit limits on the acceptability of the policy in the decision-making structure; and on the nature of the decision involved-for
alternatives chosen by .their national leaders. The silent majority's support may have example, as Wolfers (1959) proposes, whether it is a decision in response to a
been specific to what they perceived to be the President's policy of disengagement in perceived opportunity rather than to a perceived threat to national survival. In
Vietnam and probably could not Ee transferred indiscriminately to whatever policy general, personality dispositions of important decision makers are probably less likely
he might choose. to have an effect on the overall direction of a decision than on its qualitative char-
As long as the decision maker adopts policy alternatives that lie within the acteristics, such as the style in which the decision is.communicated and carried out.
implicit limits acceptable to the relevant population, he may kncounter no difficulty These qualitative characteristics, in turn, may have a wider impact by affecting the
in mustering support for the chosen alternative or for his own reelection. On the way in which the decision issue is conceived by the public, since major political
other hand, if he violates certain pervasive popular assumptions governing the range leaders often serve as models in the formation and change of public attitudes.
of acceptable alternatives, he may be risking not only electoral defeat but also a Perhaps of greater importance to the foreign policy decision-making process
loss of the popular backing needed for the achievement of the domestic and foreign than the idiosyncratic dispositions of major decision makers and opinion leaders are
policy objectives of his party or nation (Rosenberg, 1965a). those personality dispositions, developed through similar socialization experiences,
Any one of these individual actors, whether he be a member of the foreign that are shared by a large segment of the society. Such shared dispositions are likely
policy decision-making elite, a national leader, or an informed citizen, responds to a to affect the overall approach to foreign policy of the decision-making apparatus.
wide variety of macrolevel factors whenever he forms an opinion or decides on a Furthermore, they have a bearing on the way in which the general population per-
course of action. Implicitly or explicitly, he will be influenced by ecological considera- ceives the society's needs and interests and interprets the intentions of other societies,
tions, such as the geographic position of his country and the basic resources that it and on its preference for certain means over others in the pursuit of societal goals
has available; economic considerations, such as the productive activities upon which and in the response to domestic and international challenges.
the economy is based, the general standard of living and level of technology char- Situational Pressures. A second set of social-psychological variables involves
acteristic of the society, and the current point in the business cycle; political con- the pressures that individuals feel at the point of politically relevant decision or
siderations, such as the nature of the regime, the stability of the government, and action, pressures deriving from the demands of the immediate or larger situation.
the extent of its involvement in international conflicts; and considerations of military Even such a seemingly insignificant factor as an overheated or crowded room may
strategy and international diplomacy, such as his country's involvement in power blocs, enter importantly into the final outcome of a negotiation by triggering a spiraling
military alliances, and regional organizations. These considerations-which are at process of misperception and misattribution. One negotiating team's irritability,
the heart of the traditional discipline of international relations-form the context in caused by the condition of the room, may be perceived as hostility by the other team
which opinions are formulated and decisions are made, but they do not explain how and lead to a reciprocally hostile reaction. As a result, a cooperative atmosphere-
the individuals involved perceive, interpret, or react to that context. The mediating though perhaps desired by both parties-may fail to materialize during the session,
variables that enter at the level of perception, interpretation, and reaction-that is, and both sides may find it difficult to assume a cooperative stance in future negotia-
at the level of individual functioning-may prove to have substantial impact on the tions. (For a discussion of situational effects on bargaining, see Sawyer and Guetzkow,
outcomes of interaction and at the same time provide potential leverage for irnple- 1965.)
menting change. In their simulation studies, cited earlier, Robinson, Hermann, and Hermann
Social-psychological investigation, as we have noted, has focused on the (1969) found evidence to suggest that in a crisis situation decision makers tend to
different types of variables that may affect individual functioning at the levels of abandon the search for new alternatives and instead direct their energies toward a
perception, interpretation, and reaction. Among others, these include personality search for new information about the alternatives that are already on the table. At
dispositions, situational pressures, and psychopolitical assumptions. the same time, crisis tends to.lower their confidence in whatever alternative they
Personality Dispositions. A wide variety of personality factors-authoritar- finally select. Janis (1971, 1972), in his analysis of several major foreign policy
ianism, xenophobia, extrapunitiveness, need achievement, and self-esteem being just a decisions, uses the term groupthink to describe the mode of thinking that dominates
few of many examples-are potentially relevant to the political process. (For reviews in a situation in which a cohesive group of decision makers, insulated from outside
of relevant studies, see Raser, 1966; Terhune, 1970.) The dispositions that char- influences, is engaged in a stressful decision process. In such a situation, loyalty
 Herbert C. Kelman, Alfred H. Bloom Assumptive Frameworks in International Politics 273
272

to the decision-making group tends to become the dominant consideration, and assumption making, a social-psychological approach, which takes the individual actor
as its primary unit of analysis, can illuminate the assumption-making process itself.
pressures toward conformity and consensus override critical thinking. Another source
of pressures is the presence of authorities or symbols of authority, which often dis- Furthermore, it pennits a more differentiated view of the actors (the foreign policy
courages individuals from personally appraising the issue: and pushes them in the public, the parties to an international conflict, or the operative units in the inter-
national system) that these assumptions tend to homogenize; it also permits a more
direction of adhering to the prescribed interpretations and role definitions.
. At the level of the general public, situational pressures may also play a major detailed examination of the interaction processes that these assumptions often ignore.
role in the formation and expression of opinions on foreign policy issues. An atmo- In the remainder of this chapter, we shall attempt a social-psychological
analysis of three types of assumptive frameworks in international politics, to which
sphere of national crisis, for example, tends to engender widespread reluctance to
engage in independent political thought and action. Such reluctance is often re- we have already alluded: assumptions concerning the role of public opinion in
foreign policy, the nature of international conflict, and the structure of the inter-
inforced by deliberate attempts on the part of political authorities-concerned with
maintaining an image of national unity-to impose a narrow definition of proper national system.
political attitudes and behavior and to equate dissent with subversion.
Psychopolitical Assumptions. Individuals-whether they are national decision Role of Public Opinion
makers, opiqion leaders, or involved citizensLbring to the political arena a complex
Most students of public opinion in the foreign policy process would probably
of underlying assumptions, conceptual frameworks in terms of which they formulate
agree with Etzioni's (1969, p. 576) formulation: "While public opinion does not
specific opinions and arrive at decisions on issues of international politics. Often these
participate effectively in many specific foreign policy decisions, it does serve as a
assumptions represent an unanalytic acceptance of the givens of national and inter-
context that sets significant limits on the maneuverability of the national decision
national systems and an undifferentiated view of the various actors in these systems.
makers." The decision makers' freedom of action, therefore, is less dependent on
One of the consequences of such assumptive frameworks is to impose unnecessary
the distribution of public attitudes toward the specific issue than it is on the broader
limits on the range of alternatives considered .in the formulation and execution of
context within which the issue is perceived and to which the specific attitudes are
policy. An analysis of assumptive frameworks, therefore, offers a unique opportunity
linked. As Etzioni (p. 577) points out:
for those students of international relations who are particularly concerned with re-
conceptualizing and changing the processes of foreign policy making and diplomacy.
We would expect the elites to have considerable freedom of action but
Many common assumptions in international relations,, if left unanalyzed and
only within the limits of established contexts, and only so long as their
unchallenged, present obstacles to peaceful resolutions of international conflicts specific actions do not disorganize these contexts or create a counterton-
and to creative responses to new global realities. Because they often assume that text. Thus Kennedy was relatively calm about the 1961 Cuban fiasco
public support for foreign policies depends on their conformity to certain static (and public opinion in his support even rose after the crisis), but he also
principles that are (presumably) widely held and deeply felt-such as the national realized in 1962 that another fiasco, this time in the missile crisis, might
interest, the national honor, national superiority, or military strength-decision associate him with an appeasement or "softyycontext. .. . When a new
makers may underestimate their freedom to pursue innovative, peace-oriented poli- event occurs for which there is no ready context, such as Soviet expan-
cies or overestimate their freedom to engage in militarist adventures, arms races, or sionism in ,1946-1947 or Soviet orbiting of a satellite in 1957, the national
expansionist policies. Because they often assume that international conflict consists decision makers seem relatively free to interpret it. But once a bit is
entirely of a series of competitive moves and countermoves, whose outcome depends placed in a context or a context is established, the public feeds it back
on the relative military strengths of the antagonists, decision makers may select .
to the leaders . . who, under most circumstances, are boxed in or are
ineffective and dangerous influence strategies, (and miss opportunities for conflict compelled to invest much effort to alter it.
resolution based on a more differentiated model of conflict. Because they often
Decision makers often seem to assess their freedom of action by reference to
assume that the nation-state is the paramount-indeed the only relevant-unit in the
the distribution of public attitudes on the specific issue under consideration, thus
international system, decision makers may arrange for the needs of their populations
underestimating or overestimating the freedom they actually enjoy. Many decision
(for exampl'e, in the areas of military security or economic development) in less than
makers, however, are quite aware that the limits within which they operate are
optimal ways and block the evolution of a more functional, ,peace-prone global
broad and that public reactions to a specific issue may vary with the context in which
system consisting of multiple, cross-cutting units. As long as we fail to subject these
this issue is presented. Yet even these decision makers tend to make some question-
assumptions to critical analysis and to question their validity and appropriateness,
able assumptions about the nature of the contexts within which the public judges
they may continue to have a pervasive and often deleterious effect on international
foreign policy actions. While they recognize that public views on specific issues may
politics.
change-and that they, as decision makers, can facilitate such changes-they usually
Social-psychological concepts and methods can make a valuable contribution
conceive the contexts to which specific attitudes are assimilated as largely static
to identifying these assumptions and to increasing our understanding of the processes
entities. These contexts, in their view, are formed by certain basic underlying prin-
by which they are developed and maintained. Since the individual is the locus of
274 Herbert C. Kelrnan, Alfred H. Bloom Assumptive Frameworks in International Politics 275
ciples to which the public expects foreign policy actions to conform; any deviations probed more deeply into the public's feelings about the war than the earlier polls
from these principles would be unacceptable. Specifically, they assume that the had done. The results indicated a more differentiated and in a sense more flexible
public expects its national leaders to be militant in protecting the national interest orientation toward the war across both elites and the mass public than would have
and upholding the national honor; to maintain a superior military, diplomatic, and been expected on the basis of previous poll data. Although 61 percent of the re-
economic posture; and to expand their political and ideological influence throughout spondents approved of President Johnson's handling of the situation in Vietnam,
the world. As we shall try to show, this assumptive framework is based on an in- the data, reminiscent of the Patchen (1964) finding3 on attitudes toward China,
complete analysis and an undifferentiated view of the way in which public opinion suggested that a good part of that 61 percent had answered affirmatively out of a
on foreign policy issues is formed. commitment to the President rather than a commitment to his specific policies. This
Static Model of Public Opinion. Decision makers who assume a static public would in turn imply that public opinion allowed the President much greater freedom
opinion, which judges foreign policy actions on the basis of their militancy, may well to alter his course of action in Vietnam than the previous poll data had presumably
misread the mood of the public to conform with their particular model. Thus, they indicated. This conclusion is further substantiated by data showing considerable
may overestimate the public's support for belligerent and expansionist moves and openness (as long ago as 1966) to various moves to negotiate a settlement of the
underestimate the public's readiness to accept more conciliatory policy alternatives. war: 88 percent of the American people were willing to negotiate with the Viet
Such misreading can easily occur because decision makers may be biased in selecting Cong; 54 percent agreed that free elections should be held, even if the Viet Cong
the pronouncements that they regard as truly representative of public opinion; might win; and 52 percent were willing to accept a coalition government including
because public reactions are generally formulated within a limited set of officially the Viet Cong (p. 320).
recognized alternatives; and because the meaning of these reactions and the depth The misreading of the public mood on specific foreign policy issues illustrates
of commitment they represent are rarely explored. one of the drawbacks of the static model of public opinion: It exaggerates the fixity
The standard public opinion polls often reinforce faulty readings of the public of the context within which the public is likely to judge a particular action, assuming
mood. For one thing, because of their methodological shortcomings, polls may create that people will support an action if it shows "strength" and reject it if it shows
a false impression of the stability and content of opinions on a specific issue. As "weakness." In reality, however, there is no one-to-one relationship between the
Rosenberg (1965a) points out, both the motivation to please the interviewer and apparent militancy of an action and the public's reaction to it. How the public
the motivation to appear to one's self as a politically involved citizen cause poll reacts depends on how the particular action is defined-in what context it is placed.
respondents to exaggerate the extent to which they actually hold opinions on the This is a dynamic process, in which the national leadership plays a decisive role. If
issues raised and the fixity of those opinions, and to bias the opinions they express we grant for the moment that decision makers are constrained by the broader con-
toward what they perceive to be the socially acceptable consensus. The perceived texts within which the public views foreign affairs, their ability to define and re-
consensus usually reflects the decision makers' own expectation of a preference for a jefine the context of specific issues often gives them greater flexibility than they
militant posture. Furthermore, opinion polls present respondents with a limited set realize or admit. To be sure, as our earlier quotation from Etzioni (1969) indicates,
of options, usually consisting of the alternatives formulated by the decision makers the context they set at one time may limit their freedom at a later time.
themselves. Thus, the polls generally describe opinions that are already "processed'' Most decision makers, of .course, do not take a passive stance toward public
to a considerable degree, in that they are expressed in terms of the options formulated opinion. They may argue, and actually believe, that they are constrained by public
by the national leadership and with an awareness of what the leadership considers opinion when these alleged constraints are in keeping with their own policy pref-
the proper response. Such processed opinions are hardly an independent source of erences. On the'other hand, when they perteive public opinion to be opposed to or
information about what the public expects and prefers. Certainly they do not indicate insufficiently supportive of their preferred policies, they do engage in efforts to
what alternative courses of action, not yet entertained, might be ,acceptable to public change it. These efforts themselves, however, usually conform to the static model of
opinion. A related and inajor limitation of the standard opinion poll is that it ascer- public opinion assumed by most decision makers. If public opinion does not provide
tains only the percentage who approve or disapprove of a particular policy action ;ufficiently strong support to certain militant foreign policy moves, decision makers
without exploring the meaning these opinions have for different respondents, the normally try to appeal to the set of static principles that they assume to be im-
context within which they are held, and the depth of commitment they represent. portant to the public: the need to protect a vaguely defined national interest, to
Thus, if a poll shows, for example, a high rate of approval of a given policy, the uphold the national honor, to maintain national superiority, or to expand national
decision maker is provided no perspective for gauging the significance of this finding. influence. If they are worried about public support for certain conciliatory foreign
If it conforms to his assumptive framework, he is likely to misread it as representing policy moves, they try to redefine the context within which these policies are placed
stronger public support for the policy and less willingness to consider alternatives -showing that these are, indeed, consistent with the national interest, the national
than it does in fact. honor, or the nation's global power position. In either event, decision makers are
The misreading of the public mood concerning the war in Vietnam during working within an assumptive framework that exaggerates the fixity of the principles
the Johnson administration provides a case in point. In their survey of American that serve as the context for public reactions to foreign policy actions.
attitudes toward the war, conducted in 1966, Verba and his associates (1967) The assumption of a static public opinion, favoring militant postures, ignores
 Herbert C. Kelman, Alfred H. Bloom Assumptive Frameworks in International Politics 277

two major considerations: ( 1) Public adherence to doctrines of national interest or relationship of the citizenry to the' national system may also lead to a neglect of some
superiority is not universal or stable and certainly not inevitable. These principles potentially effkctive means of mobilizing support for foreign policy innovations.
derive from the nationalist ideology that characterizes the modern nation-state. This Determinants of Foreign Policy Orientations. To analyze public opinion in
ideology is fostered by the national leadership and, in established nation-states such foreign policy as a dynamic process, one must explore the ways in which different
as the United States, is widely accepted in the population-albeit with considerable individuals and groups within a population relate themselves to broader foreign
variations in intensity and precise content. Decision makers assume that "the national policy issues. These orientations, in turn, have implications for the possibility of
interest" and national superiority represent important concerns to the public, in mobilizing their support for various foreign policy actions-whether on the basis
of appeals to the national interest and superiority or on the basis of other appeals-
large part because they are important to the decision makers themselves and because
they offer them a relatively flexible basis for regulating public responses to foreign and for the conditions under which such mobilization can take place. A first step
policy issues. Very often these concerns are indeed important to major sectors of the toward a more dynamic analysis of public opinion is to identify and account for
different foreign policy orientations. How do different orientations fit into and vary
public-to some because of their own ideological commitments, to most because they
are told by their national leaders that they are important. The extent to which with an individual's general personality dispositions, his position in society, and-
people attach importance to these principles, however, and the depth of their com- perhaps most important-his relationship to the political system?
An illustration of the link between foreign policy attitudes and general per-
mitment to them vary widely for different groups and at different points in time.
sonality dispositions is provided by McClosky's research (1967a). McClosky sampled
(2) The formation of public attitudes toward a specific issue is a dynamic process,
over 5000 individuals in an effort to arrive at a more differentiated understanding
in which the interaction between individual citizens and decision makers plays a
of isolationist orientations toward United States involvement in world affairs. He
major role. The shape of public opinion emerging out of this interaction depends
found isolationism to be part of a cluster of attitudes, including.classica1and welfare
very heavily on the nature of the relationship of various segments of the population
conservatism, radical doctrines of the right and left, and attitudes critical of demo-
to the national system and to its current leadership. If the public supports militant
. cratic beliefs and practices. This cluster of attitudes, in turn, is linked to a character-
actions promoted by the leadership, it may do so more because of its underlying
istic cognitive style and underlying personality dispositions.
trust in the leadership than because of a c~mmitmentto principles of national
militancy. In the absence of trust, the leadership might find it difficult to mobilize
Despite its strong chauvinistic overtones, isolationism is frequently asso-
support for militant actions; given a basis of trust, they might be equally successful
ciated with feelings of disappointment in one's own society and disaffec-
in mobilizing support for conciliatory actions. tion from the political institutions of one's country. The isolationist
The static model may be a good predictor of public reactions under many, orientation parallels closely other forms of belief that rely heavily upon
perhaps most, circumstances. Decision makers often succeed in mobilizing public dichotomous thought processes, that lack breadth of perspective, and that
support by appealing to the principles-of national interest, honor, and superiority, seek to exclude whatever is different, distant, or unfamiliar. . . . Like
and they may maintain themselves in power by affecting a militant foreign policy other deviant orientations, it signifies for some of its proponents a failure
posture. The model cannot, however, account for those occasions in which anticipated of socialization and an inadequate internalization of the norms. I t is more
public support for a militant policy breaks down (as in President Johnson's Vietnam common among those who are, by any criterion and for any reason, paro-
policy) or earlier anticipated public opposition to a conciliatory policy fails to mate- chial and less common among those who are open to experience and cos-
rialize (as in President Nixon's China policy). It cannot account for the successful mopolitan in their perspective [McClosky, 1967a, p. 1071.
efforts by citizen groups to mobilize American public opinion against fallout shelters
and in favor of the test-ban treaty (Etzioni, 1969, p. 578). Above all, it limits the Isolationists and nonisolationists respond very differently to a variety of specific foreign
exploration of potential innovative approaches to foreign policy, which might provide policy issues, such as immigration, foreign aid, level of tariffs, participation in NATO,
alternatives to existing policies. and defense spending. Though these specific opinions may shift in the face of differ-
At the policy level, as we have already pointed out, the static model intro- ent pressures, the general pattern of findings suggests quite clearly -that individuals and
duces a systematic bias toward overestimating existing and potential public support groups differing along the isolationism-nonisolationism dimension (and the broad
for militant actions and underestimating support for conciliatory ones. It may thus cluster of attitudes and personality dispositions that it represents) are likely to differ
encourage decision makers to persist in policies that the public will not back up and in their ~eadinessto support various types of foreign policy actions.
Starting from a more sociological perspective, Galtung (1969) attempted to
to reject prematurely alternatives that the public might well have supported. More-
over, insofar as the static model ignores the role of trust in the leadership as a central link the cognitive processes underlying an individual's foreign policy orientation to
determinant of public support, assigning greater weight to the appearance of mili- his social position. He distinguished between a gradualist approach to foreign policy
(characterized by a stable, inductive, pragmatic, means-end intellectual style; a
tallcy, decision makers may be inclined to risk their credibility in the pursuit of
militant policies and thus forfeit long-run support (which is apparently what hap- differentiated' acceptance and rejection of the status quo; and a gradual orientation
pened to President Johnson in 1968). Failure to give proper weight to the ongoing to change) and an absolutist approach (characterized by a volatile, deductive,
278 Herbert C. Kelman, AIfred H. Bloom Assurnptive Frameworks in International Politics 279
moralistic, subsumptive intellectual style; a total acceptance or rejection of the Table 1
status quo; and an all-or-nothing orientation toward change). Individuals who oper-
ate at the center of the system enjoy, by virtue of their position, easy access to
important communication channels on foreign affairs. They are thus able to gain a
more differentiated picture of the international situation and the practical alterna- I Manner of Integration into the System
I
tives available to their nation; and they tend to develop a sense of the utility of I Role Participant Normative
I Ideological
pragmatic action in the field of foreign policy making. It is these individuals who,
according to Galtung, assume a gradualist approach to the foreign policy process. By Commitment to Commitment to. Acceptance of
contrast, individuals at the periphery of the system remain cut off from the imme- cultural values the role of na- demands
reflective of na- tional and its based on
diate flow of information and are denied the opportunity for personal experience
Sentimental tional identity associated sym- commitment
with the world of international relations. The result is that a moralistic ideological bols to the sacred-
thinking style combines with secondhand informational inputs (see Katz and Lazars- Source of ness of the
feld, 1955) to produce what Galtung has labeled the absolutist orientation toward Attachment state .
the foreign policy process. The two groups differ markedly and predictably in the (Loyalty) to Commitment to Commitment to Acceptance of
types of foreign policy actions they are inclined to accept or reject (Galtung, 1967). the System institutions pro- social roles me- demands
Personality dispositions and social position take on special importance because motive of the diated by the based on com-
they affect the individual citizen's overall relationship to the political system. The Instrumental needs and in- system mitment to
terests of the law and order
nature of this relationship, as we proposed earlier, greatly influences the conditions
population (principle of
under which the individual will provide or withhold support to various foreign policy equity1
moves, the kind of support he will give, and the kinds of appeals from decision
makers to which he will be responsive. Thus, the patterns of relationship to the Source: Adapted from Kelman (1969),p. 280.
system and its current leadership that characterize different segments of the popula-
tion are a crucial determinant of who can be mobilized (and when and how) to viduals can participate to their mutual benefit and have some assurance that their
provide what kind of support for various policy options. needs and interests will be met. Each type of attachment may be channeled in
In earlier work (KeIman, 1969), we presented a framework distinguishing three different ways, depending on the manner in which the individual is integrated
different patterns of personal involvement in the national political system. This into the system (as shown in the columns of Table 1).
framework illustrates one approach to the determinants of foreign policy attitudes An individual who is ideologically integrated (first column in Table 1) is
in terms of the interaction between individual citizens and the national leadership. bound to the system by virtue of sharing some of the cultural and/or social values
Such an analysis should be particularly helpful in illuminating major changes in on which the system is based. He supports the system because-and to the extent
public mood vis-A-vis foreign policy issues and in assessing the potential of alterna- that-he sees it as consistent with these values, which he has internalized and in-
tive policy choices and of new approaches to the foreign policy process. corporated into a personal value framework. This value framework generally defines
Model of Personal Involvement in National System. As summarized in . for h i the range of political alternatives he is willing to accept and provides the
Table 1, the model distinguishes six patterns of personal involvement in the national basis upon which he evaluates and decides in the political arena. Ideological inte-
system, The rows represent two sources of attachment or loyalty to the system- gration, then, manifests itself in a more or less rational and abstract evaluation of
sentimental and instrumental. The columns represent three means of integration of policies and institutional arrangements-an evaluation based on their effectiveness
the individual into the system-ideological, role participant, and normative. In in expressing the identity of the population and in providing for its needs and
other words, the rows distinguish, essentially, two types of motives that lead the interests. The cognitive processes involved in such evaluations are probably quite
individual to cathect the system. The columns, on the other hand, distinguish three similar, regardless of whether the ideological integration is sentimentally or instru-
components of the system via which members may be bound into it. mentally based. The vaIues against which policies are judged are more specifically
An individual is sentimentally attached to the system when he sees it as ethnic-cultural values in the former case and more universal values in the latter case,
representing him-as being, in some central way, a reflection and extension of him- but these may often be fused among ideologically integrated individuals. (It should be
self. For the sentimentally attached, the system is legitimate and deserving of his noted, in general, that sentimental and instrumental attachments are not mutually
loyalty because it is the embodiment of a group in which his personal identity is exclusive, although in some individuals one or the other may predominate.)
anchored. An individual is instrumentally attached to the system when he sees it as In one sense, ideological integration implies a highly conditional form of
an effective vehicle for achieving his own ends and those of other system members. support for the system. The ideologically integrated individual expects the system
For the instrumentally attached, the system is legitimate and deserving of his loyalty not only to conform to a set of basic values in its institutional structure and its
because it provides the organization for a smoothly running society in which indi- approach to policy issues but also to evolve new institutions and approaches as these
 Herbert C. Kelman, Alfred H. Bloom Assumptive Frameworks in International Politics 281
i
values confront changing circumstances. For example, an ideologically integrated adopt a conception of the "proper" foreign policy stance that the elite was propagat-
i
1 American may feel that a laissez-faire conception of the role of government may ing earlier just when the elite itself has abandoned this conception in favor of a
i have been quite consistent with the society's basic values in the 1880s but that-in new approach. Under these circumstances, rather than deviate from what he still
I light of changes in economic conditions, social needs, or government resources- considers the normatively required posture, he may question the legitimacy of those
such a conception is inappropriate for the 1970s. Thus, to assure the continuing who advocate the new approach.
support of the ideologically integrated individual, the leadership must provide for Normatively integrated individuals may also withdraw their support when
a periodic renewal of the system's commitment to its underlying values. they feel that the authorities have failed to fulfill their part of the bargain on which
In another sense, however, ideological integration can be seen as the most the individual's relationship to the system is founded, and that the basis of their
stable basis of support to the national system. Since commitment in this case rests integration is therefore threatened. A recent study of public reactions to the trial
on basic values rather than on the sanctification of specific norms, procedures, or of Lieutenant Galley (Kelman and Lawrence, 1972) suggests that normatively
symbols, the individual is prepared to support novel procedural and institutional integrated individuals may have been particularly indignant about the Calley con-
alternatives as long as he perceives them to be in line with system values. Moreover, viction because they considered it unfair to hold a man personally responsible for
he is likely to extend trust to the regime during transfers of power, as long as the actions taken in obedience to orders from legitimate authorities. They accept the
incoming administration is (in his eyes) committed to maximizing the system's values, obligation to obey orders without question, but, in return, they expect the authorities
regardless of its charismatic appeal or its relationship to a previous charismatic leader. to take full responsibility for the consequences of the actions they ordered.
Ideologically integrated individuals may thus play an especially important role in Role-participant integration (second column of Table 1) falls somewhere
what Weber refers to as the institutionalization of charisma-the attempt by national between the ideological and the normative orientations. Individuals integrated in
leaders to channel, upon the death of a charismatic leader, the loyalties that had this manner resemble the normatively integrated in that they do not measure system
developed toward the man into loyalties to a continuing national system. performance against a set of basic values. They are closer to the ideologically inte-
At the other extreme, the normatively integrated individual (third column grated, however, in their sense of "ownership" of the system. They are bound into
in Table 1) does not take it upon himself to evaluate policies and institutional the system by virtue of their personal identification with roles within the system.
arrangements in terms of their consistency with a set of basic values. He is bound to These roles enter significantly into their self-definition, and they thus have a personal
the system by virtue of his unquestioning adherence to its rules. He accepts the stake in maintaining them and in living up to their requirements. They are inclined
system's right to prescribe the behavior of its members within a specified domain to give ready support to the system provided the relevant role is made salient in the
and has learned, through his socialization and life experiences, to regard compliance situation in which that support is elicited. The supportive behavior is essentially
with the system as a highly proper and valued orientation. If his attachment to the a way of meeting the expectations of a personally significant role-the role of good
system is primarily sentimental, his acceptance of the system's right to unquestioning national citizen or some subsystem (occupational, organizational, community) role
obedience is most likely based on a commitment to the state as a sacred object in its mediated by the national system. In meeting role expectations, the individual re-
own right. If his attachment is primarily instrumental, his obedience most likely re- affirms his links to the national system and confirms the self-definition anchored in
flects a concern with preserving law and order as a necessary condition for the effi- that relationship. An individual integrated via role participation is likely to maintain
cient and equitable functioning of society. his allegiance unless he feels that the regime is systematically undermining the in-
The normatively integrated individual usually obeys demands from legitimate tegrity of the roles to which he is committed.
authorities automatically, without analyzing their value implications. His readiness Individuals integrated via role participation are likely to react differently to
to comply may, however, depend on the extent to which these demands are authori- certain specific policies depending on the source of their attachment to the system.
tatively presented as the wishes of the leadership or the requirements of law. One Sentimentally based role participation involves a strong emotional identification
primary indicator of the authoritativeness of a particular demand would be the with the role of national and its associated symbols; the individual sees the national
existence of clearly visible and specified sanctions to ensure proper behavior. system as the collective expression of his personal identity. Under the circumstances,
Ironi'cally, the very readiness of the normatively integrated individual to he tends to regard that system as the exclusive protector of his needs and interests
comply with system demands may at times impose serious constraints on the flexibility and to conceive international conflict in competitive, zero-sum terms. He is thus
of decision making within the national system. He is inclined to accept without suspicious of any arrangements that might imply an erosion of the autonomy of the
reservation the authorities' definition of what is a proper and an improper nationaI national system (for instance, by relinquishing control to a supranational organiza-
posture, particularly in the foreign policy domain. Having accepted this definition, tion), and he favors a militant posture in relations with other nations. By contrast,
however, he is likely to resist basic changes in national policy, since such changes instrumentally based role participation involves a personal commitment to various
would disrupt his sense of propriety and order and threaten his integration in the social roles whose effective performance depends on the national system; the indi-
system. Since, as Galtung (1969) has pointed out, there is generally a lag in com- vidual sees the national system as a useful framework within which to pursue his
munication of basic foreign policy orientations from the center to the periphery, the economic and other personally significant interests. He is thus quite willing to support
normatively oriented individual (who typically resides at the periphery) may finally policies that improve the system's effectiveness in meeting his needs and interests, even
 282 Herbert C. Kelman, Alfred H. Bloom Assumptive Frameworks in International Politics 283
I
I
if these policies entail some reductions in national autonomy or some concessions to mitted, were better educated, had higher incomes, were more often in professional or
I
I other nations. technical occupations, had fathers who were better educated and more often in
Some empirical findings relevant to the distinction between sentimentally and
I' instrumentally based role participation can be gleaned from the study by DeLamater,
Katz, and Kelman (1969) of national role involvement in an American community.
professional or technical occupations, were more likely to describe themselves as
"middle class" (rather than "working class"), and were more bureaucratically
oriented. The symbolically committed were more likely to have grown up in farming
This study distinguishes three types of commitment to the national system-symbolic, families, to have lived on a farm for ten years or more, and to have lived in the area
I functional, and normative-two of which bear on the distinction made in the present for ten years or more. The two groups also differed on a variety of attitudinal items.
1 I scheme: Symbolic commitment corresponds quite closely to sentimentally based role The functionally committed showed greater openness to other cultures and systems,
'I participation, and functional commitment corresponds to instrumentally based role
participation. Data were obtained through intensive interviews, consisting of a variety
greater tolerance for deviant political positions, and greater support for liberal
causes. The symbolically committed tended to favor a more militant stance in Arneri-
I
of questions about the respondent's conception of his national role, as well as a series can foreign policy and were much less willing than the functionally committed to
of attitudinal and demographic items. A scale measuring each of the three types of turn over power to international organizations.
j commitment was constructed out of items that were theoretically relevant to that
orientation and that also seemed to hang together statistically. The scale for symbolic
These findings are nothing more than first approximations, since the study

1 commitment included eight items that tapped the respondent's emotional involve-
ment with national symbols-his personal attachment to these symbols and his sensi-
was not designed to test the present model. Yet they are generally consistent with
our view of the differences between instrumentally and sentimentally based role
participation and thus suggest the usefulness of further empirical investigation of the
1 tivity to any indication that they are being slighted. A respondent would receive
positive points on this scale if he indicated that (1) anyone who criticizes the govern-
model with the use of indices specifically tailored to its parameters.
The type of model of personal involvement in the national system that we
ment in time of national crisis is not a good American, ( 2 ) anyone who does not have outlined should make it possible to specify the nature of the policy moves that
stand during the playing of the national anthem is not a good American, ( 3 ) he individuals and groups of each modal type can be expected to favor or reject, and
owns an American flag and displays it on national holidays, (4) the American public the most likely conditions and means for mobilizing their support. The support of
I pays insufficient respect to the flag, (5) he disapproves of Americans who take no the ideologically integrated individual for a specific foreign policy alternative should
I pride in America's armed forces, (6) he would consider it an insult if a foreigner depend on the value implications of that alternative, defined by his vision of the

I laughed at the Peace Corps, (7) he would be insulted or angry if a foreigner criti-
cized racial segregation in the United States and attacked the free enterprise system,
(8) he is "first, last, and always an American." According to the present conceptual
scheme, a high score on this scale would indicate a strong sentimentally based attach-
ment to the national system, channeled through identification with the national role
appropriate national goals within the international system and his assessment of the
policy's probable consequences. In the long run, ideologically integrated individuals
provide the firmest support for the system, but at any given point they may call for
a reappraisal of national priorities in terms of certain basic cultural and social values.
In the short run, the most reliable support comes from the normatively integrated
and triggered by the presentation of national symbols. individual. He is likely to accept any foreign policy move without question, as long
The scale for functional commitment included six items that tapped the as he perceives it as reinforcing existing authority patterns. His support is most
respondent's orientation to the economic benefits of American society and his empha- readily mobilized by appeals that link demands for action to legitimate sources of
sis on citizen participation. A respondent would receive positive points on this scale authority. Finally, role-participant integration provides support that is more active
if he indicated that (1) the things that particularly remind him of being an American and enthusiastic than that based on normative integration but less selective and

I
include factors relating to opportunity, (2) to be a good American a person ought conditional than that based on ideological integration. The support of an individual
to participate in public affairs, ( 3 ) apathetic persons are among those whom he who is integrated via sentimentally based role participation can be mobilized by the
regards as "not good Americans," (4) people refer to affluence and related matters introduction of national symbols, which heighten emotional arousal, and by implied
I when they talk about "the American way of life," (5) one of the most important threats to national sovereignty. Support of an individual integrated via instrumentally
things that makes America different from other countries is its level of opportunity, based role participation can be mobilized by co-opting his subsystem roles in the
(6) one of the most important things that makes America different from other of the national system. He should be responsive to appeals directed at his
countries is its level of affluence. In terms of the present conceptual scheme, a high economic and other subgroup interests and ready to accept a wide variety of policies
score on this scale would imply a strong instrumentally based attachment to the (militant or conciliatory, supportive of or antagonistic to international organizations)
national system, channeled through the person's entanglement in various social roles that appear to promote these interests.
that depend on the effective functioning of that system. From the point of view of national leaders, each type of integration represents
Comparisons were made between respondents with high scores on one of the '
a unique set of resources for mobilizing public support. They can draw on ideological
three types of commitment and low or medium scores on the other two. The "high commitments to support major policy reorientations and novel institutional arrange-
symbolic" group and the "high functional" group turned out to differ on many ments; on role-participant commitments to elicit special efforts and sacrifices in
dimensions. The functionally committed were younger than the symbolically com- periods of national crisis or of rapid economic development and social change; and
 284 Herbert C. Kelman, Alfred H. Bloom Assumptive Frameworks in International Politics 285

on normative commitments to assure the smooth operation of the system in periods the only appropriate way to settle the conflict is by physically stopping him or deter-
of relative calm and to reassert their traditional authority in periods of social unrest. ring him, by whatever means available, from further aggression. The possibility that
Considerable research is needed to clarify and refine the different patterns the types of interactions generated by the conflict may themselves be contributing to
of personal involvement in the national system distinguished by the present model; to its escalation and perpetuation is rarely if ever considered in the planning of succeed-
identify the foreign policy orientations characteristic of each pattern; and to explore ing policy moves. Rather, each side assumes that it must concentrate its efforts on
the role that these differentorientations, in turn, play in the articulation, mobilization, projecting an image of strength, backed by a visible display of military power, thus
I
and impact of public opinion on foreign policy issues. As research based on this and lending credibility to its threats and deterring its adversary from continuing in his
cc aggression."
related efforts to probe the dynamics of public opinion in the foreign policy process I
accumulates, we will be in a better position to predict the degrees of freedom of More generally, the aggressor-defender mode1 promotes the view that the only
decision makers on various foreign policy issues, the most probable sources of support viable means of guaranteeing international peace is the maintenance, on the part of
for and resistance to different actions, and the repercussions these actions are likely each major power, of a military force of sufficient strength and preparedness to deter
to have among different segments of the population. It should also become possible I any would-be aggressor. A substantial amount of effort in recent years has focused
to specify the kinds of long-term efforts at public education that would help to build on testing the validity and examining the implications of this view. (For more com-
a public opinion more favorable to innovative, constructive approaches to foreign prehensive discussions of the psychology of deterrence, see Deutsch and Krauss,
policy. I 1960, and various of the chapters in Pruitt and Snyder, 1969, and in Swingle, 1970b.)
Schelling and Halperin (1969) call attention to the irony of a deterrence strategy
Nature of International ConBict in a world whose weaponry is capable of overwhelming destruction. The existence of
offensive weapons with the power to destroy the retaliatory capacity of another
Like the common model of the role of public opinion, the assumptive framework nation-state puts a marked premium on being the first to attack or, as Schelling and
with which decision makers and the public commonly approach major international Halperin put it, at least "a close second." Thus, if a nation even remotely suspects
conflicts often.stands in the way of cooperative modes of conflict resolution. Conflict that another nation is building up its capacity for a possible attack, it will feel pres-
as a process tends to be viewed entirely as competition between two rational parties, sured into at Ieast contemplating a preemptive act of aggression in order not to lose
the one seeking justifiably to protect its interests and the other engaged in a "diaboli- the distinct advantage to be gained by attacking first. Similarly, any hesitation in
cal" effort (White, 1968) to prevent the first from realizing its goals. Pruitt and 1 reacting to what is thought to be an indication of aggression may involve an enor-
Gahagan (1972) have labeled conceptual approaches of this type as the "aggressor- I mous loss of comphrative advantage. The net result is that the use of the deterrence
defender models" of conflict. strategy may very well increase the chances of preemptive or accidental war rather
Aggressor-Defender Model. According to this model, "one party (the I than the chances of a reduction in international conflict.
'aggressor') is assumed to be the originator of the conflict and the major contributor In light of this drawback of a mutual-deterrence strategy, Raser and Crow
to its continuation. His behavior is typically explained in terms of the motives and (1969) suggest that one way of restraining international conflict might be a retalia-
emotions satisfied by the conflict and the failure of deterrents against aggression" tory force that is not vulnerable to initial attack by other nation-states. In order to
(Pruitt and Gahagan, 1972, p. 19). test this hypothesis, they employed multiple runs of the Inter-Nation Simulation, in
Ralph White's (1968) analysis of misperceptions in Vietnam suggests that which one nation, "Omne," was provided with the capacity to withstand any attack
the war was continued at least partly because both sides operated within this un- and to retaliate decisively. The results indicate that providing Omne with this. retal-
differentiated assumptive framework. Decision makers on each side tended to project I iatory capacity did indeed reduce the incidence of preemptive war, but at the same
an exaggeratedly diabolical, aggressive image of the other while at the same time time it encouraged Omne to assume a more belligerent and aggressive stance, thus
maintaining, by means of rationalizations, an exaggeratedly self-righteous perception I causing a complementary increase in the occurrence of strategic war. Although
of their own conduct. According to White's analysis, each side affected a virile na- generalizations from a simulation study to the real world require great caution, the
tional self-image, disguising its own desires for power, prestige, and self-aggrandize- study points to at least one possible unintended psychological consequence of a deter-
ment, and subscribed to a distorted view of the situation in Vietnam and of its own rence strategy, even under the conditions of an invulnerable retaliatory force. Swingle
military capacity. The resulting black-and-white conception of the conflict precluded (1970a) suggests two additional unintended reactions to deterrence, with potentially
the formation of more differentiated images of one's own and of the adversary nation; hazardous consequences. If the capacity to punish, especially in the case of bilateral
obscured areas of common interest, in which cooperation would be feasible; ,and
1 punishment, is extremely strong, "an opponent may not believe a threat to use such
inhibited the deveIopment of mutual trust. Within this context of lessened sensitivity, punishment" because "it is not in the interests of the threatener to execute the threat
l
it was extremely difficult to arrive at a peaceful and lasting solution to the conflict. .
at any time. This, in turn, gives rise to a policy of encroachment in which each
Since, moreover, according to the aggressor-defender model, the outbreak and per- I
I infraction is not large enough to justify the execution of the threat." Furthermore,
petuation of conflict are entirely due to aggression on the part of one's adversary, "severe thrkat tends to put a premium upon antagonists acting as though they are
 Herbert C. Kelman, Alfred H. Bloom Assumptive Frameworks in International PoIitics

irrational, since the appearance of irrationality increases the credibility that the the conflict spiral: Conflict is escalated as each party's aggressive moves are recipro-
threat might actually be executed" (p. 248). cated by the other; Osgood proposes to deescalate it through the initiation and
Canpict-Spiral Model. One of the consequences of the aggressor-defender reciprocation of conciliatory moves. In short, the strategy addresses itself to the
model is a conception of international conflict as a zero-sum game, in which one dynamics of conflictual interaction, which are often responsible for expansion of the
side's gains are matched by the other side's losses. Various social science writings have conflict and for the difficulty in discovering cooperative solutions.
challenged this assumption, reconceptualizing international conflict as at least poten- Osgood's analysis has received some support from laboratory studies. In the
tially a positive-sum game, permitting outcomes that leave both sides better off, Basic context of the Inter-Nation Simulation, Crow (1963) demonstrated the success of
conflicts of interest between the parties may, of course, exist; but even when they do, a GRIT strategy both in reversing a trend toward escalation of tensions and in
there are often some bases for cooperation that would accrue to the mutual benefit of inducing an initially hostile and hesitant nation to reciprocate the conciliatory ges-
the parties. The question is: Why do parties in conflict usually find it so difficult to tures of the initiator of GRIT. Pilisuk and Skolnick (1968), using an extended ver-
discover these positive-sum solutions? sion of the standard Prisoner's Dilemma task (see Komorita, 1965; Rapoport and
To answer this question, we must analyze the dynamics of the conflict process Chammah, 1965), found that the combination of a conciliatory strategy with honest
itself: What is it about the conflict situation that blocks the discovery of mutual inter- prior announcement of moves can be effective in inducing reciprocal cooperation.
ests, and by what means could these be made to surface? The aggressor-defender model The effectiveness of the GRIT strategy in these two rather different laboratory
is not a t all useful here. Not only does it obscure the possibjlities of cooperation situations gives added weight to Osgood's proposals, although one must, of course,
between the conflicting parties, but-since it analyzes conflict entirely in terms of the keep in mind the limitations of generalizing from the laboratory to the international
motives of the antagonists-it ignores the ways in which conflictual interactions system.
themselves may reinforce and enlarge the conflict. To correct for these shortcomings We concur with Pruitt and Gahagan's (1972) assessment that the conflict-
of aggressor-defender models, social scientists have developed models focusing on spiral approach represents a definite improvement over the more commonly held
the dynamics of the interaction between the conflicting parties. Pruitt and Gahagan assumptive frameworks about the nature of conflict. As a dynamic and two-sided
(1972, p. 23) have labeled these the "conflict-spiral models'' : "The basic assumption model, it corrects for some of the major limitations of the various aggressor-defender
of these models is that conflict develops and is perpetuated through vicious circles
models. At the same time, it has .some limitations of its own. Pruitt and Gahagan
in which each party's conflictual action is a reaction to the other party's recent
point out, for example, that it tends to be oversymmetrical in its treatment of the two
behavior, This reaction may be punitive or defensive; but regardless of which, the
conflicting parties; that it treats the parties as undifferentiated units, ignoring the
other party then reacts with more conflict behavior, continuing the circle." Percep-
composition of each party and the possibility that this may change over time; and
tions of one's own nation and of the other nation play an important role here. A
that it does not account adequately for the observation that conflict, once begun, often
may arm for what it considers legitimate reasons of defense; B, convinced of its own
perpetuates itself even in the absence of immediate provocation. As a result, this type
peacefulness, perceives A's action not as a defensive move but as evidence of A's
of model seems "overly optimistic about the possibility of reversing conflict processes
aggressive intent; B therefore arms in defense against possible aggression from A; A,
once they get started" (Pruitt and Gahagan, p. 27). Another limitation of the con-
in turn, perceives B's action as further evidence of B's aggressive intent and therefore
flict-spiral model is that it views the conflictual interaction entirely in terms of alter-
proceeds to increase its own level of armament; and so on. Conflict-spiral models have
been central to a number of recent analyses of international conflicts and approaches nating actions and reactions-moves and countermoves-by two independent parties,
to conflict resolution. (For fuller discussions, see Richardson, 1960; North, Brody, ignoring the possibility that in some respects the two parties (or at least segments of
and Holsti, 1964; Pruitt, 1965; Shure, Meeker, and Hansford, 1965; Osgood, 1962, each) can be viewed as a single system, with a shared definition of the situation.
1969.) As a result, this type of model may overlook conflict-resolution approaches that rely
Perhaps the best-known proposal for conflict resolution based on a conflict- on a redefinition of the relationship between the parties and of the nature of their
spiral model is Osgood's GRIT strategy (Graduated and Reciprocated Initiatives in conflict.
Tension Reduction). The strategy starts with the assumption that solutions accept- Structural-Change Model. A family of approaches that correct for some of
able to both sides can best be discovered in an atmosphere conducive to the develop- the limitations of the conflict-spiral model is described by Pruitt and Gahagan (1972)
ment of open channels of con~munication,mutual trust, and the mutual extension of as "structural-change models." These models assume that "certain changes take place
empathy. Osgood proposes the use of unilateral conciliatory steps, preceded by explicit in one or both parties that tend to perpetuate the conflict well beyond the initial

r
announcements of intent, as a means f fostering the desired atmosphere and thus
triggering a gradual deescalation of ten ions and ultimately a deescalation of military
investments. Although these initiatives entail some sacrifice for the nation that takes
motives or escalative sequence that began it, These changes may occur in the social
structure of the parties; in images, attitudes, motives or feelings; or in the salience of
issues or commitment to their solution" (p. 28). Thus, for example, the conflict may
them, they do not involve a significant impairment of its military capacity. Initial . create or bring to the fore organizational units, social movements, or population
steps are small, but they are gradually expanded as the other side reciprocates with groupings that have a vested interest in the conflict itself and thus contribute to.its
tension-reducing actions of its own. Essentially, this strategy is an attempt to reverse perpetuation. Similarly, the conflict may engender new goals, or sharpen negative
288 Herbert C. Kelrnan, Alfred H. Bloom Assumptive Frameworks in International Politics 289
images of the adversary, or bring new issues into focus, which give it a life of its own, content of the interactions constituting resolution efforts need to be designed in ways
independent of the initial precipitating incident. that allow such structural changes to occur within each side. That is, they need to
Structural-change models are not necessarily inconsistent with conflict-spiral provide inputs that generate new attitudes toward the other side, new images of
models. Often the processes highlighted by these two approaches work simultaneously possible relationships between the conflicting parties, and new goals that depend on
and in conjunction with each other (Pruitt and Gahagan, 1972, pp. 35-36). For resolution of the conflict; and they need to foster the emergence of organizational
example, at various points in a conflict-spiral process, B's escalatory response to A's units with investment in conflict resolution. Of even greater potential significance,
provocation may be accompanied by structural changes in B: a particular faction perhaps, is the promotion of structural changes in the relationship between the two
within B may gain prominence as leader of the struggle; new attitudes toward A may conflicting parties. The dynamics of conflict often create a symbiotic relationship
become crystallized. These structural changes may have a multiplier effect on the between the conflicting parties (or Eetween certain subunits of them), whereby both
conflict, increasing its intensity and duration to levels beyond those produced by the sides join in a shared definition of the conflict as intractable and in a shared commit-
spiraling processes themselves. ment to its perpetuation. It may be possible to direct conflict-resolution efforts toward
What implications does a structural-change model have for conflict resolution? the creation of a new joint system, involving both parties in a shared definition of the
I n stressing the self-perpetuating nature of conflict, this model reminds us "that conflict as a common problem and in a shared commitment to collaborative efforts
conflict is very hard to reverse once structural changes have begun to take root" at solving this problem. Such structural changes would not eliminate the competitive
(Pruitt and. Gahagan, 1972, p. 34). I t is important, therefore, to introduce tension- relationship between the conflictiilg parties; rather, they represent a shared commit-
reducing steps early enough to prevent the emergence of new social and psychological ment to collaborative problem solving that exists alongside the competitive interests
structures. On the other hand, Pruitt and Gahagan's assessment of efforts to reverse causing the conflict. Structural changes of this kind depend on the development of
conflict processes may be unduly pessimistic. It seems reasonable to propose that such contexts and procedures for conflict resolution that place a premium on adopting
efforts at deescalation as the GRIT strategy may produce structural changes in their a collaborative role and a problem-solving orientation.
own right, just as escalatory steps do. Thus, systematic conciliatory moves, announced The structural-change model does not provide ready-made formulas for
and initiated by A, may bring a peace-oriented faction within B into prominence effective strategies of conflict resolution, but it does offer some conceptual guidelines
and may help to improve A's general image within B. These structural changes may for the discovery and evaluation of such strategies. Though we cannot point to full-
have a multiplier effect on conflict-resolution processes, accelerating them beyond blown strategies derived from this model, recent experiments in conflict resolution
B's specific reaction to A's specific initiative. incorporate some of the insights of the structural-change approach. Notable among
Even at best, however, once structural changes have occurred, unilateral these are the exercises in "controlled communication" of John Burton (1969) and
initiatives cannot resolve the conflict; they can only contribute to creating an atmo- his associates at the Centre for the Analysis of Conflict at University College, London,
sphere in which conflict resolution becomes possible. As for conflict resolution itself, and the Fermeda Workshop, organized by Leonard Doob (1970, 1971) and his
the structural-change model has two general implications for the development of associates at Yale University. In both of these efforts, representatives of nations or
effective strategies. national (ethnic) communities involved in an active conflict were brought together
(1) T o overcome the organizational and psychological commitments to the for face-to-face communication in a relatively isolated setting, free from governmental
perpetuation of the conflict, an effective strategy of conflict resolution must take and diplomatic protocol. Discussion, following a relatively unstructured agenda, took
specific account of these commitments. Thus, selection of participants for any uni- place under the guidance of social scientists-knowledgeable both about group process
lateral, bilateral, or third-party effort at conflict resolution must be based on a and about conflict theory-and were designed to produce changes in the participants'
differentiated view of the composition of the parties, recognizing that some elements perceptions and attitudes and thus to facilitate creative problem solving.
within each party are more amenable to resolving the conflict and others more One of the exercises organized by Burton (held in a university setting in
committed to its perpetuation. I t may be easier to reach agreements by working with London over a one-week period in the fall of 1966) dealt with the conflict between
the former, but more difficult to actualize such agreements if the latter are ignored. the Greek and Turkish communities in Cyprus. Top decision makers of the Greek
Furthermore, efforts at conflict resolution must provide a context in which the con- and Turlcish Cypriot communities selected two representatives to the exercise, but
flict-induced structural changes within each party can be explored and counteracted. the representatives participated essentially as private citizens rather than in an
That is, each side must be able to gain some understanding of and address itself to official capacity. The relatively unstructured discussions were designed basically to
the new images, attitudes, and goals that the conflict has generated on the other side, encourage the participants to share their definitions of the conflict, their perceptions
as well as the different types and degrees of investment that various groups within of the goals and actions of both sides, and their assessments of the costs and benefits
the other nation have made in the continuing conflict. of various alternative approaches to conflict resolution. A panel of social scientists
(2) T o reverse the structural changes that reinforce perpetuation of the intervened periodically in an attempt to move the discussion away from mutual
conflict, an effective strategy of conflict resolution should aim toward creating struc- accusations toward a behavioral analysis of the causes of conflict, its escalation, and
tural changes of its own, in the form of organizational and psychological commitments its perpetuation, as well as toward exploratory efforts at possible solutions.
to the resolution rather than the perpetuation of the conflict. Both the form and the The Fermeda Workshop focused on the border disputes in the Horn of
290 Herbert C. Kelrnan, Alfred H. Bloom Assumptive Frameworks in International Politics 291

Africa, between Somalia and its two neighbors, Ethiopia and Kenya. Six nationals of intermediate distance from the decision-making apparatus. Those who are closest to
each of the nation-states involved were invited to participate in the workshop, held the locus of decision making may be less able to shed their official roles; they may be
in a Tyrolian resort hotel. Unlike their counterparts at the London workshop, they less likely, therefore, not only to experience substantial changes in attitude but also to
were mostly academics with no direct involvement in the foreign policy process. inject changes they do experience into the policy process. To assess the value of
Although some time was spent at the Fermeda Workshop in discussing theoretical workshop approaches, analysis and research must identify the points in the interna-
models of conflict and illustrative cases, these topics received much less attention tional poIitical process at which individual perceptions and attitudes have major
than in the Burton exercise. By contrast, greater attention was given to the ongoing impact, both on short-term decisions and on long-term trends, and to specify the
process of interaction and interpersonal behavior in an effort to enhance the mem- types of individuals who are located at these points and the nature of their contribu-
bers' self-awareness and communication skills, thus providing the tools and the tions to the policy process. Such information would provide a systematic basis for
atmosphere for a problem-solving approach to the conflict. determining the occasions on which workshops may contribute to conflict resolution
These approaches are, in various ways, consistent with a structural-change and for selecting the participants appropriate to a given occasion.
model. For example, in defining the appropriate participants for a controlled com- Even at best, problem-solving workshops are merely inputs into a more com-
munication exercise, Burton recognizes the different interests and commitments plex resolution process. They are not alternatives to diplomatic and political negotia-
represented by different elements within each of the conflicting parties. Thus, the tions, but supplementary or preparatory to them. They do, however, Tepresent a
exercise is focused on elements most immediately involved in the conflict on a day-to- promising approach to breaking down the social and psychological structures that
day basis (for example, the Greek and Turkish communities in Cyprus), rather than perpetuate conflict and to redefining the nature of the conflict and the relationship
those more remotely involved (for example, the governments of Greece and Turkey). between the conflicting parties.
Both Burton and Doob try to create an atmosphere that permits mutual understand-
ing and exploration of each side's perceptions, intentions, and goals, and of the Structure of International System
psychological and organizational commitments generated by the conflict. Such ex-
changes provide a sounder basis for determining the barriers to conflict resolution In opening this chapter, we referred to the assumption of the nation-state as
and for testing the potentialities of various moves toward resolution. Furthermore, the sole actor in the international system. We pointed5 out that this assumption can
these approaches are designed to reverse structural-change processes conducive to be and has been challenged. Indeed, our entire discussion is based on the view that
perpetuating the conflict (just as the GRIT strategy is designed to reverse conflict- one can enhance the study of international politics by focusing on other actors,
spiral processes conducive to the perpetuation of conflict). They create an atrno- including individual actors. In closing the chapter, let us return briefly to this issue,
sphere that increases the participants' access as well as their receptivity to inputs that which lies at the heart of the potential contributions of social psychology to the
may generate new attitudes and goals. Above all, they provide a normative context analysis of international politics.
and use a set of techniques that encourage participants to move from a combatant Nationalist Model. Both national decision makers and their constituencies
role to a collaborative role and to engage jointly in a process of conflict analysis and largely approach the international system with an assumptive framework derived from
problem solving. Thus, at least within the context of the workshop, a new relation- nationalist ideology. According to this framework, the nation-state is the relevant
ship between the conflicting parties may emerge, with a shared commitment to a and legitimate unit in the international system, in which paramount and ultimate
common effort and to a redefinition of the conflict. power is vested. It stands at the pinnacle of power and is entitled to overrule both
Workshop approaches can be expected to produce changes in perceptions smaller and larger political units. It is commonly perceived as a natural unit, and
and attitudes among the participants; for example, participants may come away with its exercise of final authority is regarded as self-evidently right. The cognitive and
a better understanding of their own side's contribution to the conflict and of the goals affective dominance of the national system in the minds of national leaders and popu-
and intentions of the other side. There is evidence that the London and Fermeda lations makes for an excessively narrow conception of the structure of the interna-
efforts were successful in producing such changes. There is considerably less evidence tional system. The organization of the world as a set of paramount nation-states has
that these workshops were successful in producing innovative ideas fbr resolving the such a powerful hold on our thinking that we find it almost impossible to think-
conflict, but there are good theoretical reasons for proposing that workshops are in and particularly to think creatively-about alternative structures.
~rinciplecapable of such outcomes. The real question is when and how the products The perspective of the nation-state dominates our thinking to such a degree
of conflict-resolution workshops-in the form of changed attitudes and innovative that even professional students of international relations are generally unable to
solutions--can be transferred to the policy process. The problem of transfer repre- escape it. They tend to use the nation-state as the basic unit of analysis and as the
sents the most serious limitation of the workshop approach. (For a fuller discussion primary actor in the international system. Even those whose observations are based
of tEis issue, see Kelman, 1972, pp. 195-200.) In general, it is reasonable to assume on the behavior of decision makers tend to assume that the nation-state is the
that feedback from the workshop to the policy process is greatest if the participants decision-making unit and that the decision makers speak for it. These assumptions are
are close to the center of decision making, as in Burton's model. It can be argued, often justified; but if we build them into our conceptualization, we are unable to
however, that the ideal candidates for participation are individuals who are at some discover the conditions under which an alternative set of assumptions for organizing
 Herbert C. Kelman, Alfred H. Bloom Assumptive Frameworks in International Politics

the global system would be equally or more appropriate. Thus, our very way of phenomenon is not fully recognized. Other important functions seem to require
studying the international system is a major factor in determining and reinforcing its transnational organization if they are to be effectively fulfilled, but this necessity
character. has not been translated into action. In either event, an asshptive framework that
The pervasive and refractory assumptions of the nationalist model have been views the nation-state as the only operating unit within'the international system and
particularly inadequate in view of the important changes that the nature and as the only natural and self-evident basis for organizing and carrying out the func-
functions of the nation-state have been undergoing in recent years. To conceptualize tions of a society does not appear adequate to current realities and necessities.
the international system as operating entirely through autonomous, supreme nation- Functionalist Model. An alternative model of the international system takes
states does not fully conform to current reality, nor does it provide an adequate as its starting point the concept of a global society, in contrast to the conception of
model for meeting many of the needs of the world's population. a world consisting of so many sovereign nation-states. There is no implication here
There is no doubt that the nation-state remains the dominant unit in the that, in an empirical sense, the world functions as a single, integrated social unit.
international system. After all, this is the basic assumption on which the international Rather, for analytical purposes, the model treats the world as a total system, con-
system is organized, an assumption that is constantly reinforced by national and sisting of various interdependent units organized in different ways to fulfill a variety
international institutions. This does not mean, however, that the typical nation-state of functions. I t makes no prior assumptions about the relative power of different
is as independent or self-sufficient as the ideal model of the nation-state would units and about the manner in which their functions are carried out. Analysis focuses
imply. There is increasing penesation of national states by other states and by on the whole range of actors in the international system, including not only the
international organizations (Rosenau, 1966). Even powerful states are constrained, nation-states but also the many individuals, groups, and corporate bodies-subna-
not only in their international activities but also in their domestic affairs-in the tional and transnational, governmental and nongovernmental-that are engaged in
political, economic, and social realms-by events and reactions in other countries. global interactions. This model thus puts the nation-state into perspective. I t permits
Regional groupings are playing a larger role in the affairs of component states. Inter- a functional analysis of the global system, revealing the functions that are and those
governmental agencies, staffed by international civil servants, carry out various that are not being adequately met by nation-states and interstate units, the functions
important nonpolitical functions (for instance, in such areas as health and welfare or that are already being carried out by a variety of transnational units that have
economic development), at least for smaller states. developed alongside the nation-state, and the functions that seem to call for new
In many important ways, therefore, both national governments and their forms of transnational organization.
individual citizens function in a transnational society, or a series of transnational I t seems appropriate to refer to this alternative framework as a functionalist
societies. This reality is also reflected in the trends toward the development of a model, not only because it is based upon a functional analysis but also because it
genuine world community. Especially among the youth of all nations a common overlaps considerabIy with the various functionalist theories of international integra-
universal culture, with a common set of values and tastes, seems to be taking shape. tion. (For fuller discussions of such theories, see Mitrany, 1966; Haas, 1964; Burton,
Furthermore, because of common problems and increased facilities for cross-national 1972. Haas also presents a detailed case study of the International Labor Organiza-
communication, such phenomena as student rebellions spread rapidly over a number tion, to explore whether and how a functional international organization contributes
of societies. International contact for many segments of national societies is extensive to international integration.) Functionalist theories are guided by a normative and
and-particularly in professional fields, such as science, medicine, the arts, and prescriptive orientation as well as an analytical one. Essentially, they start with a
various areas of scholarship-national lines have become increasingly meaningless critique of the nation-state as an effective vehicle for meeting the social and economic
for organizing the business at hand. needs of the world population, and with the assumption that these needs can best be
In addition, the nation-state is no longer capable of serving some of the met through functional (nonnational) rather than political (national) organization,
functions that it was designed to serve. Foremost among these is the function of They do not propose to mount direct attacks on natibnalism and national sovereignty,
military security, which no state-no matter how powerful-can fulfill on a unilateral but rather to build transnational functional institutions concerned with specific
basis today. Newer and poorer states, in particular, cannot entirely rely on their own human tasks. I t is assumed that, as the number and significance of such institutions
resources to carry out the functions of economic development and of meeting the increases, new loyalties-toward the specific institutions and toward the new world
health and welfare needs of their populations. Higher education, scientific research, order they represent-will gradually take shape. Whether or not such .developments
and technological development are among those functions that will probably have would eliminate the nation-state, they would certainly reduce its paramountcy, and
to be organized on a transitional basis to an increasing degree. I t should be possible they would be conducive to a more peaceful world order-peaceful not only in the
to test empirically whether some of these functions can be discharged more adequately sense of preventing war but in the sense of actively working to eliminate the roots
by national or by transnational arrangements. of war.
A major implication of the new developments is that we are living in a world To develop functionalist alternatives to structuring the international system,
that, in important ways, deviates from the nationalist model of the international we must first abandon an assumptive framework that postulates an international
system. Already, various societal functions are being organized and cultural values system entirely composed of sovereign nation-states. Of course, the functionalist
and tastes are being expressed on a transnational basis, even if the extent of this model in turn makes a series of assumptions that need to be carefully examined, both
 I

294 Herbert C. Kelman, Alfred H. Bloom I Assumptive Frameworks in International Politics 295

theoretically and empirically. Central among these are the assumptions ( 1 ) that I interpretation in competitive terms. According to our model of personal involvement
multiple loyalties can coexist, ( 2 ) that new loyalties emerge out of functional involve- in the national system, openness to functional transnational involvements should be
ments, and ( 3 ) that cross-cutting commitments and loyalties promote the integration greater to the extent that loyalty to the national system is rooted in instrumental
of a social system. There is considerable support for these propositions at various attachments. Such attachments are compatible with a more pragmatic approach to
system levels, but there is a great need for empirical research to explore their validity the organization of societal affairs. Thus, individuals who are instrumentally attached
at the level of the international system as such. Such research must take account of to the national system should be ready to support transnational institutions if these
the special character of loyalty to the national system as it has evolved in the modern are seen as having functional value-as, for example, pragmatic bureaucrats in
world. Western Europe were prepared to support the European Economic Community
The development of functional, transnational institutions does not require a because they saw it as an effective vehicle for carrying out their economic functions.
displacement of national loyalties, but it does imply a tolerance for multiple Once individuals become functionally involved in transnational institutions, loyalty
loyalties, for permitting the existence of transnational loyalties alongside the national to these institutions-and to the world order they represent-should spontaneously
ones. There is no necessary conflict between the two. Multiple sets of loyalties are, in evolve in due course.
principle, completely compatible with one another, as long as the groups to which We would further hypothesize that international integration is likely to in-
these loyalties are directed serve different functions and apply to different domains crease as more and more important segments of national societies become entangled
of a person's behavior (see Guetzkow, 1955). According to nationalist ideology, how- in a network of transnational commitments that are relevant to some of their vital
ever, loyalty to the nation-state enjoys an exclusive status; therefoie, many segments of needs and interests. The best example of such entanglements is the participation of
a national population are inclined to view national and transnational loyalties in individuals in transnational organizations that enable them both to enact their profes-
competitive terms-particularly when loyalty to the nation-state is rooted in senti- sional roles in a personally meaningful way and to work for the benefit of groups
mental attachments (as discussed earlier in this chapter). with which they are positively identified. The nature of these organizations may vary
Whenever a situation is viewed as a competition between national and trans- widely: they may be concerned with pl-oblems of business, labor, agriculture, health,
national loyalties, the latter almost invariably lose out. Transnational institutions welfare, education, science, literature, the arts, or religion; they may range from
typically lack legitimacy because they are not supported by important sentimental and intergovernmental agencies, such as those linked to the United Nations, to private
instrumental links for most individuals. They do not have at their disposal the kinds organizations set up on a completely nonnational basis; they may be, at one extreme,
of mechanisms of concerted socialization that are available to the nation-state, in highly organized and formal (such as an international mechanism for conflict resolu-
which major societal institutions collaborate from -the very beginning in building tion) or, at the other extreme, totally unofficial and unorganized (such as the world-
loyalty to the system. Educational efforts geared to "world citizenship" are distinctly wide youth movement that serve3 to express the common concerns and values of
limited in their capacity to create sentimental attachments and cannot substitute for the emerging generation). What is critical in all of these organizations is that they
the emotional conditioning and the many-sided reinforcements that underlie national create, for their participants, strong functional ties to a global society. Participants
loyalty. Deliberate efforts to develop international symbols and to create a sense of become committed, in one area of their lives, to a transnational definition of the
international identity are unlikely to succeed in the face of the power of national world, because such a definition is instrumental to meeting specific needs and
symbols and national identity; in fact, they may intensify national commitments if interests that have personal significance for them.
they are perceived as threats to national 'integrity. Similarly, transnational institutions From the point of view of international integration, these functional commit-
do not have at their disposal-given the structure of the international system- ments are important te-ause they represent a principle of organization that cuts
mechanisms for enbangling individuals in social roles and for creating functional across the division of the world into national units. In making these commitments,
interdependencies. Thus, for most individuals, international authority (as epitomized, individuals are not expected to transfer loyalty from a smaller to a larger unit but
for example, by international law) is an abstract, artificial construct; in contrast, to entertain multiple loyalties in keeping with their multiple roles in a variety of
national authority has for them a concrete and existential meaning. The comparative cross-cutting functional systems. Such loyalties to functional, cross-cutting entities can
disadvantage of transnational institutions does not represent merely a quantitative more readily bypass the dominance of the national system, They can counteract the
problem. I t is not only that national commitments are stronger; they are in a quali- polarizing effects of a world system in which all relationships are subordinated to a
tatively different position from other kinds of commitments, both to larger and single basic cleavage along national lines. Finally, they can reinforce the development
smaller units, or to units organized on a cross-cutting principle. of a broader, more realistic, and more imaginative assumptive framework, which
Under these circumstances, what are the conditions under which loyalties to recognizes the possibility of various ways of organizing the global system to meet the
transnational institutions are likely to evolve? Along with the functionalists, we needs and interests of the world population.
would hypothesize that such loyalties are most likely to emerge out of involvement
in transnational organizations, designed to fulfill specific functional purposes. Such
functional involvements are less threatening to the sense of national identity than
a diiect appeal to internationalist sentiments might be, and they are less subject to
 Contributions of Psychobiography 297

psychobiographer are in no sense unique but are shared by biographers with other
predilections as well as by social scientists in general.

History

Victodan biographers were concerned mainly with getting the facts. They
gathered together a mass of details and arranged them chronologicaIly, purporting
to reproduce, without bias, a man and his times. Thus, Pierce, in his four-volume
Memoir and Letters of Charles Sumner (1878-1893), attempted to document his

CONTRIBUTIONS OF subject's every act and gesture, as Sumner himself observed them. Similar efforts are
evident in works such as Monypenny and BuckIe's life of Disraeli (1913-1920),
Morley's life of Gladstone (1903), and the ten-volume work by Hay and Nicolay on
Abraham Lincoln (1886) . (For histories of Victorian biography, see Johnson, 1912;
Nicolson, 1927; Garraty, 1957.)
Replicating the past, however, is not simple. Even the most scrupulous of the
traditional biographers had to rely on some frame of reference in selecting and inter-
preting their data. Although the principles of selection and organization might be
implicit, the biographer's approach was governed, generally, by conventional notions
Betty Glad of what is important in a life history.and how it should be presented (see, for
example, Hartman, 1922-1923; Schneider, 1937; Langer, 1958). In the typical
treatment of a great man, for example, there was usually an introductory chapter on
his ancestry, in which his greatness was traced to his inheritance (Johnson, 1912).
The simplest laws of heredity were distorted in a frantic "cytological search for
famous chromosomes" (Tozzer, 1933). In recounting his subject's childhood, the
biographer also looked for-and usually found-signs of his future greatness-some
indication that his destiny was determined at an early age (Johnson, 1912; Fuess,
1933).
Personality descriptions, moreover, were intuitive and ad hoc; and the inter-
Psychobiography is, essentially, any life pretations based on them were conventional and unsystematic. In presenting his
history which employs an explicit personality theory-that is, a perception that in- subject's character; the biographer generally accepted the idealized self-image.offered
dividual behavior has an internal locus of causation as well as some degree of struc- by the subject himself and supported by his closest admirers (Johnson, -1912; Jones,
ture and organization (Levy, 1970, p. 56). The specific theoretical framework 1932). The subject was almost always portrayed as rational and high-minded; and
employed may be a traditional psychoanalytic one, growing out of the work of any actions based on narrow self-interest were either ignored or justified in lofty
Freud, or it may be based on the writings of theorists such as Jung, Horney, or Sulli- moral terms (Langer, 1958). And if the subject experienced an emotional break-
van. I will not attempt here to choose between these possible frameworks but rather down, it was attributed ordinarily to "overwork" (see Bain, 1927, p. 37) .l
will deal generally with all forms of depth psychology-those conceptions of per- In the works of Lytton Strachey (Eminent Victorians, 1918), however,
sonality that contain a theory of unconscious motivation. Despite certain brilliant another kind of biography was attempted. (For a history of earlier attempts at in-
achievements, psychobiography as a research form is just now beginning to win depth biography, see Guedalla, 1939; Trueblood, 1939.) Strachey and his followers,
general acceptance in political science. After a short history of its development, I no longer content with "depicting the shell of outward eventsJ' (Mumford, 1934),
outline and then evaluate several assumptions which have reinforced this scholarly pulled aside the public masks of their subjects, revealing the complex and contra-
skepticism. The main purpose of the chapter is to show that a better understanding
of the approach will enhance its possible contributions to political science. Further- * Many contemporary biographies retain these tendencies toward the use of ad hoc,
impressionistic principles for selecting and interpreting data. As Howard (1969, p. 12) points
more, I argue that the problems of data collection and interpretation which face the
out in his survey of judicial biographies, the authors make rather conventional interpretations
of personality "No Freudian inferences jar these pages, though we know that Daniel suffered
melancholia, Doe hypochondria, and Barbour insanity."
298 Betty Glad Contributions of Psychobiogaphy 299'

dictory human beings underneath (Jones, 1932; Fuess, 1933; Josephson, 1940) , historians such as Mazlish (1972) and psychologists such as Erikson (1969) and
Rejecting the multivolume commemorative biography, Strachey (1918, p. v) sug- W. C. Langer (1972).
gested a more subtle technique. The wise biographer, he explained, "will attack his
subject in unexpected places; he will fall upon the flank, or the rear; he will shoot Prevailing Assumptions
a sudden, reveaIing searchlight into obscure recesses, hitherto undivined. .. . He
This early neglect of psychobiography as a research tool may be partly ex-
will row out over that great ocean of material, and lower down into it, here and
there, a little bucket, which will bring up to the light of day some characteristic plained as a reaction against the somewhat cavalier attitude of earlier biographers
specimen, from those far depths to be examined with a careful curiosity" (cf. DeVoto, toward documentation (Garraty, 1957, pp. 139-145). More important, however, are
1933). certain closely connected assumptions that have governed the thinking of proponents
In some hands, this new form deteriorated into the "debunking" biography. as well as critics of the psychobiographic form. They have generally taken it for
Woodward (1926), for example, took pains to point out that George Washington's granted that personality theory is useful primarily for explaining pathological or
ancestry was undistinguished; that he dealt in slaves and indentured servants and irrational behavior; that analyses of individual pathologies should be followed by
ejected squatters from his lands; that he started wearing false teeth in 1789; and that proposals for therapeutic interventions on either an individual or a societal basis;
he had no children, possibly due to an earlier attack of mumps. Such an approach, that psychobiographic studies necessarily deal with the idiosyncratic traits of indi-
however, distorts the complexity of character as much as the idealization it seeks to viduals; and that the approach is an alternative to the political and sociological
replace: "Seeking to deflate . . . the extravagant reputations of the past, [these explanations of political behavior and, as such, is competitive with them. These
trends can be seen in several recent psychobiographic studies, as well as in the litera-
biographers] often completely neglected the' realities upon which they were founded.
Besides, they lost an important clue. The mask itself is as important an aspect of a ture dealing with these studies.
The emphasis on pathological behavior is evident in several major works.
life as the more devious tendencies it conceals. To tear off the mask and to throw
Alex ~Gttfried(1962) sees in Anton Cermak's drive to power in Chicago an under-
it away was a little like tearing off the face of a clock on the hypothesis that if one
lying dependency, for which Cennak compensated with action of an overaggressive
wanted to tell time correctly one must get nearer the works; it abandoned the very
and independent nature. Edinger (1965, p. 297) attributes Kurt Schumacher's
part of the instrument that recorded the action of the works" (Mumford, 1934, p. 7).
failure, in his political competition with Konrad Adenauer in postwar Germany, to
Psychobiography, which was developing concurrently with these other forms
compdsive strivings "to relieve inner personality tensions by means of adaptive
of the "new biography," showed similar characteristics-toward deeper analysis but
mechanisms that had proven effective in the past for maintaining the equilibrium of
with tendencies to accentuate the negative. I t differed in that an explicit personality
his personality system." Wolfenstein (1967) holds that Gandhi, Trotsky, and Lenin
theory was used. Initiated in 1910 by Freud in his analysis of Leonardo da Vinci,
carried into their adult lives unresolved conflicts with the parental generation (each
psychobiography at first was used mainly by psychoanalysts. Typical were Abraham's
man struggling with a primitive image of authority). Forrestal's performance as
study of Amenhotep IV ( 1912), Jones' work on Louis Bonaparte ( 1913), and Hart-
Secretary of Defense, according to Rogow (1963), was much influenced by severe
man's study of the abolitionist William Lloyd Garrison (1922-1923) . Shortly there- inner conflicts and guilt feelings, due to his earlier relationships with a dominant
after, several full-length psychoanalytically oriented biographies appeared-for mother and an indulgent father. Fred Israel (1963, p. 132) suggests that Key Pitt-
example, Johnson's Randolph of Roanoke (1929) and Clark's Lincoln (1933). (For man's growing alcoholism during his years as chairman of the Senate Foreign Rela-
surveys of this literature, see Dooley, 1916; Fearing, 1927; Schmidl, 1962; Kiell, tions Committee was due to "promotion depression"-his responses to advancement
1964.) Lasswell (1930) went beyond these works in his attempt to use depth psy- to a position that he felt incompetent to handle. Alexander and Juliette George
chology for the development of social theory. Through in-depth studies of several (1956) attribute many of Woodrow Wilson's adult role failures to his youthful un-
political actors, he sought to demonstrate that "political movements derive their healthy relationship with his exacting father; conflicts over that relationship .con-
vitality from the displacement of private motives upon public objects" (pp. 202-203). tributed to Wilson's breakdowns whenever his most intensely held commitments
Political scientists, however, did not build on these early beginnings for some were challenged.
time. An occasional political biography-for example, DeGrazia's 1948 article on Although the foregoing biographers have dealt with subjects who showed
Gandhi-drew upon the Freudian framework (compare McConaughy, 1950) ; but strong unconscious conflicts, each has also looked at his subject's strengths and skills.
the systematic use of psychobiography for the exploration of the interface between Other studies, however, have been almost totally denigrative. For example; Zeligs
personality, attitudes, and political behavior did not really begin until 1956, with ( 1966) in his analysis of Whittaker Chambers seizes on every error and every lapse of
the publication of the Georges' book-Woodrow Wilson and Colonel House-and memory to prove that Chambers was a pathological liar, a cheat, and probably an
Smith, Bruner, and White's Opinions and Personality. There were followed by the overt homosexual. Freud and Bullitt (1967) in their analysis of Woodrow Wilson
works of Lane (1962), Gottfried (1962), Rogow (1963), Barber (1965, 1972a, make certain that "little Tommy" gets away with nothing (Erikson, 1967). There
1972b), Glad (1966), and Wolfenstein (1967). Similar work has been done by was, apparently, something contemptible in Wilson's desire to be respected and in
300 Betty Glad Contributions of Psychobiography

the breakdowns that he suffered. His achievements are belittled (he had to become pendent of all circumscribing systems, classifications, or blueprints. He is a bei
a statesman in order to be a "man") ;and the views of contemporaries who found process, always unfinished, a being of only partially expressed potentialities
him gracious, or a warm and inspiring teacher (for instance, Allen Dulles, 1966), mental health consequently is to be defined in the infinitely varied terms of
are simply ignored. individual himself-that is, in reference to his own possibilities. His ego is t
This assumption-that personality theory is specifically a tool for the study understood as creatively integrating internal needs and external demands" (p, 3
of the pathological-is made explicit by Rogow (1968; see also Merriam, 1925; There is, in short, a tendency in the psychology and politics literatu
Fearing, 1927) : "While most political leaders neither require nor merit a psycho- assume that personality theory is applicable onIy to the atypical. Though
biography, the form is particularly appropriate when we are dealing with odd or atypicality is ordinarily associated with the pathogenic-the deviant who is ju
deviant political careers" (p. 605). Rogow goes on to suggest that the approach be "neurotic"-in some instances the deviant may be especially strong, possibly a "
applied to left- and right-wing political extremists, ilcoholics, suicidal types, "born of history." But even here personality theory may be viewed as having little predi
losers," individuals with psychosomatic illnesses, and individuals who make erratic value, merely proclaiming that the subject is free and therefore able to mak
political shifts. own choices. For general discussions of the issue of free will versus determinism
Many psychobiographers have also felt obligated to supplement their analyses Fromm (1947, pp. 221-227), Garraty (1957, pp. 5-7), and Holt (1961b).
with proposals for therapeutic intervention. For example, this is a salient theme in The fourth assumption-that the psychobiographic approach is an altern
Rogow's (1963, pp. xv) work on Forrestal: "Forrestal's illness and suicide were not to more traditional political and sociological explanations-is evident in se
inevitable; the illnesses and suicides of other world leaders are not inevitable; and it works. Lasswell (1930) set the stage for the competition between the two niod
is not inevitable that the world community become ill and commit suicide. But there explanation in his use of psychoanalytical methods to show that political movem
will be other Forrestals and other wars if attempts are not made to prevent, detect, cannot be understood in terms of manifest political intentions, deriving, instead,
treat, and cure those illnesses that affect rational mental processes in decision-making vitality from the displacement of private motives upon public objects. Wolfen
environments." To this end Rogow suggests a change in public attitudes-a recogni- (1967), though he explicitly rejects this form of reductionism, tends to ignore
tion that public office imposes considerable strain on an individual and that mental social and economic factors that led to Trotsky's revolutionary role choice-that
illness may be an occupational hazard. To keep the obviously ill from holding high that Jews were restricted to a certain territory, could not own land, were forbi
office, he also recommends psychological tests as a condition of employment; to keep to enter certain professions, and were limited by quotas in the universities.
them from breaking down while serving, he recommends early diagnosis and therapy Today, however, it is mainly the critics of psychobiography who are guil
(see also Lasswell, 1948, ch. 6; Klineberg, 1964, pp. 65-66). this form of reductionism, often ignoring explicit disclaimers from the authors. T
Related to this emphasis on psychopathology, but somewhat distinct from assume that if other relevant factors can be demonstrated, personality examina
it, is a third tendency-the equation of the "individual" with the "idiosyncratic." are not required. For example, James Q. Wilson in his review (1962) of Gottfr
Thus, Morton Kaplan (1964) has written that the extent to which psychological Boss Cermak of Chicago ignores the sophisticated political analyses in that w
explanation will apply, even marginally, to international politics "will hinge on the concluding that the explanation of the transformation of Democratic party po
psychological reactions of particular individuals with particular personalities in in Chicago will "require more than a clinical look at the boss's bowels" (an emp
particular political roles in particular political circumstances." Even Fred Greenstein that Gottfried explicitly rejects). Morton Kaplan (1964) questions the whole u
(1971) seems to be governed by this assumption. In an excellent essay that clears psychological explanation in political science : "Did Churchill project his own hos
away many objections to the use of personality theory in the analysis of political onto the Germans when he urged Britain to oppose Nazi aggression? To pose
behavior, he highlights those factors that make idiosyncrasies count. Thus he holds question is to demonstrate that adequate analysis requires a political analysis of
that personality factors are relevant only in situations where personal differences elements of Nazi policy. That is, one must know whether Germany was prob
might have an impact-for example, in the actions of a top political leader in a genuinely aggressive or ChurchilI probably genuinely hostile. And, if one makes
crisis situation. Joining this with the first assumption above, he assumes that the analysis, the policy consequences will generally be elucidated apart from psycho
"individual" is also likely to be pathological: "Wherever the circumstances of politi- cal analysis."
cal behavior leave room for individuality, the possibility exists for ego-defensive The preceding quotations from outspoken critics of psychobiography refl
aspects of personality to assert themselves." more widespread attitude among political scientists-an assumption not only
The equation of the "individual" with the "idiosyn~ratic'~takes a somewhat institutional or environmental analysis is an alternative to individual psychology
different form in the writings of Henry Kariel (1967). For him the unique individual that it has greater explanatory power. Even the literature dealing with leade
is likely to be creative and forward-looking, Thus, he sees the psychiatrist's objective shows this bent. As Tucker (1965; see also Rutherford, 1966) has pointed out, t
as the freeing of the individual from all formulas. The implication is that the mature is a tendency to assume that the American political system selects individual
person is unique in the sense that he is free-that is, undetermined and unpredictable. high public, office who will perform their function in accordance with the req
"By nature, man is obstinately in rebellion against social constraints. He is therefore ments of the situation; pathogenic types are sorted out because they are unab
to be comprehended as an irreducible phenomenon, as existing ultimately inde- build sufficient social support to get to the top. A discussion of indivihal c
302 Betty Glad Contributions of Psychobiography 303

acteristics, it follows, is not required for most political explanations. Even the litera- men are to be really understood. Barber (1965, pp. 252-253) sees certain "law-
ture on totalitarianism is influenced by this point of view, with Arendt (1951) and makers" as especially productive and well-adjusted personalities. In my work on
others arguing that this phenomenon can be explained in bureaucratic terms rather Charles Evans Hughes (Glad, 1966), I show that his early development of certain
than by an individual leader's pathology (Gilbert, 1950; Tucker, 1965; Rutherford, ideological structures and skills was significant in preparing him for his very success-
1966). In short, as Edinger (1964, p. 437) has said, when we consider conduct to be ful career in American politics. His political life cannot be viewed as the displace-
the result of political environment, "the behavior and personality of the individual ment of infantile needs into the public arena; rather, it must be seen as an expression
tends to be all too often obscured-if not factored out." of a basic personality structure which was well geared to his chosen field of action.
The therapeutic slant of some of the literature may be dealt with quickly.
Critique Although personality theory has historically been wedded to medicine and its proofs
dependent upon specific clinical responses, therapeutic intervention is not a requisite
The conceptions of psychobiography discussed in the previous section have of its use as an explanatory framework. Indeed, Freud himself attributed greater
limited unnecessarily the utility of this form for the analysis of political behavior. significance to psychoanalysis as a science of the unconscious than as a therapeutic
Reflecting the close connection, historically, between personality theory and clinical procedure (see, for instance, his letter to Jung, Dec. 6, 1906, in Jones, 1955; see also
practice, these assumptions are based on a neglect of contemporary ego psychology, Greenacre, 1955; Hitschmann, 1956). In the social sciences generally, an under-
as well as of certain scientific commonplaces. The source and nature of these mis- standing of the relationships between variables does not obligate the investigator to
understandings may be delineated and certain clarifications made along the following control a given social system as a demonstration of the validity of his ideas. Scientific
lines. investigation requires only that the variables be isolated and manipulated in the
The assumption that the p~~chobiographic form is especially suited to dealiig laboratory or through comparative studies and that hypotheses be checked against a
with the pathological or the irrational subject has its roots in clinical psychology. broader framework of scientific knowledge. In short, the application of personality
Depth psychology, traditionally, has been concerned with the neurotic, whose self- theory to the histories of political actors does not obligate the author to put his sub-
defeating tendencies can be traced back to unresolved emotional conflicts of his jects onto the couch and solve their problems or even to propose ways to keep them
early years. It is this tradition from which most psychobiographers have drawn. As from infecting the broader society with their neuroses.
Edel (1961, p. 465) has aptly observed: "I do think that certain papers in applied The equation of the "individual" with the "idiosyncratic" is also based on a
psychoanalysis have lost much of their value because their authors have enjoyed their misunderstanding of the nature of the clinician's emphasis on subjective experience
underwater snorkeling to such an extent that they never once looked up to see the in the therapeutic situation. I t is true that each person uniquely (that is, subjectively)
.
great .glittering exposed mass of iceberg. . . To be sure, the submerged part deter- experiences his own anxieties, dreams, and fantasies. Further, his emotional fixations
mines the shape of what is above. Nevertheless, it is the visible shape which confronts and defenses can be traced back to specific encounters during the formative years
.
the world . . and it is the relationship between the submerged and the exposed with his family, peers, and teachers. Moreover, any basic change in his personality
which is all-important." may require a specific type of encounter with one psychoanalyst-an encounter
Contemporary ego psychology, however, offers a conceptual apparatus for enabling him to recall and relive those specific emotional experiences that led to
dealing with the entire human being, a way of looking at the connections between repressions and symptom formation.
the submerged and the more visible aspects of personality. Though it does not ignore These experiences and adaptations, however, are not singular but are more
the more primitive drives, ego psychology is concerned with the mature, the rational, or less patterned (Trueblood, 1939; Dollard, 1935; Fromm, 1941, pp. 277-299). All
and the adaptive or creative response to contemporary situations. Erikson (1963), human beings have some common emotional and intellectual structures and poten-
for example, has emphasized the synthesizing function of the ego in developing stable tials-fbr example, a capacity for affectionate attachment and an ability to learn
conceptions of the self through a series of stages that extend throughout life. Maslow through abstract symbols. These in turn are shaped by some common human ex-
(1954) has been concerned with the characteristics of the "healthy," creative human periences, such as a long period of dependency on parents or other adults and the
being. And Kubie (1961, pp. 58-59) has pointed out that neither a rigid insistence gradual construction of an identity through the learning of symbols and the recogni-
on rationality and control nor the mere projection of unconscious impulses into tion of boundaries between self and others. These experiences may be specific, in
action leads to creativity; only when the formerly unconscious impulses are touched some respects, to an individual's culture or subculture-for example, the quality
by and reorganized by the rational faculties are new solutions found. Such a process of affection given or attitudes toward food and sex. Indeed, the precise structure
is possible only for an individual who can afford to rlsk the unknown, to engage in of the family, its values, its conflicts, -the techniques by which it socializes its members,
the "freely searching, scanning, shaking-together process which we call free associa- the acceptance or rejection of various defenses against anxiety-all can be viewed as
tion.)) embodiments of a particular culture. But whether universal or culturally specific,
Recent psychobiographies reflect the influence of revisionist ego psychology. the adaptations and subjective experiences of each individual, important as they are
Erikson's works on Imther (1958) and Gandhi (1969), for example, show that the to him, follow more general lines. His inner life and his relations to his family-the
strengths, 'and not merely the weaknesses, of great men must be considered if these private realm-are patterned to a considerable extent and are therefore predictable.
 Betty Glad Contributions of Psychobiography 305

I t is true, of course, that the arrangement of human and culturally specific brought home to him by one of the mothers, who explained: "Well, I think they
traits in any one individual is in some sense distinctive. His personality, like his just went ahead and did it, because it was there for them to do. And their spirits,
thumbprint, has certain unique swirls. These differences, however, usually are much they held up because they knew they was doing something good, and it made a lot of
less significant than the similarities; the truly unique aspects of personality ordinarily sense to them."z
consist of only minor variations on a few basic patterns. The very development of Even a demonstration that certain political behavior has its roots in infantile
depth psychology is predicted on the assumption that the most basic human char- fixation does not negate the need for political explanation. Wilson's bargaining
acteristics are widely shared. Indeed, personality theory would be impossible were this behavior at Paris cannot be evaluated solely in terms of his unconscious needs; the
not so. (For the debate between "ideographic" and "nomothetic" views see Allport, political results of this bargaining style must be judged in political terms. Forrestal's
1938; Beck, 1953; Eysenck, 195qb; Holzberg, 1957; see also Chaper Two in this (Rogow, 1963) inclination to take rigid bargaining stances may also be understood
volume.) in terms of his pathology, but these stances served him well when, as Secretary of
The ernphasis on the "individual" in psychoanalytic thought, then, should not the Navy, he had to negotiate with Air Force Secretary.Symington over the establish-
be equated with the study of the atypical. Each person is seen, more appropriately, ment of the Department of Defense. Even the demonstration that the Soviet threat
as an exemplar of all human life at one level of abstraction, and of specific cultural had become a central thenie in Forrestal's paranoic syrnptomology does not tell us
or subcultural characteristics at another. The distinctive aspects of this configuration much about the reality of that threat to the United States. And, as Gottfried (1962)
for most individuals are apt to consist of only minor variations. has shown us in his study of Boss Cermak, the very pathology that is damaging to a
The assumption that the psychobiographical approach is competitive with man personally may prove useful in his career as a power seeker and have significant,
more traditional political and sociological explanations is based, in part, on the perhaps even constructive, political results.
misconceptions just discussed. But it is also related to the emphasis which depth In short, the emphasis on unconscious motivation in traditional personality
psychology has given to unconscious motivation and to its consequent tendency to theory has sometimes led to a depreciation of the importance of situational (eco-
underplay situational variables. nomic and political) variables for human behavior and their outcomes. Though the
This emphasis on the unconscious is in some respects legitimate, based as it is importance of unconscious motivation and infantile fixations cannot be denied, poli-
on a recognition of the processes of externalization, whereby the neurotic masks his tical scientists must also look at structural variables within which an individual acts
own self-defeating tendencies by attributing his failures to outside forces. But this and must weigh their impact on his behavior.
emphasis also reflects a tendency among clinical psychologists to generalize from The presumed conflict between psychological and environmental analyses
their experiences with a relatively select clientele to human beings in general. The can aIso be traced to confusions in our scientific thinking-the failure to take into
anaylsand generally comes from the upper middle class and is impeded in his goal account developmental sequence and to distinguish between explanation at different
attainment by his own buried conflicts and self-defeating defense mechanisms rather levels of analysis (Greenstein, 1967). I t is true that personality is largely shaped by
than by such factors as poor education, lower-class status, poverty, or serious physical prior environmental forces and it is to these that the social scientist usually refers
ailments. Cure, it follows, depends less on social reform or a lucky change in indi- when he speaks of "social characteristics." But though social characteristics can
vidual circumstances than on new subjective experiences-that is, on an ability to cause psychological characteristics, they do not eliminate the significance of the
bring to consciousness emotional conflicts repressed by the individual early in his Iatter, as some political scientists have assumed. Rather, the characteristics at these
life and expressed through symptom formation. two levels continue to interact with each other.
There is a tendency, then, to see the analysand's problems as isolated from The confusion is due, in part, to the inappropriate application of certain
the social and economic crises of the day. Lynd (1965) provides an example from ;tandard techniques for the elimination of spurious correlation, controlling for third
a case cited by Elisabeth Hellersberg. An intelligent young man who had been a Factors.
left-wing social critic in college later developed a serious mental illness. His recovery
was manifest in a decrease in his radical interests and in his ability to get a raise. Unless one reached the most primitive original cause that initiated the
One day he simply lifted the telephone receiver and asked for an appointment with sequence that led to the independent variable under study, one would
his superior, which he was immediately given. The result was an instant salary in- always be able to find at least in principle an antecedent condition that
crease of 100 percent. As Lynd wryly comments: "The implication seems clear: really was responsible for the effect. In this sense, all demonstrations of
the healthy, adjusted person does not need to question the society; he can demand
and get whatever economic support he needs from it." When a social scientist relates individual economic and political failures and "stress"
to individual pathology, he is apt to go on and equate the healthy personality with the
More recently, Coles (1967) has concluded that a similar middle-class bias successful one. This may account for the reluctance of many social. scientists (Rutherford,
made it difficult for him to understand southern black children who were going 1966) to admit that pathogenic individuals may get to positions of power in society. Actually,
through mobs in order to attend previously segregated schools. He was puzzled by his the paranoic type of personality may well work his way into positions of considerable political
failure to find any "psychopathology" in these children. The truth was finally authority, as Devereaux (1955) and Tucker (1965) have pointed out.
306 Betty Glad Contributions of Psychobiography 307
relations . . . are spurious. Now in practice one can never reach the done in a way that is complementary rather than competitive with the other social
original cause in the developmenta1 sequence. . . . But in addition, the sciences,
concept of spuriousness cannot logically be intended to apply to antece-
dent conditions which are associated with the particular independent
variable as part of a developmental sequence. Implicitly, the notion of an Scientific Utility
uncontrolled factor which was operating so as to produce a spurious find-
ing involves the image of something extrinsic to the . . . apparent cause. Given a proper understanding of the psychobiographic form, one can see its
Developmental sequences, by contrast, involve the image of a series of potential for theory building in political science. I t should yield a deeper insight into
entities which are intrinsically united or substitutes for one another. All historical events and, in addition, should promote empirically based generalizations
of them constitute a unity and merely involve different ways of stating the about the connections between political and psychological phenomena. This potential,
same variable as it changes over time [Hyman, 1955, pp. 255-2561. only partially realized now, may be outlined in general terms, as follows.
Depth psychology is likely to yield more significant interpretations of the
But how does the analyst determine which antecedent conditions are intrinsic personality structures of political actors than does traditional biography. For example,
parts of a developmental sequence? Hyman (1955, p. 256) suggests the following the circumstances under which Woodrow M'ilson's "stubbornness" would be evoked,
guide: "Instances where tlie 'control' factor and the apparent explanation invoIve and the specific form it would take, can best be elucidated through the usage of
levels of description from two different systems are likely to be developmental se- concepts from depth psychology dealing with compulsive behavior (George, 1971).
quences. For instance, an explanatory factor that was a personality trait and a Anton Cermak's hostility, suspiciousness, and doggedness (Gottfried, 1962, pp. 336,
control factor that was biological such as physique or glandular function can be 373) take on a deeper meaning in the light of Franz Alexander's theory of organ
conceived as levels of description from different systems." neurosis. With the knowledge of Cermak's colitis, Gottfried was able to impute to
Another way of clearing up this confusion is to look at sociology as the study Cermak (and to verify in other ways) attributes of the typical colitis patients studied
of stable patterns of behavior between individuals within a group and psychology by Alexander-underlying feelings of weakness, dependence, and helplessness, to
as the study of the intrapsychic systems of individuals. Psychology is useful in pre- which the ego reacts with feelings of shame and guilt. Karen Horney's (1950) model
dicting either the responses of different types of individuals to comparable stimuli of the perfectionistic personality helps to explain Charles Evans Hughes' apparent
or the responses of similar types of individuals to diverse or changing stimuli (see self-confidence and high levels of performance along with his fits of depression, de-
Smelser and Smelser, 1965). The emphasis in political science up to this point has fensiveness, and avoidance of iintimacy with others (Glad, 1966, pp. 112-1 14). Only
been on the former concern-for instance, how different individuals might react to in the light of Horney's model do a11 of Hughes' characteristics fit together into one
a particular political candidate or issue. Of equal concern, however, should be such conceptual framework. And Hitler's "strength," as W. C. Langer (1972) points out,
matters as the use of terror in political rule or the impact on individuals of traumatic was fundamentally a reaction formation created to cover the weakness he despised
events such as atomic warfare (Beradt, 1968; Lifton, 1968). in himself.
Briefly, then, psychoanalytic explanations are not competitive with economic, The psychobiographic approach has another advantage. Relying on an explicit
socioIogica1, or political analyses. All human behavior is the result of multiple causa- conceptual framework, it enables an author systematically to compare one individual
tion, and explanation can be sought at several levels of analysis (see Chapters Three with another. Traditional political biographies, with their ad hoc and impressionistic
and Twelve in this volume). The human psyche is involved in all political behavior interpretations, make such comparison impossible, as Howard (1969) discovered in
-including routine and conformist responses. The psychic makeup of any individual his survey of judicial biographies. By way of contrast, Wolfenstein (1967) was able
is the result of his physical heredity and his earliest encounters with the culture. His to make meaningful comparisons between the lives of his three subjects (Trotsky,
behavior at any one moment is the result of his psyche as it has evolved to that point Lenin, and Gandhi) because he used an explicit conceptual framework. It helped
in time, as well as those factors in the environment with which his psyche interacts. I t him to isolate in each man those key personality traits that might have had relevance
is at this juncture that psychology connects with sociology and political science. for revolutionary leadership. Others using explicit personality theories in comparative
The argument to this point might be summarized as follows : Psychobiography biography include Smith, Bruner, and White ( 1956), Lane (1962), Barber (1965,
has been generally restricted to the explanation of pathological and idiosyncratic 1972a, 1972b), and Glad (1969).
The psychologicalIy oriented biography also lends itself to a holistic approach
behavior, often slighting the environmental factors that influence behavior and
-that is, to the study of the complex entity that is an individual life. As Alexander
sometimes recommending therapies at the societal level which are politically na'ive.
and Juliette George (1956, p. 318) have said: "The variety of ways in which given
With these misconceptions cleared up, however, the approach may prove to have a
personality factors may express themselves in a political leader's behavior . . . emerge
broader utility. Psychobiography may well go beyond the usual concern with the
only when the career as a whole, not merely a few isolated episodes from it, is
idiosyncratic and pathological behavior of individuals to explore the impact of
examined in detail." This type of biography has been explained in detail by Edinger
modal and rational personality structures on the political process. And this can be
(1964, pp. 668-669) :
308 Betty Glad Contributions of Psychobiography 309

For our purposes, a holistic approach to the dynamics of personality devel- biographical study of one political actor to broader political patterns, however, can
opment and expression-which gives consideration to antecedent as well be dealt with as folIows.
.
as contemporaneous events, . . to the "inner manJyas well as the im- Insofar as it can be demonstrated that an individual is typical (in important
pact of sociopolitical "outside" variables in the patterning of behavioral respects) of a "class," generalizations to other members of that class can be drawn
.
characteristics-seems most appropriate. . . In considering the psycho- from his biography. These initial classifications, of course, are difficult to make. As
logical development of our subject, we must seek to examine not only indicated earlier, the psychic makeup and the career of any individual is in certain
formative childhood experiences but personality development through basic respects typical of human beings in general and in other respects typical of
adolescence and adulthood to the point of any particular political act, others in his particular culture. (For discussions of typological analyses see Hempel,
including traumatic experiences and other personality crises which may 1965; Lasswell, 1968.) But subcultural variations also occur. In a complex society
have significantly influenced subsequent behavior. Nor must we neglect the proliferation of social classes, social roles, and ethnic groupings creates subcultures
the impact of socialization on personality development, the continuous which can lead to wide variations in career and personality structures. Furthermore,
process by which the individual adopts patterns of adjustment and re- some individual idiosyncrasies will manifest themselves. Any mobile individual in a
sponds to his social and political environmental stimuli.
complex social structure such as the United States encounters ideas and values that
Psychobiography also lends itself to the study of political cultures. Survey he ignores or rejects because they appear unworkable, 'create anxiety, or promise him
no satisfaction. Nor can the creative leap of the "new hero of historyy' be ignored.
methods have been used to study isolated attitudes, personality traits, and their con-
For better or worse, Napoleon, Bismarck, Hitler, and others fused together various
nection to specific political acts such as voting. The psychobiographic approach, how-
cultural strains in a way that bore the imprint of their unique personalities.
ever, is better suited to the exploration of systems-organically connected institutional
There is a problem, then, in distinguishing between idiosyncratic and modaI
and psychic variables in an ongoing political culture. It provides a manageable re-
personality and career patterns, and between characteristics of human beings in
search entity-the individual and his career-for analyzing these complex relation-
general and characteristics of subcultures. (See Chapter Twelve in this volume.)
ships. Starting with the subject's psychic system and empirically exploring the past
But if a biographer is to generalize, he needs to know what his subject might typify.
and present relationship of that psyche to environmental inputs and political outputs
Proofs, then, must be devised to enable him to make these distinctions.
(that is, his personality and career in relationship to a particular milieu), the investi-
The problem of the idiosyncratic is minimized when the individual chosen
gator is able to cut into a complex set of interacting variables-to remove and for study is neither a psychotic nor a creative g e n i u e a charismatic leader who
empirically examine a slice of culture (see, for example, Dollard, 1935, pi 4; Aberle, breaks old molds, It is easier to generalize from the experience of a moderately
1961; Holt, 1962). successful leader who was widely esteemed as embodying the virtues of his day
My work on Charles Evans Hughes (Glad, 1966) is a case in point. American (Trueblood, 1939). As Lovejoy (1936, pp. 19-20) has pointed out, it is on these
foreign policy failures since World War I have been attributed to the "American sensitive souls of less than the greatest creative power that contemporary ideals
approach"-that is, to the tendency to view concrete problems through a moral- record themselves with clarity.
legalistic framework, which has often distorted reality and courted political failure. We may assume that an individual is an exemplar of his culture or subculture
Impressionistic surveys drawn from many sources might have been used to substan- if we can show that he has the following qualifications: (1) He is performing with
tiate this charge. Scientific inquiry, however, requires a more systematic approach to wide acceptance in a high-status role to which he has had routine or easy access,
data; yet the topic must be kept sufficiently narrow to remain within the grasp of (2) This success was obtained without great psychic costs to himself; in other words,
the researcher. Psychobiography is one such approach. A study of the "American he is an adjusted personality within his culture. (3) His basic values are also mani-
mind" through Hughes-the major foreign policy decision maker of his day and a fest in key institutions which he has encountered in the socialization process. (4) He
representative of the officialculture-provided the limits necessary for a specification has been rewarded from many different sources on the basis of these values. (5) His
of the subject matter, as well as a proper subject for the inquiry. After establishing associates (teachers, colleagues, followers) hold similar values and also fit easily into
through an analysis of Hughes' personality and career that he was a man of his times, the culture, performing in influential roles and being rewarded by others in the
I was able to explore the complex connections between his policies and thought as a culture. These constitute proofs that the individual has been successfully integrated
key to the nature of these connections within the broader American culture. into his culture or subculture and that his personality and career embody at least
Psychobiography is particularly useful in formulating hypotheses about the some significant aspects of the culture.
interactions between social structure and psychic mechanisms. These hypotheses can The individual or subcultural idiosyncrasies that such a person may maintain
then be validated by more rigorous forms of investigation. But it is also possible to can be empirically explored through comparative biographical studies. Thus, the
generalize to broader political phenomena from the experiences of a single individual. researcher can contrast one subject's personality and career with others who have
The general guidelines for the appropriate use of the case-study method in science had different social experiences and perform different social functions. Trait clusters
need not be repeated here (see, for example, Campbell and Stanley, 1963; Horst, can then be delineated-some overlapping, some discrete, some characteristic of men
1955; Becker, 1968). The particular problems encountered in generalizing from the in all cultures, some shared by the individuals of one broad culture, some specific to
310 Betty Glad Contributions of Psychobiography 311
particular class, ethnic, regional or religious groupings, and a few, perhaps, altogether My work may be used to illustrate these considerations. As Secretary of State,
unique. Charles Evans Hughes (Glad, 1966) met both his personal needs and the expectations
of significant others (for instance, the president, congressional leaders, subordinates
Specific Contributions in the department, leading journalists, and foreign ambassadors). He was not able,
however, to direct American foreign policy along the substantially new paths sug-
At this point, it might be useful to outline specific areas where the psycho- gested by the radical changes in the international situation of the United States after
biographic method has been employed in political science and to suggest possible World War I. A successful leader in terms of traditional values and definitions of
contributions in the future. The following areas of investigation will be discussed: his office, he was not a "hero of history," not a man capable of manipulating the
role performance and recruitment; the nature and cqnsequences of political attitudes; political situation to produce creative solutions to new political problems.
the relationship of attitudes to background characteristics and socialization processes; Borah and Fulbright (Glad, 1969), unlike Hughes, grappled with the bound-
the psychic foundations, in general terms, of political ties. aries of their offices (as chairmen of the Senate Foreign Relations Committee),
Psychobiography has been most fruitful in the analysis of role-performance employing new strategies and ideas in order to change policy outputs to meet (in
characteristics, Barber (1969, 1972a, 1972b), for example, has shown that subtle their terms) the requirements of a changing situation. Both men, with quite different
differences in aspects of presidential role-performance style (types of rhetoric used, personality structures, exhibited a certain independence from the opinions of others.
ways of relating to significant others, techniques of decision management) can be But Borah's independence was based on an emotional detachment from others and
related to underlying personality characteristics. Wedge (1968) has done a similar an ability to externalize conflicts. Insensitive to the reactions of those about him,
analysis for Khrushchev, as has Mazlish (1972) for Nixon. with no conscious doubt about the correctness of his position, he was able to integrate
Furthermore, overall performance may be evaluated in terms of the fit his energies around one core device-political attack justified as moral superiority.
between role expectations, individual style, and personality. Edinger (1964) has And this ability to stand alone on strategic and ideological matters-untroubled by
explored some of these relationships. There are leaders, he notes, whose personality moral doubt and seemingly unaware of external political complexities-was the key
characteristics drive them to behavior that is dysfunctional for their office. Driven by to his success. Fulbright, on the other hand, appreciates the ambiguities of the politi-
unconscious goals or the desire for immediate gratification, they unconsciously invite cal situation and has a measure of sensitivity to the opinions of others. Unlike Borah,
defeat. Woodrow Wilson and Kurt Schumacher, Edinger suggests, were two such he does have doubt and ambivalence, which may slant his blows and deter him from
individuals. Adaptive role performance, he continues, requires other personality the maneuvers that might give him more immediate impact. Yet his role as dissenter
characteristics: "Men like Franklin Roosevelt and Konrad Adenauer . . have in . has been less compulsive than was Borah's and is based more on his perceptions of
common a keen sense of 'reality' which enables them to adjust their behavior to the changing political realities than on inner compulsion. In other words, Borah's style
expectations of their salient counterplayers and thus to acquire and exercise politicaI seems to have been based on a somewhat rigid ego structure and a need to dissent,
leadership. Their personality characteristics apparently permit sufficient outward while Fulbright's seems to have its source in a relatively flexible and reality-testing
orientation to satisfy their personal needs as well as the expectations of those who ego.
could facilitate or block their desire for political leadership" (p. 655). Aside from suggesting interesting connections between personality and role
Edinger also recognizes another possibility-one in which the role player meets performance, studies such as the above are helpful in developing theories about
the expectations of others and at the same time influences the very definitions of his leadership. My study of Borah and Fulbright, for example, suggests that stress situa-
office and its outputs. The great leader, he says, handles some of his own inner tions can elicit innovative leadership from at least two different personality types.
problems by adapting to the expectations of others and by changing his surroundings. The analysis also supports Hollander's (1958) thesis that any leader has a limited -
"Striving to find his own identity, he strikes out against his environment and succeeds idiosyncratic credit against which he can trade. Insofar as he performs well and
in stimulating a satisfying response in the public arena because he is the right man at meets expectations in certain areas, he gains a kind of credit, a n ability to deviate
the right timeJJ (p. 666). elsewhere. But if he deviates from role expectations too often, he may lose the sup-
The various possible responses of a political leader to his office may be char- port of others in his political environment.
acterized as adaptive, maladaptive, or innovative (Glad, 1969). In the adaptive Psychobiographic sketches also have been useful in dealing with the selective
response, the officeholder successfully meets the work, self-control, and other require- recruitment of personality types for specific political offices. Analyses of the careers
ments associated with the office; in the maladaptive performance he falls short of of Woodrow Wilson and Anton Cermak, for example, support LasswelI's (1930,
these expectations. In both of these responses, the traditional conception of the office 1948) hypothesis that politics is likely to attract power seekers who are compensating
and of its proper political output remains intact. In an innovative adaptation, how- for underlying feelings of weakness. Wolfenstein's (1967) biographies of revolu-
ever, the officeholder is sufficiently sehsitive to the expectations of others to maintain tionary leaders suggest that the choice of a revolutionary role is related to psycho-
his following, while simultaneously employing new strategies and ideas to change logical traits. Repressed aggressive feelings toward the father may find a suitable
the output of the office, His purpose may be to articulate and integrate new demands target in the repressive authoritarian regime which permits the individual to "exter-
into the political system or to expand his own power. nalize his feelings of hatred." Other work could be done along these lines. For
 Betty Glad Contributions of Wychobiography

example, one could compare and contrast the personalities of bureaucrats with party (1960) pioneering studies based on his concentration camp experiences (see Beradt,
leaders, diplomats with generals, crisis with noncrisis leaders, or middle- with high- 1968; Dicks, 1972). And the nature of panic behavior could be further explored
level political managers (see Smelser and Smelser, 1965, pp. 255-322). through the analysis of individuals who have faced sudden and severe public crises
Psychobiography has also been used to explore the relationship of personality (Stanton and Perry, 1951; Langer, 1958; Lifton, 1968).
to political attitudes. Through in-depth analyses of ten men, Smith, Bruner, and
White (1956) showed that the relationship between opinions and personality is a Data Collection and Analysis
very complex one. Support for or opposition to the publicly formulated alternatives
on current issues appeared "in the guise of final common paths"; that is, individuals In addition to the traditional problems in obtaining data, the psychobiogra-
could arrive at common policy points from diverse private routes. Intellectual and pher faces special difficulties. As one historian has noted, "It is almost impossible to
temperamental traits were related (Smith, 1958), however, to politically relevant reconstruct his subject's inner life and at the same time remain true to the frag-
items such as cognitive style. Personality seemed to be most clearly related to mentary evidence available" (Duberman, 1967; see also Kiell, 1964, pp. 14-2 1) .
differences in concreteness or abstractness of thinking. "More extensive appraisals The psychobiographer, ordinarily, cannot verify his interpretations by giving his
can be envisaged .. . of the degree to which the attitudes prevalent on the topic subject projective tests or scientifically designed interviews, nor can he get his subject
in a defined population serve one or another of our three functions [the management on the couch or place him in small-group experiments. If the subject is alive and in
of emotional conflicts, social adjustment, and reality testing]. Such partial but ex- power, he will probably be less than candid in interviews, and he is not likely to
tensive inquiries into opinion dynamics look particularly promising as avenues to the consent to any serious probes into his psyche or his political strategies. And if he is
understanding and perhaps prediction of shifts in public opinion in response to events dead, as Rogow (1968) points out, he has literally taken "his dreams and fantasies,
and to different styles of leadership or manipulation" (p. 16). his Oedipus complex and identity crisis, with him." If he has been a major public
Personality and ideology may also be related to background characteristics. figure, he has also taken care to erase the traces of his secret life. Politicians, as
Thus, Levy (1948) found that anti-Nazi Germans, as contrasted to typical Germans, Rogow states it, "are a notoriously secretive lot, and any inkblots they had inad-
had escaped rigid and conventional family structures. Lane (1962), in his biographies vertently left behind are usually wiped clean by their loyal posterity."
of fifteen working-class men in "Eastport," attempts to explain the sources of the Yet one can gather from the traditional biographical sources sufficient informa-
ideology of the "common manJyin terms of his culture and experiences. Furthermore, tion on an individual to warrant a psychological explanation. Memoirs, letters,
much of the current work on political socialization (for instance, Greenstein, 1965) speeches, artistic productions, public documents, newspapers, interviews with asso-
could be tested and extended via in-depth biographies. ciates, and oral histories can all provide clues to the subject's personality which the
Psychobiography's greatest potential may reside in the testing of several astute observer can then piece together into an explanatory framework (Allport,
theories that deal with the emotional bases of the political bonds between people and 1942).
a delineation of the meaning of "political culture." The character of the charismatic Aside from the usual handling of these data sources, the psychobiographer
tie has been studied through analyses of charismatic leaders and their supporters or might find it useful (see Edinger, 1965) to apply such tools as content and factor
subjects (Gilbert, 1950; Tucker, 1965; Willner, 1968, 1969; Stark, 1968, 1969). analysis to speeches, letters, and other personal documents. Through the usage of
Sereno's (1962) views on the leader-follower symbiosis and his skepticism regarding such techniques, a subject's basic ideology may be more clearly revealed. An inter-
the very possibility of decision making and myth making by an elite could be investi- action analysis of the subject's immediate entourage may also yield a new perspective
gated through analysis of the relationship of a "hero of history" to selected indi- on his personality and leadership qualities. Furthermore, the biographer might choose
viduals in his life. Perry's theories (1957) of national identification might be explored to observe P. contemporary subject in action by observing his behavior at press con-
through comparative biographies of traitors and patriots. (See Chapter Ten in this ferences, on the platform in public hearings, and in legislative bodies. Through
volume.) The psychological bases of the perception of and response to an enemy direct observation, the biographer can document, for example, the subject's gestures,
could be studied through the comparative analysis of parochial and cosmopolitan his speech habits and bearing, and his responses to stress. On occasion, researchers also
leaders (cf. Holsti, 1969). The impact of "national character" on policy making has have been able to administer projective tests to their subjects (Lasswell, 1930;
been investigated through the analysis of "typical" policy makers (Glad, 1966) ; and Smith, Bruner, and White, 1956; Gilbert, 1950), although influentials are not likely
the dynamics and political consequences of cultural lag have been investigated to submit to these procedures. In some instances the subject's artistic productions-
through the analysis of individuals socialized in one historical period and exercising his writings, paintings, compositions-may be used in lieu of the foregoing (see, for
political responsibilities in another (Glad, 1966). instance, Freud, 1910; Mumford, 1934; Allport, 1942; Edel, 1961) .And though seldom
There is also much to be learned through an analysis of the behavior of indi- available, medical and psychiatric records may be used. Rogow (1963) and Weinstein
viduals in extreme situations. Such an examination helps to define the conditions of (1967) have shown how such documents can be used for interpreting the inner life
behavior so ordinary as to be assumed "natural" (for instance, compliance behavior of a subject.
or passivity in extreme situations). Thus, the efficacy of fear and torture in de- Armed with personality theory, the biographer wiil find significance in data
moralizing and ruling a subject population might be extended beyond Bettelheim's that the traditionalist will often ignore as too private, too personal, or simply ir-
314 Betty Glad Contributions of Psychobiography 315

relevant. Thus, Cermak's colitis became central to Gottfried's (1962) interpretation and direct observation. Thus, I followed an analysis of relevant published materials
of Cermak's personality. In my work on Hughes (Glad, 1966), several "unimportant" (interviews, biographies, the Congressional Record, newspaper stories, speeches,
items proved crucial in interpreting his personality: for example, his attitudes toward committee hearings) with my own observations of Fulbright's public appearances (in
an old schoolmate in financial need, his devotion to Swoboda's exercise regimen, his the summer of 1969) and interviews with Fulbright and significant others in his
ordering of clothes from Brooks Brothers, his desire and ability to take mountain- polftical environment.
climbing vacations by himself, and his defensive response to A. Lawrence Lowell There are problems in this approach, of course. When the processes of data
and George Wickersham when they attacked his reversal on the United States' entry collection and interpretation cannot be systemized in accord with specific criteria and
into the League of Nations. Similarly, certain details in Fulbright's public presenta- quantitative measures, the observer's own characteristics will influence the data he
tions during the summer of 1969-for instance, his handling of his own mistakes, gets. This is evident in both interview and direct-observation techniques. The status,
including lapses in memory, and his misuse of psychological terms, which showed a age, and sex of the interviewer will affect his abiIity to gain access to influential
relative lack of defensiveness-were useful to me in interpreting his personality and persons, as well as their willingness to be candid with him (Dexter, 1970, ch. 6 ;
role-performance style (Glad, 1969) .8 Gorden, 1969, pp. 127-137). The kinds of data obtained will also depend on how
"Much of this may appear to be 'nit picking,' " as Rogow ( 1968) has pointed the interview is conducted. The old notion that the interviewer should be as un-
out, "but it is well to keep in mind that it is the nits in psychoanalysis, the seemingly obtrusive as possible can be questioned (Dexter, 1970, pp. 24-29). If he plays a
trivial episodes to which, at the time, little importance may be attached, that often neutral role, he may get only neutral, guarded responses. For example, in my inter-
supply the ~rincipalclues to personality development." views with decision makers for the Borah-Fulbright project (Glad, 1969), I sometimes
In collecting and analyzing biographically relevant data, some researchers drew routine and uninformative responses. In these circumstances, I occasionally
have emphasized the need for objective measures and quantitative techniques (Bald- asked provocative questions or suggested interpretations in order to evoke unguarded
win, 1942; Dollard and Mowrer, 1947; Holsti, 1969) . In terms of scientific rigor, responses. This tactic often yielded new information-bringing forth anecdotes and
the approach has much to recommend it, but as a research technique it is likely to attitudes which had been absent earlier in the conversation. Similarly, in direct-
have a limited use in psychobiography. In covering an entire life, the biographer is observation situations, the observer's own personal and intellectual characteristics are
required to draw from a wide variety of sources; intuitively, he may discover that relevant to what he sees and hears. The ability to perceive slips of the tongue and
some are more important than others (Garraty, 1957, p. 219; Molt, 1961). In the changes in voice tone or speech patterns, for instance, is based on a capacity for
Hughes study, for example, themes apparently unimportant by quantitative measure sensitive perception and an understanding of the implications of nuances.
(for example, the number of references to "reason"in his speeches) were shown- 1
Furthermore, any biographer 1s bound to have an attitude toward his subject
through an investigation of the teachings of his parents, church, and school author- --often positive but sometimes not (Edel, 1961; Erikson, 1968; Roazen, 1968). The
ities-to be the keystone of his ideological system. This finding could not have been psychoanalytic profession has long been concerned with the negative effect which
determined by any formula for collecting and handling the data announced at the biased practitioners might have on their subject, developing procedures for its mini-
beginning of the project; it was based, instead, on an interplay between theoretically mization. Recognizing this problem, some psychoanalysts have suggested that only
based "hunches" and flexibility in the choice of trails leading to data. Furthermore, experts in their profession are qualified to engage in p~~chobiographic studies (Hitsch-
in comparative biography one may have to rely on data types that are not compar- mann, 1956).
able in quantitative terms. Any available source must be used when ofie is trying to This proscription, however, is founded on a basic confusion. 'The psycho-
capture relevant but slippery facts about diverse individuals, who may have left quite biographer is not. engaged in therapy (Edel, 1961). Though there may be aspects of
different tracking behind them. countertransference in his perceptions of his subject, an intimate relationship is not
This concern for getting the data wherever it might be found was the basis of established with the subject, making the usual procedures for controlling transference
my decision to approach Borah and Fulbright in different ways (Glad, 1969). In and countertransference somewhat irrelevant. Rather, the psychobiographer is en-
Borah's case, I was able to obtain relevant data from his letters and scrapbooks in gaged in historical and scientific endeavors; and if the usual scientific methods are
the Library of Congress and from certain interviews in the Oral History Collection followed, interpretations about personality and role performance will be subject to the
at Columbia University. Though this kind of material was not available for the usual tests that govern social and political science (Garraty, 1957; Edel, 1961).
study of Fulbright, he and others in his environment were available for interviews Because the psychobiographer may attend to some unusual and varied data
forms and because his personality may influence what he can get and how he inter-
3 Boder (1940), Baldwin (1942), and Dollard and Mowrer (1947) have shown that
quantitative techniques can be used to make interpretations from such item's as theme se- prets it, it is particularly important that he follow traditional rules for data collection
quences or adjective-verb usages. For general discussions of the relevance of minor details to and management. In interpreting data which is often fragmentary and sometimes
personality interpretation, see Mumford (1943), Guedalla (1939)' Garraty (1957, pp. 113- internally contradictory, he is well advised to follow guidelines such as those sug-
146), Bushman ( 1966); cf. DeVoto (1933) and Boyd (1932). Boyd, in criticizing what he gested by Allport (1942) in his "Use of Personal Documents in Psychological
considers to be an overemphasis on sex in psychobiographical works, quotes George Bernard Science." (See also Malone, 1943; Gottschalk, Kluckhohn, and Angel1 [n.d.]; Garraty,
Shaw to prove how unimportant the matter is. This is an attitude that any psychobiographer
would explore further in a study of Shaw. 1957, ch. 8.) Equally important is the requirement that his sources be given, so that
 316 Betty Glad Contributions of Psychobiography 317

the credibility of his sources and his relative objectivity in employing them can be Aside from these practical problems, the biographer may be restrained by his
~hecked.~ own moral and aesthetic standards from revealing all the private peccadillos of his
Some psychobiographers have not met these standards, showing an almost subject. Dumas Malone (1943) experienced such problems in his work on Thomas
cavalier attitude toward facts (as noted by Garraty, 1957, pp. 128-135; and Langer, Jefferson. In the course of his research, he came across certain "lurid" stories about
1958). The Freud-Bullitt (1967) book on Wilson is a very good example of what not Jefferson's personal life, most of them emanating from an editor who had turned
to do in this regard. The details of authorship are unclear; indeed, Erikson (1967; upon his original benefactor. Jefferson, moreover, had admitted the truth of one
see also Hofstadter, 1967) finds it hard to believe, on stylistic grounds alone, that such story. What should the biographer do? Malone dealt with the problem by trying
Freud could have written almost any portion of the work as published. Bullitt's to depict the situation as honestly as possible, while at the same time not accentuating
informants were anonymous; and, in addition, statements are made which no it. "Much depends on the length of the biography I am writing, but at all events I
biographer could possibly substantiate-for example, that 'little Tommy" never had must try to depict the man in the proportions which I have perceived." Similar stan-
a fist fight in his life. Freud has been criticized also for his reconstructions of Leo- dards, Malone also points out, were used in dealing with potential entries into the
nardo da Vinci's life (Garraty, 1957, p. 114), as has Phyllis Greenacre (Wyatt, Dictionary of American Biography, each case being judged on its merits. Insofar as
1956) for her work on Jonathan Swift and Lewis Carroll. Greenacre, in particular, there was any rule at all, Malone contends (pp. 135-136), "it was that dereliction in
shows an open disregard for the usual tests in authenticating her "facts," using the matters of public concern was more important than private morality."
psychoanalytic framework to originate data. Childhood wants, she argues (1955, p. In dealing with these problems, Guedalla (1939, pp. 932-933) suggests that
107), can be "reconstructed from known characteristics, problems, and repetitive the standards of good taste are primary and that these standards vary with the sub-
. actions supported by memory traces." Indeed, "the experienced psychoanalyst knows ject.
just as definitely as the internist observing later sequellae of tuberculosis . . that.
the deformity is the result of specific acts upon the growing organism." That which is relevant to the portrait you are trying to paint, even if it
This kind of retroactive reasoning, which gives up any independent verifica- .
is grotesque . . is not vulgar. If it is irrelevant, it is unpardonable
tion of data, ignores certain important differences between therapy and history. In the comic relief of a music-hall order, and then you are behaving in an in-
therapeutic situation, interpretations are derived from the free association of the
.
ferior fashion. . . Let me give an example. If what you are trying to
write is the life of a lyric poet, his emotional life is of the very core and
subject and verified by his concurrence in them and by his subsequent emotional center of your theme. In that case it is material . . . which young lady,
growth as he comes to accept them. When this form of verification is given up, as .
seated upon which stile . . was the first young lady. If, however, you
it must be outside the clinical situation, the only checks left on the inquiry are those are writing the life of a governor of the Bank of England . . such .
used, generally, in the historical and social sciences. matters as these are irrelevant and unpardonable comic relief; and telling
Another kind of problem arises when one is dealing with sensitive topics, as us about them, that is no proper part of your business.
psychobiographers frequently do. Access to sources is often granted on the condition
that certain kinds of information not be used. Garraty (1957, p. 171) cites several This distinction between men in different roles is too simple a distinction for
instances where prominent biographers were confronted with this kicd of problem. the psychobiographer. Even a governor of the Bank of England has a personality
Robert T. Lincoln, for example, withheld his father's papers from a series of biog- which influences his role performance. And insight into his personality may be gained
raphers, including Ida Tarbell and Senator Beveridge. A single exception was made by going into matters such as his love life. Affairs of the heart should be given their
for his father's secretaries, John Hay and John G. Nicolay, but even here he exercised proper proportion, it is true; but if their relevance to significant aspects of personality
"plenary blue-pencil power'' over sections of the work in no way related to the is obvious, they should be included in a biography (cf. Boyd, 1932).
Lincoln papers. The italicized sections in the quotation below represent one such Certain special access, coverage, and documentation problems in interview
deletion at Robert Lincoln's request: "Thomas, to whom were reserved the honors techniques might be discussed at this point. The grapevine approach is useful in
of an illustrious paternity, appears never to have done anything else especially procuring access to interviewees. In my work on Fulbright and Borah as chairmen of
deserving of mention. He was an idhe, roving, ineficient, good-natured man, as the the Senate Foreign Relations Committee, for example, the initial interviews in
son of a wid0.w is apt to be according to the Spanish proverb. H e had nv vices so Washington were arranged through two key Senate aides, and from those initial
far as we can learn but he also had no virtues to speak of. He learned the trade of interviews over fifty others were arranged. Approximately 80 percent of the indi-
a carpenter but accomplished little o f it. He was an easy-going person." viduals contacted in this manner responded favorably to a request for an interview
(see also Dexter, 1970, pp. 28-36).
4To check against possible bias in case studies Paige (1968) has suggested that A questionnaire is of value in conducting interviews because it provides for
analyses be kept separate from data presentation, the language in the latter being "decon- the efficient and systematic coverage of the items to be discussed and ensures some
taminated" through the deletion from the account of words rich with connotations. Aside kind of comparability in the data collected. However, questionnaires also have dis-
from a possible scientific loss from such usage (the loss, for example, of "subtler" data), there
are obvious aesthetic costs. Most biographers think it important to evoke moods through their advantages. If rigidly adhered to, they are likely to lead the discussion into topics
selection of details and choice of words. See Fuess ( 1933), Malone (1943), Hughes (1964), about which the interviewee is either ill informed or unwilling to talk, thereby leav-
Garraty (1957, ch. lo), and Kennan ( 1959). ing his special knowledge and possible contributions untapped. Furthermore, a set
 Betty Glad Contributions of Psychobiography 319

format may create a stilted atmosphere in which sensitive information is not likely better suited to studies where little is known about the dreams, fantasies, or childhood
to flow (see Gorden, 1969, pp. 51-57). experiences of the subject. It is somewhat easier to describe character structure, with
In attempting to resolve these problems in my Borah-Fulbright project, I used its reality testing and ego-defending mechanisms, than to attempt an analysis of un-
open-ended questions concerning the work, style, and other role-performance char- conscious material (see, for instance, Glad, 1966; cf. Bushman, 1966). Relatively
acteristics of my subjects. T o promote spontaneity and candor in the subjects' re- traditional sources-letters, statements by coworkers, public speeches-all can be used
sponses, I memorized these questions before each interview and brought them up in to show how the subject manages power, handles his work, and relates to others.
a conversational manner at an appropriate time in the interview. Furthermore, ad- Analysis of the subject's speeches is particularly useful in this context, since it reveals
justments were made for the special interests and knowledge of each interviewee. his tacit assumptions about reality and the ideological prisms through which he views
One staff member of the Senate Foreign Relations Committee, for example, whom I the world. The extent to which ideological structures are deeply rooted in the per-
had asked about Fulbright, turned out to have anecdotes to tell about Borah and sonality system has been pointed out by Schilder (1936), who has made "ideological
Key Pittman, which proved most useful. analysis" a part of his therapeutic work. Interviews also can produce information
There is also an inherent conflict between gaining information and document- about the public aspects of character structure. More sensitive questions about the
ing sources in interviews, and the weighting of these concerns will influence how one subject's relationship to his mother or wife or about his sex life are Iikely to yield
makes and keeps interview records. Tapes have obvious advantages, since the diver- .defensive and unproductive responses rather than useful information.
sion of note taking is avoided and documentation is nearly perfect. Because of these Although there is some data loss in this kind of analysis-the etiology of the
advantages, my original intent in the Borah-Fulbright project was to tape the inter- psychological mechanisms observed and their connection with specific fantasies and
views and deposit them in the Oral History collection at Columbia university, tg be emotions-this loss does not present a problem of major proportions. The ego-defense
opened under conditions specified by the interviewee. I soon discovered, however, or ego-integrative aspects of behavior, as observable in adult functioning, are the
that most interviewees felt inhibited by the tape and preferred not to use it. I t immediate cause, in personality terms, of adult political behavior.
seemed, then, that I had to take notes. To retain spontaneity and information flow Maslow's (1954) description of the self-actualizing person may be useful in
in the interviews, however, I decided not to take notes during an interview. Instead, explaining those who manifest in their careers some form of creative accomplish-
each exchange was reconstructed from memory immediately after the interview ment, including unusual leadership ability, Rather than veiling the great leader in
through dictation into a tape machine (cf. Garraty, 1957, pp. 208-210). the hero-of-history mystique, these theories provide explanatory frameworks, per-
The information was often given to me with the understanding that it was not spectives from which political performance can be analyzed in terms of its style and
for attribution, and in some instances the material was "privileged." T o provide a situational components (cf. Holt, 1961). Thus, the capacity for innovation may be
means of documenting these sources, I devised the following compromise. Attribu- related to a certain kind of psychic development (Maslow, 1954; Kubie, 1961), while
tions in the final paper were made not by name but by code letters which indicated the ability to create in a political context may require, in addition, a special way of
in general terms the role and the political inclination of each individual. These relating to people (Edinger, 1964). The innovative leader must have a sensitivity
code letters also referred to the typescripts, now in the University of Illinois library, to the expectations of others in order to maintain a following, and an ability to stand
which gave all the details of the interview. alone in order to try new methods or see new solutions to problems. The essential
features of a leader such as Abraham Lincoln are more likely to be caught in net-
Interpretations works such as these, rather than in a worm's-eye view (such as that of Clark, 1933)
that emphasizes his infantile fixations, ego-defensive maneuvers, and his "immature"
When approaching a biographical study, the researcher may choose from a religious beliefs.
number of psychological frameworks (Jones, 1932). The choice should be governed In addition to personality theory, the biographer may also use conceptual
by the nature of the available research materials, the purpose of the study, and the models from sociology dealing with such matters as the socialization process, role-
feasibility of validating the conclusions (see Trueblood, 1939). performance expectations, and organization theory. The breadth and depth of the
If information about the subject's childhood is available, the Freudian prin- analysis, as well as the choice of research techniques, will depend on the researcher's
ciples might be employed. Diaries or other autobiographical reports, information primary concern. When the objective is to relate a specific political characteristic to
from the subject's parents and others about their educational techniques, psycho- key emotional conflicts or to ego-defensive maneuvers, it may be best to focus on
analytic or medical records where available-all lend themselves to this kind of select and more obviously relevant aspects of the careers and personalities of several
analysis. When these reports contain information about the dream and fantasy life of key individuals. On the other hand, if the objective is to weigh the emotional, ideo-
the subject, the biographer is doubly fortunate. For contemporary subjects, published logical, and situational concomitants of an individual's role performance, then a,
sources may be supplemented with interviews, and in some rare instances projective broader analysis of his life and self is called for. A biographer can use the entire con-
techniques may be used (see Chapter Fifteen in this volume). Gilbert (1950), for ceptual apparatus of the social sciences to relate his subject's political position to
instance, was able to use projective tests on certain Nazi leaders, because they had the subject's personality, physical and intellectual makeup, and current and past
lost a war, were in prison, and were required to accept Gilbert as their psychologist. environmental influences.
The neo-Freudian frameworks of Erik Erikson and Karen Horney may be In choosing conceptual frameworks, the biographer should be aware of certain
320 Betty Glad Contributions of Psychobiography 321

problems in timing. An early decision to employ a particular theoretical system can 01 modal and rational (adaptive, reality-oriented) personality structures on the
provide focus and a kind of efficiency in information gathering; it may help the political process and to do it in a way that is complementary rather than competitive
researcher more quickly to separate significant from insignificant data. If a concep- with sociological and political explanations.
tual framework is chosen prematurely, however, the investigator may try to force As a form, psychobiography has scientific merit. Not only is it likely to provide
the data into inappropriate categories. As a check against this, the biographer must a deep and systematic study of personality; it also permits a holistic approach to the
remain flexible enough to be able to "break sety' should the data so require (True- personality and politics field and the building of political generalizations. Through it,
blood, 1939; Guedalla, 1939). (For a criticism of Freudian rigidity along these lines, insights can be gained into such phenomena as the relationship of personality to
see Wyatt, 1956; Schmidl, 1962; Hughes, 1964, pp. 44-53.) background characteristics, role-performance characteristics, political attitudes and
For example, I initially perceived of Charles Evans Hughes as a political perceptions, and patterns of behavior in panic and crisis situations.
realist. In rejecting the principles of collective security and in emphasizing the Though the psychobiographer cannot eliminate elements of subjectivity in
primacy of American national interests, he seemed to fit Hans Morgenthau's model his work, certain restraints and principles of validation should govern his approach.
very well. Given this perception of him, I was puzzled by his dogged commitment Depth psychology, for example, may direct him to details, many of them quite elusive,
to projects such as the World Court and the Pan-American Arbitration treaties. which are ignored in traditional studies; nevertheless, he must be guided by the
Then I began to notice the references in his speeches to the rule of reason and to usual tests for documentation followed in history and the social sciences. The sources
perceive his tacit assumption that reason is in some way related to evolutionary of his data must be given, so that his facts are susceptible to an outside check; and
change. I t was after I had gone back to the writings of his father and his favorite his explanatory framework should be explicitly presented, so that it may be judged
teachers that I began to see connections. Only then did Hughes's basic tenet-that by other scholars in terms of its ability to handle the data in an economic and
only evolutionary political change is rational-become clear. A commitment to internally consistent way that accords with the broad body of social science theory
collective security, then, with its assumption of a radical reordering of human and data.
institutions and values, would be irrational. Modest change, as reflected in proposals
for the establishment of the World Court, however, could be supported as a develop-
ment in accord with reason and guaranteed by history.
The worth of the interpretative framework used in a psychobiography can be
judged by its efficiency and economy in ordering a wide range of significant data as
well as by its prior scientific standing. In other words, the interpretative framework
must explain the given details in an economical, consistent, and meaningful way
that accords' with depth psychology as well as with the general body of social science
data and theory. (See Knutson, 1972.) This test, which Farber (1950) calls plausi-
bility, is described as follows : "For a scientifically trained person plausibility probably
involves some swift mental manipulations which attempt to fit the hypothesis into
accepted frames of reference, thus roughly testing it, and finding that it does violence
to none of these frames and indeed fits some fairly well." But there is another factor
making for plausibility. " T h a t is, that a single concept appears successfully t o sub-
sume a number of discrete social phenomena, or at least to interrelate them.)'
In this sense, then, the evaluation of psychobiographic interpretation differs
in no substantial way from those for any broader-gauge theory in political science.
The psychobiographer is simply borrowing an explanatory framework from another
field in order to explore causal factors which have been hitherto slighted but are
of considerable importance to the analysis of political behavior.

Conclusion

In disposing of certain key assumptions-common to psychobiographers and

their critics alike-psychobiography will have a greater utility for political science.
Revising his assumptions to take full account of contemporary ego psychology and
scientific philosophy, the psychobiographer is able to go beyond the usual concern
with the idiosyncratic, the pathological, and the therapeutic to explore the impact
 Surveys in the Study of Political Psychology 323
An essay by Katz, in the same volume with McClosky, reduces this ambiguity.
According to Katz (1967, p. 149), "survey and field research can contribute heavily
to the development of a more adequate psychology." Furthermore, "a naturalistic
description of the behavior of human beings in various social contexts is a prerequisite
for the understanding of man" and is readily provided by survey but not by labora-
tory methods. Katz's argument is persuasive and documented by many examples that
enlarge and elevate psychological thought. The title of his essay indicates that Katz
is describing the "potential of survey methods in psychological research" as well as
"the practice" of surveys in current psychology. These methods may not be omni-
present or omnipotent as yet, but the paper certainly fills one with expectations of
the better psychology that could be founded upon the creative use of survey methods.
Evidence from so judicious a psychologist as Katz cannot be construed as
special pleading. However one might define the particular field of "political psy-

STUDY OF POLSCSPICAL chology," it surely is located at the intersection of political science and psychology.
Thus, the McClosky and Katz essays taken jointly suggest how powerful survey
methods can be for political psychology specifically, and we shall draw freely on their
many ideas in this paper.

PSYCHOLOGY The skeptic may still reserve judgment. He might argue that McClosky and
Katz, as contributors to a volume on survey research in the social sciences, necessarily
presented the method in too prominent a light. Then he should weigh the evidence
in David Sears' chapter on political behavior in the Handbook of Social Psychology
(1969). That assignment did not compel any special attention to survey research.
Herbert H. Hyman The essay is a substantive presentation of findings on electoral and public opinion
processes treated from a social-psychological perspective. Yet Sears remarks, "Almost
all the studies cited are suruey studies rather than experimental studies" (p. 316,
italics added). The past accomplishments of the method, its proven power to illumi-
nate the field, compelled his attention. A sense of the historical growth of survey
method and its intimate relation to the development of political psychology is pro-
vided by comparisons of the three editions of the Handbook of Social Psychology,
spanning more than a thirty-year period. The 1954 edition contained a chapter,
"The Psychology of Voting: An Analysis of Political Behavior" (Lipset and others,
O m n i p r e s e n t and omnipotent, if it were Vol. 2, pp. 1124-1175), which also leans heavily, but not exclusively, on evidence
not idolatrous, might be an apt description of the survey method in political psy- from surveys. The first edition, published in 1935 (K. C. Murchison, ed.), con-
chology. McClosky (1967b, p. 65) hails it as "the most important procedure in the tained no chapter whatsoever on "political behavior" although i t did contain a classic
'behavioral' study of politics" in a review of survey research in political science. His essay on attitudes by Gordon Allport. In part the discipline was oriented then toward
reference to "the handful of large-scale voting surveys that have in three decades different substantive problems, but any attempted chapter would have had little
taught us more about the act of voting than was learned in all previous history" empirical evidence to draw upon, since the rise of the modern sample survey (in the
(p. 71) surely gives a sense of the method's awesome powers. Iate 1930s) postdates that edition.
McClosky is but a single judge, perhaps prejudiced in our favor; but a careful One should ponder the next passage in Sears' preface: "Experimental research
reading of his decision will show how well documented it is and how truly versatile on political behavior is quite feasible, and has been attempted on occasion. I t is to
survey research can be in illuminating the problems of modern political science. Yet, be hoped that those occasions will become more frequent" (p. 316). A chorus of
to be precise, he is writing about the "behavioral study of politics,)' not about "poli- readers may well chant: A paradox, a paradox, a most ingenious paradox. If experi-
tical psychology." Some subtle shades of meaning may attach to these vague, un- mental research on political behavior is "quite feasible," why has it been so Gcca-
defined terms, and perhaps surveys are not so all-powerful in whatever special realm sionaI in the past? Perhaps the mystery is easily solved. Perhaps it is so far from
is political psychology. feasible in its classic laboratory form that most scholars do not attempt it, and those
that do often produce something trivial or grotesque. A reading of the essay by Katz
(1967) will suggest the inherent obstacles to a laboratory-based and experimentally
324 Herbert H. Hyman Surveys in the Study of Political Psychology 325

oriented social psychology. His analysis is equally applicable to a political psychology. population or universe under study with a minimum of error. Although the universe
In his words, "Many psychological problems do not lend themselves to laboratory can be a specialized, relatively homogeneous, and geographically concentrated popu-
manipulation, either because of the time dimension, the complexity of variables, or lation, often it is a widely dispersed, heterogeneous mass of people. Under these
the power of the social manipulation" (p. 150). But the roots of the difficulty, in conditions rigorous sampling is essential if only to make it possible to cover the
his judgment, lie deeper yet-in the classical experimentalist's fundamental concep- multitudes efficiently, economically, and accurately. Even with the economies intro-
tions, which lead him to emphasize purity and aridity of formulation and to dis- duced by sampling, it usually is a large and costly enterprise. The large numbers also
regard variations in subjects and problems of sampling people and their environing pe-?nit one to refine or differentiate the descriptions, to break down the mass and
conditions. No doubt it is these many obstacles that account for Sears' inability to describe the varied patterns in different social contexts and groups, and to use
find many useful experimental studies to draw upon. further modes of analysis developed for examining the processes and causes underly-
Sears indirectly sharpens our awareness of a singular virtue of survey research ing phenomena.
on political psychology-its feasibility. The virtue, of course, could be a vice if politi- T o approach large numbers of people spread far and wide in their normal
cal psychologists become enticed by the sheer feasibility into the habitual and exces- settings, the investigator must employ a large field staff-each with his own idiosyn-
sive use of a method for which they were inadequately trained or which itself suffered crasies and remote from central control. Consequently, the procedures must be
from serious deficiencies. Such dangers cannot be ignored. There have been many explicit and standardized-partly to make the unwieldy apparatus operate efficiently
surveys that were poorly conceived but all too easily undertaken and badly executed. but mainly to instruct and ensure the proper functioning of the staff. The standardi-
However, such studies as T h e American Voter (Campbell and others, 1960), zation of procedures also reduces uncertainty, ensuring that all know the operational
T h e People's Choice (Lazarsfeld, Berelson, and Gaudet, 1944), Negroes and the definitions of the entities being studied and that the original and other investigators
Netor Southern Politics (Matthews and Prothro, 1966), T h e Kesponsible Electorate can subject the same problem to repeated study, thus providing confirmatory tests
(Key, 1966), Union Democracy (Lipset, Trow, and Coleman, 1956), Personality and evidence on processes of change over time. The standardization of measurements
and Social Change (Newcomb, 1943), Communism, Conformity, and Ciuil Liberties on the individuals in the single or repeated surveys ensures that one can evaluate
(Stouffer, 1955), T h e Development of Political Attitudes in Children (Hess and comparisons unambiguously, confident in his judgment that any differences are real
Torney, 1967), Children and Politics (Greenstein, 1965), T h e Civic Culture (Al- and not artifacts of varied procedures applied to the contrasted persons.
mond and Verba, 1963)-and these are only a small selection-surely establish that Any study of a large number of individuals is bound to generate a large
the method does not suffer from severe inherent deficiencies. As we shall argue, in volume of data. But surveys typically study many variables or characteristics for each
concurrence with the three judges and as these vivid exhibits convey, surveys are unit, ensuring a multifaceted description and tests of the multiple and complex
indeed powerful weapons. They are omnipresent because their feasibility and desir- causes of a phenomenon. The volume of data increases geometrically. Quantitative
ability make them attractive to many scholars. measurement is essential if only to compress the bulky material into comprehensible
and manageable form, but it also makes for exactitude of description and permits a
Nature of Surveys great many statistical and analytic procedures to be applied.
What about methods that lack some of the specified features? If they are
Exactly what has been adjudged so powerful? Survey research seems to rule conducted in a natural setting, they certainly are "field methods." They have that
the realm of political psychology. Perhaps this is only an illusion arising from a loose- m;ch in common with a survey, but they should be labeled accurately-for instance,
ness of definition, the accomplishments that properly belong to other methods being as case studies, community studies, or field experiments. By our definition, they are
credited to a "survey" because we have assimilated them all under one broad and not surveys. A study with so few cases that it is almost an invisible entity might be
false label. By a detailed and restrictive definitian of a survey, we shall appraise its called a "minisurv~y" but it is really no survey at all, just as a miniskirt at its best
powers better. is almost no skirt at all. A field study involving substantial numbers of cases but
A survey is an inquiry of a large number of people, selected by rigorous lacking the rigor of sampling or standardized procedure or quantitative measurement
sampling, conducted in normal life settings by explicit, standardized procedures might be called a ccquasi-survey,'abut if it is devoid of all such safeguards, it is really
yielding quantitative measurements. I t belongs to the larger class of field methods no survey at all and should be labeled a "pseudosurvey."
and is in sharp contrast to the experiment carried out in the artificial setting of Minisurveys and quasi-surveys, even pseudosurveys, may lead to discoveries.
the laboratory. I t is the total constellation of features of this particular field method Indeed, history documents the lasting contributions that such methods, on occasion,
that makes it powerful. All work together to ensure the desirable property, and all have made to political psychology. Lasswell's pioneering study Psychopathology and
of them are essential. Politics (1930) provides a vivid example. At best it was a minisurvey of some un-
The ability to study large numbers of people in their varied and normal life known but small number of cases, but it was a work of discovery and deep explora-
settings yields that "naturalistic description of the behavior of human beings in various tion. By our ?tringent definition, the contributions of all such field methods to politi-
social contexts" that Katz stresses as important for the growth of psychology. The cal psychology cannot be credited to the accounts of the survey method. But the
rigorous sampling, plus the large numbers, permits one to generalize about the deficiencies of such methods should be acknowledged, That they are surrounded by
 326 Herbert H. Hyrnan Surveys in the Study of Political Psychology 327

uncertainty and risk cannot be doubted, and the safety and generalizability that The inexpensive minisurvey and the costly giant survey on an entire nation
accrue only when the true survey in its entirety is operated should be recognized. are but two polar types. The national survey-better yet, the survey of several na-
Those who launch their own small voyages of discovery and exploration, tional populations-approaches the study of mankind in all its generality, describes
equipped with inexpensive and flexible methodological tools, might liken the survey humans over the widest range of social contexts. Thus, it is a form of inquiry to be
to a giant fishing expedition lumbering across the sea of humanity, creaking along on cherished. But if one cannot afford five thousand cases, a national survey limited to
a regimented inexorable schedule. The fishing net might seem so gross and crude a sample of one thousand may be more than adequate for one's purposes and within
that much would slip through its grid and be lost to science; strung so close to the one's means. If one's resources are too meager for that, but the problem calls for
surface that nothing at a subterranean depth would be caught; stretched so rigidly only a few measures, the institution of tlze amalgam or carauan national survey now
along a predetermined course that it could never be cast in some new direction to provides a practicable approach. One buys only a little piece of the periodic national
make an important catch that had been unexpectedly sighted. A master fisherman surveys that the major agencies conduct, paying only a small portion of the total
despite limited resources might spot some promising little pool in that great sea, drop cost, which is divided among the several parties who have amalgamated their
a line to a deeper depth, and with patience, flexibility, and skill make a great scien- interests.
tific catch. If one cannot afford the out-of-pocket costs for a piece of a new national
In terms of the image, Lasswell was a master fisherman. Indeed, an ardent survey, or if one's problem is not amenable to that design, the strategy of secondary
fisherman can move from pool to pool, enlarging the richness and variety of his analysis provides an even more economical solution. Within the thousands of surveys
catch. Riesman and Glazer (1952) in their case studies of character and politics already conducted and conveniently accessible in various archives-many of them
did just that-fishing for a while in Vermont, then in the backwaters of Chicago, national in scope-may be contained the data appropriate to one's problem, and
and then off the East River at 99th Street in New York City. A diligent and patient the costs of fielding a new survey are completely obviated. But those investigators who
fisherman can even return again to the same little pool, spending hours exploring its are forced or wish to conduct a new minisurvey or larger survey of a specialized group
deeper depths. Lane (1959a) in his case study of political belief did just that, spend- may also profit greatly by combining their narrower study with the secondary analysis
ing fifteen hours with each of fifteen men drawn from a special pool in an eastern of national surveys, a design that one might label semisecondary analysis, since it
industrial city which contained only white, married, native American, working- blends the old and the new together. The national data provide statistical norms for
and lower-middle-class men. evaluating one's crude or limited sampling and empirical points of comparison that
The metaphor, however, conveys too rosy a picture of these other methods. stop the investigator from making pseudocomparative inferences, simply using his
In 1930 no one knew in what exact pool Lasswell had done his fishing and how he imagination to fill in the cell that is contrasted with his own cases but that he has
had drawn and caught the fish. No one could anchor his findings. They simply not in fact studied.
'
floated somewhere in the great sea of humanity, and only through the accident of Old surveys may also provide sampling frames for the selection of individuals
an "After-Thought Thirty Years Later" (Lasswell, 1960) do we obtain some vague from what would otherwise be a most elusive population to locate and sample on
clues to their social location. His little pool of cases was only a drop in the ocean of one's own. By using the old surveys as such a "locator device," one may sometimes
humanity. What if he had fished in other waters? incidentally obtain measurements over time for a panel study, the first wave of
Minisurveys and pseudosurveys are at best suggestive. Whether they are deep measurements costing one nothing. By these and other applications of semisecondary
and wide explorations that yield true discoveries must be confirmed by more rigorous analyses, the lone, poor scholar can use old surveys to strengthen his own studies. (See
methods. That crude methods can stimulate ideas--some of which later prove to be Hyman, 1972a.)
good and others misleading-and that minisurveys should be buttressed by major The riches of the archives far surpass most investigators' dream of wealth.
surveys are the points to be noted. Make the extreme assumption, however, that none of the old surveys can provide the
particular sampling frame or statistical norm or empirical point of comparison that
Variety of Surveys an investigator needs. Then (as mentioned above) the scholar can buy into an
amalgam survey, "piggybacking" the few questions at little cost just to create the
The genuine survey should not be seen in terms of one stereotyped image. special sampling frame or norm or empirical point of comparison he needs.
Much latitude and a great variety of surveys are possible. The size of sample, the Strategic Universes. The national survey-whether old or new, giant or
nature of the universe, the research design, and the instruments can all be chosen so medium in size, financed in one way or another, used exclusively or as an auxiliary-
3s to create that one special inquiry best suited to orie's purse and purpose. One is is but one large class of surveys. A genuine survey, just like a minisurvey or case
not limited to a single survey. A set of surveys, varied in character so as to complement study, can be sharply focused on a narrow geographical area, a community, or a
each other or phased over time, may form the component parts of a larger inquiry. specialized population of individuals, whether concentrated or spread throughont the
Nor is one limited to surveys alone. Other methods and surveys can be combined body politic. One then studies his problem in its most strategic site but by the rigorous
into a multiform inquiry, yielding the combined advantages of the several methods of a survey. Everyone who knows that classic survey T h e People's Choice
and attenuating the errors peculiar to each. Consider some of the possibilities. (Lazarsfe1d, Berelson, and Gaudet, 1941) remembers the findings and the panel design
328 Herbert H. Hyman I Surveys in the Study of Political Psychology 329
employed, but few recall or ponder the fact that it was restricted to several panels of only after someone else has provided a norm. Scientists who pioneer in surveying
about six-hundred individuals living in Erie Coqnty, Ohio (mainly in the little city limited, however strategic, universes-if they are conservative-would seem to have
of Sandusky), in 1940, What an odd choice. What strategy was behind it? "This to wait to generalize until they extend their own surveys, or until someone else
county was chosen because it was small enough to permit close supervision of the subsequently does it for them. But they can also follow the strategies already men-
interviewers, because it was relatively free from sectional peculiarities, because it was tioned-creating their own norms through piggybacking, or finding them through
not dominated by any large urban center although it did furnish an opportunity to diligent searching of the archives and secondary analysis. Indeed, much evidence on
compare rural political opinion with opinion in a small urban center, and because for the political ideologies of children, based on nationwide sampling, did exist years
forty years-in every presidential election in the twentieth century-it had deviated
before this survey and had, in fact, been used in an early secondary analysis of
very little from the national voting trends" (p. 3 ) . The findings were strengthened political socialization (Hyman, 1959).
by the large sampling and rigorous methods applied to this strategic universe. Never- One restricts a survey to one geographical site often with the intent to de-
theless, the investigators were concerned about their generalizability beyond that
scribe some political-psychological pattern in a strategic group and then to relate that
narrow place and point in time, So they chose a second restricted universe, Elmira, pattern to its sociocultural context by supplementary observations of the community.
New York, and surveyed its population in 1948 (Berelson, Lazarsfeld, and McPhee,
In this respect, such a study combines the rigor and breadth of a survey with some
1954). If one large-scale survey has that much vulnerability, we must recognize that
of the depth of the classical community study. The inquiry partakes of the kind of
a minisurvey would be even more vulnerable. political social psychology or the political anthropology one finds, for example, in
If we accept the risk of narrowing our findings in order to sharpen them, there
Banfield's (1958) study of the culture and politics of an Italian community, although
are many possibilties for surveys of restricted but strategic universes. Depending on
the observations are not based on a prolonged period of field work. Indeed, a
the availability of the sampling frames, the geographical concentration of the units,
community survey that does not provide such an interpretive background is not
and their accessibility, the costs of inquiry may be modest. One example will make
exploiting its full advantages. So, for example, the changes among the individuals
the case in its true complexity. Jaros, Hirsch, and Fleron (1968) conducted a survey
surveyed in Elmira could be seen not only in terms of the larger temporal context of
of a sample of 305 children drawn from the universe of rural public school children
the 1948 presidential campaigns, as revealed by the panel design, but also in the
in grades 5-12 in 1967 in Knox County, Kentucky. Certainly the individuals within
context of the local institutions, as revealed by conventional field work in the com-
the scope of the inquiry were geographically concentrated, easily accessible, and
munity (Berelson, Lazarsfeld, and McPhee, 1954). A survey researcher has much
measured by simple means that were most economical: standardized paper-and-
to gain by this particular multimethod design. In this particular form, however, the
pencil instruments administered in the classroom by the teacher. Indeed, to make it
single-community survey leaves one uneasy and for the very same reasons as does the
.
feasible, "a few schools, not accessible by road, did not participate. . . The cost
conventional community study. The connections are simply woven together by the
of including them would have been very high" (p. 566n). Cheap it was, but what
discerning analyst into a plausible pattern. I t is proof by juxtaposition, and really
an exotic choice of universe. What commended it was that it typified the isolated,
another example of the pseudocomparative design, since the analyst is making the
rural, impoverished region of Appalachia, a subculture sharply contrasted in its
assumption that the behavior of the individuals located in a contrasted cell or com-
family structure and ideology with majority America; there, the investigators ex-
munity which he has not studied would be different.
pected to find a pattern of political socialization at variance with those earlier re-
The empirical survey of a second community, or a third, would make it truly
ported-a pattern with a much more "malevolent" content. The further restriction
a comparative design, and the scientific gains would certainly repay the increased
to the children in the area was apposite with the developmental concerns of the study.
costs. But if the investigator cannot afford the expenditures for such a semirestricted
For the sake of the argument, turn history back and assume that this survey
inquiry, there are ways to operate the single-community survey so as to strengthen
had been done ten years earlier, prior to the time when Greenstein (1965) studied
the tests of sociocultural factors. The community is a whole only because the investi-
the political socialization of children in New Haven, or Hess and Easton (1960)
gator has perceived it that way. Look at it another way. I t is composed of neighbor-
studied it in Chicago, or Jennings and Niemi (1968a, 196813) studied it in national
hoods, each with its distinctive institutions and subculture. And since any survey
sample surveys. Then, however strategic the universe and rigorous the survey, the
involves a large sample, the population of each type of little community within the
restriction to the one site with its emphatic cultural tone would have created the risk
larger community has presumably been sampled adequately, or by proper planning
of a pseudocomparative conclusion, in which the investigators filled in the other cells
can be. The investigator really has a series of surveys; by collecting the background
with imaginary data, much more "benevolent" in tone. Or, conceivably, they might
information on each smaller context, he can thus conduct the comparative tests
have run the risk of overgeneralizing their findings, seeing Appalachia as the micro-
required. In this fashion, Katz and Eldersveld (1961) examined the influence of the
cosm of the nation and projecting upon the nation as a whole the malevolent patterns
precinct political organization and of the ethnic and class composition of the neigh-
they had documented in one area. The existence of the other surveys was what pro-
borhood on political behavior in a single survey restricted to Detroit. The example is
tected them, and they properly located their findings by reference to the statistical
chosen advisedly, since such a conglomerate as Detroit obviously contains many
norms provided.
contrasted communities and is perceived as a unity only because of the grossness of
One might point the moral that the best time to start a restricted survey is
our vision and the narrowness of our focus on a few overarching institutions.
 I
330 Herbert H. Hyman Surveys in the Study of Political Psychology

The amount of background information required may seem to make this vey, no matter how big the sample, the cases are spread too thin to catch many
design too costly. But it can be carried out in an efficient and relatively economical partners. One may feel that even in a community survey the yield of partners is
way: the same interviewers who conduct the survey can function as a large crew likely to be too few if simply left to chance. Special sampling designs, however, have
of field workers to collect the background data on their respective neighborhoods. been developed for increasing the yield (Coleman, 1958) from a given size of sample.
And if one chooses his community wisely, ail the background information on the The instruments and coding must also be designed so as to locate the partners, and
neighborhood units may already be available in convenient form, and no costs or can be sharpened so as to enrich the evidence on the relations between such indi-
labors for special collection of the information are involved. For example, a wide viduals and their perceptions of one another.
array of such information for small neighborhoods in Los Angeles County is now on Sometimes investigators restrict their surveys to one or more communities
file in the UCLA Political Behavior Archives (Marvick and Bayes, 1969). Indeed, not in order to see the psychological patterns in the population from the perspective
the existence of such information makes a particular community a strategic site for of particular environments but for the opposite reason: to see some community event
a survey, apart from any theoretical preconceptions about its being a fruitful place or occurrence against the background of the population's psychological patterns.
to examine some problem. Both the practical and the theoretical are coordinates for For example, the fifteen cities included in surveys sponsored by the Kerner Commis-
locating strategic sites for community surveys. sion had had varied experiences of recent civil disorders, and the surveys were
Just as one can reorganize his perception of a single-community survey, one quickly set in motion to reveal some of the psychological forces that had created
can perceive a nationwide survey in a new light. I t is not only a single survey of one these events-although, to be sure, the longer prior history of each community was
giant collectivity but also many surveys of the series of smaller collectivities contained also used as context to explain the patterns in the population (Campbell and Schu-
within it-of sixty counties, six giant cities, or fifteen or thirty largish cities. By man, 1968).
subdividing and then regrouping these territorial units into types, one has many cases Thus far, our illustrations of strategic universes for study have been in terms
for testing contextual hypotheses and a wide range of values on the contextual of their geographical extent and location. Obviously, other principles guide one to a
variables. The background information on the different places can be collected strategic universe. A few illustrations suggest both the variety of other possibilities and
efficiently by the regular crew of interviewers from local documentary sources, the ways in which a survey can be focused on a specified group. A political scientist
observation, or from interviews with key informants (see Hyman, 1945), or may comes naturally to the thought that the power a group exercises make it a strategic
already exist in some convenient compendium about the nation and its subdivisions. universe for a survey. Surveys of political elites seem more consequential than are
By sensitivity to such possibilities, W. Miller (1956)-simply by subdividing national surveys of ordinary men, and the political, as well as the theoretical and the practical,
survey data already in hand by county units-was able to demonstrate that the provides a ,principle for locating strategic universes. One can easily make elites the
political behavior of individuals with the same personal characterltics varies accord- exclusive object of attention in a survey (see, for example, Hunt, Crane, and Wahlke,
ing to the party that dominates their particular environment. Such tests can now be 1964), but such an approach then relegates the rank and file to an object worthy of
accomplished economically and efficiently by secondary analysis for contextual units no attention from surveyors. Sensibly, the two groups are often incorporated into the
of particular size and for particular characteristics of such units. Turk's (1969, 1970) same large inquiry, the findings about each being regarded as complementary parts
"Large City Data File" at the University of Southern California contains information of the total picture. Thus, Stouffer (1955) surveyed political intolerance in the
on over 300 variables that characterize 130 American cities with populations of over 1 9 5 0 ~not
~ only in a sample of about five thousand ordinary Americans, representative
100,000. The archives of the Interuniversity Consortium for Political Research con- of the national adult population, but also in a sample of about fifteen hundred
tain information on all counties in the United States, including their historical char- individuals who held offices in the local governments and major voluntary associations
acteristics over the last 150 years. The national survey may thus be regarded as in each of the cities where his rank-and-file sample lived. In the analysis, the two
doubly desirable. I t can describe patterns with utmost generality, over the broadest groups often were compared, just as the children of Appalachia were compared with
kind of environment, and it can also serve to cxamine the way processes operate in children in New Haven and Chicago. But the implicit model is somewhat different.
varied smallish environments. The restricted survey can be subdivided also, but It is not as if two surveys of independent, contrasted groups had been done, simply
there is no magic by which it can be made to grow beyond its limited boundaries. for some informative test. The elite and the mass in the same town presumably
The survey of the restricted site does, however, have one hidden advantage interacted with each other, complemented each other, in some complex way that
for the political psychologist, if only he is sharp enough to note a subtle feature of jointly determined the patterns of intolerance manifested in the 1950s. Surveys of
the design and to exploit valuable data that are normally buried within such a survey. complementary samples thus have a similarity to community surveys of interacting
Without benefit of small-group research, he can study small-group interactions within partners, although the former can be operated on a grand scale across very wide
his narrow survey. Since the large sample of one small community tends to saturate geographical universes and the interactive processes under study do not necessarily
the population with interviews, by chance he has probably sampled many interacting operate through the same direct intimate relations.
individuals in that community, several members of many of the primary groups. Other vivid exhibits show the varied models of complementarity that can
Automatically he has evidence from the parties in such social transactions, if only he I shape the design. Shortly after the 1958 election, using a national sample of all
constituencies, Stokes and Miller (1962) conducted a survey of the electorate and
meshes the data from those who are in fact partners. By contrast, in a national sur-
332 Herbert H. Hyman Surveys in the Study of Political Psychology
complementary surveys of the incumbent congressman and other major party candi- rather than about themselves. Admittedly, the investigator can end up on somewhat
dates from the same set of districts. The model is explicit: "Through these direct shaky ground. One can easily establish that humans are fallible observers and that
interviews with the persons playing the reciprocal roles of representative government, perceptions are guided by need. But one can validate such reports in surveys in four
this research has sought careful evidence about the perceptual ties that bind, or fail to ways: by choosing as a strategic universe those whose credentials establish some
bind, the congressman to his party and districts" (p. 532). special competence and history of opportunity to acquire the requisite knowledge;
A national survey of the Turkish peasantry (Frey, 1963) provides an exotic ex- by drawing large samples of informants, whose reports can be consolidated and used
ample of the general design. Within each of the approximately four hundred villages as checks upon one another; by creating the mental set toward objectivity through the
from which the peasant sample had been drawn, four elite individuals were chosen line of questioning; and by using internal tests of the degree to which these inform-
and surveyed: the secular official who headed the village and his wife; the religious ants' reports are correlated with measures of their psychic dispositions, rather than
head and his wife. The implicit model was one of influence and interaction that informed by their respective levels of competence (see, for example, Hyman, Levine,
flowed through several channels, women being linked to the elite mostly via other and Wright, 1967) .
women, and men and women being influenced by both the religious and secular elites. Surveys of informants remind us that the method is flexible enough to be
A last and somewhat different example is provided by the national surveys of operated in ways that resemble the classic procedure of anthropology and com-
political socialization (Jennings and Niemi, 1968a), where-in addition to the main munity study, although the informants are drawn by a rigorous sampling design, their
sample of high school youth, complementary subsamples of the teachers, fathers, and number is much larger, and the investigator's contact with each informant is briefer.
mothers of the children, elites of a sort, were also surveyed. Thus, both the subjects There is no survey that does not contain at least some minimal element of the role
and the agents of socialization were studied, and the several sets of findings could be of informant. But the role is rarely maximized in survey research. Even in surveys of
dovetailed to provide a more comprehensive understanding of the process. By draw- elites and populations that have some special competence, the respondent role is
ing unbiased subsamples of the various complementary groups, rather than samples generally made to predominate. Parents and teachers in surveys of political socializa-
of equal size, the investigators introduced an economy into the design without tion could be construed as competent informants about the children, about the
jeopardizing the conclusions. transactions in the family unit or in the classroom, but more frequently are treated
These several studies, particularly Stouffer's survey, suggest a valuable set of as respondents whose own ideologies are measured. One could, of course, have a
distinctions and will lead us to a fourth principle for choosing strategic universes, and different mix within each interview in such surveys, or one could use separate sub-
to even more general matters of research design and questionnaire construction. Al- samples as informants and respondents respectively. For example, Hess and Torney
though Stouffer's leaders were elite (since they exercised control over many other (1967) drew a complementary sample of about four hundred teachers of the classes
individuals and large organizations, and since they possessed unusual competence), from whom the children were drawn, and their "attitudes were measured to com-
instructions and questions put to them did not cast them symbolically into a per- pare their responses with those of their students and to assess [inferentially] their
formance of their public roles or a rendition of their knowledge about governmental influence on the attitudes of the younger subjects" (p. 237). In addition, a subsample
and other institutions and constituencies in the community. They were reporting on of 169 teachers were cast as informants, and the "role of the school in socializing
themselves, about their private feelings and thoughts and preferences. Admittedly, children1' was determined by a "curriculum questionnaire," which contained such
they were also asked about their perception of the world; but this, like the other questions as how much time was spent on each of a series of topics, whether the
variables, was treated as psychological stuff, not as objective information. By a strict teaching was planned or incidental, or whether flags and other patriotic symbols
interpretation of these data, one must say that we have learned only about private were displayed in the classroom, the reports implicitly being accepted as factual
character, not public performance. T o be sure, this is in no sense unimportant evi- descriptions of the schools.
dence. I t was an advised research decision to seek out this level of data. Certainly, The last example will also suggest some additional valuable distinctions. Sur-
as Lasswell suggested long ago, the psyches of these very important persons must in veys are usually conccived of as studies that focus on individuals, taken singly or in
some degree be irrepressible and invade their public life; but it is also true that great numbers. But the individuals sampled may be only the vehicles of the study,
"good" people can be inhibited and corrupted by the social forces pressing on them carrying us toward an ultimate target universe which has never been directly meas-
when they have to function in some public arena. The tolerance toward nonconform- ured. That target universe may be another population of individuals, or it may be a
ity of Stouffer's elite might have been quite different if the entire inquiry had been set of collectivities (family units, organizations, or institutions), as in the example of
structured so as to make the public part of their lives salient to them. the classroom and the schools. The use of informants as vehicles permits surveys to
Stouffer's elite also had unique and prolonged opportunities to observe many describe collectivities. The sampling of respondents who are members of a collectivity
other local people and the organized life of the community, and in some degree were permits us, by addition, to describe it, if we use the model that it is an aggregation.
very superior persons (gifted, educated, trained to be sensitive) and had unusual Instruments, The discussion of the role an individual assumes in a survey
access to recorded sources of information. If the inquiry had been structured dif- brings us directly to the topic of questionnaire design. Although our definition re-
ferently, they could have been cast in the role of informant, not respondent, instructed quires that erery survey incorporate explicit standardized procedures that yield
to tell as objectively as possible--even with the aid of their records-about others quantitative measurements, there is still room for great variation in the instruments
334 Herbert H. Hyman Surveys in the Study of Political Psycholog 335

employed. No one should be misled by the stereotyped image of a "poll," in which populations. Many verbal instruments are listed, critically reviewed, and findings
only a few questions in fixed-alternative form are asked. Not all surveys are that therefrom summarized conveniently in a set of three handbooks developed by Robin-
rudimentary. Although there are limits to the patience of people, it is common and son and his associates (1968, 1969).
practicable for surveys to include sixty to one hundred questions, some with sub- The following examples illustrate the variety of verbal instruments that can
questions. By the power of words and memory, the respondent's past, present, and be incorporated into surveys of the general population. Thorndike and Gallup (1944)
view of the future can be encompassed in the lavish list of questions. Different realms developed two parallel forms of a steeply graded, multiple-choice, twenty-word
of experience-the cognitive, affective, conative-can be explored and many varia- vocabulary test which they were able to use on a sample of 3000 adults as a measure
bles studied simultaneously. If one question will not suffice for valid and reliable of verbal intelligence. Morgan and his associates (see Robinson, 1968, pp. 154-158)
measurement of a variable, a battery or scale can be constructed. Our definition developed a thirty-five-item composite instrument, containing four subscales, to mea-
calls for quantitative measurement but does not demand that all the data be trans- sure "a modernism syndrome" or a "concern with progress." Stouffer developed a
lated into that form. Surveys commonly supplement quantitative description and fifteen-item Guttman-type scale to measure tolerance of political nonconformity
analysis with qualitative case materials-background reports by interviewers, com- (Robinson, 1968, pp. 163-164). McClosky (see Robinson, 1968, pp. 170-1 78)
ments, protocols, brief profiles, illustrative quotations from individual questionnaires developed a sixty-five-item composite instrument, containing subscales, to measure a
-to enrich the findings. The questionnaire can combine various procedures. Some syndrome of "democratic or antidemocratic attitudes." Matthews and Prothro
questions may be closed and force all respondents to expresi their individuality in developed a Guttman-type scale to measure political participation; this scale was used
terms of the few alternative categories provided, but some are open-ended and in a survey of eleven southern states (Robinson, 1968, pp. 429430). Campbell and
supplemented by probes so as to allow individuals to express their views in depth his associates developed four- and five-item scales .to measure sense of citizen duty
and detail and in their very own words and style. The! survey researcher of long and political efficacy; these scales have been used in a series of Michigan national
ago who invented that classic opener "What you are about to say interests me very surveys and thus provide many norms (Robinson, 1968, pp. 459-4452). Short versions
much" was being witty, but he was also making a serious point. The stimulus clearly of the F scale to measure authoritarianism, based on from four to ten items, have
is standardized but sure leaves plenty of room for individuality. The responses could been used successfully by the Survey Research Center and also the National Opinion
even be quantified or counted under categories, although it would be hell to build the Research Center on a series of surveys (Robinson, 1969, pp. 262-270). Finally,
code. Christie's "Machiavellianism" scale has been adapted for use in national samples
A single interview with an elaborate questionnaire can be long and searching, (Robinson, 1969, pp. 506-516) .
but recall also that some designs involve more than one encounter with a respondent. Survey researchers, however, have not limited themselves to conventional
We tend to think of the seven interviews with the panel of respondents in Sandusky verbal instruments. All sorts of other approaches have been applied successfully.
as discrete points of comparison over time, but if we add all seven together it makes Vicary (1948) adapted that classic instrument the word-association test; by giving
one very lengthy unified inquiry into a person's views. We tend to focus on the core his test to subjects a second time, be was able to assess stability of performance. The
inquiry in Almond and Verba's The Civic Culture (1963), in which about one ancient tradition of constitutional psychology inspired Stapel (1947-48) to obtain
thousand respondents in each of five countries were asked almost one hundred ques- interviewer ratings of body type, and corollary ratings of temperament, on six na-
tions, not an insubstantial instrument. But we may forget that small subsamples of tional surveys in the Netherlands. The estimates were highly stable from survey to
about one hundred respondents in each of the countries were reinterviewed at a later survey, and. the ratings of temperament were not biased by the interviewer's own
date and asked an additional 130 questions in a "political life-history interview." temperament. Although this is an unusual example, the use of interviewers to make
Although a survey researcher must make his procedures explicit and standardized all sorts of observational judgments and ratings of the respondents is exceedingly
for his interviewers, they do not have to be uniform over the grand total of all re- common, and represents an assessment based generally on many clues, verbal and
spondents. They should be uniform over those subclasses whose measured character- behavioral, in the course of the interview. By contrast, survey researchers sometimes
istics will be compared, so as to reduce ambiguity in the inferences. As long as the base such ratings on the overall reading of the entire interview after it is turned
investigator prescribes what variations will be introduced when, he can safely into the office, treating it as a kind of clinical protocol but without benefit of first-
interpret his findings. There is flexibility, but it is controlled by him. By their design, hand observation. An unusual example ~f the joint use of both "holistic" rating
Almond arid Verba reached the depths they sought, but economically and without methods can be found in the surveys of the German population at the end of World
burdening all respondents. They needed only a small staff of skilled depth inter- War 11, where they were employed successfully in the assessment of the respondent's
viewers to cover the subsamples. identification with Nazism (Peak, 1945).
Survey researchers have been unusually creative in inventing or adapting Robinson and Hefner (1968) applied a procedure reminiscent of classic
instruments for use with large samples of ordinary individuals. The political psycholo- concept-formation tests in experimental and clinical psychology to a large sample in
gist should borrow liberally from what is already available, thereby saving himself Detroit. Subjects were given a list of nations and asked to group together those that
the hard labor of instrument design and getting in the bargain an instrument whose they judged to be similar, thus describing the ordinary man's "perceptual map" of the
properties have already been established plus statistical norms from other times and world. By supplementary questions and inference, the investigators established the
 Herbert H. Hyman Surveys in the Study of Political Psychology 337

dimensions or concepts underlying the groupings. 'Various forms of the instrument, instruments can explore the psyche of the ordinary man. One should not think, how-
listing different nations, were rotated over a series of equivalent subsamples. This ever, that a survey has to be excessively psychologistic or that the measurements must
split-ballot design has many useful applications. When contrasted types of respondents be exclusively subjective in origin. As has already been noted, in designs that are
are examined, procedures (as earlier noted) must be uniform. Otherwise, both instru- contextual objective information on the individual's environment obtained from
ment variation and respondent variation confound the interpretation. But if respon- independent sources can be merged with his subjective reports. And in surveys that
dents are made equivalent, varying the ballot provides an unambiguous test of the sample strategic populations of known character (for example, government officials,
contribution of the instrument. This one methodological use of the general procedure school children, registrants.in particular parties, members of organizations) objective
is enough to suggest that the measuring instruments in surveys can be ca1ibrated.l information on each individual sampled, known from records or simply by virtue of
I n this same study, and in many other surveys, the Semantic Differential (Osgood, membership in the specified universe, can also be merged with his subjective report.
Suci, and Tannenbaum, 1957) was also found to be applicable in surveys of the The examples presented are not typical of the routine, unimaginative survey.
general population. But in the hands of ingenious investigators, the survey is an instrument that is most
For use in a national survey, Veroff and his associates (1960) adapted the pliable. There are constraints, however. It should be stressed that the ability to in-
Thematic Apperception Test for assessing need for achievement, afiiliation, and trude upon ordinary people in the course of their normal lives and to elicit the
paver. Greenstein and Tarrow (1970) used a story-completion method in a cross- measurements is dependent on consent and cooperation. (In only very rare instances
national survey of political socialization. Graphic materials (for example, in the shape -such as the United States Census, and even here the sanctions are very rarely
of thermometers or ladders) have been used to obtain ratings and judgments of applied-are respondents legally compelled to participate. Thus, one might define
'
magnitude. Pictorial materials have been used, sometimes as projective tests (see almost any survey as "an inquiry involving voluntary cooperation.") There are limits
Chapter Fifteen in this volume) but also to reduce ambiguity surrounding an ordi- to what the people allow, and ethical constraints on what the surveyors should
nary question, to present an otherwise ineffable stimulus to the average respondent attempt or perpetrate.
or a complex concept to a verbally unsophisticated population, and to test aesthetic What is allowed varies with time and place and topic but does set limits on
preferences (see, for example, Hess and Torney, 1967, pp. 234-235; DeFleur and the power of survey methods to explore particular problems. A recent report, some-
DeFleur, 1967; Noelle and Neumann, 1967). Aesthetic preferences may seem remote what far afield, provides an example. For attempting a survey among university stu-
from our topic, but such preferences for given kinds of pictorial stimuli could serve dents in Madrid to discover what young people know about sex, Nicholas Caparros
as an index of cultural modernity, of "bohemiani~m,'~ or of ethnic identity, which was accused of conducting a "shameful survey," subjected to a police investigation,
would be useful to the political psychologist. charged with creating a public scandal, and then faced trial and a six-month jail
The split ballot creates opportunities for designing another kind of projective sentence (Rome, Daily American, December 15, 1971, p. 4: 7). Such a case is perhaps
test, as illustrated in a study by Haire (1950). A brief shopping list of groceries was irrelevant to the progress of political surveys, but consider a more relevant datum. In
presented to subsamples of respondents, all being asked to describe the woman their cross-national inquiry, Almond and Verba (1963, pp. 117-118) found that
shopper who had supposedly made the list. The lists were systematically varied over 32 percent of their sample of Italians interviewed in 1959 refused to report how
the different but equivalent subsamples, a given commodity being rotated in and they had voted in the last national election, and 31 percent refused to report their
cc
out of a list and other commodities being constant over all subjects. One is reminded usual local vote." The refusals were distributed about equally in all social strata,
of Asch's classic experimental studies of impressions of personality. The paradigm suggesting a prevailing societal norm. By contrast, in the American sample the mag-
might well be used not with shopping lists but with a list of attributes or actions of nitude of such refusals was between 1 and 2 percent. Fortunately, on the many topics
nations or government officials, the subjects being asked to assess the importance of of central interest other than voting, refusals were rare; and the refusals to report
one component in the total impression. voting preferences could even be exploited to shed indirect light on one matter of
This partial list of examples will suggest the many ways in which survey central interest, 'the emotion and suspicion pervading Italian political culture. But
those political psychologists who focus their surveys on voting would no doubt see
*Other designs than the split ballot can, of course, be used to assess the quality of such a situation in a gloomy light, and should treat it as a cautionary example.
survey instruments. Panel studies, for example, provide automatic evidence of repeat
reliability. Although opinions are changeable, some characteristics (such as age, place of Surveys as Experiments. In reviewing the variety of surveys, although they
birth, past education) are by definition unchangeable. Repeating the face-sheet items on a have often been compared with other field methods, we have made no reference to
second wave of a panel thus reveals the amount of error in such measures, and such studies field experiments. However, surveys can be incorporated into an experimental design,
in the literature provide comforting evidence. On a single-shot survey, an investigator may evcn though they take place outside the laboratory. Panel or trend surveys may be
deliberately reinterview a small subsample after a very brief interval of time, in order to spaced deliberately around some major event, thus providing a natural experiment
estimate the stability of response, making the assumption that attitudes and other psycho-
logical characteristics, although changeable over long intervals, should not be subject to on a complex, prolonged, and powerful set of variables whose effects are inferred
capricious short-term fluctuations. Following this design, Hess and Torney (1967) retested from the changes measured in the successive surveys and from collateral evidence
a 10 percent subsample of their original twelve thousand children after an interval of reported by the respondents. Where the "treatment" is sufficiently circumscribed (for
from four to fourteen days, and computed stability coefficients.
 Herbert H. Hyman Surveys in the Study of Political Psychology 339

another untreated community, which functions as a control group, or available trend more in no way deny the argument. If the political psychologist points his survey
data from national surveys can be exploited for control group estimates. in the right direction, then 3t will describe individual differences that are relevant.
Occasionally, surveys may even be incorporated, into true field experiments, Moreover, a great many completed surveys that seem devoid of political-psychological
when some powerful agency deliberately manipulates some set of independent vari- content should not be dismissed too quickly. The theorist who pauses long enough
ables or program of social intervention in collaboration with a researcher who is over them may sense some hidden significance and make a contribution to political
geared to measure the effects by his surveys. The treatment may be allocated in some psychology by the secondary analysis of what are trivia to less thoughtful minds.
special arrangement, so that experimental and control or comparison groups are The individual differences described by any inquiry are obviously dependent
created, all to be surveyed as part of an elegant experimental design for measuring on the sample that is measured. With small or biased samplings and restricted or '

the net effects of manipulation. Indeed, evaluations of programs of social intervention downright peculiar universes, an investigator may underestimate the variability in
might well be described as field experiments combined with survey measurements. the population, judge something to be common when it is truly rare, or regard
In these two ways, an experimentally oriented surveyor can test major con- something as aberrant when it is truly ordinary among common men. He may miss
temporary variables which are beyond his personal powers of manipulation. But his important clues about its social location and correlates and spin out theories and
control over the apparatus of the survey itself provides a direct way to create his research that are misguided. The survey, by virtue of its rigorous and large-scale
own experiments within the course of a survey. A classic mode of manipulating sampling of the general populstion, automatically yields accurate evidence on the
variables in experimental social psychology is by verbal stimuli and instructions shape of the distribution of individual differences, on the extent of variability, on the
presented to the subjects. Surveyors occasionally employ the very same vehicle within modal pattern in the large and the differentiated patterns in various social contexts.
a split-ballot design, applying variant stimuli to equivalent subsamples, cornmunicat- The instruments, of course, are as' essential as the sample to the accurate
ing a passage of text to one group but not to another, introducing a proposition with description of individual differences. The mode may be displaced by a biased instru-
the name of a prestigious and credible source in one group but not in another. The ment, and the variability that can be documented is limited by the insensitivity of
split ballot, although usually employed for methodological experiments on the in- the measurements-people seemingly all the same or cut on a few models simply
fluence of question wording (Noelle-Neumann, 1970), is readily converted to such because the instrument cannot register the fine differences between them. But survey
substantive experiments. But although the approaches of the experimentalist and instruments are not as unrefined as some might think. Inspect the distribution of
the surveyor can converge, in usual practice the surveyor moves in an opposite political activity among American adults, described by Woodward and Roper (1950)
direction, exploiting his method not for an experiment but for its unique advantages: from a national survey more than twenty years ago. The scores, shown in Table 1,
for discovering things he had never dreamed of, obtaining reliable and generalizable represent a composite of voting, membership in political organizations, communica-
descriptions of the endless variety of social, psychological, and political patterns, and tion with public officials, political-party campaign activity, and talking politics with
weighty, if not final, evidence on the processes moving through long courses of time friends. The instrument clearly distributes the population by fine gradations, and
and complex chains of causations that explain the patterns he has described. the obtained national distribution says a good deal about the nature of political man
Now consider the ways in which surveys can be exploited specifically for the in America at mid-century.
study of political psychology. Classic psychology, however, has concerned itself with man's enduring nature,
although some branches are concerned with man's changeability. Political psycholo-
Individual Dierences and Total Political Character gists, whether guided by the one or the other concern, would follow the same line
of inquiry, ,using their instruments on repeated surveys spread over time. What is
Political psychology, however one might elaborate the details of the field, stable and what is changeable are but the two sides of the same question, and the
surely must mean the study of politics from the perspectives of psychology. The answer is readily bought with the same methodological coin. Table 2 presents partial
distinctive concern of psychology has always been the individual. Indeed, whole distributions for various discrete political activities obtained in American national
branches of the discipline (differential psychology, psychometrics) have been erected surveys conducted by the Survey Research Center during four presidential campaigns.
simply on the basis of the study and measurement of individual differences. Then it Apart from the suggestion of a tiny secular trend, such data reveal that low levels
must follow-whether we realize it or not-that we are engaged in political psy- of political activity are indeed a stable feature of modern American man. Such a
chology a great deal of the time when we conduct surveys, since they automatically finding also was implicit in the Woodward-Roper survey and was bought for less
yield reliable evidence on the distribution of individual characteristics. coinage than trend surveys demand. The question asked in a single survey can have
The individual differences described in surveys do not necessarily have poli- a much broader temporal referent than the .moment of inquiry. Thus, for example,
tical relevance or psychological significance. For their own good reasons, some have Roper's respondents were asked whether they had ever communicated with a public
surveyed individual differences in the taste for fancy cars or rock music; in belief in official,not whether they had communicated only during that particular month.2
the devil, the malevolence of scientists, or the frailty of women; in preference for day-
2 We gloss over important questions of the reliability and validity of such measure-
light saving time, pets, southern cooking, football, or hunting; in such behavior as ments. An investigator who broadens the temporal referent of a question must, of course,
buying on the .installment plan or going to church. These examples and thousands ponder the risk that memory error is introduced. And any form of questioning presents
340 Herbert H. Hyman Surveys in the Study of PoliticaI Psychology 341

Table 1. DISTRIBUTION
OF POLITICAL
ACTMTYOF Table 2. POLITICAL
ACTMTYAMONG
AMERICAN
ADULTSSINCE1950

Score on Woadward- Percent

Ropsr Scale Percent
Belong to Civic or Political
High Activity 12 .1 Organization 2 3 3 4 3
Worked in Campaigns 3 3 6 5 6
Attended Political Meetings 7 10 8 9 9
Talked Politics with Friends 27 28 33 31 33
Votes 73 74 74 78 76

Source: Robinson and others (1968, p. 591).

I
Table 3. POLITICAL
TALKAMONGGERMAN ADULTS
OVERA TWENTY-YEAR
PERIOD

Often . Occasionally Hardly Ever

Low Activity 1 19.1
Percent

1952-April 17 37 46
July 19 41 40
Psychology also has long been concerned with uniuersal man, although, here October 16 , .38 46
again, such branche3 as social psychology are especially concerned with the way 1953-September 18 44 38
society and the group modify man. Again, these are but two sides of the same ques-
1955-May 15 41 44
tion. Cross-national surveys readily provide the comparative distributions needed,
although such surveys are very expensive. By way of illustration, Table 3 presents 1956-March 17 37 46
contrasted distributioils for one discrete form of political activity, "talking politics," June 13 38 49
in another society, Germany, based on many surveys by the Institute for Demoscopy.
1957-August 14 41 45
These data are not strictly comparable with the American findings, because of varia-
tions in the instruments, but are presented because the long time series provides an 1958-June 13 41 46
unusual illustration of the stability of these distributions. Despite all sorts of vicissi- 1959-May 16 41 43
tudes in German society and politics in the years 1952-1964, this feature of German 1960- July 14 39 47
political man persisted. As he entered the 1970s, however, a substantial change seems
1964--December 19 43 38
to have occurred, judging by the one recent point in the time series. Whether politi-
cal conversation will remain a t this new higher level is a n empirical question that 1972-March 28 45 27
can be answered only by future surveys, although other evidence suggests to the
Source: Noelle and Neuman ( 1957, p. 45) ; Noelle-Neumann, personal
analysts that a new political climate has emerged. communication.
I t is possible to show the contrast in national distributions from truly c0mpa.r-
general problems of appraising error. Certainly the stability of the data in Table 2 suggests able surveys. For example, in Almond and Verba's (1963) inquiry into The Civic
high reliability and little sampling error. The dangers of a bias toward the systematic Culture, a question about the frequency of talking about public affairs was asked in
inflation of reports of such "socially desirable" behavior have also been subject to empirical five countries. I n June 1959, 39 percent of their German sample answered "never."
study. They are far less than one might have thought; and, for some'modes of analysis, (This is a more extreme position than provided in the question used by the Institute
constant errors do not distort the inferences. (See, for example, Campbell and others), for Demoscopy and thus likely to be characteristic of a group smaller than those who
1960; Cahalan, 1968.)
 Herbert H. Hyman Surveys in the Study of Political Psychology

"hardly ever" talk politics, found to be 43 percent in the May 1959 entry in Table the pattern that prevails in the population is hard to understand if one sees the polity
3.) In contrast with their 39 percent in Germany, Almond and Verba found that as the simple sum of the characteristics distributed in the general population. To
only 29 percent of the British sample and 24 percent of the American sample, but determine collectively the functioning of the system, therefore, one is forced to
66 percent of their Italian sample, "never" talk politics (p. 116). develop more subtle models of the way individuals are aggregated. The differentiated
These differences were obtained from one set of surveys describing the respec- distributions for various social groups are very helpful in this connection. Perhaps
tive populations at that particular time, although the question clearly has a much a rare pattern typical of only a particular small group in the population is effectively
broader temporal referent. This safeguard may not satisfy those psychologists who multiplied because of the great weight that group has in the governmental process.
are interested in national political character and who therefore wish to distinguish Thus, when Key (1961, pp. 536-537) "frets with the puzzle of how democratic
between differences that for sure are enduring and those that are transient reflections regimes manage to function," he locates the "missing piece" in "the upper layer of
of the circumstances in a society at a given moment. Note, therefore, the crucial activistsJ' and in the peculiarity of their "subculture" rather than "in the qualities
finding in Table 3. In twelve different surveys between 1952 and 1964, the propor- of the people that may he thought to make democratic practices feasible." This
tion of Germans who "hardly ever" discuss politics never reached as low a figure as model suggests an obvious agenda for research and underscores some of the points
Almond and Verba reported for England and the United States, nor as high a already made. I t calls for a series of surveys of elites and analyses of the obtained
figure as they obtained in Italy. Ideally, cross-nationa1,surveys should be replicated distributions to see whether they are both distinctive and homogeneous enough to
over time; but since that is extremely expensive, a single comparative inquiry can support the notion of a peculiar subculture characterized by certain themes.
be buttressed by available surveys, another instance of the utility and feasibility of There are other models. Perhaps a pattern that is exceedingly rare within a
semisecondary analysis. Indeed, the German time series through 1-960 was available certain group but essential to their political survival becomes weighty because it is
for all to exploit, conveniently compiled in the periodic handbooks of the Institute multiplied by the vast population size of that group. Thus, for example, when
for Demoscopy. Woodward and Roper (1950) reported the distribution of political activity, they
The findings, even though extended over time and space, are nevertheless also were able to eflgage in the following political arithmetic: While the individuals
fragmentary in describing only one feature of man, his Ievel of political activity. on the level-A economic level were three times as likely to have a "very activeJ'
Thus, they may not satisfy those who are interested in the total political character. score as those on the level-C economic level, 54 percent of the active people came
The characteristic involved, however, refers to a broad domain of activities. The from level C and only 13 percent from level A. The poor made up by their numbers
single dimension along which individuals are scored is already an abstract and (far more than 3: 1) for their relative inactivity and thus could make their weight
general one, and the distribution on this variable is critical for all sorts of theorizing. felt.
Contemplating such distributions in the course of his trade, Roper (1957) was led One may try to build these models by armchair theorizing, but it is a most
to distinguish, within the Amerikan public, the "politically inert" and the "participat- difficult mental exercise. Instead, one may start inductively and empirically searching
ing citizen" ;and Almond (1950) was led to his forinulation of three political strata: for clues by inspecting the distribbtions within and between societies obtained from
the "mass public," the "attentive citizen," and that rarefied group given for its many past surveys. It may turn out that very different polities exhibit aggregate
excellence the title "opinion elite." The statistical norm obtained routinely from distributions that are much the same in many respects, and different onIy in some
such a distribution-far toward the pole of inactivity-diffcrs sharply from what is particular parameter; or it may be that the difference inheres only in the distribution
defined as normative in traditional theories of democracy. That finding has either obtained for some particular subgroups. How far one can go with aggregate distribu-
supported radical critiques of democracy as incompatible with man's political nature tions from one society is suggested by Key's (1961) treatment of the "patterns of
or has led to a revised and more realistic conception of the requirements for democ- distribution." Building on conventional descriptive statistics, plus the classic insights
racy (Berelson, Lazarsfeld, and McPhee, 1954, ch, 14; Almond, 1950). of Floyd Allport, Key examined such patterns as bimodality and a J-curve and
It is a quick and easy step in survey analysis to break down the aggregate determined their implications for the functioning of the American polity.
distribution into a set of differentiated distributions and thus describe the patterns in
different social groups. Inactivity is not randomly distributed but located pre- Relevant Variables and Political Character
dominantly in the lower social strata. Thus, applied political psychologists with a bent
toward reform are helped to see where and how they'might *ark to ameliorate the Although surveys of single characteristics and their distributions may not
situation. Other political psychologists, who might be too quick to conclude that gratify some psychologists, for all the reasons mentioned they deserve considerable
inactivity represents willful neglect, are forced to see that it reflects social disad- inquiry. But what particular variables are especially relevant and important? We
vantages and to look for its roots in the right places. shall list some of the general areas within political character and suggest fruitful lines
Still other political psychologists, from inspecting such national distributions of exploration.
and juxtaposing them against their observations of the political system, are forced Political Action. I t is self-evident, even without recourse to our earlier
to consider what Greenstein (1971) has called the problem of "aggregation." How examples, that the action of individuals produces, if anything can, some effect on
some political systems can operate at all or as well as they do under the burden of the political system. Political psychology must either begin or end its inquiries at this
 344 Herbert H. Hyman I Surveys in the Study of Political Psycho~ogy 345
point if it is to have true relevance. The level of general political activity and the could be a persistent force and even make a public nuisance of himself by penetrating
frequency of specific classes of acts warrant the most careful description and attempts into places where he was not wanted and without any invitation.
at explanation. Properly, .these variables have been the focus of a great deal of survey As one contemplates the many surveys in which opinion leadership of one
research. type or another has been studied, a most obvious line of inquiry seems to have been
Given the prevalence of inactivity that the distributions document, those neglected. No one seems to elaborate upon the initial questions by probes so as to
actions that bring inactive and active strata into contact, and thus provide for the describe the content of the transactions. "What, specifically, did the advisees ask?"
arousal of otherwise passive individuals, assume special importance. Similarly, actions "How long did you talk?" "What did you say to convince him?" The format of an
that enlighten otherwise ignorant individuals are important in light of the widespread initial question followed by probes is a common and feasible procedure; if one is
. political ignorance that has been documented. Thus, the substantial literature on overwhelmed by the cost or thought of coding so much open-ended material, an
personal influence or opinion leadership or the diffusion of information is of critical adaptation of Schuman's (1966) random-probe technique could be employed. The
significance for understanding some of the dynamics of political behavior (see, for probes could be applied to only a subsample of the instances or different probes
example, Lazarsfeld, Berelson, and Gaudet, 1944; Berelson, Lazarsfeld, and McPhee, allocated to different subsampIes.
1954; Katz and Lazarsfeld, 1955; Bell, Hill, and Wright, 1961; Rogers, 1962). Political Afect. The variables of political action and, as we shall soon see,
American national surveys of these patterns are still needed, however, since many political attitudes have been incorporated in endless surveys. Ironically, surveys of
of the classical inquiries are limited to particular communities where the process and political afect, descriptions of the distributions of various kinds of emotions and
magnitude of such influences may well be peculiar. Comparative surveys in other sentiments that accompany political thought and behavior, are almost completely
nations might also reveal different findings. Acts of influence might well vary in absent. So fundamental a category of experience, one of such classic interest to psy-
character and magnitude depending (1) on ecology and the opportunity for con- chologists and one that may well either arouse people to action or inhibit their ac-
tact; (2) on the hierarchical structure of the society, making some influentials more tivity, seems to have been singularly neglected. To be sure, there are some pallid
visible and notable but perhaps more distant and unapproachable; (3) on the sense indicators of an investment of some kind of emotion into politics in routine questions
of trust that pervades interpersonal relations; and (4) on the presence of specialized about interest or attention, but the kinds of data presented in Table 4 from Almond
institutions that formalize and channel communications. and Verba's (1963) comparative national surveys exist in almost splendid isolation.
Two forms of leadership, the responsive and the self-propelled, should be Respondents were asked how often they felt each of a series of emotions. The statis..
sharply distinguished in the questions used and should not be neglected in the anal- tics tabled contrast those who report such feelings any of the time and those who
ysis. In The People's Choice (Lazarsfeld, Berelson, and Gaudet, 1944), these re-
spective forms of leadership were measured but not distinguished. That is, each Table 4. ABOUTVOTING
FEELINGS AND ELECTION
CAMPAIGNS
respondent was asked whether he had ever been asked for advice and also whether IN FOURNATIONS
IN 1959
he had ever tried to convince someone else politically. Individuals who had engaged
in either act were treated as "opinion leaders" without any distinction, and nowhere
in that analysis can one even find the intercorrelation between the two activities. But
I
I
United
States England Germany Italy
frbm Robinson's (1952) subsequent factor analysis of these data, we know it to be for
Sandusky a mere .17, the one behavior hardly predicting the other (see also Agger Percent
and Ostrom, 1956).
In a subsequent major study of personal influence (Katz and Lazarsfeld, Sometimes find election
campaigns pleasant and
1955), conducted in Decatur, IIlinois, the indicators of political influence were con-
enjoyable 66 52 28
fined solely to measures of the "responsive" pattern, respondents reporting whether
they were asked for their advice at all and, if so, how frequently. When those re- Sometimes get angry
spondents were then asked, "Do you know anyone around here who keeps up with during campaigns 57 41 46 20
the news and whom you can trust to let you know what is really going on?" "about
half, in fact, were unable or unwilling to name anyone within their acquaintance
I Sometimes find campaigns
silly or ridiculous 58 37 46 15
whose competence and trustworthiness in public affairs they accepted" (p. 140; Never enjoy, and never
italics supplied), Almost identical findings were obtained in a subsequent study of a get angry, and never
Wisconsin community (Lowe and McCormick, 1956). Truly passive people, it feel contempt during
appears, will not bother to solicit advice. Indeed, they might avoid it (Wright and campaigns 12 26 35 54
Cantor, 1967). Even if they felt impelled to seek advice, they might be too ignorant 970 963 955 995
to know where to turn for it or might well seek it in the worst quarters. Such factors
seriously limit the role of the responsive leader. The self-propelled leader, by contrast, Source: Almond and Verba, 1963, table 13. Reprinted by permission of Princeton
University Press.
346 Herbert H. Hyman Surveys in the Study of Political Psychology

never have them. Almond and Verba use the last construct in the table as an index new primary surveys, Almond and Verba obtained evidence on the more enduring
of political apathy. They define apathy as the absence of all feeling rather than as features of German national character and on the situational determinants of such
inactivity, the definition implicitly given to it in much survey research. Some in- sentiments. Now, if one were to juxtapose the time series on trust that has become
active people may have strong feelings of disgust, despair, fear, and the like, about available from a series of national surveys by the Survey Research Center, one could
politics and in that respect may be far from "apathetic." document empirically that the distinctive American pattern is also an enduring one
The questions used by Almond and Verba dealt specifically with the affect (Robinson and Shaver, 1969, pp. 529-532; Campbell, personal communication).
accompanying elections, certainly a moment when emotions might be temporarily Other dimensions of affect have direct relevance for political psychology, and
heightened. The good example of Almond and Verba could be followed and extended the distributions can and should be plotted regularly. Long ago, for example, Almond
to many other political spheres and times by the construction of appropriate questions (1950) asserted that Americans are subject to wild swings of mood in relation to
and the use of other instruments. Every interviewer of long experience can recall foreign affairs. Since there is no stable investment of feeling in that realm, policy-
those rare encounters where a respondent literally exploded with moral outrage or makers have only a shifting base of support on which to rely. It should be noted that
some other violent passion when some political question was put to him, but he can the original formulation and a subsequent critique many years later (Caspary, 1970)
also recall endless conversations on the most earthshaking questions that remained were both based on secondary analysis of national surveys.
utterly bland in tone. Interviewers' ratings of affect could readily be collected on a Political Knowledge. Men are also guided, if only imperfectly, by their cog-
systematic basis, and their spontaneous observations codified. nitive processes and levels of knowledge. This region has been subject to a great deal
The study of political affect may also contribute to our understanding of of research. Sears ( 1969, pp. 324-329) provides a convenient summary and secondary
self-propelled opinion leaders. They seem so essential and yet have been found to be analysis of such findings. Although some items of information are more readily
too few in numbers, in the occasional studies, to manage the magnitude of their absorbed than others, political ignorance is widespread and applies to many topics.
assignment in the body politic. Unless they are motivated by hope of gain, or recruit Further analysis, however, has mapped the distinctive social geography of the pattern.
themselves from the ranks of aristocrats who have both the leisure and the altruistic There are persistent concentrations of ignorance in certain places-notably, among
and moral impulse to elevate the mass, or are recruited and trained as professional those with little education. By intercorrelating the answers to long batteries of ques-
cadres, they would have to spring from the ranks of agitators, those who are propelled tions, the analysis also speaks to the problem of the generaIity or specificity of
to act unceasingly because of the depths of their own passions. Lasswell (1936a, knowledge. There appear to be "chronic know-nothings" politically and those who
p. 488) reminded us long ago that the "agitator has come by his name honestly, for "know it all7' (Hyman and Sheatsley, 1947). In some degree, these respective types
he is enough agitated about public policy to communicate his excitement to those of persons are socially determined, being prevalent in particular social groups, but
about him." Apart from old case studies, we know little about modern agitators or the analysis also suggests that idiosyncratic factors are involved. The problem is
their numbers because we have not charted the distribution of emotions in politica1 important to the welfare of democracies as well as to political theorists and psy-
surveys, nor singled out those a t the extremes for special study. chologists. Further survey research of a somewhat distinctive design could sharpen
The study of another dimension of affect, although not anchored in the our understanding.
immediate political environment, would add to our understanding of influence Obviously, life has made it easy for some people to learn almost anything,
processes. As earlier noted, individuals may be deterred from seeking advice not only whereas for others learning has been made exceedingly difficult. That is what the
by passivity or ignorance but also by the fears and suspicions that may pervade inter- basic analysis underscores. But the surveys also yield instances of deviant individuals
personal relations. Indicators and scales for the measurement of such variables have who despite their advantages end up as know-nothings, and of other deviants who
been found practicable and applied on a national scale and comparatively. surmount their handicaps and become know-it-alls. The surveys, if we only take
In their comparative study, Almond and Verba (1963) applied a scale for note, routinely function as locator devices to obtain unbiased samples of such
the measurement of "faith in people" previously developed by Rosenberg (1956, specialized, rare, and otherwise elusive types, and yield them in sufficient numbers, if
1957). Rosenberg found that scores indicative of "misanthropy" predicted anti- need be by slow accumulation and the pooling of many surveys. Subsequent intensive
democratic and anti-internationalist attitudes, but his results were based only on surveys of these special groups would be strategic for theorists and practical men.
samples of American students. Almond and Verba found that distrust was much Political Beliefs and Attitudes. Since men's political acts are guided not
more characteristic of the German and Italian samples in 1959 than of the British only by knowledge and feeling but also by their conceptions of and orientations
or American samples. Such feelings, of course, can be aggravated or allayed by cir- toward features of the political environment, this vast region of the political psyche
cumstances. In 1948, for instance, a sample of German citizens were asked whether, has been the subject of endless and varied surveys. T o be sure, in only a few realms
in their opinion, people can be trusted (NORC, 1948). In those most threatening can the individual's inner directives be directly translated into specific individual acts
times, distrust was almost universal among Germans, only 6 percent expressing trust, toward the objects in question (for instance, toward nonconformist neighbors or
By 1959 trust had risen considerably, to the level of 19 percent. Nevertheless, a toward minority groups in his town). With respect to the distant and overpoweringly
dramatic difference between Germans and Americans persisted. By semisecondary important features of the political environment, actions are reserved for those in
analysis, the combination of old data obtained incidental to the original inquiries and power, and ordinary men can express their orientations and beliefs only by voting
 348 Herbert M. Hyman Surveys in the Study of Political Psychology

their preferences or supporting some policy stand. In this connection, it should be educated strata being peopled by noneconomic liberals and economic conservatives,
noted that many findings in this realm are bound to be confusing and paradoxical. and the lower strata showing the opposite profile. In subsequent studies (Hero,
The irony of the ordinary man's fate is that he may have but few and unsatisfactory 1969; Campbell and others, 1960) the two basic clusters of foreign policy attitudes
choices to make for the political representation of his attitudes and beliefs. All sorts and domestic economic attitudes have been isolated repeatedly, but the two spheres
of men must stand under the same political banner simply because that is the only seem almost independent of one another, linked by only the most feeble of correla-
one available and though a poor representation of their views, the best they can tions. In perhaps the most definitive of such studies, Hero, by the secondary analysis
find. Given these circumstances, one must expect to find some heterogeneity among of national surveys spanning a thirty-year period, does document, however, that the
fellow partisans and low correlations between partisanship and single attitudes since linkage between the spheres has varied over time and that there are a few narrower
men cannot find anything or anyone to represent their total constellation of attitudes patterns of attitude whose contours do cross both domestic and foreign issues.
and must sacrifice some preferences to advance other desires. The structures revealed by statistical methods applied to arrays of measures
Within this broad region, surveys have been employed not only to describe necessarily have a somewhat abstract, intangible character. The elements do inter-
the content of beliefs and attitudes but also other formal dimensions-for example, correlate, hang together within the sample; but the investigator has no sense of the
the intensity with which beliefs are held. One such dimension that helps bring order psychic cement that binds them. He does not find a visible constellation of belief
into this domain, and is of special interest to political psychologists, relates to the and attitude in a concrete person. Some investigators therefore have adopted a more
matter of structure. Are specific beliefs and opinions and attitudes of common men clinical approach, examining the answers of each respondent to see how he articulates
part of larger constellations, perhaps even organized so thoroughly and elaborately the connections between his ideas and to determine whether there is an explicit pat-
that they form total belief systems, coherent ideologies about the entire world of terning. By such an analysis of surveys conducted during the 1956 presidential elec-
politics? Alternatively, are they discrete, so that there is no consistency in an indi- tion, Campbell and his associates (1960) concluded that only a small minority of
vidual's attitudes across related situations and toward features of the environment Americans, perhaps less than 10 percent, exhibit an ideology in the sense of a struc-
that in fact have much in common? (See Chapter Three and Four In this volume.) ture formulated in terms of some modestly abstract set of concepts that political
This is one of the classic problems of psychology and is obviously of special interest scientists and sociologists of knowledge would accept as ideological.
to political psychologists, either because they believe that there is some unity to the Such an approach tells us a good deal about the clarity and style of political
person or because they see politics as the clash of ideologies or as devoid of ideology. thought of the common man. It does not deny the existence of attitude structures
As empirical researchers, they would find their task greatly simplified if they could and belief systems, but it sharpens our realization that such structures are cemented
count on the existence of comprehensive attitude structures or belief systems. Then not by abstract ideas but in some other ways. Indeed, statistical analyses replicated
they viould not have to engage in endless itsy-bitsy measurements and could simply over the 1956 and 1960 elections (Campbell and others, 1960; Converse, 1964)
locate individuals along the few major axes that define the basic structures within revealed the two main subsystems of attitude found by the other investigators. This
which the specific entities are contained. Alternatively, if there were a great deal of clinical approach also bears on the larger question of intellectual functioning and
specificity, they might be urged to caution in testing theories about personality presents us with a paradoxical problem. Ideologies, being pushed and sold by
determinants of political attitudes, since so much would then seem to be determined ideologues and elites, pervade the political environment; judging by the 1956 findings,
by the situation rather than by the inner man. And they would have to move in the however, the common man shows little receptivity or capacity to absorb such ideas.
direction of many measures of attitude in the hope of transcending the specificities Perhaps the explanation is that few such ideologies were being pushed during that
and finding some thread of unity to tie on to their personality measures. particular bland period. Field and Anderson (1969) had this very hypothesis and
That the level and nature of meAtal organization do vary in different universes tested it with surveys conducted in 1964, when Senator .Goldwater was aadramatic
has been empirically documented. Thus, surveys of the problem in the general popu- symbol of conservatism and stimulated a sharp clash of ideologies. Ideological thought
lation are especially welcome and fairly plentiful. For example, by intercorrelational had expanded to about one third of the population by then; but about two thirds of
methods, sometimes carried to the point of factor analysis, applied to a series of the population still did not function in this way, despite the fact that the scoring
heterogeneous attitude measures, G. H. Smith's (1948aJ 1948b) early finding that for "ideology" was lenient and the stimulus most compelling.
there are "two kinds of liberalism" has been confirmed; there may be even more than Such findings permit us to study a few relatively broad systems of political
two (Ailinsmith and Allinsmith, 1948; Williams and Wright, 1955). For sure, in beliefs and attitude, to locate and describe individuals along a few general dimen-
America in the 1940s there was no total ideological system in the common man that sions. But we are not instructed by the studies of structure to ignore narrower topics
disposed him toward liberalism or conservatism of every kind. But there was a con- for inquiry in the region of political attitudes. Little subsystems of belief, even discrete
siderable degree of organization: domestic economic attitudes clustered into a con- attitudes, may be of critical importance and strategic objects of investigation to which
sistent pattern of liberal or conservative views; and attitudes toward foreign affairs we may wish to devote whole surveys or strings of surveys. Beliefs about and popular
and noneconomic domestic matters, such as civil liberties and the, treatment of support of war, beliefs and attitudes relating to civil liberties for political nuncon-
minorities, formed a liberal or an illiberal cluster. In Smith's studies, at the level of formists, attitudes about the president and the way he performs his duties-these are
ecological correlations, the linkage between these two systems was negative, the more examples of sdch specifics. Indeed, these very examples have been studied through
 Herbert H. Hyman Surveys in the Study of Political Psychology 351

surveys and often accomplished simply through secondary analysis (see Mueller, 1970, no sense of duty to participate. The former is truly the "high-minded citizen," since
1971; Hamilton, 1968, Hyman and Sheatsley, 1953; Lipset, 1964; Lipset and Raab, he believes he has nothing to gain from his participation.
1970). Consider as another example the joint distribution created by measuring
A last set of beliefs deserving of special study are the beliefs about other political knowledge and political attitude. We arrive at some estimate of those who
people's beliefs and attitudes. Floyd Allport (1924) and Katz and Schanck (1938) might be labeIed the "political irresponsibles" (people who hold a strong position in
talked long ago of pluralirtic ignorance, the false ideas that many of us have about the absence of knowledge) and of the "political Hamlets" (people who know a lot
each other's ideas. In mass society, the barriers to correct knowledge may be great; and yet have lost all sense of where they want to go politically). In these ways surveys
but pluralistic ignorance is also found in the little community. I t may even be found
among those who are members of the same association (Breed and Ktsanis, 1961). I can provide a many-faceted and comprehensive description of the individual.

All sorts of whimsical or tragic processes may flow from pluralistic ignorance, with
consequences for politics. Those suffering from what Allport called the "illusion of I Explanations of Political Behavior

universality" may fight on, encouraged by the belief that they have an army of mil- Surveys can describe the politics of men, but the task of explaining their
lions behind them-in fact, a phantom army-and may validate their own views by behavior still remains and may appear to some to be beyond the powers of survey
a consensus that is only the product of their fantasy. Others, suffering from another methods. It should be noted immediately that one really has moved far in the direc-
variety of pluralistic ignorance, may surrender their position in the false belief that tion of explanation when one has achieved adequate description. Catching the right
no one shares their views, when in fact they are quite an army. hypothesis is half the battle; testing it in the right way is only the second half.
Political Character. By the systematic description of the characteristics of Obviously, it is much better to catch it quickly rather than to find it only after a long
individuals on these major types of variables, we arrive ultimately, adapting Katz's string of negative tests of fruitless hypo these^.^
phrase, at a "psychological taxonomy of political behavior." Along the way, to achieve How shall one catch the right hypothesis and put it to a good test? We are all
comprehensiveness, we may explore within our surveys new regions and plot addi- familiar with the contribution of serendipity to the growth of theory. That needs no
tional specific dimensions in greater detail. For example, some may wish to study review, but the special relevance of surveys to the experiencing of serendipity does
a special region of political values. Whatever one might add to ensure the realization deserve brief mention. Anomalous observations would seem to be chance occurrences
of the taxonomy, survey methods can provide the descriptions of the populations for that rarely come our way. Yet by casting the wide net of a descriptive survey over
many times and places. Yet some psychologists might find the uItimate product un- a total population, the investigator is more likely to capture such anomalies, and in
satisfying. They might regard it as only a catalog or inventory of parts of men, in no turn catch a good hypothesis.
way describing the whole man, the separate characteristics never being joined together Katz remarked that "a naturalistic description of the behavior of human
to form the political character. Yet this would be to underrate the degree of syn- beings in various social contexts" is a prerequisite to understanding. The ability of
thesis that can be reached in survey analysis. When we present a description of a a survey to work as a locator device helps in many ways in hypothesis catching. The
general variable, a composite index of political knowledge, or political activity, or distributions in the aggregate, over time, and for major social groups provide a large
liberalism, or ideologically flavored thought, we have already reached a halfway stage set of statistical norms for locating a particular pattern and sensing what it means.
between the description of a discrete characteristic and the total political character. Knowing that it is deviant or normative in the population at large and in specified
We can move still further toward complex description, toward typological constructs groups, persistent or transient, parallel to other patterns or a discordant feature of
which convey a profile or combination of the different political features of a man. a larger profile, we are less likely to concoct arbitrary assumptions and wild theories,
Instead of locating individuals along one dimension, to produce a univariate distri-
and more likely to move in the direction of sound hypotheses. If one is still confused,
bution of the population, we can locate them by reference to scores on two dimensions
the initial surveys can serve to locate unbiased samples of individuals who exhibit
(producing a bivariate distribution) or by reference to three dimensions (producing
the mysterious entities for subsequent deeper exploration, leading ultimately to clari-
a multivariate distribution) .
fication.
To cite one example, the Survey Research Center developed scales of "citizen
Within the ordinary survey, the analysis usually involves both descriptidn and
duty" and "political efficacy," to measure two dimensions of belief that underlie
explanation. Ironically, the mix these days is all in the direction of explanation;
political activity, and described the separate distributions of the American population
description, despite its utility, is regarded as pedestrian and only a lowly exercise of
on each of these dimensions over time. Eulau and Schneider (1956) simply com-
the powers of a survey. A special methodology, elaborated for testing causal hypoth-
bined the two scales into a superscale of "political involvement." This approach
yields a more comprehensive description of the forces that affect an individual's
3 This is not to deny the value of eliminating hypotheses by successive disconfirma-
actions, and we are then in a position ,to isolate interesting types for further study. An tions and the possibility in some instances that a negative finding indirectly supports another
individual with little efficacy who is nevertheless impelled by a high sense of duty is hypothesis, which is the only alternative explanation. For such possibilities in survey analysis,
a very different moral type from the individual who despite his feeling of efficacyhas see Hyman (1955, pp. 226-241).
352 Herbert H. Hyman Surveys in the Study of Political Psychology 353

eses with survey data, is codified in easy steps in the literature (Glock, 1967; Rosen- nevertheless, can help them locate the psychological causes and links and bring those
berg, 1968; Hyman, 1955). Whatever hypothesis may guide each political psycholo- to ultimate tests. A recent monograph illustrates the sequence. Reed (1972) estab-
gist to measure some favorite independent variable, all, as survey analysts, can follow lished by the secondary analysis of many surveys that there are enduring differences
these steps toward tests that are logically sound and built upon the solid foundations between southerners and northerners in political and other respects. One might say
of good sampling. Because thd single survey is a spacious vehicle that can carry many that being southern is a powerful determinant of behavior; but one might ask, as he
variables, no analyst need risk his whole inquiry on an exclusive and possibly .mis- did, exactly what it is about southern life that causes the patterns. By multiple
guided hypothesis. By a kind of benign Parkinson's Law, he is led to fill the vehicle to standardization-in effect, the simultaneous control of a set of variables-he estab-
capacity with many variables, and thus is guided naturally toward complex models lished that the differences cannot be explained away by the relative lack of industriali-
of causation. He can examine one or more individualistic determinants of some zation, urbanization, and education in the lives of southerners. Multiple standardiza-
phenomenon, but observe how their influence is modified in different social contexts. tion, defined more precisely, yields the hypothetical distribution that describes what
He can examine influences out of the distant past, but as they have been overlayed southern behavior would be like if southerners had the same amounts of urbaniza-
by recent stimuli. He can trace political behavior back to its deeper roots in some tion, education, and industrialization as northerners. In one sense it is description, but
fundammtal motive or cognitive or affective entity, some style of personality or it also constitutes an explanatory test of a complex of causes and, in this case, led to
intellect-and not just to a single such source; but he can also trace the links in the conclusion that some other subtle and residual features of southern culture are
tf;e long causal chain that connect the private and political spheres. Some of the the true causes. Reed then theorized that the South is a cherished normative reference
variables previously presented only as ways to describe the political sphere indeed group and that patterns of socialization help perpetuate its distinctiveness. He was
may be treated as just such causal links-for example, political beliefs and attitudes then able to make direct tests to support these social-psychological hypotheses.
as underlying political action, and political knowledge as shaping beliefs and attitudes. The paradox that groups often do not produce the patterns one would expect
Initial surveys, as earlier noted, can also serve to locate unbiased samples in most of their members was a major stimulus to the rise of reference-group theory.
of rare and elusive groups who exhibit a set of independent variables implicated in Differentiated descriptions, the distributions for various social groups yielded by
some hypothesis, whose influence is then tested. In this connection, one may suggest surveys, provide a continuing source of nourishment for the theory by showing how
the rare potentiality for locating individuals whose particular political and psycho- much variability exists within a group and how much uniformity between groups.
logical characteristics at a much earlier stage of their lives were measured by surveys One may conceive of the variability within a group as stemming not from the fact
out of the distant past and whose subsequent long-term development is traced by that some individuals adopt strange reference groups but simply as the product of
impaneling them in a new survey years later. Ethical and technical difficulties stand their varied multiple memberships. But, as already noted, surveys can classify indi-
in the way, but designs involving such selective expe~imentalimpaneling should be viduals along multidimensional lines and then document how much variability per-
regarded as a real possibility. (See Hyman, 1972aJpp. 58-61; see also Chapter Two sists. And special measures of the normative and comparative reference groups that
of this volume.) individuals seleci are readily incorporated into surveys, and the effects of such psy-
In these specialized ways, surveys can move from description toward explana- chological variables put to direct test (for such examples, see Hyrnan and Singer,
tion, toward good hypotheses and good tests. But it may also be argued that one 1968).
already has arrived at an explanation when one has established securely the differen- In these various ways, descriptive surveys lead toward explanations of politi-
tiated descriptions of political characteristics. For example, the contrasted distributions cal behavior. Some political psychologists may be creative enough to catch a good
of political activity among young and middle-aged people or poor and middle-class hypothesis right out of their own imagination or from out of some prior body of
people-distributions described time and again by dozens of surveys in many countries theory. Then, of course, they can move immediately to an expianatory survey con-
--establish that age and class are causes of political activity. Some analysts would taining measures of the appropriate variables. There is, however, still another avenue
find such a notion of "cause" too crude and not psychological enough for their taste; to explanation that can be explored in survey research: the study of processes of
but it is a perfectly legitimate assertion as long as it is protected against spuriousness, learning and development as explanations of political behavior. Nothing could be a
and that is readily done in survey analysis. If one wishes, however, to rnove beyond more cefitral concern of p~ychologythan learning and development, the way indi-
this gross level, refinement is easily achieved. One can easily test the effect of combi- viduals are altered by experience over the course of their lives and naturally grow
nations of social characteristics working simultaneously to reinforce each other or to and change and decline. Surveys can reveal such processes over the long course of
create cross pressures; or he can test the effects of status sequences, movements over time and often also isolate their sources, although, as always, there are causes behind
time from one social position or milieu to another. causes.
Some might still find this level of causal analysis too "social," seeking instead The many surveys of political socialization need no review here (see Chapter
to isolate those psychological factors that regularly accompany a s t a h or follow from Five). Clearly, the young are a population strategic for study. Special surveys of
it, or those that impinge in no socially ordered fashion upon individuals and "truly samples at various "early" age stages are readily conducted and provide a quasi-
cause" a process and an effect. What they would regard as pure social description, longitudinal picture of an important segment of the longer process of political devel-
354 Herbert H. Hyman Surveys in the Study of Political Psychology 355

opment-although, of course, there are technical difficulties in conducting conven- demonstrate and quickly interpret in a unidirectional fashion may sometimes reflect
tional surveys among the very young and in sampling young adults, who often are a process moving in the opposite direction. A political system can invade and alter our
inaccessible in such institutions as the army and residential colleges. psyches, just as it can reflect our prior dispositions. A legal system can legitimate our
Surveys among adults always enumerate age. This datum provides the basis highest or basest impulses, and it can reflect them.
for quasi-longitudinal analysis of a long segment of the process, either in a specially We emphasize the reverse perspective not just to urge caution in our inter-
conducted inquiry or through the secondary analysis of a great many surveys. Some pretations but to elevate neglected problems to a higher place in our research agenda.
investigators may reconstruct features of the distant past and even try, as Almond We would then revive a great, if uncommon, tradition in psychology and a classic
and Verba did, to obtain entire political life histories-producing, so to speak, a concern of political science. Recall Lewin's interest in the effects of authoritarian
complete longitudinal inquiry. These several approaches do present technical diffi- leadership and the memorable questions that political theo~istshave asked about
culties. Problems arise in the interpretation of comparisons between age groups whether a given political system is bad for men. Recall that Lasswell (1960), who
drawn from within the same survey, in deciding whether the differences reflect the had dealt with the psychological sources of political behavior in Psychopathology
general aging process or the particular experience of different generations. But these and Politics, finally remarked that "the purposeJ' of psychological analysis "is to
problems are not insurmountable. Ironically, the secondary analyst often is in a much reveal the significance of any institutional practice . . . for human values," to
better position than the primary analyst, since he can draw together a larger body of appraise it "in terms of importance for man'' (p. 276). Psychological data in abun-
data from different time periods. Sometimes, he can apply the method of cohort dance in ordinary surveys tell us of man's sensitivities and satisfactions under a
analysis to a series of available trend surveys, producing definitive evidence on the particular political order. Trend, and comparative contextual, and cross-national
general process of aging, and also on the influence of the specific experiences of surveys could systematically document the psychological effects that varying political
different cohorts or generations (For a summary of such studies, see ~ ~ r n a1972a, i, institutions produce. We could and then would deal with noble problems,
ch. 7.)
Some may find such approaches too inferential and too fraught with techni-
cal difficulties.I t should be noted that truly longitudinal surveys, special panel studies
of the changing politics of adults, have been maintained over periods as long as five
to six years, providing evidence on a fairly lengthy segment of process (see, for
example, Campbell and others, 1966). By the strategy of selective experimental
impaneling, it would be possible to bring an even longer segment under scrutiny.
It is ironic that special surveys of the very old are very rare, while surveys of
the young are fashionable and multiply every day. The old are an equally strategic
population for the student of learning and development. They too should be sur-
veyed despite the technical difficulties in sampling the full population, some of whom
are contained within institutions, and in interviewing those who are feeble and
handicapped. Here, too, it is comforting to note how much can be accomplished
through secondary analysis by highly skilled investigators, and one can only hope that
more investigators will expIore these farthest reaches of time.

This chapter has reviewed the many ways that surveys can describe and ex-
plain the political behavior of individuals and in turn help us to understand the
political system. That is enough of a contribution. But we cannot conclude without
mentioning a rare contribution that surveys can also make.
Examining politics from the perspective of psychology is somewhat like the
task of the subject in old-fashioned experiments who was shown a picture where
figure and ground kept reversing themselves in his perception. The political psy-
chologist faces a similar situation, although it has not been so apparent to him. He
too can reverse his normal rigid perspective. Instead of looking at politics only as
the figure against a background of psychological variables from which it emerges,
he can focus on the psychological as the figure against a background of the political
,qystem that has shaped it. The study of the relation between psychology and pditics
can move in both directions, Indeed, it is likely that the empirical relations we
 Experimental Research 357

hope of further advancing this trend to incorporate experimentation into political

psychology, this chapter is written with a very definite readership in mind: those
behavioral scientists interested in political psychology who wish'to add experimental
techniques to their repertoire of methodological skills.
Because of space limitations, not every aspect of experimental design and
execution can be covered. Therefore, those that are covered in detail represent aspects
of experimental methodology which (in my opinion) need the most emphasis for the
above readership, given the current state of the experimental art in political psychol-
ogy. Readers wishing more details about certain aspects should turn to Campbell and
Stanley ( 1963) , Aronson and Carlsmith ( 1968), Kiesler, Collins, and Miller ( 1969),
and Winer (1971). Readers should be warned, however, that after reading this
EXPERIMENTAL chapter and the above works, they will not yet be good experimenters. To learn to
be good experimenters, behavioral scientists must start doing experiments.

A Sexy Illustration

We turn immediately to an illustration from a realm which older generations

saw as nonpolitical but which has recently become the subject of much political
debate (see Millett, 1970; J. Q. Wilson, 1971) and even served as the topic of a
John B. McConahay presidential commission study (Commission on Obscenity and Pornography, 1970).
In their monumental studies of sexual behavior, Kinsey and his colleagues
(1948, 1953) found that 77 percent of the males and 32 percent of the females in
their samples reported being aroused by "photographs, drawings, motion pictures,
and other portrayals of sexual action." Kinsey used a self-selected or volunteer sample
and "in-depth" clinical interviewing techniques, which involved open-ended questions
with multiple probes. Using a national probability sample of adults and straight-
forward "survey interviewing" techniques, Abelson and his associates (1970) found
that 23 percent of the males and 8 percent of the females were aroused by sexual
materials. Finally, Berger, Gagnon, and Simon (1970) found in a sample survey of
F o r most of its history, political psychol- American college students that 46 percent of the males and 14 percent of the females
ogy has relied almost exclusively upon the sample survey and the in-depth interview were aroused by exposure to "pornographic photographs, drawings, comic books,
as its chief empirical methods. However, recently we have seen an awakening of movies, and writings." The results of these surveys-using three different types of
interest in a method dominant in other branches of psychology for years, the con- samples, two different types of sampling techniques, and two different methods of
trolled experiment. A few examples of this emerging interest are the establishment interviewing-appear quite consistent with Kinsey's conclusion that men are more
of a laboratory for political research at the State University of New York at Stony likely than women to be aroused by sexually oriented materials.
Brook; the founding of a new scholarly journal, Experi~nentalStudy of Politics; the When males and females were exposed to such materials under experimental
popularity of the experimentally oriented Program of Advanced Training in Psy- conditions, however, a different set of findings emerged. Jakobovits (1965) found
chology for Political Scientists, now in its fifth year at Yale University; the increasing that males and females are equally aroused by erotically "realistic" stories and that
number of "experimental" articles appearing in established political science journals females are significantly more aroused than males by "hard-coreJ' stories. Similarly,
(see S. R. Brown, 1970) ; and the appearance recently of symposia devoted to Sigusch and his associates (1970) found that women are more aroused on the average
experimental research in political science (see Laponce and Smoker, 1972). I n the than men by exposure to pictures of heterosexual intercourse.
Work on this chapter was supported in part by USPHS training grant 5 TO These data illustrate two of several pitfalls of a political psychology based
1MH11526, Advanced Training in Psychology for Political Scientists, to Yale University, J. exclusively upon the sample survey as a method of research and two of several
B. McConahay, Director. The author wishes to thank Diane Straus and Shirley F. valuable contributions that experimental procedures can make to the discipline
McConahay for their research assistance and Noreen O'Connor for typing multiple drafts (other gains will be discussed later). First, the findings of survey research may
of the manuscript. simply reflect widely held myths about human behavior or the sample's consensus
about what 'the socially desirable answer to the interviewer's question should be.
 John B. McConahay Experimental Research 359

Thus, when asked if he is aroused by sexually oriented materials, a male respondent both psychological predispositions and the stimulus properties of the immediate
might say "yes" because he believes that most males are aroused by such stimuli or situation. Survey research, though conducted in "natural" settings, is rarely able to
because he thinks that the interviewer or researcher will regard him as less than mas- take into account the specific context of the behavior reported by interviewed sub-
culine if he responds that he is not aroused. Similarly, a woman might answer "no" jects. Hence, one might overlook the contexts controlling respondents' various reports
when asked the same question because she believes that women in general are not of their behavior and, as in the above example, conclude that women are not as
aroused or she fears that the interviewer will regard her with contempt for admitting aroused by erotica as men in any and all contexts. ~ x ~ e r i m e n t research,
al though
to a response that nice girls and nice women are not supposed to have. frequently conducted in "artificial" settings, makes the environmental setting quite
A second pitfall is that the results might be interpreted as reflecting basic explicit and examines the ensuing behavior on the basis of the joint effect of both
laws of political psychology when they merely reveal a picture of the way people environmental and predispositional factors. If we are going to make theoretical and
are currently behaving in response to existing political and social stimuli. Thus, in empirical advances in political psychology, we must take both factors into account
our example, the differences between men and women in the various surveys might and consequently must use both experimental and survey methods (along with
have resulted from the fact that pornography (until very recently) was an exclusively others).
male domain. I t was produced by men for male entertainment. Hence, to both male
Experiment Defined
and female respondents, the terms sexually oriented or pornographic might have
suggested pictures of nude females, striptease shows, and the Playmate of the Month. We have so far used the terms experimental and experiment rather freely, but
Such materials would arouse heterosexual males but have very limited impact upon what exactly do we mean by them? In this chapter, we shall use the terms experiment
heterosexual females. Consequently, even without the distorting effects of mythology or experimental design to mean a situation in which certain variables are manipulated
and social desirability, males would report having been aroused by such materials in such a way that the subjects cannot choose or select the level or type of treatment
while females would remember being unaroused. they receive; then the results of these manipulations upon other variables are ob-
The laboratory experiments overcame both of these difficulties by using phys- served. This definition is quite similar to that of others (Brownlee, 1960; Campbell
iological as well as self-reported or subjective measures of arousal and by presenting and Stanley, 1963; Cox, 1958; Edwards, 1960; Fisher, 1925, 1935; Winer, 1971)
pictures or stories intended to arouse heterosexual women as well as men. The except that most of the others, writing in the agricultural or laboratory psychology
physiological measures, in addition to providing an independent check upon the tradition, assume that the experimenter controls the independent manipulation and
arousal level of the males, called the attention of both males and females to their determines which subject receives what level or type or treatment. While this pre-
physiological state and made them less likely to give false levels of arousal on self- experimental equalizing of treatment groups by random assignment is often possible
reported scales. (The subjects may have thought the investigators would know how in political psychology (and, as a technique for equalizing, is superior to matching on
aroused they were anyway.) The pictures of nude men and the stories about nude one, two, three, or n dimensions), the researcher does not always have this power.
men and heterosexual intercourse were sexual stimuli that the women would rarely There may be occasions when God, a bureaucracy, or fickle fate assigns subjects to
have encountered under commercial or "natural" circumstances. Hence, they reached treatment conditions while the researcher bravely takes advantage of these ongoing
heights of arousal from these materials that they would never have achieved looking experiments to collect data and share the results with his or her colleagues. The
through the pages of Playboy. essence of an experiment in this chapter, then, is not that the experimenter (E)
controls and randomly assigns the treatment (X) but that some random or chance
Need for Multiple Methods mechanism beyond the control of the subject (S) is used to achieve pretreatment
equality of experimental groups upon the relevant dependent variables. Therefore,
The point of this long illustration is not that experimentation is the answer experimental in this chapter is not synonymous with laboratory, since the random
to all research problems in political psychology. Other methods-for example, pro- assignment of subjects to treatment conditions may occur in a laboratory or in a
jective techniques, corrections for social desirability, and cross-cultural surveys used field setting.
as additional techniques-might have helped Kinsey, Abelson, Berger, and their Where pretreatment equality of groups is attempted or assessed by some means
associates to avoid their pitfalls. Furthermore, sample surveys are indispensable for other than random assignment, we shall, after Campbell (1968; Campbell and
developing a picture of the ways in which people are currently behaving in our Stanley, 1963), refer to the research design as quasi-experimental. These designs are
political or social culture. Perhaps the greatest contribution of the survey researchers not ideal but may be the best available and may yield valid data to a sensitive re-
has been the destruction of social myths about the .ways people conduct their sex searcher.
lives, make up their minds about electoral choices, and participate in urban violence
Previous Studies
(Sears and McConahay, 1973).
Experimentation, then, is not the answer to all research problems in political Researchers wishing to do experiments in political psychology will either have
psychology, but it is an indispensable additional approach. As many of the contribu- countless or very few examples to guide them, depending upon the criteria for
tors to this volume emphasize, behavior in a specific instance is a joint function of selection. Since all of psychology is important for understanding political behavior,
 John B. McConahay I Experimental Research 361

the political psychologist could start with studies of psychobiology, animal learning, In the second experiment, Eldersveld (1956) replicated this turnout effect for
and autonomic conditioning. Indeed, the laboratories for political research at the propaganda and also explored several alternative techniques of propaganda. In this
State University of New York at Stony Brook are among the finest physiological experiment, conducted in 1954, residents of Ann Arbor who had voted in national
psychology facilities at any university. and state but not local elections were exposed to different types of propaganda urging
Of more direct relev*nce, perhaps, are the studies by psychologists which them to vote in an upcoming local election. This hard core of five hundred local
have rather obvious poiitical implications even though their authors were not neces- election apathetics were assigned at random to one of six experimental groups or to
sarily interested in pursuing these implications. Here we might include (among many a control group. The experimental groups received the following propaganda treat-
others) the classic studies of attitude change and mass communication (Hovland, ments: (1) personal contact by students, (2) personal contact by party canvassers,
Lumsdaine, and Sheffield, 1949; Hovland, Janis, and Kelley, 1953), the studies of (3) telephone contact, (4) contact by students and by mail, (5) "rational" appeals
selective exposure (Freedman and Sears, 1965), the social-psychological studies of sent through the mail, and ( 6 ) "moral" appeals sent through fie mail. The control
leadership (Fiedler, 1965), the experiments in violence and aggression (Berkowitz, group was not contacted at all. The results (based.upon a check of the city clerk's
1969), and the work of Christie and Geis (1970) on Machiavellianism. records to see who actually voted) replicated the turnout effect for personal contact,
Special mention should be made of the work on obedience by Milgram which differed significantly from both the noncontacted groups and from the mail-
(1965) and on deindividuation by Zimbardo (1969). Milgram found that Yale contact groups. Mailed propaganda did not produce a significant turnout effect as it
students (and later, ordinary people off the streets of Bridgeport, Connecticut) hgd in 1953. Although one might find some fault with certain aspects of the experi-
would give supposedly lethal levels of electric shock to persons simply because they mental techniques used in these two experiments, they still stand as well-executed
were ordered to do so by an authority figure. Zimbardo found that people, when experiments and are even more awe inspiring when one considers the state of the
deindividuated, would give increasing levels of shock even to persons they liked, once experimental art in the mid 1950s.
others started doing it. The implications of these experiments for a political psy- The third "true" political psychology experiment executed before 1970 is the
chology of the rnilitary and for an understanding of My Lai are obvious and fright- work of Kamin (1958). His dependent variable was voting not in an actual election
ening. (See Chapter Six for a discussion of some of these implications.) but in a fictitious electoral contest. That is, Kamin took advantage of the low level
Political psychologists wishing to learn from experiments with manifestly of information in most mass publics (Hyman and Sheatsley, 1917; Converse, 1964;
political content will, at present, have few examples to guide them. Some field studies Sears, 1969) to present his subjects with a choice between fictitious candidates for
of campaign techniques or propaganda and electoral choice (Blydenburgh, 1971; a fictitious o%ce in the guise of asking opinion-polling questions. The subjects were
Hartmann, 1935) are not experiments by our definition (even though they have the English and French Canadians, and the fake opinion poll asked for their choice of
words experiment or experimental in their titles). They tend instead to be quasi- candidates in an election that pitted ca~zdidateswith French surnames against those
experiments, since they do not use random assignment to determine ievels or types of with English. The party affiliations of the two fictitious candidates and the position
treatment. Other studies (for instance, Barber, 1966) are sometimes loosely refered of names on the ballot were also varied at random. The results showed that, in the
to as experiments but are more appropriately designated as simulations (Abelson, absence of party labels, the English preferred English candidates and the French
1968bj. Finally, many studies did incorporate some form of random assignment into preferred French Eut that names last on the ballot (regardless of ethnicity) were
their designs but failed to have sufficient replications (Reback, 1971) or used the preferred least. When party labels were present, both English and French subjects
wrong number of degrees of freedom in their data analysis (Gertzog, 1970). preferred the candidate of their party regardless of the candidate's ethnicity. A
Three experiments have been chosen somewhat arbitrarily for review here procedure somewhat similar to gamin's was used in 1966 by Lorinskas, Hawkins,
because they are ail in one field, electoral behavior, and are among the best-executed and Edwards (1969) to study ethr~icityand party identification among urban and
experiments in any field. The first, a study by Eldersveld and Dodge (1954), at- rural voters in Illinois with Polish surnames. It is not reviewed here because of cer-
tempted to assess the effects of different propaganda techniques upon voter turnout tain difficulties in the urban-ru.ral comparisons and because of errors in their data
and preference. In 1953, subjects in three wards in Ann Arbor, Michigan, who had analysis.
previously expresed ignorance of or hostility toward a proposed city charter revision Despite the valuable contributions of these experiments to political psychology
were assigned at random, and as individuals, to receive one of three treatments: and the success of some social-psychological experiments in illuminating certain
(1) mail propaganda, (2) personal contact by students posing as campaign workers processes important for a theoretical understanding of aspects of political behavior,
in support of the charter revision, and (3) no mail or personal contact by the re- in recent years a number of manuscripts reporting upon experimental studies were
search team. The data, gathered in postelection interviews, suggested that both mail rejected (quite unfairly, in the opinion of many investigators) by the established
and personal contact will increase voter turnout (turnout efect; see Kramer, 1970) political and behavioral science jourcals. Thus, to provide an outlet for further
and make voters more likely to vote in the advocated way (preference efect; see research into political phenomena and to stimulate further experimental research in
Kramer, 1970). Furthermore, personal contact produced a greater turnout effect this field, a new journal, Experimental Study of Politics, was founded in Febniary
(statistically significant) and a greater preference effect (marginally significant) 1971. The appearance of this new journal means that those interested in experi-
than did the mailed propaganda (both statistical tests based upon my calculations mental approaches to political psychology will cot be forced to publish only in
from their data). psychology journals, where the study of politics is considered an applied rather than
362 John B. McConahay Experimental Research 363

a basic research field. Furthermore, researchers will now have an outlet that might generally untestable by any technique. (For an exposition of this problem in the
catch the eyes of more persons concerned with "nonpsychological" aspects of politics. area of democratic theory, see Cnudde and Neubauer, 1969.) Without middle-range
The articles thus far published in Experimental Study of Politics (ESP) are theories, both psychologists and political scientists have sensed that the disadvantages
not yet as methodologically sophisticated as those in the established psychology jour- of the experimental approach outweigh the advantages.
nals (Journal of Personality and Social Psychology and Journal of Experimental The major disadvantages usually cited by critics of experimental approaches
Social Psychology, for example), but they are improving with each issue as political to psychology and politics are the artificiality of the manipulations; the unrepresenta-
scientists learn to do experiments by doing them, and they are not as sterile as articles tiveness of the samples studied; the impossibility of generalizing from the controlled,
in the established psychology journals either. bivariate or trivariate laboratory to the multivariate world; and the cost of over-
coming this artificiality, unrepresentativeness, and isolation by moving toward field
Advantages and Disadvantages experiments or bigger, more elaborate, better-staged laboratory experiments. (For the
details of these criticisms, see Chapter Twelve in this volume.)
As Sears (1969) has pointed out, there is no reason in principle why experi- These criticisms are generally but not entirely valid when applied to laboratory
mental studies of political psychology should not prove fruitful. Yet, as we have seen, experiments. Moreover, the charges of artificiality and isolation are frequently based
very few experiments have been attempted in political psychology or more general upon a misunderstanding of what the laboratory experimenter is trying to do, a con-
political behavior. Perhaps political scientists and sociologists simply are unfamiliar fusion of "mundane" with "experimental" realism (see Aronson and Carlsmith,
with experimental techniques; or perhaps academic psychologists-fearing that they 1968). Furthermore, experimentalists are not the only researchers who are guilty of
would be thought of as applied psychologists by their academic colleagues-have making theoretical leaps from possibly artificial or weak empirical operations. The
shied away from studying political behavior; or perhaps the subject matter of poli- survey and correlational studies of alienation provide a classic case in point, as their
tical psychology simply does not lend itself to experimental study. Most probably, critics have pointed out (Keniston, 1960; Kraut and McConahay, 1971; Seeman,
however, existing political-psychological theories have made experimentation in- 1971). Finally, the high cost of overcoming these difficulties by moving toward field
appropriate to this date. experiments or bigger, more elaborate, better-staged laboratory experiments is also a
As with any other research technique, the experimental method has both its factor in correlational research. A sample survey of adults in the United States or
advantages and disadvantages. The productive researcher should choose a technique even south central Los Angeles is a very expensive task. Consequently, many correla-
that best suits his problem, one whose advantages in that specific situation outweigh tional analyses use dubious operationalizations of theoretical concepts imposed upon
its disadvantages. What are the advantages of the experimental approach? Advocates somebody else's questions, which are then reanalyzed and reinterpreted. And many
of experimental methods usually agree upon the following advantages, summarized correlational analyses combine the worst of both worlds by giving paper-and-pencil
by Aronson and Carlsmith (1968, p. 10) : "In sum, the major advantage of the questionnaires to captive audiences of college students or other unrepresentative
laboratory experiment is its ability to provide us with unambiguous evidence about samples. Thus, the problems of social desirability, face validity, and unrepresentative-
causation. Second, it gives us better control over extraneous variables. Finally, it ness of subjects are combined with the inability to untangle questions of causation.
allows us to explore the dimensions and parameters of a complex variable." The principal advantages of an experimental approach to the study of politi-
Aronson and Carlsmith, writing for an audience primarily composed of true cal psychology may be summarized as follows: ( 1) The experimental approach per-
believers in the experimental approach to social psychology and of those who would mits greater confidence in causal inferences than with other procedures. (2) There is
soon be socialized into the state of being true believers, could get away with making greater control of extraneous variables. (3) There is better opportunity for precise
what other observers might regard as an exaggerated claim for the experiment's exploration of the dimensions and parameters of a complex variable, so that one can
ability to settle questions of causation. In theory an experiment may be able to study process as well as outcome. (4) The approach affords greater control of social
provide unambiguous evidence of the direction of causation; but social science desirability than with other procedures. (5) In conjunction with appropriate theories,
studies, using any technique, rarely provide completely unambiguous evidence, and one is able to create, explore, and evaluate conditions that have not yet evolved or
in many if not most complex social and political interactions the direction of causa- developed "naturally" in the existing political and social world.
tion is rarely one way. Nevertheless, as compared with correlational and observa- The disadvantages of an experimental approach to political psychology may
tional techniques, the experiment does allow for relatively more control of extraneous be summarized as follows: (1) The manipulations and settings, especially of labora-
variables and does permit detailed exploration of complex variables. In short, experi- tory experiments, appear artificial. (2) The samples used are generally haphazard
ments are invaluable tools for theory building and theory testing. As indicated, (Kish, 1965) and unrepresentative of any specifiable population. (3) The control of
surveys and other correlational techniques are great for discovering and documenting extraneous variables (deemed an advantage above) introduces bias, which limits
things as they currently exist or changes as they take place, but their utility diminishes the generalizability of experimental findings to the multivariate "real" world. (4)
as we attempt to explain why and to examine possibilities that have not yet occurred. Experiments that might overcome the above disadvantages are impossible in some
Middle-range empirical theories are an aid to both of these tasks. Unfortunately, cases and impossibly expensive in most others.
most of the theories in political psychology are not middle range but are global and When we have appropriate middle-range empirical theories to test, the
364 John B. McConahay Experimental Research 365

advantages of experimentation will, I believe, frequently outweigh the disadvantages. threat to internal validity if all the subjects in a single session receive the same
As yet, however, we do not have an abstract formula for deciding when an experi- experimental treatment.
ment is appropriate. The complete and flexible researcher should be familiar with Maturation, processes at work in the individual which operate independently
more than one technique, so that he does not have to formulate every problem or of any historical or experimental events, poses a serious threat to political socializa-
theoretical model to fit one procedure. Just a.s social ,psychology suffered in the past tion studies. Included here are aging or maturing, fatigue, increasing hunger, and
because experimentation was viewed as the only path to truth, political psychology the like.
will be retarded if we behave as if only sample surveys, only multiple regression, or Testing the effects of taking a test the first time upon responses the second
only introspection is the path to truth. time, can be a real threat to attitude and opinion research (see especially the dis-
cussion in Kiesler, Collins, and Miller, 1969). Familiarity with the questions may
Basic Methodology increase performance on proficiency tests (reading, mathematics, grammar), or the
mere act of giving an opinion or a response to an attitude question may set into
In this section, we shall examine some of the basic methods and issues com- motion cognitive and emotional processes which produce a change in behavior quite
mon to all experiments involving human subjects. The approach will be mainly apart from any "communication" or other manipulation introduced by the experi-
intuitive and expository rather than mathematical. There are a number of good menter.
mathematical treatments of this subject (Fisher, 1925; Hays, 1963; Winer, 1971), Instrumentation changes pose a special threat whenever human judges are
and the would-be researcher should have an understanding of the mathematical called upon to rate subjects on a set of behaviors or dimensions. For example, the
foundation of this field; but there is a large gap between knowing the mathematics first few subjects might not have appeared nearly so radical to the judges after the
and being a good experimenter. In addition, since much of the mathematics and judges had been exposed to the entire sample of subjects.
statistics of the experimental design are a special case of regression analysis, I Statistical regression occurs as a result of the random error inherent in any
assume that most aspiring political and social scientists already are familiar with the existing psychological measuring device (personality scale, attitude scale, intelligence
general if not the specific mathematics involved. test). I t should not be confused with regression analysis (Blalock, 1960) or with re-
Internal and External Validity. Following the tradition of Campbell and gression in the psychoanalytic sense. Because the measuring instruments are imperfect,
Stanley (1963), we shall distinguish between the internal and external validity of persons who score extremely high or extremely low on an instrument the first time
empirical studies. Internal validity refers to the question "Did in fact the experi- have had their extremity from the mean accentuated by random error. Hence,
mental treatments make a difference in this specific experimental instance?" (Camp- second measurements using the same or correlated instruments should find these
bell and Stanley, 1963, p. 175). Internal validity, then, is necessary if an experimental extreme scorers moving back, regressing, toward the mean, since the randomly
study is to be causally interpretable. External validity deals with the question of distributed errors are now operating in another direction. Since researchers in politi-
generalizability: "To what populations, settings, treatment variables, and measure- cal psychology are usually interested in groups or persons with varying degrees of
ment variables can this effect be generalized?" (Campbell and Stanley, 1963, p. 175). some psychological property (self-esteem, need for power, authoritarianism, con-
As with other inductive inferences, the question of external validity can never be servatism), a confounding of the experimental effect with regression effects is a very
answered completely. Though complete closure is impossible, some designs will potent danger. Therefore, the rule of thumb should be to look out for regression
approach external validity more closely than will others. Generally, with a large effects whenever subjects are chosen because they are extreme on some measure and
number of specific exceptions, psychologists are more concerned about the internal the dependent variable is an identical or correlated measure.
validity of a research design while political scientists are more concerned with the Selection biases occur whenever subjects can choose the treatment condition
external validity. they will be in. This is true even when the subjects do not know that they will be
Campbell has identified nine classes of extraneous variables which threaten in an experiment at the time they make the choice. For example, people who choose
to confound all experiments (Campbell, 1957a, 1969; Campbell and Ross, 1968, to attend early-morning classes (as well as those who teach them) are different in
Campbell and Stanley, 1963). We shall examine each quite briefly here. A thorough many ways from those who choose to attend afternoon classes. If the experimenter
discussion of the first eight factors may be found in Campbell and Stanley (1963) ; assigns one treatment to the morning classes and another to the afternoon, then he
the last factor is discussed in Campbell and Ross (1968). confounds the differences between morning and afternoon people with his treatment
History, any specific event that may occur between the observations on the de- difference.
pendent variable, is the first threat to internal validity. A president may be assassi- Experimental mortality, a differential loss of subjects between treatment
nated, a war may escalate, a political scandal may occur; any one of these events groups or treatment and comparison groups, is a major threat whenever the study
might affect an experimenter's dependent variable quite independently of the experi- covers a long time period or whenever subjects leave an experimental group after
mental manipulation. As the time increases between measurements and as isolation it is established. For example, as Kelley (1967) has noted, the somewhat counter-
from the real world decreases, this threat becomes more and more of a reality. The intuitive dissonance effects produced in forced-compliance studies (Festinger and
specific events within a single session, intrasession history, may also become a serious Carlsmith, 1959; Carlsmith, Collins, and Helmreich, 1966) are trivial if many
366 John B. McConahay I Experimental Research 367

subjects have refused to emit the counterattitudinal behavior in the low-reward There is virtually no way to deal with this threat except by drawing a random sample
condition. Because they were aware of this, researchers in this field have always taken from the population of interest (adult Americans, adult Americans and Western
great pains to keep subjects in their experiments. Europeans, adults from all over the world, all peoples) or by successive replications
Selection-maturation interaction, biases introduced by differential rates of in various subsamples of all peoples.
maturation or other autonomous changes across treatment groups, can occur when- External validity is further threatened by the reactive effects of experimental
ever subjects are selected by the experimenter for treatment conditions on any basis I arrangements, generally known as the Hawthorne effect (Roethlisberger and Dickson,
other than random assignment. For example, if the special participatory democracy 1939). People who know that they are in an experiment may respond differently
course in a political socialization study is given only to students in the fourth grade from those who do not. One solution to this problem is to experiment upon people
while students in the fifth grade are used as the control, the resulting differences may who do not know that they are in an experiment, but such a procedure runs into
be a function of the different rates at which fourth- and fifth-grade students mature. serious ethical problems.
Finally, instability in the data-resulting from the unreliability of measures, Multiple-treatment interference is a threat to external validity whenever sub-
fluctuations in sampling of subjects or units, and other spontaneous or autonomous jects are used who have participated in other experiments. This is an especially
errors in repeated or "equivalent" measures-may also bias the results. This is the significant hazard when researchers use college students or students from university-
only threat to internal validity to which statistical tests of significance are applicable affiliated "laboratory" elementary and secondary schools or when advertisements in
(Campbell and Ross, 1968; Campbell, 1969). the newspaper attract a number of professional subjects, who earn or supplement their
In true experimental designs, random assignment to conditions by the experi- income by participating in studies of every sort-from tests of new drugs to political
menter or some other generator of chance assignments will effectively neutralize these psychology studies.
threats to internal validity. Other confounds resulting from the experimental manipu: The final two threats to external validity result from the complexity of the
lations themselves will require more complex designs based upon random assignment measures and treatments involved in social and political experiments. The measures
(Underwood, 1957). that we use are not "pure" but contain systematic as well as error components which
Any number of factors could affect the generalizability of the results of a study are irrelevant to the dimension we wish to measure. Statistical tests can allow esti-
and hence its external validity. Only the six discussed by Campbell (1969) will be mates of the distortion due to errors, but the irrelevant responsiveness of components
reviewed here. Other threats to external validity will be discussed later in this chapter. I
of our measures may produce results that are mistaken for experimental effects.
The reactive or interaction effect of testing, a major threat to external validity, Similarly, irrelevant components of the experimental treatments may produce replica-
may increase a subject's sensitivity to the treatment or independent variable so that tions (or failures to replicate) that bias our understanding of phenomena. Therefore,
one cannot generalize from pretested to unpretested individuals. This effect, which is we need multiple measures with as little in common as possible except for the theo-
an interaction between the testing and the experimentally manipulated independent retical variable they are supposed to measure, and we need experimental replications
variable, is a threat to external validity and should not be confused with the simple that are equally diverse. As Kiesler, Collins, and Miller (1969) urge, we need a
effect of testing, which is a threat to internal validity. For example, if an experimenter "heterogeneity of irrelevancies."
wanted to test the effects of a certain antiwar communication upon a certain popula- Control of Independent Variables. In the typical laboratory experiment,
tion, he could control for the simple effects of giving the pretest by pretesting attitudes
toward war in the entire sample and then dividing them at random into an experi-
I whether in the biological, agricultural, or physical sciences, the experimenter has
virtually complete control over his independent variable. The experimenter decides
mental and control group. Since both groups were pretested, differences between the which rat gets the severe shock, which plot gets the hybrid grain, and which combina-
two groups could not be attributed to testing; internal validity would still hold. At I tion of chemicals gets the calcium phosphate. As a consequence, direct causal in-
the same time, exposing subjects to the pretest might have made them especially ferences are relatively easy to make and the experimental designs need to be made
sensitive to the communication delivered to them, so that they paid closer attention I
I complex only to increase efficiency (Campbell and Stanley, 1963; Fisher, 1925, 1935).
to the message than they normally would have. Thus, it would be very difficult to In the typical laboratory experiment involving human subjects, the control is less
generalize from pretested to unpretested subjects. Though internal validity was high
in this example, external validity was threatened. This sort of threat can be handled I and the designs must become correspondingly more complex; but there is still more
control than one has in correlational research, where the control is only a mathemati-
by making the pretest disguised, removed in time, unobtrusive, or otherwise non- cal approximation at best.
reactive (Webb and others, 1966), or by use of a posttest only or of an experimental I To illustrate why control is so important, we shall examine the problems
design that does not involve a pretest or that has a means of assessing the distortion raised by a study of campaign or propaganda techniques. Let us suppose that we
due to pretesting (Campbell and Stanley, 1963). 1 want to do a piece of research dealing with some aspect of local-level campaign
The threat to external validity most frequently cited by political science activity (such as voter registration or personal contact or mailed leaflets) upon voter
critics of experimental research is the potential interaction between the subjects I behavior. If we attempt to use straight correlational techniques in our study, as has
selected and the experimental arrangements. In other words, our manipulations been done frequently (see, for instance, Crotty, 1971; Cutright and Rossi, 1958; Cut-
might work on college students and white rats, but they may not work on people. 1 right, 1963; Kramer, 1970; Price and Lupfer, 1973; Wolfinger, 1963), by assessing
368 John B. McConahay Experimental Research 369

the type or level of activity and then correlating it with voter turnout or voter matically makes all other variables orthogonal or uncorrelated with the independent
preference, we are left with the nagging possibility that either the direction of variable. Thus, in the true experiment the researcher controls not only for those
causation is opposite to that which we wish to infer or that some unknown or unanti- factors that he or his colleagues thought were plausible (before or after the data were
cipated third, fourth, or fifth factor or combination of factors caused both the high gathered) but also for all possible third- or higher-factor correlations with the inde-
turnout or preference effects and the level or type of campaign activity. pendent variable: those which his critics might dream up and those which research
For example, the campaign managers, unable to propagandize every district fifteen years later reveals to be of crucial importance.
or person in their electorate, might have chosen the high-turnout or high-preference In many instances this distinction between all possible correlated factors and
wards, precincts, or individuals for special contact because they knew that they those that are plausible might appear impractical. We might be able to advise a
would get results there; that is, the high turnout and preference rates in these wards political candidate to invest his money in personal or telephone contacts rather than
caused them to be propagandized. A more complex three- or four-factor interaction emotional mail appeals without being certain that all possible artifacts had been
might also be responsible for producing a turnout or preference effect in the propa- ruled out (see Campbell, 1969). Nevertheless, when we wish to move toward the
gandized wards. For instance, the ethnicity, ideology, life styles, or other personal construction of rigorous empirical theories, we should at least strive for the perfection
characteristics of the campaign directors or available volunteer workers might have that experimental control gives us..
made them choose certain wards, precincts, or persons to propagandize and exclude Randomization. In the preceding discussion, the term control was used in
others because the campaign personnel were more at ease with some groups or more two related but somewhat different senses. First, it referred to the power of the
concerned about the interests of some groups than others. Here the control might experimenter to determine the fate of the subjects or treatment units with regard to
flow from certain issues or candidate styles, which produced both the turnout and the application of the independent variables (or to the lack of power of the subjects
preferences of the campaign workers and the voters in the target wards. or treatment units to determine or self-select their fate). I t also referred to the ability
In the campaign-activity literature, there have been some brilliant attempts of the researcher to rule out other factors related to both the independent and the
to introduce statistical controls which rule out reversals or third factors in the causal dependent variables. The first type of control could be exercised in either experi-
relationships. (See especially, Grotty, 1971; Kramer, 1970.) Regression or analyses mental or quasi-experimental designs (although in some quasi-experimental designs
of covariance using previous election results can shed light on the direction of the subjects' power is much greater than in others or in true experiments). The
causality, and various controls for plausible third and fourth factors can rule them second meaning of control is present only in true experiments, because random or
out or assess their effect. However, these mathematical controls have their limitations. chance mechanisms for assigning treatment units to conditions is necessary if all
First, the investigator can control or rule out only those factors that he recognizes as possible third- and fourth-factor correlations between independent and dependent
potential threats. Usually these are the most obvious social and demographic factors variables are to be controlled or ruled out.
and a few structural variables for which data were available. Second, the researcher When we speak of randomization, two possible misinterpretations may occur,
may be able to think of other potential factors, but he cannot deal with them because depending upon the background of the reader. For those unskilled (or at least un-
the data were not gathered or were not available or because his sample was so limited schooled) in the techniques of survey research, random assignment may be interpreted
that he has used up his degrees of freedom with the controls previously introduced. to mean haphazard or even sloppy. This is not the case at all. As we shall see, the
Finally, as Simon and Rescher (1966) have demonstrated formally, no correlational procedure is explicit and exact, quite the opposite of haphazard, careless, or sloppy.
method can resolve the causal question. For those readers approaching political psychology from a survey research back-
In experimental studies of campaign techniques (Eldersveld, 1956; Eldersveld ground, randomization in assignment to treatments may be confused with the pro-
and Dodge, 1954) there is no question of the direction of causality. The experimenter cedure for drawing the sample or selecting the levels of the independent variables.
controlled who got the contact, the leaflets, the emotional appeal, or the telephone Random selection of samples and treatment levels is important in experimental as
calls. Hence, he or she caused the voting effects; the voting effects did not cause the well as survey research. The samples used in experimental research may be equal-
experimenter (as opposed to the party chairman) to choose a given ward. While the probability samples (Kish, 1965; Young, 1966), in which case randomization is an
direction of causation in quasi-experimental studies of voting techniques (Blyden- integral part of the selection procedure; or the samples may be haphazard, con-
burgh, 1971; Gertzog, 1970; Reback, 1971; Gosnell, 1927; Hartmann, 1935; Orbell, venient, or not even a sample at all, as in cases where the total population of interest
Dawes, and Collins, 1972) is also usually clear cut, differences between the groups is used in the experimental design. The nature of the sample affects the external
caused by third and fourth factors are not so easy to rule out as in true experiments validity of the study but does not affect the control of independent or third-factor
because the matching or analysis procedures may still leave possible factors which variables.
cannot be eliminated or assessed. As with mathematical controls in correlational The procedure for choosing the levels of the independent variable affects the
studies, the researcher can assess or rule out orily a limited number of factors in a or statistical model chosen for data analysis (fixed versus random,
quasi-experimental design. On the other hand, true experirnents-using randomiza- finite versus infinite). The nature of the appropriate rnodel determines the size of the
tion or chance mechanisms to assign treatment units to conditions-can rule out all error term used to test the significance of the effects and also influences the external
possible third- or fourth-factor correlations because the randomization process auto- validity to some degree; but, again, it does not affect the control of extraneous
 John B. McConahay Experimental Research

factors. (For discussions of the mathematical models required for tests of significance different independent national samples: one drawn by the University of Michigan's
in experimental research, see Cornfield and Tukey, 1956; Ferguson, 1959; Stanley, Survey Research Center and interviewed both before and after the presidential
1956; Winer, 1971.) election of 1964; the other drawn by the United States Bureau of the Census and
In order to achieve control over extraneous factors, a random or chance interviewed only after the election. By some very clever analyses, Clausen was able
procedure is a must. Just as random selection is necessary to ensure that a sample is to rule out noninterview biases and response invalidity as explanations for the 3.5
drawn from a population in which every subject or unit was equally likely to have percent larger turnout reported in the SRC sample over the Census Bureau sample.
been drawn, so random assignment is necessary to ensure that every subject or unit Hence, he concluded (tentatively) that the preelection interview had stimulated
is equally likely to be given any treatment or combination of treatments. people who normally would not vote to vote in the election that followed. Neverthe-
When subjects or units are randomly assigned to treatments, the following less, Clausen (and his readers) could never be certain that the turnout rates of the
benefits accrue to the researcher. First, all third- or fourth-order correlations between two samples would have been equal had they both received either preelection and
the independent and dependent variables are eliminated. Second, the treatment postelection interviews or postelection interviews only. The samples were drawn by
groups are equated on all dependent variables at the time of randomization. Third, different organizations, were of different sizes, were intended to represent two different
postexperimental differences between groups can be tested for significance by the populations (all voters versus all adults), and therefore might have differed in their
appropriate mathematical models developed to estimate the expected values of the turnout rates even without the preelection interview. Clausen's quasi-experimental
error variation in dependent variable scores. design provides suggestive evidence of the stimulation effect of a preelection interview,
When the analysis of variance model, based upon random assignment, can but it is weak evidence to use in starting a chain of theorizing about psychological
be used for assessment of significance, smaller sample sizes are generally required mechanisms to account for this stimulation effect.
than in correlational work to achieve the same statistical confidence levels in the An experiment by Kraut and McConahay (1972), planned and executed in
differences between treatment groups. This is because, as S. R. Brown (1971) pointed blissful ignorance of Clausen's work, provided much less ambiguous evidence that
out, the estimate of variation due to error in correlational work is a theoretical the interview effect is real and rather powerful in low-salience elections. In this
valued based upon sample size. Hence, in correlational work, error is reduced study, an equal-probability sample of registered voters with Italian surnames was
principally by increasing sample size. In experimental work, using the analysis of drawn from a middle-class ward in New Haven, Connecticut. The experimenters
variance model, the estimation of variation in dependent variable scores due to divided the entire sample in half by assigning every other name on the alphabetical
error is an empirical matter. Although error can be reduced in experiments by list of persons in the sample to group A or group B. Then, by flipping a coin, they
increasing sample size, it can also be reduced by manipulating other factors directly designated group B as the experimental group and group A as the control group.
under control of the experimenter in the concrete situation. While this reduction Subjects in the experimental group were subsequently contacted and interviewed by
may represent an artificial magnification of the experimental effect, reducing external graduate students. (The subjects knew that the interviewers worked for Yale but
validity and introducing a hazard rather than a benefit, I think that it is a benefit to were unaware of their status as graduate students.) The interviews involved a modi-
have the estimate of error determined empirically in each situation and that the fied version of studies (Kamin, 1958; Lorinskas, Hawkins, and Edwards, 1969) in
option of devoting resources (time, money, effort) to aspects of the research other which subjects were asked some standard party-identification and political-efficacy
than the procurement of subjects is also a benefit. questions and then given a series of hypothetical electoral choices pitting candidates
Matching on one, two, or more dimensions is the method of assignment to with Italian and Anglo-Saxon surnames against one another. When additional subjects
treatment conditions usually posed as an alternative to randomization. This procedure were needed to meet the requirements of the ethnicity study (because some experi-
has several drawbacks, however. First, as we have indicated repeatedly, though the mental subjects had moved or died) ; a new probability sample of Italian-surnamed
researcher might match on a large number of dimensions, he cannot match on all subjects was drawn. This sample was twice as large as needed for the ethnicity study
dimensions except by randomization. Second, it is frequently impossible to match on and was split on an even-odd basis into two groups, which were then assigned to
certain combinations of dimensions. For example, Hartmann (1935) matched his experimental and control conditions by the flip of a coih. Party officials (unaware
wards on population density (38.44 versus 37.28 persons per acre) but then dis- of who was an experimental and who was a control subject) identified the persons
covered gross inequalities in wealth in the wards ($2070.12 versus $871.72 mean in the total sample who had voted in an election two weeks after the interviews
assessed value per capita). Although modern multivariate techniques can reduce (May) and three months after that (August). In both elections, interviewed subjects
some of this bias, they cannot eliminate it. Third, matching on other extraneous voted at a higher rate than noninterviewed subjects.
variables, even when possible, does not ensure that the experimental groups are Noninterview biases and response invalidity, the two major alternative expla-
matched on the most important variable of all: the dependent variable. nations for Clausen's findings, were handled automatically by randomization in the
As an illustration of this last point, let us consider two studies, one quasi- Kraut and McConahay experiment (1972). Differences in turnout rate due to
experimental (Clausen, 1968) and the other experimental (Kraut and McConahay, differences between the experimental and control groups in the number of persons
1972), of the effect on subsequent voting behavior of subjects interviewed by poll- who were incapacitated, dead, or absent from the community were ruled out because
takers. Clausen compared the voting turnout rates reported by respondents in two these persons were equally likely to have been assigned to either group, and the turn-
 John B. McConahay Experimental Research 373

out rates were computed on the basis of the entire experimental and control groups- the experimental condition if the second number is even, and the third subject or
not simply on the basis of who was actually contacted. Similarly, response invalidity treatment unit to the control condition if the third number is odd. This procedure
was not a possible rival hypothesis because errors in recording and/or deliberate could be used to randomize not only subjects or treatment units but time of day, day
tampering with the voting records were as likely to have happened for subjects of week, date of month, or any other factor that might vary in such a way as to con-
assigned to the control as for those in the experimental group. Finally, there could found the assignment of treatment units to conditions.
not possibly have been any subtle, undetected, unthought-of, or unexpected differ- Readers should note that I have used the phrase random assignment of sub-
ences between the experimental and control samples in the second study (differences jects or treatment units rather than the simple phrase random assignment of subjects,
that might have made for differences in turnout rates even without an interview) I have done so because one of the most frequent methodological errors made by those
because the experimental and control subjects were drawn as part of the same sophisticated in correlational techniques who attempt to do experimental research is
sample, using identical procedures from an identical population. to think they have randomized an entire subject population when they have ran-
The Kraut and McConahay study illustrates one method that can be used for domized only certain units within the population. This error occurs repeatedly
making random selections. There were two treatment conditions; and since there throughout the pages of Exprrimental Study of Politics (see Dawson, 1971; Dyson
was no reason to suspect that an alphabetical list of Italian surnames would have a and Scioli, 1971; Lamare, 1971; Reback, 1971) . The error usually occurs under one
systematic pattern of variation which coincided with the choice of every other one of two conditions. In the first, subjects are assigned at random as individuals to one
for group A, this method of systematic assignment was chosen. However, one must of two experimental sessions. For example, in one session one hundred subjects see
be careful in using this or other systematic procedures. As Kish (1965) has pointed an antiwar film and in the other session one hundred subjects hear an antiwar
out, a monotonic trend or an unfortunate systematic fluxuation among subjects in lecture. A second condition under which this error is likely to occur is when groups,
the sample may bias the assignment. For example, if an experiment were being wards, or other units of subjects are assigned at random to experimental treatments.
conducted in a California housing development where two basic styles of "uniquely For example, certain wards may receive a great deal of party activity and others
individual" houses were available, one costing $20,000 and the other costing $30,000, receive very little or none at a11 (Gertzog, 1970). Mere assignment of subjects to
an assignment procedure placing every other house on a list of street addresses in conditions in one of these fashions is not the error. The error occurs when subjects
the experimental group might bias the assignment if the developer had used a are assigned in these ways and then the data are analyzed by pooling the dependent
similar procedure to intermix his "deluxe" and "superdeluxe" houses. Thus, much to variable scores for the individual subjects and doing statistical tests on the basis of an
the experimenter's chagrin, he might later discover that the experimental subjects inflated number of degrees of freedom.
were more affluent than the control subjects. Let us examine these two error-prone assignment techniques in some detail. In
This type of bias could also occur if subjects were coming to a laboratory doing so, we will examine hypothetical rather than real cases, so as not to single out
setting at various times of the day and the experimenter decided to assign all the any one of these new experiments in political psychology for special criticism.
morning subjects to one treatment condition and a11 the afternoon (or even worse, Let us suppose that an experimenter wants to test the effects of a one-sided
all the evening) subjects to a second condition. He might think he had avoided the antiwar television program versus those of a two-sided program upon the attitudes
problem by assigning the 10, 12, 2, 4, and 6 o'clock subjects to one condition and of the general public. He discovers that the American Broadcasting System has made
the 9, 11, 1, 3, and 5 o'clock subjects to the other. However, he is in trouble if he two pilot films--one film blatantly antiwar and antimilitary, the other film balancing
does not realize that ROTC meets only on the even-numbered hours. Thus, his "pro- much of this same antiwar material with a rebuttal by -"General George I. (G.I.)
military'' communication (given during odd-numbered hours) mrght get, on the Joseph." The experimenter draws a random sample of citizens living within a radius
average, an especially sympathetic reception for reasons having nothing to do with of fifty miles of his town (in order to cut down on the proportion of academics in his
the "emotional" quality of its arguments. sample and study some "real people"). By offering his subjects fifty dollars for a
The moral, then, is that systematic assignment by a procedure of every nth two-hour session, he gets all of the potential subjects to agree to come down to the
subject or unit is a relatively easy and straightforward technique provided the re- theater he has rented (so that they will not associate his study with the local college)
searcher has thought about possible confounding trends in the sample. on a night he has designated. Unfortunately, he has spent so much money on securing
The best, most elegant, and only slightly more inconvenient procedure is to the film, drawing a true random sample of all adults, seducing them into coming to
use a table of random numbers (for instance, see Lindley and Miller, 1958) or of his special showing, and renting the theater that he can show the film only two
random permutations (for instance, see Winer, 1971). In this procedure, subjects or times, once on Monday and once on Tuesday. He is not worried about this, how-
treatment units are assigned to treatment conditions according to how they coincide ever, because his subjects agree to come on exactly the night he tells them, and he
or correspond with a number on a table. For example, if the experimenter has two assigns them to nights at random (using a random-number table).
treatment conditions (experimental and control), he goes to a table of random E shows the one-sided film to one hundred subjects on Monday and the two-
permutation of some even number of digits (16 or 32, for instance), chooses a sided pilot to one hundred subjects on Tuesday. After each showing, he asks the
column, and then assigns the first subject or treatment unit on his list to the experi- subjects the same set of unobtrusive, cleverly disguised, highly valid, and highly
mental condition if the first number is even, the second subject or treatment unit to reliable questions which tap their antiwar feelings, antimilitary attitudes, and pacifist
374 John B. McConahay Experimental Research 375
inclinations. He also asks them a set of more direct policy-oriented questions about decided by a combination of counterbalancing and randomization. In this procedure,
the ordering of American priorities. Finally, his graduate students call the subjects at the one hundred subjects exposed to the one-sided film would have been distributed
home two months later and ask them some questions as part of a public opinion poll, across five sessions of about twenty subjects each. The same would hold for the
which is not perceived by the subjects as in any way related to the experimenter's subjects exposed to the two-sided pilot. The data analysis should in this instance be
research. based upon computing the average score for each session and then computing the
He then analyzes his data. He computes the mean score for each of the two average of the averages for the five one-sided sessions and the five two-sided sessions.
groups (one-sided and two-sided) and discovers that the differences between the The t test on the differences between these two final average of averages should be
two groups are not very big. Hence, he has to do statistical tests in order to convince based upon eight degrees of freedom.
journal editors and journal readers that he has discovered something more substantial To repeat and summarize, this experimenter's difficulty grew out of his con-
than just a chance fluxation. Accordingly, he does a series of t tests on the differences founding of the nature of his communication (the properties of his independent
between the means of the two groups, each having 198 degrees of freedom, and variable) with intrasession history. Since the subjects on Monday night had in
writes an article concluding that a one-sided communication is more effective than common both the one-sided communication and whatever happened in their session,
a two-sided communication in producing attitude change in this population. The they had to be treated as a single unit for the data analysis unless the experimenter
journal editor rejects the article, noting that if the experimenter wanted to conclude wanted to conclude that it was this combination that was responsible for the observed
that the different messages did make a difference, he should have used zero degrees difference. The problem, then, is not that the experiment was run in only two ses-
of freedom in his t test (impossible, of course). If he wanted to conclude that "some- sions. I t could have been run in only one session and the data analyzed on the basis
thing which happened in the Monday session" produced mori attitude change than of individual scores if each subject had received his or her treatment as an individual.
"whatever it was that happened on Tuesday night," then he could still use 198 If, for example, subjects had read one- or two-sided newspaper editorials distributed
degrees of freedom. at random throughout the same group, the data could have been analyzed at the
This experimenter's problem was that in his procedure he confounded intra- individual level because all subjects would have had the same intrasession history.
session history with the independent variable (one-sided versus two-sided). If he Hence, only the difference in reading material would account for the observed
had wanted to conclude that it was the difference between the messages (rather than differences in attitudes, feelings, or opinions. The rule of thumb should be to analyze,
an obnoxious drunk in his Army Reserve uniform at the Tuesday session or his own as a unit, all subjects who have both a common intrasession history and a common
more confident and relaxed manner on Monday or the fire drill in the building next experimental treatment. Since this analysis is usually accomplished by averaging an
door on Monday or the breakdown of the boiler on Tuesday or some other more individual subject's scores within units (which increases the reliability of a unit
subtle, perhaps unspecifiable difference between what happened in the two sessions) score) and then averaging across units, the increase in reliability generally offsets the
that made a difference in the subjects' antiwar attitudes, he could have used one of decrease in degrees of freedom.
two strategies. First, he could have arranged for each subject to see the one- or two- The above error results from a confound within the experimental design. The
sided film alone. He could have used many television sets in a number of isolation second error, assigning units at random and analyzing on the basis of individuals,
booths or had subjects come to the theater at one- or two-hour intervals all week long. results from a violation of the assumptions underlying the mathematics of randomiza-
In this case he could have taken the average of each individual's score on the depend- tion. This error occurs in most studies of voting behavior where whole wards or
ent variable and use a t test with 198 degrees of freedom. He could then conclude that precincts are assigned to some treatment condition-for example, when one set of
it was the message alone and not the combination of message and session that pro- precincts is flooded with emotional campaign materials and another with rational
duced the difference because each subject's individual session would have been materials (Hartmann, 1935). I t could also occur, however, in studies of political
unique. The irrelevant events-such as drunks in uniform, the fire drill, his own socialization, in which whole classrooms or schools are assigned the same curriculum.
changes in mood (presumed to vary independently of the film he was showing), the For experimental and quasi-experimental voting studies where this error did not
changes in temperature in the building, and so on-would have been distributed occur, see Blydenburgh (1971) ; where the correct degrees of freedom were used; and
across all sessions in a fashion which did not coincide exclusively with one of the Eldersveld and Dodge ( 1954) and Eldersveld ( 1956), where the randomization was
communications. That is, the unique intrasession histories would have been randomly on the basis of individuals rather than wards or precincts.
distributed across conditions. As another hypothetical example, let us suppose that an experimenter wants
As a second possible strategy, he could have had smaller numbers at each to test the effects of endorsement of a candidate by a well-regarded or a poorly
session and multiple replications of each session. For example, he could have assigned regarded public figure upon voter preference. With the consent of the candidate,
subjects to sessions on Monday, Tuesday, Wednesday, Thursday, and Friday eve- he picks thirty wards in which to distribute the literature and (using a table of
nings, held two sessions each night, and shown each film at one or the other of these random permutations) assigns the wards to one of three conditions: (1) Leaflets
sessionseach night. The one-sided film could have been shown at the first session on are distributed telling of the candidate's great qualities of leadership and probity; the
three nights, and the two-sided film could have been shown at the first session on the leaflets are signed by the incumbent president of the United State;, a person loved by
other two nights. Which film would be shown first on which night could have been everyone in this electoral district. (2) The same leaflets are distributed bearing the
376 John B. McConahay Experimental Research

signatures of a famous local murderer, rapist, and child molester, a person uniformly plicity of presentation. The logic, however, applies to multilevel and multifactor
hated by everyone in the district. (3) The same leaflets are distributed without a experimental designs as well.
signature or other source identification. 'The experimenter then waits two weeks and One last point that should be raised before I leave this discussion of random-
sends legions of graduate students out to interview an equal-probability sample of ization is that matching and randomization can be combined to gain statistical
voters drawn from the three districts. By means of Thematic Apperception Tests, precision in some circumstances, especially where the number of experimental units
Semantic Differentials, and some clever tests of his own creation, he is able to obtain is small. Thus, in our last example the experimenter could have matched as nearly
a highly reliable and valid interval scale of preference for the candidate. I as possible on the vote for his candidate in the preceding election (or the nominee of
Because of his cleverness, his interviewers' resourcefulness, the huge amount of his candidate's party) by creating ten groups of three wards, each ward having
money he spent, and a tremendous amount of luck, this experimenter, then, obtained approximately the same voter-preference rates. Then one ward from each trio could
measures of candidate preference from 100 percent of his original sample, so that he be assigned at random to one of the experimental conditions. This procedure is
had no sampling bias; and he secured data from one hundred subjects in each of called blocking in the statistical literature and is discussed intuitively in Campbell
his three wards. Since the leaflets were assigned at random and the sample was un- and Stanley (1963) and mathematically in Cox (1957), Feldt (1958), and Lindquist
biased, he was certain that the three groups were equally likely to prefer the candidate (1953).
before he bombarded them with his propaganda. Hence, he performed a three-level Randomization i n Field Settings. In the hypothetical examples above and
one-way analysis of variance upon his data, using F tests having 2 and 297 degrees in many actual experimental studies of political psychology, internal validity is
of freedom. After examining the results, he concluded that it was better to be en- frequently sacrificed to attain a certain measure of external validity. Films acquired
dorsed by a dearly loved incumbent president than by a hated murderer, rapist, and from networks at considerable expense are used instead of less expensive television
child molester but (interestingly enough) that it was better to have been endorsed by tapes or written materials created by the researcher. Random samples are drawn from
the latter than to distribute unsigned literature. The journal editor, unfortunately, a population of voters or all adults in a district. The research is conducted in field
concluded that the experimenter should have used F tests with 2 and 27 degrees of settings, where real elections or real committee decisions can be observed. Unfor-
freedom in his various analyses, because unit of randomization and unit of analysis tunately, since all research is conducted within the limits of a finite (though occa-
must be the same for tests of this sort. Since the experimenter had randomized on the sionally large) budget, the resources devoted to increasing external validity will
basis of wards, his total N was 30, not 300, and his degrees of freedom had to be reduce the resources that an experimenter can devote to internal validity. This brings
reduced accordingly. Unfortunately, when the new analyses were run, he found us to the researcher's dilemma: Increasing external validity may reduce internal
himself with no statistically significant results. validity to the point where the results are uninterpretable; but introducing more
Rather than wade any further into the mathematics, let us attempt to get an and more control to maximize internal validity may introduce more and more iso-
intuitive grasp of the problem. The reason for randomization in this experiment was lation, sterility, and artificiality-which reduces external validity to the point where
to equate the three treatment groups on the dependent variable and all other variables the results are clear but ungeneralizable.
at the time of randomization. Since the unit of randomization was the ward rather As with most real-life dilemmas, there is no ideal solution. Not only are we
than the individual, the wards composing the high-prestige treatment group. were unable to conduct "crucial" studies in which two competing theories are pitted against
equal to the wards composing the other two treatment groups. In other words, the each other and one of them destroyed by the empirical results (Campbell and
high-prestige treatment group-which equaled the other two treatment groups-was Stanley, 1963), but we can rarely conduct perfect experiments in which both internal
a group of wards, not a group of individuals. and external validity are maximized. What we will need, then, is programmatic
The experimenter could have avoided this error if he had averaged scores rather than one-shot research. In programmatic research, a series of experiments (or
I
within wards and then averaged the averages. He could also have recorded the studies) are conducted in which we compensate for flaws in one study by designing
percentage of the vote the candidate got in each ward and then averaged the per- a second study which is free of those flaws but has a different set of weaknesses,
centages across wards within treatment groups. The rule of thumb, then, is that the which can be compensated for in a third study, and so on. Furthermore, in such a
unit of randomization (persons, groups, wards, classrooms, legislative committees) program the internal validity of each individual study should be maximized while
and the unit of data analysis must be identical. the weaknesses in external validity should be compensated for by different studies.
In the two examples illustrating these common errors in randomization, the In fact, an experimental study (specifically excluding quasi-experimental studies)
hypothetical experimenter used parametric statistics ( t and F tests) to assess the which does not have perfect internal validity is not worth doing. In experimental
significance levels of his results. His errors could not have been avoided if he had research, excuses for design flaws that jeopardize internal validity but are justified as
used nonparametric tests (Chi squares, for example). The error was in violating the I necessary to increase external validity must be categorically rejected.
logic of randomization for equation of experimental treatment groups and would be If the researcher is going to attempt an experiment, with all the effort and
I
reflected in the number of degrees of freedom chosen for any type of statistical test. cost entailed in such research, he ought to emerge with a result that is clearly
The examples also involved only two- and three-level single-factor designs for sim- interpretable in a causal sense. However, once the internai validity has been achieved
 John B. McConahay Experimental Research 379
I 1967), was that on a posttest a year later the first group showed the least decline
by proper randomization and the necessary control groups, then efforts should be
made to increase external validity in a fashion that does not jeopardize internal in F scores while the third group showed the greatest decline.
validity. Political psychologists may have other opportunities for field experiments
A true experiment in a field setting, while very difficult to perform, does which occur as a result of our roles as consultants or social reformers. Frequently,
begin to approach an ideal solution to the dilemma. Frequently, researchers have we are asked to advise a school board, legislative committee, or governmental admin-
overlooked the opportunities to do field experiments because the laboratory scientists istrative agency. The current emphasis upon evaluation and accountability by the
were more comfortable with the ease and extreme control of the laboratory, and United States Office of Education may provide excellent opportunities for research
survey researchers simply did not think in experimental terms, finding it easier to into political socialization and the political psychology of educational reform.
measure differences in the variable of concern than to create them. Both types of Along with Campbell (1969), we would stress the importance of pressing for
researchers, however, should at least entertain the possibility that they might study random assignment under circumstances where not every unit (person, ward, school,
a certain research question by means of a field experiment before settling for their classroom) can be subjected to the change at the same time. Random assignment
most familiar method. is generally the fairest way to distribute experimental treatments in addition to having
Field experiments may consist of simply transposing the laboratory into the the properties of creating the conditions for a true experiment. For example, I was
field, taking the experiment and experimenters to the subjects rather than waiting involved in planning and evaluating an alternative high school program in New
for the subjects to come to the laboratory. Such studies have both internal validity, Haven, Connecticut (Hawley and others, 1972). Many more students wanted into
through randomization and control of experimental variations, and external validity the program than could be admitted. Some school administrators and school board
insofar as they are in natural settings and the subjects are usually unaware of their members initially favored establishing a screening committee to choose students by
participation in an experiment. For example, studies of bystander intervention (Bryan some set of criteria to be established by students, parents, teachers, administrators,
and Test, 1967; Piliavin, Rodin, and Piliavin, 1969) generally involve a team of and other interested persons. The social scientists, however, were able to convince all
researchers who fake an accident or an argument in a public place while another concerned that a lottery (stratified on certain racial groups to match the racial
experimenter notes who intervened and in what manner under different experimental balance of the entire secondary school system) would be the fairest method of choos-
conditions. The above-cited studies of campaign techniques and of ethnicity and ing students from the list of applicants. The acceptance of the lottery method
satisfied a number of interests peculiar to various subgroups within the population
electoral choice are of this sort. In these field experiments, the randomization follows
involved in planning the new school and its program. It saved time, reduced or even
a procedure similar to that in laboratory experiments.
eliminated political intrigues concerning who would be admitted, reduced the
But still other possibilities for field experiments offer great opportunity for
potential for a struggle over the admissions criteria, minimized the potential damage
research and have not been exploited to any extent as yet. These take advantage of
to the self-esteem of rejected students, and, incidentally, equated the admitted and
"naturally" occurring random assignments to gather meaningful data that are
nonadmitted groups of students on all independent and dependent variables (at the
internally and externally valid. By "naturally" occurring random assignments we
time of the lottery at any rate), and-especially important-it created a control
mean those set into motion independently of the experimenter.
group of nonadmitted students whose initial motivation to participate in the program
The best example of this type of field experiment is the study by Siegel and
was equal to that of the admitted students.
Siegel (1957) of the effects of reference and membership groups upon attitude
Naturally occurring experiments and those engineered by advisors as a means
change. These researchers took advantage of the fact that Stanford University
of equity have two great advantages over laboratory experiments. First, they have
utilized a lottery to pick the women students who would get to live in highly desired
powerful variations in the independent variable; actually living in a row house does
"row housing." The women who wanted into row housing and those already living
not compare in any way with role-playing such a relationship or being assigned to
in row housing were higher on a modified F scale (Adorno and others, 1950) than participate in a group for an hour. Second, they have high external validity. In addi-
other females on campus, but most Stanford women (regardless of housing) tended tion, they are superior to survey studies because of their high internal validity. Field
to score lower on the F scale the longer they were in school. The experimenters pre- experiments of this sort may also have a disadvantage vis & vis laboratory experi-
tested the women who applied for row housing at the start of an academic year ments. Frequently the exact stimulus producing the change is hard if not impossible
and then let the lottery create three experimental groups: (1) those who wanted in to specify. For example, the observed differences may result from exposure to an
and did get into row housing; (2) those who wanted in and did not get in, but entire school curriculum or to only one course in that curriculum (Hawley and
applied the next year; and (3) those who wanted in, did not get in, and did not others, 1972) . Similarly, differences observed after an experiment in racially inte-
apply again the next year. So far as F scale scores were concerned, this created (1) a grated education might result from the intergration experience or from the fact
group whose membership and reference groups were both high F; (2) a group that some students rode buses to school. These ambiguities may retard the develop-
whose membership group was low F but whose reference group was high F; (3) a ment of middle-range theories in some cases; in others the theories are such blunt
group whose reference group and membership group were both low F. The result, instruments that eliminating even gross alternatives can only help in sharpening them.
consistent with reference-group and other social evaluation theories (Pettigrew, The rule of thumb, then, ought to be: Researchers should be on the lookout
380 John B. McConahay Experimental Research 381

for field settings in which experiments can be performed, in which experiments are differences, and this confound is a possible threat to the external validity of the
occurring "naturally," or in which the investigator can suggest randomization as the study. Good experimental procedures, therefore, should include both multiple
best and fairest means of deciding who gets what treatment. Whenever we hear or methods of assessing' the same individual-difference dimension of independent vari-
read of an instance where a chance mechanism is used to affect people's political ables and multiple methods of measuring the dependent variables (see Campbell and
lives or behavior, we should ask if this opportunity could be utilized for research Fiske, 1959).
purposes. Social psychologists, for example, have already started taking advantage of Advocates of psychological control in experiments (variations in the experi-
the selective service lottery to examine its effects upon social psychological attitudes mental stimuli of the same treatment condition to increase "experimental realism"
(Apsler, 1972). Surely, political psychologists might be interested in whether the and create a uniform psychological state or perception) should be especially careful
nineteen-year-old subjects in this experiment also have their political behavior to use multiple "take" measures or checks upon their manipulations, since method
affected. and hypothesized psychological state are confounded in single-method procedures.
Qz~asi-Experiments. When pretreatment equality of groups cannot be On the other hand, advocates of a uniform stimulus presentation or stimulils control
achieved through randomization, quasi-experimental or patchwork designs may still should Ee careful to replicate, using very different stimulus manipulations, since
yield internally valid data with high external validity. The general principle involved method and concept are confounded in a stimulus-control procedure. (For discussions
is to introduce as much randomization as possible and to compensate for various of stimulus control versus psychological control, see Aronson and Carlsrnith, 1968;
threats to internal validity by finding comparison groups which enable an experi- Kiesler, Collins, and Miller, 1969.)
menter to infer that a specific threat to internal validity is highly implausible even Experimenter Egects and Bias. In the 1960s, experimental researchers dis-
if he is unable to rule it out entirely as a possible threat. A number of these quasi- covered the experimenter. As a result, new possible threats to both internal and ex-
experimental designs have been suggested in some detail (Campbell, 1968; Campbell ternal validity must be dealt with. The threat to internal validity was first called
and Stanley, 1963), and specific applications of them to concrete research questions to our attention by R. Rosenthal's work (1966, 1969) on experimenter bias. This
have also been published (Campbell, 1969; Campbell and Ross, 1968). kind of bias is a threat to internal validity because the various persons who run the
experiments for the researcher (influenced by the researcher's hypotheses or their own
Complicating Factors hypotheses regarding the outcome of the experiment) may act differently in different
experimental conditions. This different, often quite subtle, behavior is thus confounded
We could stop our discussion at this point inasmuch as the basic principles of with differences in the experimental manipulations across treatment conditions.
expzrimentation have been presented. However, political psychologists should be Discussion of procedures for controlling experimenter's bias (such as automation and
concerned with generalizing from specific studies to broader theoretical issues, ar,d we "blind" experimenters) and criticisms of Rosenthal's original research may be found
usually (though not always) use human individuals or human institutions as the in Aronson and Carlsmith (1968) and Kiesler, Collins, and Miller (1969).
subjects and the experimenters in our studies. Because of this, the world of the The threat to external validity results from the obvious fact that different
experimenter is even more complex than the above presentation implied. In this experimenters are different. Different experimenters, quite independently of their
section, we shall look quite briefly at these additional factors. Though our examina- expectations regarding the hypothesized outcome of the experiment, may produce
tion will be almost cursory (because of space limitations), these factors are important different results; consequently, the results obtained by one experimenter may not
and the complete experimenter will have mastered them as well as other aspects of generalize to those of other experimenters. Survey researchers have been aware of
experimental design. the different effects of various interviewers for some time (see Hyman and others,
Multiple Measures and Methods. Frequently political psychologists wish to 1954), but experimental researchers began systematic examination of such effects
include individual-difference measures as part of their experimental designs. For much later (McGuigan, 1963; Kintz and others, 1965). As with experimenter bias,
example, Morrow (1972) included self-esteem and personal efficacy, two concepts there are no across-the-board panaceas for the problem of experimenter effects. The
measured by psychometric scales, along with an experimental manipulation of the best general approach is to consider the experimenter as a context variable (Kiesler,
status of the frustrator in his test of a social-psychological model of aggressive politi- Collins, and Miller, 1969) and to employ more than one experimenter, each running
cal behavior. In fact, probably because of our heavy reliance upon correlational all conditions of the experiment. This procedure permits assessment of the experi-
methods in the past, most of our political-psychological concepts are operationalized menter effect by including the experimenter as a random factor in the analysis of
by paper-and-pencil individual-difference measures (authoritarianism, liberalism- variance.
conservatism, dogmatism). Demand Characteristics. Experimental arrangements and design may com-
The variance associated with measures of this sort has at least three com- bine to pose a threat to internal validity (as well as to external validity, previously
ponents: random-error variance, method variance, and variance contributed by discussed). For example, Orne (1962) has proposed that subjects enter an experi-
differences in individual's standing on the trait or dimension. Error variance can be ment with a problem-solving orientation aimed at discovering what the experimenter
assessed and controlled by statistical tests of significance, but the other two are hope- wants them to do. Because the students who serve as subjects in most experiments
lessly confounded in a single study that uses only one method of measuring individual value science,'they will then do what they think the experimenter's theory calls for-
 John B. McConahay

their notions being based upon the demand characteristics of the experimental situa-
tion (their analysis of the design, the cover story, and the like) and not upon cues
given off by the experimenter. A similar formulation is Rosenberg's (1965b) concept
of evaluation apprehension, which also assumes that subjects will do what they per-
ceive as the normal or proper thing for them to do once they have figured out the
experimenter's intention. Attempts to minimize the effects of this phenomenon have
ranged from increasing the deception of the subjects in some way (through separated
posttests, "preexperimental" experiments, elaborate cover stories) to careful analyses
of the design for demand characteristics. These attempts are discussed in detail and
different solutions proposed in Aronson and Carlsmith (1968) and Kiesler, Collins,
and Miller ( 1969).
Ethics. Concern about the reactivity of experimental situations, demand
characteristicj, and the increasing experimental sophistication of their college stu-
SIMULATION:
dent subjects led many psychological experimenters to move into field settings, where
the subjects had no idea they were in an experiment, or to build ever more elaborate
hoaxes into their experimental designs. As a result, pretesting of experimental manipu-
ATTEMPTS AND
lations and elaborate debriefings (or "de-hoaxings," as they are called in one
laboratory-oriented psychology department) became increasingly necessary. (For the
best description of careful pretesting and debriefing procedures, see Aronson and
Carlsmith, 1968.) Eventually the manipulations began to rival the wildest and
POSSIBILITIES
fondest dreams of the mad scientists who inhabit television's late, late shows. Under-
standably, a number of psychologists have become uneasy about the ethics involved
in doing experiments in which there is a great deal of deception or stress created for Rufus P. Browning
human subjects, especially when it is the subjects who suffer and "science" or the
scientists who benefit (Baumrind, 1964; Ring, 1967; Kelman, 1 9 6 8 ~ )The . debate
thus far has been mainly speculative and rhetorical, but the ethical issues involved
in doing experimental or any sort of research should not be too readily dismissed
simply because of the form in which they have been raised. Political psychology is an
especially sensitive area, not only because of the manner in which our data might be
gathered but also because of the use to which our findings might be put.
I
Final Note
T h e scientist strives to see behavior more
I t would be a gross exaggeration to claim that the discipline of political clearly, to tell us more about it, and to weave ever stronger links between what he
psychology will not advance until every scientist in the field is an expert in experi- sees and what he says. The chapters of this volume on surveys, experiments, and
mental technique. But it is also a serious mistake to dismiss out of hand experimental projective techniques all focus on distinctive ways of observing or seeing behavior.
methodology as a tool for political-psychological research. To restate a theme that In the case of simulation, this chapter examines some special ways of saying-ways
has run throughout this chapter, we need to develop testable middle-range theories of describing behavior and of finding out what the descriptions signify.
and a stance toward research which allows us to use multiple methods: intensive Simulation methods may be used to describe and study the behavior of any
interviews, projective tests, sample surveys, cross-cultural comparisons, content system. The following social and individual processes have in fact all been studied
analyses, standardized tests, questionnaires, and other procedures still to be invented. with the help of simulation: referendum campaigns in local communities (Abelson
Laboratory and field experiments should take a prominent place among these other and Bernstein, 1963) ; the decision making of a legislature faced with a series of
methods and probably will as more and more and more political researchers sharpen bills (Cherryholmes and Shapiro, 1969) ; the responses of a politician to a flow of
their experimental skills by doing experiments.
I
1
information and questions about his politically salient environment (Abelson and
Carroll, 1965) ; and the decisions and interactions of nations (Guetzkow and others,
384 Rufus P. Browning Simulation: Attempts and Possibilities 385

1963). In contrast to other descriptions, however, simulations are working models of can be used to explore the implications of models that are precise but entirely non-
the processes under study. All descriptions aim to communicate the main features of mathematical in conception, a radical innovation in fields where behavior has
behavior. Simulations go beyond such descriptions in an effort to reproduce ob- appeared impossibly intricate to trace and not describable in mathematical form (for
served behavior, even of complex systems, with a high degree of precision. The ability instance, thinking).
to reproduce the behavior of a human system would imply a strong descripxion of its Computer simulation is sometimes too demanding a method. Suppose a re-
processes; thus a working model of the system is desirable in itself. It is also a searcher has a set of hypotheses about some processes of a system and about its basic
necessity because often the only way to understand a complex description fully is to structure but lacks confidence in his knowledge about other parts of the system. I t
carry out the cuinulative implications of every statement in it. may be possible to construct a simulation in which people, programmed in part by
Unfortunately, carrying out the implications of a complex descriptive model the structure and predetermined processes of the game, play out the roles of hypo-
is frequently an impossibly burdensome task for a person to accomplish unaided. thetical actors in the system to which the simulation refers. Especially where it is
The solution is to rely on the technology of the computer (computer simulation) or genuinely diEcult to observe closely the behavior and thinking of actors in the
on ways of involving human actors to play out the assumptions and conditions set reference system, it may be fruitful to obtain observations about the behavior of
forth in the model (games or mixed, man-machine simulation). To operate the people in a simulate. This is one of the motivations behind the sustained and fruitful
computer model or the game with human players is to simulate the processes of the (and continuing) development of the Inter-Nation Simulation (Guetzkow and
system represented in the model: "A hypothetical stream of behavior is generated others, 1963; Guetzkow, 1968, 1969; Smoker, 1968a, 1968bJ 1970). The use of the
that can be compared with the stream of behavior of the original system" (Clarkson term reference system rather than real system records the fact that we can compare
and Simon, 1960, p. 920). In this manner, the implications of what is said about a model only to some reference information about the real world, never to reality
behavior can be compared precisely with what is observed, and better descriptions itself, which we know only through selected and interpreted statements (see Guetz-
can be devised to fill the gaps uncovered by the comparison. kow, 1968).
Capabilities for describing complex processes precisely and for reproducing Human or mixed man-machine simulation may suggest new explanations of
the behavior of a human system are indeed desirable features of a method for scien- behavior; may be used to replicate particular reference systems (for instance, particu-
tific description. The reader should keep in mind that simulation, like other methods, lar time periods in international relations) ; may generate alternative futures or pasts
has its costs as well as its benefits, and also the deeper warning that simple insights, that shed light on how such alternatives might occur; and may afford an unusually
even ambiguously stated, may fruitfully guide our thinking about behavior. The rich setting for laboratory experimentatiori (see Chapter Thirteen in this volume).
researcher must decide for himself whether a precise working model of a complex Such simulations or games may be used also for teaching purposes. (For additional
process is what is needed and whether it is worth the costs of development and use. information, see Inbar and Stoll, 1971; Tansey and Unwin, 1969; Boocock and
Conceptually, simulation models depart sharply from other ways of describing Schild, 1968.)
behavior. A large part of our knowledge about human behavior is stated in the form An experimental approach may be fruitful for computer simulations as well
of static, correlational propositions: variable Y is said to be related to variable X. (Naylor, 1969, 1971). Experiments with a suitably valid simulation model may help
Considerations of the sequence and timing of events are frequently left implicit or the researcher understand the dynamics of the reference system; in policy applica-
ignored altogether. Not so in simulation: the aim of generating a stream of behavior tions, the effects of alternative policies can be evaluated.
over time requires constant attention to the sequences in which events can occur and In short, simulation is a set of methods for constructing hypothetical models
the time periods involved. A high correlation between attitudes of political efficacy of behavior and for generating their implications, in the search for better explanation
and, say, participation in politics is interesting, but a simulation requires a statement and theory or for useful prediction. Simulations may be used to exzmine the con-
about what happens to participation when the attitude changes-whether participa- sequences of alternative assumptions about behavior; for validation against particular
tion changes as well, how soon, and in what way. Much more information is needed real systems; or for the generation of data. The operations of a model may be carried
than is given by the correlation. Simulation requires strong statements about the out by hand, by computer, by human actors, or by combinations of these.
relationships among a great many such events and conditions. Simulation has had a brief but vigorous history. Raser (1969, ch. 3) traces
The need for precise description often entails the use of mathematical func- the roots of modern simulations back to several centuries of military games, but
tions in simulation models. Sometimes it is possible to derive by logical methods the more closely related developments in small-group experimentation, decision theory,
main consequences of such a model. Especially for complex models, however, deriva- and systems analysis occurred much more recently. Dutton and Starbuck (1971,
tion is not feasible, and computer simulation provides the only way to determine how ch. 1-2) note that work on machine and other physical models of behavior appeared
the model behaves. as early as the last years of the nineteenth century, but almost all of the references
Sometimes mathematical descriptions are partly or entirely inappropriate for in their bibliography on machine simulation of behavior have appeared since 1945-
the phenomena under study. Before the advent of the computer, verbal models of that is, since computers became available for research.
complex behavior, models precise enough to reproduce behavior, were scarcely con- Herbert A. Simon's work stands at the origins of several strands of develop-
ceivable. The computer offers a general logical and symbol-processing capability that ment of both computer and human simulations in behavioral science.,With others,
386 Rufus P. Browning Simulation: Attempts and Possibilities 387

he developed and introduced to several fields the elements of a theory of decision approximates the behavior of human subjects? To motivate an answer to this query,
making, by individuals and in groups and organizations (Simon, Smithburg, and we must have in mind some political actor whose thinking or choosing processes we
Thompson, 1950; Simon, 1947; March and Simon, 1958). The theory created a would like to describe in detail, with a view to explaining the stream of behavior.
strong impetus toward the systematic study of decision making. Simon's associates Official as Information-Processing System. To elucidate why and how we
at Carnegie Institute of Technology included Richard Cyert and James March, who might go about studying the problem-solving process of a political officeholder, build-
published in 1963 A Behavioral Theory of tlze Firm. Their theories of organizational ing on the theory and methods already developed in the context of formal problem
decision making were made operational in the form of computer simulations that solving, consider the problem facing the head of a large government department at
generated major decisions of various business organizations over time. Harold Guetz- budget time, when he receives requests from the chief officials of the bureaus in the
kow collaborated with March and Simon (1958) and went on to lead the mixed, department and must himself prepare from them a budget to submit to the legisla-
man-machine Inter-Nation Simulation project at Northwestern University (Guetzkow ture. I t is of some interest to understand just how a department head handles this
and others, 1963). Simon's students have published many computer simulations. task. We know that some departments become veritable engines of innovation through
Simon himself, in association with Allen Newell especially, participated in a radical their budget processes, while others wallow apathetically through the same formal-
transformation of the study of thinking and problem solving through a series of ities; the way that budget requests are treated, therefore, appears to be crucial
computer simulation models and a theory of human information processing (Newell (Browning, 1970). To guide our research into the department head's thinking, we
and Simon, 1956, 1972; Newell, Shaw, and Simon, 1958; and many other publica- refer to Newell and Simon's H u m a n Problem Solving (1972). All of the italicized
tions). concepts in this section are defined and discussed in detail there.
Several surveys of simulation in social science have been published. In my Suppose that revisions of requests are developed and agreed to in meetings
view, the best of those relevant to political psychology is a long article by Abelson of the department head and his staff with each bureau chief and his staff. We sit in
( 1968b), which should be required reading for anyone seriously interested in simula- on these meetings, transcribe the discussions, and go over the head's behavior with
tion. Four of the seven computer simulations that he describes in some detail are him afterward, tryifig to tease out the thinking that generated his behavior, hoping
studies in political psychology. Abelson offers an extended and wise discussion of eventually to formulate a working model of it. The budget meetings last four or five
when and how to do computer simulation; Dutton and Starbuck (1971) have hours; our research strategy leads us to work first with this relatively short-term
prepared somewhat different but also incisive suggestions in a book that includes an behavior. We are able to observe a series of meetings- with different bureaus at about
enormous bibliography and a good selection of articles on computer simulations, the same time. Thus, some features of the head's task environment are held constant,
drawn from diverse sources. Guetzkow, Kotler, and Schultz (1972) bring together and we have an opportunity to observe his performance on a number of somewhat
reviews of both human and computer simulations in many fields of social science, different problems within it (revising the budgets of the various bureaus). These tasks
plus reports of particular simulations selected from the published literature. The text are genuinely problematic, not routine, for this department head. He will have
Comijuter Simulation Techniques (Naylor and others, 1966) offers useful advice objections to the requests before him, and he will know when a bureau's budget is
and information, especially for simulation of mathematical models of a kind common finally acceptable; but its final shape, when he incorporates it into the department's
in economics. (See also Coleman, 1964c, 196513; Guetzkow, 1962; Cohen and Cyert, budget, is built up only as the meeting proceeds. It depends on what he learns about
1965.) Because so many authors have surveyed simulation methods and applications, the bureau's requests and operations and about the bureau officials themselves as he
I will emphasize suggestive uses rather than attempt an exhaustive coverage of simu- questions, cajoles, persuades, threatens, and reassures them during the meeting, and
lations of individual, electoral, small-group, organizational, and organizational system as they respond with information, objections (or acquiescence), and reconstructed
behavior. justifications.
From study of the meetings and from conversation wi,th the department head,
Individual Behavior we conclude that three goals define the immediate task for him: First of all, he seeks .
to convince the legislature that his department's budget requests should be granted,
The theory of individual cognitive processes has been more sharply trans- even though they involve large percentage increases and public funds are scarce.
formed and expanded by computer simulation than any other field in the behavioral This goal entails both modification of bureau requests and improvement of the
sciences. That emerging success story has favorable implications for the study of justifications for them. Second, he seeks to bring the bureaus into line with his goals
individual political behavior because others can build on the concepts and methods for the agency. (In the longer run, his goals are shaped by his subordinates' best
developed there; borrowing may be fruitful even though the aims of research and proposals and by experience.) Finally, he seeks to stimulate the bureau heads to
the research settings of cognitive psychology are not on the face of it very close to allocate their full energies to the attainment of his goals, which involve substantial
the objectives and settings of research into political behavior. policy and organizational change and require continuing innovation by the bureaus.
What does it mean to the student of political psychology that researchers The second and third goals pose a most delicate problem of combining obedience and
have developed the elements of a theory of human thinking and that they have been enthusiasm. All three goals could also be thought of as subsidiary to some higher
able to generate via computer simulation problem-solving behavior that closely goals for the department head, referring perhaps to the public purposes of his depart-
 Rufus P. Browning Simulation: Attempts and Possibilities 389

ment and to his own career. While this may be the case, it is not necessary to specify processes and that the human information-processing system confronted with a task
these higher goals in order to generate his behavior in the meeting. combines these processes in a finite number of programs that are triggered by particu-
The department head's behavior is generated from a problem space, a repre- lar kinds of subgoals. For example, our department head evokes a discrimination net
sentation of the task environment internal to him: facts of the request and of the to evaluate a bureau's request prior to the meeting. The net contains his major cri-
bureau as he perceives them; of his own objectives and criteria of acceptability; of teria for acceptance of requests. During the meeting unacceptable requests undergo
the methods he might use to find unacceptable requests, to determine the status of a program of finer discriminations to specify precisely the difference between the
major problems and opportunities in the bureau, to improve faulty justifications, and request as it stands and an acceptable request; then goals are evoked and methods
to evaluate and shape the budget and the performance of the bureau; and of other applied to reduce these differences, difference reduction being a common type of
aspects of the problem and possible solution procedures. problem-solving goal and method.
The department head has a rather rich repertoire of methods. Some of these The illustri~tioncan prove nothing, of course, but may suggest that interesting
are routine and trigger only small changes in the budget, like the addition of a microprocesses of individual political behavior can be observed and modeled. To
sentence to a justification in order to tie a request more explicitly to the department's help us in this work, a substantial body of theory and method, developed in the study
stated goals. One method, frequently applied, eliminates requests that the legislature of formal problem solving (playing chess, proving logical theorems), is available.
is likely to criticize. Other methods of assessment and questioning open up whole W h y Simulation? As we study the variety of department heads or other
areas of the budget for restructuring, and may lead to long sequences of questions political figures, we will probably find large differences between them in the way
and reformulations. Some of these searches for a better request will be abortive. they approach their tasks. Some department heads will not seek change, will not be
Sometimes a second approach is taken to these items, and another way of organizing interested in the problems of the bureaus, will not be sensitive to the needs of their
or justifying them developed; but some items that initially appeared salvageable turn subordinates, and will not apply criteria to their budgets sufficiently powerful to
out poorly and are dropped. avoid legislative criticism. Either way, useful information is obtained. If a department
Particular subgoals evoked and applied in the course of the meeting include head does well, studying his internal problem-solving processes gives us information
administrative considerations, such as matching a request for a new division to a rule about the task and about the psychology of successful performance. If he does poorly,
of thumb for feasible rate of growth, taking into account the administrative burden we gain information about the nature of the internal mechanisms that limit his
of recruitment, training, and startup. Other subgoals refer not to the budget per se performance.
but to the interpersonal relationships at the meeting. Praise is mingled with criticism. Newel1 and Simon suggest that a theory of individual performance must
Care is taken to avoid harsh criticism of a bureau chief in front of his staff. Both in include both an analysis of the tasks involved and an analysis of rational adaptation
the pattern of budget cuts and acceptances, and in the flow of verbal communication, to task requirements-and even, we might add, an analysis of irrational adaptations.
the bureau chief is led to feel that the department head is helping him deal with In fact, these dual concerns are already clearly present in the existing correlational
the problems of his bureau; thus, an implicit exchange is established. A third set of research on managerial styles and productivity (for example, Likert, 1967), phenom-
subgoals refers to the bureau chief's understanding of the relationships between his ena rather similar in content and setting to the interactions be.cween agency officials
bureau's structure, its budget, the major trends in its environment, and the goals of over their budgets, described above. Would it not be equally fruitful to apply a
the department for the bureau. The department head probes the bureau chief to correlational approach to the budgetary process, over a set of agencies, measuring
determine his perception of these relationships; if his perception is in any way defi- the relationship between administrative styles and the budgetary behavior of sub-
cient, time is taken to explain the relationships. ordinate officials? This would also be fruitful, I believe, and I have no quarrel with
This particular department head is in some ways a difficult subject because it. The essential argument for choosing to describe precisely the microprocesses of
his subgoal structure is so extensive and his repertoire of methods so rich. With less behavior lies in the level of explanation desired. A simulation of individual behavior
experienced, less highly motivated, and less intelligent officials, we would expect a "posits a set of processes or mechanisms, that produce the behavior of the thinking
less complex and extensive problem space. Nevertheless, we might be able with further human. . . . I t does not simply provide a set of relations or laws about behavior
study to describe not only his behavior but also some aspects of the structure of goals from which one can often conclude what behavior must be" (Newel1 and Simon,
and methods that generates it. (We need not insist that our model of his behavior 1972, p. 9). Rather than correlating a few selected features of the processes under
be capable of receiving and emitting natural language, however. A capacity for study, Newell and Simon's simulation approach seeks to reconstruct the process itself,
understanding and formulating sentences is an enormously difficult problem in itself; in model form. Obviously, this fs a more detailed account of the events than is
discretion dictates a symbolic representation of the task environment and of the customary with a correlational approach. I t is also a stronger account, setting forth
applied by the department head. Newel1 and Simon do not demand of their less ambiguous statements about many more aspects of behavior and its generators.
computer models that they mutter and stutter like the human subjects, only that they Cognitive-Process Simulation. The concentration on process and on the
make the same moves.) explication of mechanisms sufficient to generate individual behavior is a radical shift
Our reconstruction of his problem space is facilitated by the knowledge that of direction in psychology. In political psychology, too, such an approach takes a
the complexities are made of a relatively small number of elementary information deeper slice of behavior and attempts a stronger description of it. As noted, cognitive-
 Rufus P. Browning Simulation: Attempts and Possibilities

process simulation differs from a correlational approach to the relations between the nected behavior no doubt account for the tendency of simulation studies of individuals
attitudes and the actions of individuals, and the effects of actions. I t differs also from to focus on relatively short behavior sequences, in settings which permit unusually
a "whole-personality" approach, since it does not attempt to lay out all the important close observation and interviewing. Work on a formal problem is a matter of minutes;
sources of action internal to the individual; rather, only those mechanisms sufficient budget meetings last several hours; the psychiatric interview is one hour; the re-
to generate behavior in a particular task environment are posited. Third, Newel1 and sponses of the political ideologist modeled by Abelson (discussed below) might take a
Simon's information-processing descriptions of problem solving give much greater few seconds or a couple of minutes. To be sure, intensive work with these short
prominence to the intended rationality of the individual at a task than we are sequences should prepare us for equally well-specified models of longer ones; but in
accustomed to in political psychology. Newel1 and Simon were able to assume that the meantime, correlational and other approaches are still necessary if many instances
their subjects were properly motivated to perform the intellectual tasks set before of the longer sequences are to be studied at all, or if we are to reconstruct processes
them. The researchers could assume that the subjects really were trying to reach a that have already taken place.
well-defined goal, known to both subject and scientist. Though simulations are too demanding a method for modeling long or past
In the context of formal problems, the end goal is common to all subjects; sequences of behavior, or for portraying many different individuals, still we can find
the intermediate methods and subgoals may differ from one person to another. But easy continuities of theory between familiar sorts of correlational and developmental
this is not necessarily the case in political or administrative problem solving. Different studies, on the one hand, and simulations, on the other. For example, suppose we
people bring different motivations to the same political position; the behavior of measure the motives for achievement, affiliation, and power of a set of politicians in
individual officials will be generated by partly different end goals. Because different policy-making positions, and then find that certain patterns of motivation are asso-
goals create different subjective tasks in a given environment, political psychologists ciated with distinctive patterns of behavior in office (Browning, 1968a, 196813). In
will want to extend the cognitive-process approach to include an explication of tlie the parlance of problem-solving theory, we are tapping the particular goals (motives)
motivational constructs, or goals, that various individuals bring to a task. Still, carry- that these individuals bring to a given task environment, and the methods (behavior)
ing the Newell-Simon theory of problem solving over into political psychology that they apply to reach their goals. If we regard our respondents as intendedly
would probably induce a clearer recognition of the fact that political actors are com- rational, we should not be surprised to find some correlation between the goals and
monly goal-directed and intendedly rational; a greater emphasis on an analysis of the methods, between a concern for achievement and actions intended to produce
the task environment in which their goals and actions operate; and closer attention it. In the correlational research mode, conceptualization and valid measurement of
to the connections between goals and actions. behavior and of the goals and expectations relevant in the task environment are
Newel1 and Simon's argument in favor of the analysis of rational behavior crucial, and it is no small achievement for the researcher to obtain high and reliable
and of the task environment in which it occurs reflects in part simply their interest correlations. In the simulation mode, conceptualization aims not at measurement of
in formal problem solving. Nevertheless, political psychologists as well may profit summarizing factors of personality but at construction, from the directly observed
from the suggestion that they understand thoroughly and clearly the rational content behavior and thinking of the subject, of mechanisms sufficient to reproduce behavior.
of the behavior they study. In some contexts, several kinds of rational behavior may We may still wish to call a particular program of criteria for evaluating situations
be conceivable, corresponding to different subjective goals; or it may be useful to posit and choosing actions "achievement motive," but we will have gone beyond detection
a goal and a variety of rational behavior on prior theoretical grounds-for example, of a mechanism linking situation and behavior to a precise specification of it.
ambition for higher office (Schlesinger, 1966) or for higher agency budgets (Nis- Descriptions of political thinking as individuals try to adjust their actions to
kanen, 1971). the demands of their goals and of the problematic environment are not yet available
Though rational behavior is the focus of Newel1 and Simon's simulation in sufficient detail to permit precise modeling, but it is clear that many fragments of
studies and may be a useful concept in political psychology as well, it is certainly not such thinking can be obtained or constructed from interviews (Browning, 1969, ch. 4-
an inevitable focus of research using the simulation method. Any behavior occurring 5) and in some cases from analysis of the papers and public record and statements
over time may be simulated; provided the subject's states and acts are truly strongly of well-known political figures (George and George, 1956). Barber's work on state
connected and adequate access to his behavior and thinking is feasible. For example, legislators (1965) and especially his analysis of twentieth-century presidents of the
Kenneth Colby and his associates (Colby, 1965, 1968; Tesler, Enea, and Colby, 1968; United States (1972a) strives to bring out the structure of the personality mechanisms
Colby and Schank, 1973) have developed a body of theory and simulations of the that generate performance in a given task environment. Barber interprets presidential
client in a psychiatric setting, their aim being to construct mechanisms of personality psychology in terms of the demands that the task environment of the presidency places
sufficient to produce the observed verbal behavior of- the client in response to the on the president-demands commonly recognized by the occupants of that difficult
statements and questions of the therapist. The question of rationality does not arise position even when they are not able to respond in an effective way to them. The
in their work; but it is necessary, and possible in this case, that they have close access president as problem solver comes through clearly in Barber's analysis, equipped in
to the client's behavior and thinking and that his mental states and verbalizations be personality to cope rationally with some of the problems of the office, but unable in
strongly connected to each other and to the stimuli emitted by the therapist. the face of other demands to find or apply methods of dealing with them that could
The two requirements of intensive observational access and strongly con- lead to success in the light of his own goals.
 Rufus P. Browning Simulation: Attempts and Possibilities

Bardach's (1972) description and theoretical construction of highly rational simulation to study the psychology of political behavior. Even given a concern for
political entrepreneurship is not explicitly psychological in orientation but still brings I process, of what special use is simulation?

1
to light significant features of individual political thinking and behavior. What we see I t is easier to recommend the intensive study of some interesting political and
in Bardach's account of the repeal of mental commitment laws in California is the psychological process than it is to insist on the value of a computer simulation. No
dynamic mobilization and manipulation of authorities and other interests in a com- doubt many researchers would be content to rest with verbal accounts of the behavior
plex, problematic political environment. The leaders of the repeal effort foresee oppor- I they study, and there is much to study before verbal accounts become entirely in-
tunities and problems; map out strategies for building support and for adding to
their capacity to obtain additional support; and neutralize, pacify, or outmaneuver I adequate. However, there are limits to what a verbal description of the mechacisms
that generate performance can convey, because of the imprecision of words and
the anticipated opposition. An opposing figure is placed in a situation constructed because the structure of natural language does not force us to squeeze out ambiguity
so as to damage his credibility and maximize the chances that his personality will lead and to make all relationships explicit in the system we are describing. This essential
him into irrational, counterproductive statements. Another prominent authority, I ambiguity, and the lack of a sufficiently powerful method for generating the behavior
known to oppose repeal and to fight tenaciously when cornered, is carefully soothed
with personal praise and inclusion in repeal strategy meetings. The mechanisms that
generate these strategies are not completely described, but some are well specified
I described, must leave us with a large component of uncertainty that we have the
right description, the right model of the actor's problem-solving sysrem. Computer
simulation does not allow ambiguity; it does force the user to make all relationships
and others can be inferred from the detailed analysis of problem and behavior explicit; and the computer does offer a sufficiently powerful way of generating the
presented. No simulation is attempted, but the orientation of the study toward in- implications of even a very complex description.
ternal calculations and behavior process is akin to that of problem-solving simulations. Although empirical work should usually be conducted in close support of a
One might have conducted an explicitly psychological study of political simulation effort, the researcher should expect a significant shift of time and energy
effectiveness in the mental commitment repeal case, attempting perhaps to fit to the toward questions of theoretical formulation and simulation technique. For this
data a model of goai effectiveness as a function of position and of personal skill and investment, researchers report heuristic benefits as well as advantages of precision
style, and skill and style as a function of "personality characteristics." Measures of and generating power. Colby (1965, p. 501) wrote of his simulations of neurotic
"political efficacy"; of authoritarianism; of achievement, power, and asliative processes that it was too early to plan extended empirical tests of the models: "At
motivation; of manifest anxiety; of Machiavellianis~n-these or others might have the moment their main contribution has been to clarify theory by forcing the theorist
been obtained, and surely some oi them would be found closely related to one or to be explicit about his system, its entities, and relations." In the same vein, Abelson
more differences in measures of behavior among the political actors. It is difficult to (1968b, p. 292) suggests: "One virtue for the investigator in struggling with these
believe that this approach, even skillfully executed, could yield an explanation as flow-charting problems is that he may be forced into closer scrutiny of his own
satisfactory as that provided by Bardach's less quantitative but better-grounded anal- theory. . . . Once the investigator starts thinking about process timing, he must
ysis. Instead of measures of "personality" detached from the tzsk environment, think more deeply about the psychology of the processes."
Bardach is able to characterize, even if not completely, the state of mind and strategy Indeed, it may be that the prime value of most social science simulations is
of the principal actors at several stages in the unfolding campaign for repeal; the heuristic (Freeman, 1971). Perhaps they should be appreciated for this even when
links between circumstance and mind, the mind and action, are reconstructed in a a valid model is not attained. Depending on the stage of developmeot of a particular
way that is most uncommon in correlational analysis. field, an effort to pinpoint the gaps in knowledge may be precisely what is needed,
The suggested analysis of the agency head's budget behavior, Bardach's and a valid model too much to hope for.
account of political entrepreneurship, and the Newell-Shaw-Simon models of problem Newell and Simon (1972, p. 870) state that the computer was for them "an
solving have in common a further characteristic that distinguishes them from corre- important tool, for it has permitted us to carry out, through simulation, detailed
lational studies. All three concentrate not so much on individual differences and comparisons between theory and data, and to derive numerous empirical predictions
their effects-by definition the essence of the correlational method-as on explication from theory." These functions are so crucial for the specification and validation of
of common forms of thinking and behavior. We know that there are many kinds of a theory (though not for the initial statement of it) that it is difficult to see how
problem solvers, agency heads, and political entrepreneurs, but the main focus in Newell and Simon could have developed their theory without the use of the com-
these efforts is on the development of theory suitable to describe the range and puter. Even so, they assert that "the computer as hardware" was not as important as
basic characteristics of the behavior found. In effect, they seem to be saying, a better "the new concepts of information-processing systems and the new formalisms and
understanding of individual behavior is necessary before we throw our resources fully languages for describing such systems with precision" (p. 870). The conceptual inno-
into exploration of the consequences of differences between individuals. vations had roots in a growing interest in human thinking on the part of scholars
Are Simulation and Computer Essential? The studies discussed in the in diverse fields, antedating the invention of the computer; no doubt the latter greatly
previous pages are like the simulations of individual problem solving in their focus stimulated the theoretical developments. In earlier publications of Newel1 and Simon
on process, on close observation of behavior, and on reconstruction of internal and their associates, the computer models were prominent; by the time of the 1972
mechanisms that generate behavior. Obviously, one does not have to use computer book they were able to describe the human information-processing system directly in
 Rufus P. Browning Simulation: Attempts and Possibilities 395

terms more appropriate to the study of human behavior (and more comprehensible problem solving from which it springs. The goals and methods of formal problems
to human readers). The resulting theory was the goal, the simulations instrumental are defined by the problems themselves in precise symbolic notation (for instance,
in its development and testing. logic) or in terms for which a notation can be readily invented. But the elements
rrIdeology Machine." Abelson's ideology machine, a model of a closed belief of a belief system are linguistic concepts and sentences; hence, the simplification that
system, represents an approach with potentially great significance for political psy- might be achieved via a symbolic notation for words would increase the distance
chology. It is the only computer simulation extant that models political behavior of between the model and the task environment of the subject, and would constrain the
a politically interesting individual and is well grounded in psychology and (appar- use of the model to pinpoint gaps in knowledge about the effects of language on
ently) in data on the subject (Senator Barry Goldwater). The model portrays the belief. Construction of the model pinpointed certain gaps in the available knowledge
performance of an individual confronted with statements or questions about events about psycholinguistic processes. Subsequently, experiments were designed and carried
in international relations; for example, "Secretary Laird says the communists are out to fill the gaps (Gilson and Abelson, 1965; Kanouse and Abelson, 1967).
planning to attack Thailand." One version of the model (Abelson and Carroll, Abelson's effort to model one individual, and one with a well-developed and
1965) has mechanisms to test statements for credibility; if a statement is found cred- relatively closed belief system, is sensible. It would be desirable also to have a subject
ible but upsetting-inconsistent with beliefs already in the system-the mechanism one could interview; most of the ideological content of the model was developed
attempts denial and rationalization. Later reports (Abelson, 1968a, 1971, 1973) from public statements by Senator Goldwater. Interviews would bring the method of
describe a capability for generating scripts, or potential event sequences, in response to data collection closer to the Newell-Shaw-Simon method of observing subjects, ob-
questions about what will happen; these reports greatly strengthen and broaden taining their own verbal account of what they are thinking, and asking them ques-
the conceptualization of belief systems. tions about what they are doing and why.
The structure of Abelson's model is already a contribution to the theory of In research on formal problem solving, researchers could simply take as given
attitude structure and of ideology in particular. The basic elements in the system are the goals of the subjects; they were assumed to be well motivated, and their goals
about five hundred nouns or noun phrases and one hundred verbs or verb phrases. and subgoals were defined by the problem. In Abelson's belief-system model, in con-
A variety of relationships organizes these elements and provides the connections that trast, all of the motivation and all of the particular goals and subgoals of the "prob-
permit inferences and answers. First and most simply, there is a horizontal relation- lem" (of inconsistencies between beliefs) are part of the individual's belief structure
ship between elements: nouns are connected to verb-noun predicates, as in "Cam- and process, not part of an objective problem external to the system. Thus, knowing
bodia accepts U.S. aid," to establish the statements "believed" by the system. Second, the person's problem space and getting it into the system are much more difficult.
elements are connected hierarchically: most nouns and verbs are classed within The gap between observation and model is especially painful for the "evaluations"
general categories; for instance, "Hungary" would belong to the class of "communist that register affect in the model: "Every element, whether concept or predicate,
nations." I t is an instance of "communist nation," and "communist nation" is a carries a signed quantitative evaluation summarizing the resultant of all positive or
quality of "Hungary ." This relation permits the system to determine the credibility negative affects attaching to the element" (Abelson and Carroll, 1965, p. 26). Where
of a new statement by checking whether the statement is believed true of other in- these quantities come from is not clear. Yet some method for obtaining them is essen-
stances in the category (induction) or whether it is believed true of the categories tial, because important processes take the evaluations as input or operate upon them.
to which the instance belongs (deduction). A statement that includes a general verb While the treatment of evaluations may involve an empirical gap in the
category between two noun categories may define a generic event; for example, model, apparently a good deal of interesting empirical work has been done in the
"(communist nation) (physically attack) (neutral nation)." The third relation construction of the belief system; unfortunately, this work has not to my knowledge
within the belief system connects generic events to build up the scripts-sequences been reported in published form. Abelson mentions that his method is to paraphrase
of generic events. Scripts permit the system to respond to a query about what will the known beliefs of the subject and that "the paraphrasing procedure is extremely
happen following a particular event. This event is located as an instance of a generic difficult" (Abelson and Carroll, 1965, p. 30), but the rules for paraphrasing are not
event; then the script that follows from that generic event is triggered, in effect as given, nor, indeed, is the resulting belief-system model described except by illustration.
a model for what will happen in the particular case referred to by the original Even if no simulation were involved, this work would be of interest, and its public
question. A script may include alternative possibilities, with different sequences flow- availability is to be hoped for.
ing from each one. Simulation offers to students of ideology and of attitude structure and change
Abelson (1971, p. 13) gives this illustration of a script-generated answer to an opportunity to describe their processes precisely and to generate tests of the
the question "If communists attack Thailand, what will happen?'The model's validity of process models. (For discussion of these processes, see Chapters Three and
answer: "If communists attack Thailand, communists take over unprepared nations Four in this volume.) Simulation is a demanding method, however, not routinely
unless Thailand ask-aid-from United States and United States give-aid-to Thailand." or easily applied to the description of individual political behavior and thinking.
Obviously there are problems of style, but that is unimportant; the meaning of the Clearly, a prior commitment to simulation in studies of political-psychological process
answer is plausible. carries a high-risk rating. But studies of process in political psychology, as we can see
Abelson's project is more difficult in several ways than the work on formal from recent work of this type, already make a strong contribution to knowledge and
 Rufus P. Browning Simulation: Attempts and'Possibilities 397

prepare the way for the more precise and reproducible statements that are the aim altogether. (Rosenthal also explicates portions of McPhee's model that are not well
of simulation methods. defined in the original publications.) Both Rosenthal (p. 251) and Boudon (1965,
p. 13) conclude that McPhee's model makes excessive demands for data on the
Computer Simulations of Electoral Behavior microbehavior of voters-data that are in practice out of reach. Nevertheless, as a
pioneering effort at a most difficult task, McPhee's work is still worth referring to
The possibility of simulating the complexities of an election campaign on the and learning from.
computer has attracted several researchers. 3ecause the earlier work in this subfield Abelson and Bernstein's (1963) model of community referendum contro-
has been repeatedly reviewed by other authors, only truncated descriptions will be versies similarly attempts to predict outcomes of a campaign from simulation of
offered here. individual exposure, communication with others, and attitude change. (Francis,
Pool, Abelson, and Popkin (1965) attempted a simulation of the 1960 and 1972, pp. 251-256, provides an explication of the model in flow-chart form.) The
1964 presidential elections for prediction purposes (as contrasted with other simula- theoretical specification of the model is a tour de force in itself and suggests a worth-
tions reviewed in this chapter, which aim for description of process). The model while strategy for simulators: good theoretical work is interesting and important
seeks to predict the vote of 480 groupings defined by socioeconomic characteristics; even if a simulation is not operationalized or not tested (but perhaps the simulation
then knowledge of the extent to which each group is represented in the population should not be promised until it has been tested) ; and careful development of the
of each state permits prediction of the vote by state. For a first effort, the model, theory should precede the writing of the simulation model in any case. Abelson
based on data at least two years old but improved by some good guesses by the (1964) argued that available mathematical models of the distribution of attitudes
authors, did astonishingly well, achieving a correlation of .82 between the predicted subject to controversy tended to predict a movement of attitudes toward agreement,
and actual vote percentages for John F. Kennedy in thirty-two states outside the whereas the evidence from fluoridation referenda and some other controversies
South (p. 57). A similar effort for the 1964 election managed a correlation of .90 showed polarization: hence, the simulation, to obtain effects observed but not pre-
with the state-by-state vote, including the South, with a median error of 3.4 percent dicted. However, it is not clear that the very complex model set forth by Abelson and
(p. 174). The 1960 model is extensively reviewed by Rosenthal (1965). Bernstein was needed to account for polarization. Boudon (1965, p. 14) argues that
Although the Pool-Abelson-Popkin model does not try to represent psycho- "it would suffice to introduce the familiar hypothesis of selective exposure to generate
logical dynamics explicitly, nevertheless some interesting findings of psychological bimodal distributions" of opinion. If and when confrontations of the model with
relevance emerge from analysis of the model and of the huge data bank the authors data on fluoridation and school desegregation controversies appear in print, it should
assembled to support the model. For instance, they demonstrate that socioeconomic become possible to determine what degree of complexity is necessary to produce
predispositions are only additively related to attitudes on a large number of issues. observed effects. The question raised about the need for complexity suggests a
Multiplicative effects are rare except for "party-dominated" issues-those for which strategy of developing and testing simpler models against the data in the hope that
party identification is strongly related to attitude; on these issues, party identification the more complex model will not be necessary. Indeed, the model has so many
interacts with other socioeconomic factors in their impact on attitude (Pool, Abelson, parameters that it may be extremely difficult to test, as Boudon (1965) and Rosenthal
and Popkin, 1965, ch. 4) . (1968, pp. 253-254) argue.
Other voting simulations, by McPhee and associates and by Abelson and Shaffer (1972) reviews the Pool-Abelson-Popkin and McPhee models and
Bernstein, contain explicit representations of individuals and more or less elaborate reports simulations of voting in the 1961 presidential election, using two different
arrangements for contact, communication, and influence among them. McPhee's models, one based on a theory of the rational voter developed by Anthony Downs
model consists of processes for stimulation (for instance, from the media), for dis- (1957), the other on a six-component multiple regression model developed at the
cussion among voters, and for learning (modification of dispositions) (McPhee and Survey Research Center (Stokes, Campbell, and Miller, 1958; Stokes, 1966). Both
Smith, 1962; McPhee, Ferguson, and Smith, 1971). Although the McPhee model of these models attempt prediction of individual votes with no mechanisms for
was used, with modifications, for limited predictive purposes (McPhee, 1963), it was interaction.
designed to investigate via conceptual experiments macroprocesses of voter preferences Shaffer shows ingenuity in obtaining measures for the concepts of Downs'
extending over many elections. In one such application, McPhee and Ferguson (1962) model from the 1964 SRC survey data; however, students of voting behavior may
used the model to test whether and under what conditions an electorate subjected to still be less than satisfied with some of the measures used. For example, a key con-
repeated strong disturbances from political crises and charismatic appeals will cept of Downs' model is "current party differential": the difference between the
develop immunity to the effects of additional disturbances-an interesting question. utility a voter has received from the party in power and the utility he could have
But even though the model produces interesting results, and even though these results expected from the opposition party (Downs, 1957, p. 40). For a measure of this
roughly parallel certain historical experiences, the model remains at the level of a tool Shaffer (1972, p. 70) relies entirely on respondents' answers to the question
for speculation because parts of it are not well grounded either in theory or in data. '<Do you thirik it will make any difference in how you and your family get along
Rosenthal (1968, pp. 245-253) argues that some of the conclusions McPhee arrives financially whether the Republicans or the Democrats win the election?" Also, the
at have been obtained with simpler models and analyses, avoiding simulation concept of "information costs" is represented in the model by a measure of total mass
 Rufus P. Browning Simulation: Attempts and Possibilities 399

media use, a very different notion. In spite of these deficiencies in the data (which reported the development of a computer simulation of the campaign as well as the
stem from the fact that the SRC survey was not designed to test the Dolvnsian vote in presidential elections. Schauland, Naylor, and Kornberg ( 1971) have built
model), a modified version of the model, with the information-cost variable dropped a man-machine simulation; one of its purposes is to give people practice at making
altogether and a party-identification variable added to guide the votes of those simulated campaign decisions on such matters as allocation of resources, media mix,
respondents with zero "party differential," correctly predicted 66.6 percent of the and issue stands.
reported votes and abstentions in the sample (p. 107).
The SRC six-component model attempts to explain an individual's voting Group and Organizational Processes
choice as a function of his attitudes toward the Democratic and Republican candi- In the section of this chapter on individual behavior, an illustration of possible
dates as persons; his attitudes toward the parties and candidates related to group research on the problem-solving behavior of a department head was developed.
benefits, domestic policy, and foreign policy; and his attitude toward the general Ordinarily, the group context of the proposed research would lead us to think of it
performance of the parties in managing the government. It does this by regressing as research in group interaction; but it would be much too difficult at this stage to
voting choice on measures of each of the six attitudes. By this method, multiple corre- develop models for all the participants and thus to generate their interactions; there-
lations of .72 to .75 are obtained for the four presidential elections studied, and a fore, it was proposed to take the responses of the other participants as given and
correlation of .98 between estimated and actual majorities over the same four contests attempt a model only of the thinking and statements of the head. Any research that
(Stokes, 1966, p. 28). tries to develop computer simulations of the thinking of more than one problenl
The SRC model as it stands is not predictive: it takes the estimate of best- solver in a group is probably overambitious if the group task involves qualitative
fitting regression parameters from the data on voting choice in' each election, rather representations and moves or is otherwise complex, or if the members see the problem
than predicting it. The weights so determined for the effects of the six attitudes on in different terms. Accordingly, researchers observe groups performing simple tasks
voting choice are different in each election. Because Shaffer wanted to predict vote$, under experimental control. Models are developed to describe behavior in the re-
he adopted a different procedure. Values for each of the six attitude dimensions in stricted problem space so created. Most computer simulations of small-group inter-
the original SRC model were defined as "the arithmetic difference between a re- action refer to such experimental settings. Simulations of more complex processes
spondent's pro-Republican and pro-Democratic responses whose content relates to have left large parts of the model to be played by human actors, as in the Inter-
the dimension" (Stokes, 1966, p. 27). Shaffer apparently simply sums these differ- Nation Simulation (but an all-computer model by Bremer, 1971, builds on the INS
ences across all six dimensions. (Shaffer refers to summation of "net Democratic work) ; or they have yielded models of organizational decision rules without trying
preference valences" [p. 671, a term that is not defined.) This procedure produces a to trace the individual contributions and interactions. These efforts will be discussed
weighting of its own. Specifically, it would seem that Shaffer's procedure weights in turn.
most heavily those attitude dimensions on which respondents are most widely Small Groups. One laboratory situation that aims for a clear specification
polarized; that is, those dimensions on which the distribution of net partisan re- of conflict is the Prisoner's Dilemma game. The structure of rewards in this game is
sponses is most widely dispersed. Such a weighting has some intuitive appeal. such that the two players gain if they cooperate with each other; if one cooperates
Shaffer predicts that all respondents with more responses favoring the Demo- and the other defects, the defector gains even more, but the cooperator loses; and if
cratic party or candidate will vote Democratic; those with more responses favoring both defect, they both lose. Each player is constantly tempted to increase gains by
the Republicans will vote Republican; and those with equal numbers of responses defecting from mutual cooperation; but he also runs the risk that the other player
favoring each party will abstain. This rule generates correct predictions for 69.5 will defect, too, thus turning gains to losses. The rational strategy is mutual coopera-
percent of the sample's reported vote and abstention decisions (Shaffer, 1972, p. tion, but this often does not occur for a large number of plays. The situation is
121). genuinely problematic for the players.
Neither the SRC nor the Downsian model qualifies as a simulation in the Emshoff (1970) noted that one important model of the game (from Rapo-
sense of generating the behavior of a system over time. These are one-shot predic- port and Chammah, 1965) reproduced only the mean behavior of a group of sub-
tions; nothing happens over time in the models. Furthermore, they are predictions jects and was not very useful for predicting an individual game; in addition, its
from attitudinal data obtained at a time very close to the election itself. If one predictions were not sensitive to different possible reward structures. He criticized
wishes to make best predictions from survey data obtained close to election day, other studies on the grounds that they failed to come up with a "unified" model
there are better ways to do it. The SRC six-component model is probably best used that could reproduce the observed behavior. Emshoff proposed a computer simula-
as Stokes and his associates have used it; namely, to trace the net relation of each tion which would reproduce behavior in individual games, based on a general model
attitudinal component to voting choice over time. There is interest, of course, in of the behavior of the individual players. The model of the players is not an informa-
the problem of defining "rational voting" and testing'models of the rational voter tion-processing model of the Newell-Shaw-Simon type, but Emshoff did study the
against survey data (Shapiro, 1969). There is interest also in developing the theory decision processes used by players; he then developed equations in which to express
of the campaign process, which several of the simulations mentioned above gloss over. the effects of four key parameters on players' predicted moves. The parameters opera-
Working at the Survey Research Center, Coombs, Fried, and Robinovitz (1968) tionalized were competitiveness, memory,' foresight, and rigidity.
400 Rufus P. Browning Simulation: Attempts and Possibilities

The competitiveness parameter enters into equations that determine a value been much studied. The model is interesting psychologically as well; the competitive-
to the player for the payoffs stated in the game-Emshoff had found that ,some ness parameter surely refers to a factor of general importance where goals conflict,
subjects were responding not only to their own rewards but also to the possibility and the memory and foresight parameters identify basic differences in cognitive
of getting more than the other player. The memory parameter determines how far reach among individuals. The simulation is at least promising.
back in the game a player's memory extends, and what weight is given to recent Without going in great detail into studies of the PD game, we might find it
versus past moves. The foresight parameter measures the extent to which a player instructive to compare Emshoff's approach with a mathematical approach to the
places a value on the expected future consequences of each move as well as on the game by Ofshe and Ofshe (1970a, 1970b). They explicate and test in several con-
reward expected directly from the move itself. These three parameters modify the texts a mathematical model that yields predictions both of coalition choices in a
value the player attaches to each possible move, given the conditions and previous three-person coalition game and of cooperativeness in the PD game. In contrast to
moves of the game. The player then moves so as to maximize this subjective value. Emshoff, the Ofshes have made no effort to predict individual behavior; also, they
Rigidity has to do with a player's response to a cooperative move by the other player predict behavior only after it has stabilized (usually after fifteen to thirty trials in
after a string of defecting moves by both players; the rigid player will continue to their work), whereas Emshoff tried to reproduce behavior in the first thirty-five
play so as to maximize the value of his next move (as altered by his competitiveness, trials. Thus, the Ofshes' model is not so ambitious-but it is clearer, more general,
memory, and foresight), whereas the nonrigid player will follow the other's coopera- better related to the theory of utility and choice, and better validated empirically
tive move with cooperation. The rigidity parameter appears to be an ad hoc rule than Emshoff's.
introduced to correct the failure of the model to predict cooperation in that particular The Ofshes' model predicts the average performance of a group of subjects
situation. over many plays or trials of a game. In the coalition game, the measure of per-
Twenty games of thirty-five moves each were run in the laboratory and formance is the proportion of all the subject's choices in which he chooses a coalition
simulated with the model. The laboratory results were used to find the specific values with the player who is most likely to reciprocate (thus yielding the payoff for forming
for the eight parameters (four for each player) that gave the best fit for each of the a coalition). In the PD game, the measure is the proportion of all moves that are
twenty different games. Parameter values were assumed stable for the duration of a cooperative, given the degree of cooperativeness shown by the other player. These
game. For each game, the model was simulated one thousand times, using one are in both cases the choices that maximize individual payoffs in the long run.
thousand randomly varied values for each of the eight parameters. The parameter I n the Ofshes' work as in Emshoff's, a player's utility is based partly on his
values yielding the best fit to the laboratory result of the game were chosen to char- valuation of the rewards received by the other players, as well as on the payoffs he
acterize it. "Best fit" means essentially the smallest number of errors in reproducing expects for himself. The inclusion of responses to the gains of others is an important
the individual players' moves; thiry-five moves for each player, where the first two addition to models of rational behavior in experimental games. In the coalition game,
moves of the model were simply assumed to be those of the human players. the Ofshes hypothesize that the players have some concern for an equitable distribu-
When the best-fitting parameter values were used for each game, the model tion of payoffs among the three players, a concern which tempers the maximizing
was able to reproduce moves with more than 80 percent accuracy, on the average, strategy of always choosing as coalition partner the other player who is most likely
over all games. Large differences between games in accuracy of reproduction suggest to reciprocate. To the extent that the observed behavior of the players turns out to
that the model represents some games rather well but that variables not in the model be related to the equity function in the model, we might think of their behavior as
affect play in other games. Trial runs showed also that it was able to generate a corresponding to a situation in which political actors are guided by real concerns for
wide range of sequences of moves, corresponding to various patterns produced by the principle of equity, as they sometimes are. Another possibly analogous situation
human players. is the classic dual problem of forming coalitions sufficient to win both a party's
Ability to reproduce play in laboratory games is an essential and difficult nomination and the ensuing election. The coalition that wins the top prize in the
achievement. I t is still less than prediction because the values of the parameters are nominating convention is constrained to offer the losing players sufficient side pay-
taken from (chosen to give best fit to) the data reproduced. The next step in vali- ments (lesser nominations, other offices, policy commitments) to retain their support
dation would involve a determination of the stability of the parameters by first in the election. This is not a concern for equity but might be conceptualized and
estimating their values from one set of moves and then trying to predict to a second measured in the same manner and produce the same effect on behavior.
set. A further extension might measure the parameters for players independently of In the Ofshes' model of the two-person PD game, players are thought to be
plays of the game. concerned in part with gaining more than their opponents-a competitive orienta-
Computer simulations of laboratory groups are reported also by McWhinney tion that modifies the simple maximizing strategy. Here again, we are familiar with
(1962), Coe (1964), and Clarkson (1968) ; and of other aspects of social behavior the tendency of some political actors to seek not only gains for themselves but losses
by Loehlin ( 1968), Gullahorn and Gullahorn ( 1963, 1971) , and Rainio ( 1965, for others. I n addition, sometimes the zero-sum structure of political situations seems
1966). to require competitive behavior if gains are not available except by imposing losses
EmshofF's model is interesting because of the clarity with which conflict is on other actors.
defined in the Prisoner's Dilemma game, which has the advantage also of having The models for both the coalition and the PD game define a player's utility
 402 Rufus P. Browning
SimuIation: Attempts and Possibilities 403
for a move as the sum of (a) his utility for individual payoff from the move and (b)
his utility for the effect of the move on the other players-the fairness or equity of simpler, more transparent mathematical form. (See Coleman, 1964c; Gregg and
Simon, 1967; Dutton and Starbuck, 1971, Ch. 1, 3.)
the payoffs for them in the coalition game, and his relative gain over the opponent
Several promising developments in the study of real decision-making groups
in the PD game. An equation is derived from this sum. Using the equation, one can
estimate the parameters for the two sources of utility from data on play of the in government have not yet been adequately explored. They are introduced here
not because they are immediately relevant to simulation but because they could be
games.
used to provide the data on which a clear conception of the group decision process
The test of the model is its ability to generate predictions. The parameters
might be built. The study of real groups should proceed along with the study of
are estimated from an experiment with a sample of subjects. Then, to test the model,
another sample of subjects from the same population plays the game; their behavior experimental ones.
An early contribution that has never been satisfactorily followed up was the
is predicted by the model with the parameter values estimated from the behavior of
work of the Conference Research Project at the University of Michigan in the late
the first sample. Thus, no information is taken from the data on the test condition
1940s and 1950s. Researchers observed meetings in government departments and
to generate predictions for it. In addition, a change in the experimental treatment
in firms, recording a large number of categories of behavior. Some of these categories
was made from the estimation experiment to the prediction experiment; this further
referred to the interpersonal relationships and feelings in a group; others referred
strengthens the test of the model, because it is required to make accurate predictions
to the tasks that must be performed for the group to solve a problem and arrive at
under different conditions. The accuracy of the predictions obtained is astonishing.
a decision. The research produced many interesting findings that qualify familiar
For example, under various conditions and for groups ranging in size from six to
assumptions; for example, the availability and use of facts was associated with high
twenty-two, discrepancies between observed and predicted proportions of choices
levels of consensus in groups that found themselves in disagreement over the sub-
of the reciprocating player in the coalition game range from .0008 to .0328 (usually
stance of decisions but not in groups where conflict was intensely emotional and
larger discrepancies for the smaller groups, as expected from sampling error), for
personal. Neither the urgency of an agenda item nor its importance to the welfare
proportions that fall in the range from .4 to .95 (Ofshe and Ofshe, 1970a, pp. 344-
of the organization was related to the way conflict was handled or to the level of
345).
consensus reached (Guetzkow and Gyr, 1954). Numerous other findings raise ques-
Clearly, the model performs extraordinarily well at predicting the average
tions about why and how the observed relationships occurred or why they occurred
performance of a group of players over a number of plays of the games. I t estimates
in many of the conferences but not in others.
parameter values by a straightforward least-squares technique rather than through
One respect in which the observed conferences may have been underdescribed
the awkward simulation-and-search method devised by Emshoff. The parameters
has to do with the content of the problems facing them. A review of studies of
have very clear and direct utility interpretations;, although the model does not yet
individual problem solving and of small experimental groups suggests that tight con-
attempt predictions of individual behavior or of behavior on the early trials of the
trol over the task environment greatly facilitates the development and validation of
games, it appears to offer an excellent base for such efforts. EmshofPs model is also
models of behavior. What the field researcher lacks in control over his subjects must
promising, but the approach taken by Ofshe and Ofshe may well prove to be more
be obtained as much as possible via description. Knowledge about characteristics of
fruitful at the very tasks that Emshoff applied himself to.
the tasks facing real groups, and the goals and perceptions of the individual members,
The Ofshes are able to arrive at their model by further simplifying already
would facilitate our thinking about the process and outcomes of group decision.
simple game situations. Instead of setting three people to the coalition game, they
A second promising idea for the study of group decision processes is realized
began with one person facing two players whose moves were controlled by computer.
in Barber's (1966) study of real governmental groups (town boards of finance)
In effect, they first studied individual behavior in a game setting; only later did
brought into the laboratory. Barber assigned all the boards the same tasks, thereby
they study freely interacting human players (Ofshe and Ofshe, 1970b). Instead of
holding constant at least the objective task environment (though different boards and
leaving the allocation of rewards between the members of a coalition as part of the
their individual members approached these tasks with varying goals, perceptions,
game, they stipulated the reward for each member, thus removing another source of
and methods, some of which Barber describes). Many of Barber's findings depended
variability from play. Control is of course the obvious advantage of the experimental
directly on his ability to exercise control over aspects of each group's performance.
situation. With careful step-by-step introduction of complexities into the model and
Gathering better information about the nature of the problems that groups
the experiment, it should be ~ossibleto elaborate upon both, obtaining in this way
face, and observing governmental groups in both natural and laboratory settings,
models that explain more intricate and varied behavior without sacrificing accuracy or
should enhance our ability to develop more clearly defined explanations of the process
theoretical clarity and simplicity.
of group decision. (See also Weick, 1965; Scott, 1965.)
Perhaps it will still be necessary eventually to adopt a simulation approach
Large Groups. As we pass from the comparative simplicities of small groups,
to individual behavior in these game situations. However, in contrast to the problems
the researcher's problem is magnified many times. In practice, he has to write his
for which Newell, Shaw, and Simon built their models of problem soIving, the coali-
description of behavior at a level that ignores most of the details of individual deci-
tion and Prisoner's Dilemma games are deliberately constructed to permit mathe-
sion and interaction. He may decide to turn to a mixed man-machine simulation or
matical description. Hopefully, behavior in them will be more readily modeled in the
to a game played entirely by human players, in order to generate data on which
to build theory. The mixed and all-human simulations of organizations and systems
404 Rufus P. Browning Simulation: Attempts and Possibilities 405

have been used as rich experimental settings in which to trace the effects of personal- decisions in firms (Cyert and March, 1963), budgetary allodations in governments
ity characteristics on performance and outcomes. The all-computer simulations, in (Crecine, 1967, 1969; Gerwin, 1969), and decisions of national leaders in an inter-
contrast, tend to squeeze out personality factors as these are commonly understood, national system (Bremer, 1971). Particular decision rules specify what information
or to avoid identifying observed behavior with particular individuals even though about the organ'zation and its performance is attended to, how this information
that is where it must originate. Nevertheless, some of these simulations are interesting impinges on goals, and what alternatives are considered. Commonly, it is not clear
to the student of political psychology because of the suggestive use they make of whether the decision rules stated for an organizational subunit inhere in one person
psychological theory to model organizational behavior. or in a group. Nevertheless, some are certainly associated with individuals; in any
Cherryholmes and Shapiro ( 1969) simulated on the computer roll-call voting case, decision rules of the kind described could be used to characterize individual
in one session of the House of Representatives. A review of the literature suggested decision making.
that much voting could be explained as a function of factors impinging on a legisla- What kinds of decision processes are described in these models? What kinds
tor's self-interest-for instance, constituency characteristics-or predicted from his of political behavior can be brought into focus with the concepts and methods put
party membership. Most of the predictive work in the model is done by the predisposi- to use in them? Several of the simulations do describe behavior and to some extent
tion phase of the model, using these sorts of predictors. Ideology was introduced I
the thinking of officeholders such as the urban school superintendent and his staff
as a conceptual variable but measured simply by the member's past voting record. 1
in Gerwin's study, and the municipal government departments, city council, and
Cherryholmes and Shapiro predicted that members whose self-interest and alliances mayor in Crecine's work. Both studies describe largely routine decision rules for
created strong cross pressures on a particular bill would develop only weak predisposi- adjusting budgetary allocations among departments, functions, and accounts. The
tions for or against the bill. The simulation proposes that these members undertake limitation to routine behavior and the emphasis on common adaptations to the
additional communication about the bill. Via a probabilistic process, they are exposed
to the stands of other members and of the president, and these may change the
1 organizational environment make the Crecine and Gerwin simulations more obviously
relevant for the student of government budgeting than for political psychologists
voting predispositions of the cross-pressured members. The communication phase generally. One could describe and simulate a variety of modes of adaptation and
improves slightly the already good predictions of members' votes. The cross-pressure decision in organizations, with an emphasis on individual differences. The Crecine
mechanism and communication phase are consonant, in general terms, with the and Gerwin studies establish base points for such an effort, with respect to budgeting.
available knowledge about these phenomena as well as about the Congress. Cherry- Cyert and March's work on decision processes in firms and Bremer's simula-
holmes and Shapiro's success at predicting 80-90 percent of individual votes testifies tion of national decisicn making (with many formulations patterned after Cyert
to the strong and clear self-interest of the legislators and to the public availability of and March's) operationalize concepts of attention, learning, level of aspiration,
information about the institutions and other factors to which that self-interest is tied. aspiration change, hostility, and reciprocity that are more familiar to psychologists.
Congressmen's votes, at least, appear to be very largely determined by the interaction By way of illustration, one of the many models in Bremer's simulation provides that
of widely shared goals of reelection and influence with identifiable characteristics of the rate of verbal or diplomatic hostility emitted by one nation toward another is
the member's environment. a function of the rate of hostility anticipated from the other nation, multiplied by a
One further aspect of Cherryholmes and Shapiro's method deserves mention. reactivity parameter ( a psychological characteristic?). Three characteristics of the
They might have estimated statistically the weights for the various attributes of a dyadic relation may modify the expression of hostility predicted by that simple
member vis-Q-vis a bill in the predisposition phase of the model; instead, they chose equation: power differences (weaker nations suppress hostility toward stronger ones) ;
to assign undoubtedly cruder weights on the basis only of the evidence aiready alliance relationships (nations suppress normal reactions to hostility when the other
available in the research literature on the importance of the attributes. The model nation is an ally) ; and economic dependence (dependent nations suppress hostility).
with the crude, preset weights was a stronger test of the existing level of knowledge A particular nation might weight these constraints on its hostility differently than
about roll-call voting and represents a stronger theory than a statistical model, other nations (though no effort has been made yet to determine weights for actual
because the model as it stood took no information from the behavior (votes) it was nations).
intended to reproduce. A model that has to be given information about the behavior Bremer posits that national officeholders seek to make their positions secure.
it is to replicate is less powerful in an important sense than one that makes pre- The variable that operationalizes leaders' goals for security of office is an aspiration
dictions without any knowledge of the predicted behavior. (See Hanna, 1971, for level for probability of continuing to hold office. The security-of-office aspiration
a formal argument to this point and a measure of the information value of a model.) enters into decisions such as the allocation of national resources to consumer goods
Following the test of the model, it is also appropriate now to estimate statistically the rather than military forces or basic investment. Aspiration level is a function partly
parameters for the hypothesized variables, to see whether a simpler model is just of aspiration level in the previous time period, partly of the level of security cur-
as efficient and to assess the stability of the parameters over several sessions of the rently experienced, and partly of the security of officeholders in other, similar nations
Congress. ( a referende group). The possible different weightings for these three factors might
Computer simulations of organizational behavior have attempted with models represent cultural or psychological patterns. The parameter that transmits the effect
of decision rules to generate the time paths of largely quantitative decision variables of current experience is set so as to raise aspiration levels more rapidly when experi-
adjusted repeatedly by the organization under study, such as pricing and production ence is favorable than they are lowered when experiencing is discouraging.
406 Rufus P. Browning Simulation: Attempts and Possibilities 407

The organizational decision models all posit that the decision makers re- particular the international system. A simulation that includes human actors is one
evaluate and possibly change their decisions each time period, which may be a week way of doing this. If we attempt a computer simulation or mathematical model of
or a month for business units or a year or more for governmental budgeting and extremely complex behavior about which we know little, we will have to invent many
international relations. The models recycle each time period through the same struc- processes to obtain the necessary closed model. I t does not take very many inventions
ture of decisions, but under changing conditions as the decision environment is al- based neither on theory nor on data to render a model literally incredible. But if
tered. The timing and regularity of recycling may be a feature of the process under enough is known about the behavior of a system so that part of it can be defined at
study, as it is for governmental budgeting, or may be partly a construction in the least tentatively, then people can be set to work carrying out the actions posited for
model to impose some theoretical order on a less well-known or less well-structured the rest of it. This is the notion embodied in the Inter-Nation Simulation (INS) and
process, like the cycle of national decisions in Bremer's simulation of an international its many versions and descendants.
system. The models assume some stability of decision rules over time, or simple In INS, people play the roles of major leaders of nations, deciding how to act
mechanisms for altering decision rules from one period to the next. toward other nations and how to allocate national resources, such as whether to
The decision-rule formulations assume either a largely quantitative task en- spend more for consumer goods or for arms. The programmed portion of the simula-
vironment (appropriations, force levels, prices; or, by extension, votes, organizational tion, whether calculated by hand or by computer, simulates basic economic and
size) or a level of aggregation at which the researcher is willing to describe the political structures and processes that shape the behavior of the human players in
relation between decisional situations and choices in mathematical terms, as in por- their decision-making roles. For instance, one function determines the level of con-
tions of Bremer's model. If the researcher is able to quantify features of the decision sumer satisfaction that results from allocation of resources to consumer and other
maker's environment as rates or probabilities (whether the decision maker thinks of purposes; the level of consumer satisfaction subsequently affects the probability that
them this way or not), and if detailed knowledge of his internal decision process is the current set of leaders will continue to hold office. The functions of the pro-
unimportant or too difficult to obtain, then the simulations of organizational decision grammed modules of the simulation and the actions of the human players feed into
offer many useful models. Extension to other phenomena depends on finding quanti- each other. The model builder is not faced with the impossible task of formulating
tative decision processes or on researchers' ingenuity at quantification and ability to mechanisms that could generate all the behavior of the simulated nations, but he
develop the large data bases necesriary to model tfie decision maker's task environ- can use various settings he creates as motivating structures to observe the behavior
ment. of the surrogate human actors.
T o the extent that the organizational decision models do not portray simply A wide range of research efforts may be indicated by the term simulation
routine performances, they describe processes that are strongly goal-oriented and applied to all-human or man-machine models. On the one hand, simulation means
intendedly rational. Decision makers are seen as working at their tasks, motivated to some researchers an experiment in an enriched setting. This meaning is common
to achieve certain rather clearly defined goals by applying a limited number of in references to experimental studies of organizational phenomena, in which the
familiar methods. Cognitive processes come to the fore. They may have their origins experimenter creates a realistic task for a laboratory group and establishes levels of
in the interaction of past experience and the personalities of individuaIs, but the authority in it. Experienced actors as well as the usual student subjects may be
origins are not examined or de'scribed in the models. Presumably it would be possible brought into the laboratory (Barber, 1966), with special advantages and problems.
to study a number of decision makers at the same task, specify the differences in Zelditch and Evan (1962) discuss the methodology of simulated bureaucracies.
their decision processes, and proceed from there to investigate where the differences Typically, the tasks and structures invented by researchers have not been widely
came from. It should be possible also to develop different models of the same decision adopted; hence, there has been little continuity in development of theory from a
task from study of human decision makers and to assess their performance under succession of studies using essentially the same setting. Nevertheless, such enriched
various hypothetical conditions. Thus, one could evaluate the rationality of the range experiments may be efficient generators of information when control over variables
of decision rules used. An example of this approach in economics is a study by is needed and not readily attainable in field research. Coleman's Iegislative game
Baumol and Quandt (1964) that investigates the consequences of a number of simple ( 1964a, 1966b) provides a setting whose characteristics deserve fuller exploration than
quantitative rules of thumb for pricing decisions; their performance was found they have received.
surprisingly good under the limited conditions assumed. The context is distant from Both argument and evidence make a persuasive case for enrichment of experi-
political psychology, but the method is of interest. mental settings, if inferences beyond the boundaries of the laboratory are hoped for.
I n a neat demonstration of the effects of even minimal enrichment, Pilisuk and
Man and Man-Machine Simulation of Systems Rapoport (1964) gave two sets of subjects an identical Prisoner's Dilemma game, one
in abstract form, the other with the same choices labeled as degrees of disarmament.
I t is no coincidence that simulations dispensing with the computer in part or Subjects were more cooperative in the disarmament version. Where a concern is to
entirely are found associated with especially complex phenomena about which our make inferences to a real system, an appropriately enriched setting has greater
ignorance is especially great. In spite of problems of complexity and of access for "proximal similarity" to the reference system (Raser, Campbell, and Chadwick,
data collection, we still need to study the behavior of systems of organizations, in 1970; Abelson, 1968bJp. 279). This does not mean that all details of the setting need
 408 Rufus P. Browning I Simulation: Attempts and Possibilities, 409
be "realistic," but it certainly argues for the inclusion of variables thought to interact
generate conflict or diversity; inability to view a situation from another perscn's point
with other variables in the reference system. The Inter-Nation Simulation and
of view; inability to generate alternative perceptions and outcomes; and a tendency
others in the same vein induce substantial information overload, uncertainty, com-
to seek structure, avoid delay, and close thinking quickly.
plexity, and anxiety, all factors thought to affect behavior in the international system.
Driver's findings on the relation between cognitive structural complexity and
A setting that lacked these features and was not labeled like real national decisions
aggression and irrationality in INS are intriguing. Groups made up of people with
and international relations would be that much more distant in concept and behavior
structurally simple cognition performed major aggressive acts (war and unprovoked
from real international systems. Inferences from laboratory to referent would be less
arms buildup) three times more frequently than groups of cognitively complex per-
well founded.
sons (Driver, 1965, pp. 30-34). In the complex groups, aggressive behavior was
INS and its offspring are more than enriched experimental settings, however. I
instrumental to the achievement of other goals. In the "simple" groups, however,
They represent efforts to operationalize and connect "islands of theory" about diverse
aggressive behavior tended to be related to attitudes toward violence rather than to
national and international processes. Human actors are used to generate decisions
goals prominent in the task environment of INS. If members of the group tended
within the simulated nations not because the aim is to study individual behavior but
to be favorably disposed toward violence, then violence occurred irrationally and
because the difficulties in programming national decision functions are too great
even at objectively low levels of stress. Violence as a response to frustration was much
(Guetzkow, 1968, pp. 692-699). INS was created primarily to build the theory of
more likely for the simple-thinking groups than for those with more complex cogni-
international relations rather than to investigate the behavior of individuals in the
tive structures and processes (Driver, 1962; summarized by Schroder, Driver, and
simulated positions of national decision makers. There is some overlap between these
Streufert, 1967, p. 142). The line of research has been continued with a somewhat
uses, to be sure, since the behavior of the individual actors is a prime determinant of
simpler simulation, the Tactical and Negotiations Game (Streufert and others, 1965;
the behavior of the system.
Simulations of international processes are well reviewed and referenced by 1 Streufert, Suedfeld, and Driver, 1965; Streufert and Streufert, 1969; Streufert,
Streufert, and Castore, 1969). The findings constitute a strong argument for getting
Smoker (1972). One important line of development has spawned a family of mixed
the house of cognitive processes in order and applying our knowledge of them to the
man-machine simulations, devised for particular uses or to improve upon the theory
embedded in INS, including Smoker's own International Processes Simulation. I
study of political behavior.
Selecting Aclors by Personality Characteristics. Hermann and Hermann
Another line seeks to develop explicit models of the behavior of the human actors
(1967) conducted a pilot study, using INS, in which the assignment of players to
and to test the models against the data from INS runs and against data from the real
roles in the simulation was based on the extent to which their personalities matched
international system, as in Bremer's (1971) computer simulation of INS. For present I
available knowledge about the personalities of particular historic actors-specifically,
purposes, attention focuses on the use of INS and related simulations either to study
national leaders of European nations on the eve of World War I. One of the authors,
individual and small-group behavior or to study system behavior, with special sensi-
I a psychologist, studied materials on the leaders and developed a distinctive profile
tivity to the effects of individual differences. Two applications are discussed: (a) the
study of individual cognitive processes and capabilities during decision making under for each one. Categories of dominance, self-acceptance, and self-control appeared to
differentiate among all ten of the leaders selected for matching. These characteristics
conditions of stress, complexity, and information overload; and (b) the study of
and others identified in the analysis were measured in a group of peopie (high school
system behavior when actors are selected according to their personality characteristics.
(For other studies of personality and cultural factors under simulated conditions, students) by the California Psychological Inventory (Gough, 1956). Attitudes
see M. G. Hermann, 1966; Druckman, 1968; Terhune and Firestone, 1967; C. F. toward such concepts as suspicion, frankness, making decisions, peace, success, and
Hermann, 1969; Powell, 1969. See also Chapter Ten in this volume; some of the war-also found to characterize one or another leader-were measured among the
assumptive frameworks discussed there could fruitfully be studied via simulations.) I students by the Semantic Differential (Osgood, Suci, and Tannenbaum, 1957).
The players selected as matching most closely the personalities of the real
Cognitive Phenomena. Research into the information-processing abilities I

and problems of the human actors in simulations is of special interest because it has leaders were placed in two runs of the INS, modified to simulate several days in
been vigorously pursued and because it touches again upon the viewpoint expressed
I
July 1914--the few days just before war broke out. The history of developing con-
I

throughout this chapter about our need to understand cognitive processes in political flict and events from the period preceding the simulated days were given to the
players, in order to approximate in the simulation the conditions that shaped behavior
behavior.
Using the Inter-Nation Simulation to study the behavior of individuals under in the reference system. Names of countries were fictitious, and particular events
the cognitively challenging decision conditions of that game, Driver ( 1962, 1965) were altered in their details (but not in their meaning) in order to disguise the
divided players into nation-teams consisting of individuals with simple and with correspondence between simulate and reality. (The aim was to structure the players'
1 behavior by the structure of the reference system, not by their memory of the actual
complex cognitive structure and process. Perhaps the clearest way to define these
labels is to list the referents used in the classification procedure (Schroder, Driver, I outcome; nevertheless, some players did perceive a similarity between the simulation
and Streufert, 1967). The responses of cognitively simple individuals are charac- I and the 1914 setting.)
One of the two runs carried out included players closely matched to the
terized by overgeneralization; hierarchical organization; absoluteness; inability to
real leaders; the other, players less closely matched. Both in detail and in major out-
410 Rufus P. Browning Simulation: Attempts and Possibilities 41 1

comes, the run with matched players resembled the reference events much more With this perspective the model world becomes an attempt to demonstrate or show
closely than did the other run. The matched run was on the verge of war at the or reveal the way parts of reality could or should be, and differences between the
close of the simulation; the other run was preoccupied with plans for a conference two worlds are rectified by changing aspects of reality through social and political
called by one of the leaders. Numerous messages and conferences in the matched run action."
resembled closely the events of the reference period. Perhaps the very term validation implies an inappropriate finality to the
One cannot draw firm conclusions from two INS runs with high school stu- confrontation of data with model. A theoretical effort may stimulate our thinking
dents. Still, the approach was bold, the method interesting, and the result suggestive. or lead to the design of better research. I t is reasonable to expect contributions of
If replicable, the findings could validate, with respect to the pre-1914 reference this sort rather than some kind of final validity (Raser, Campbell, and Chadwick,
system, both the scenario for the simulation and, most significantly, the selection and 1970). Furthermore, Eecause many simulations are very complex, it is commonly
rating of personality traits and the matching of players to real leaders. The procedures more sensible to ascertain what parts of a simulation are invalid and to change them
for determining the personality characteristics of the leaders are of great interest in rather than to declare the whole model invalid.
themselves and deserve a more complete and public account than they have received. Complexity, Sustained Development, and Cost. More than one simulation
We are learning to assess the personalities of leaders (George and George, 1956; has been so complex as to exceed the capacity of its inventors to understand it,
Barber, 1972a), but the method applied by the Hermanns appears unique and leading to "Bonini's paradox" (Dutton and Starbuck, 1971, p. 4; referring to Bonini,
worth further development. But then to have, in addition, a simulation method for 1963) : "Because the assumptions incorporated in the model are complex and their
validating the personality analysis opens up a range of new possibilities for students mutual interdependencies are obscure, the simulation program is no easier to under-
of political psychology. stand than the real process was." Especially the computer modeler should ask: How
I
complex must a model be to portray the behavior? I t is a mistake to include in a
Caveats model every detail one can observe in the real process.
Complex models of intrinsically interesting processes demand long, sustained
The reader will perceive the dominant tone of this chapter as optimistic- development. The Inter-Nation Simulation and the Newell-Shaw-Simon work on
not for simulation generally or for all uses of it, but for some possible directions of problem solving fall into this category. Both have been developed by a large number
research. It would be inappropriate to leave the matter at that, without pointing of researchers over more than a decade; both will require more decades of revision
more clearly to major problems and requirements, and acknowledging the critics l
and extension. Both projects have commanded significant resources and strong
(Singer, 1965; Boudon, 1965; Zald and Schliewen, 1968; Rosenthal, 1965, 1968; institutional support, and these were essential.
Howrey and Kelejian, 1969; Powell, 1969; Shaw, 1971; Padioleau, 1969) . Simulation tends to be a slow and expensive way of doing research-in part,
Validation. Most models and theories are intended to refer to some real and justifiably, because of the complexity of the problems under study. In addition,
phenomena, and sooner rather than later. If this is the case, the theorist should have however, the technology is very demanding and time-consuming, especially on the
in mind at the outset a procedure for validating his model, for comparing it with computer side of the method. Many a sad tale could be told about great aspirations
the reference system. Validation of simulations is still something of a new issue, but that ended in stacks of useless computer output, obtained after anticipated weeks
a number of good statements are available (Abelson, 1968b; Naylor and others, of work stretched into months, even years, of frustration with flighty student pro-
1966; Naylor, 1969; C. F. Hermann, 1967; Dutton and Starbuck, 1971; MacRae grammers, opaque computer systems, and frequent system changes.
and Smoker, 1967) . I Publication, Communication, and Parsimony. I t is imprudent to promise a
I
Not every model need refer directly to an existing real system. A simulation simulation before it has produced intelligible results, because it may never do so, in
I
may involve a more limited examination of the implications of concepts and hypo- spite of good intentions and earnest effort. Many researchers, myself ruefully among
theses: How do these ideas behave when put together? Day and Tinney (1968), for I them, have so strongly desired the final product that they came to expect it. Un-
I
instance, examined through computer simulation the consequences of fragmented fortunately, simulations have often turned out to be more problematic than they
decision making, primitive learning, and satisficing (versus optimizing) decision rules seemed to be at an earlier stage of development.
in a hypothetical firm; their model is clearly not rich enough to represent a real A different problem arises also from the overly strong wish for a running
firm, and the authors do not intend to validate the model. Nevertheless, a useful In several simulations reviewed in this chapter, interesting empirical and
purpose is served by their demonstration. This does not mean we should avoid theoretical work has gone into the construction of a model but has not been fully
confrontation with data, but we need not undertake the confrontation every time we I described in print. Perhaps researchers should take their work and publications step
examine an idea. 1 by step, not concentrating so singlemindedly on the simulation that may eventually
A pointed use of simulation that does not fit readily into a conventional result. A gradual buildup of theory, data, and publications would also facilitate
validation scheme is to suggest and evaluate alternatives to the real system. Smoker communication about the simulation, typically a difficult problem.
(1969, p. 11) argues: "It is possible to take the complementary position and to I Frijda (1967) rightly criticizes computer simulators for failure to communi-
evaluate the 'real world' relative to the 'model world' incorporated in a simulation. cate their work adequately. Simulations are difficult to describe, and model builders
 Rufus P. Browning
- have frequently neglected to present them unambiguously. If a complex simulation
is worth doing, it must be worth describing in full detail and with careful attention
to the requirements for adequate communication that Frijda sets forth. Moore
(1968) suggests a standard format for the description of decision processes; Newel1
and Simon (1972) present a notation for information processes.
The detail incorporated in many simulations not only hinders communication
but appears on the surface to place them at a severe disadvantage with respect to
parsimony. A full exposition is not possible here (see Gregg and Simon, 1967, for a
suggestive discussion), but some points may be set forth for consideration:
1. Simulations often make much stronger statements about the processes under
study than do competing mathematical or verbal theories-more precise, more de-
tailed, more extensive. The fact that two models are about the same process does not
mean that they attempt equally ambitious explanations.
"THE NEW FRONTIER
2. The hundreds or thousands of statements in a computer program (or in
a scenario for an all-man or man-machine simulation) should not be equated with
a small number of verbal or mathematical statements in a competing model. Much
OF PROJECTIVE
of the detail in a simulation program or scenario just copes with the technology.
Gregg and Simon (1967) argue that mathematical models sometimes achieve an
illusory parsimony by leaving implicit important features of a process that are made
explicit in a simulation.
3. A theory is validated mainly by surviving tests of its predictions-efforts
to disconfirm it and replace it with a competing explanation (Popper, 1959). If the
theory is so ambiguous or general that no clear test of its predictions can possibly be Jeanne N. Knutson
developed, then its validity cannot be determined by scientific investigation-a state
of affairs that typically generates more heat than light. A theory whose statements are
precise, detailed, unambiguous, and explicit is therefore much to be preferred. While
simulation models do not necessarily represent good theory, they are likely to be
strong in the important ways noted. Some loss of parsimony may be a price well
worth paying. Nevertheless, as suggested earlier in this chapter, mathematical methods
are sometimes more suitable, especially where the environment of the modeled
behavior is quantified and simple, as it may be in an appropriately controlled
experiment.
Clearly, we stand at an early point in the use of simuIation to develop knowl- T h e area known today as political psychol-
edge about behavior generally and political-psychological behavior in particular. We ogy has been shaped by the research instruments used and by the research interests
see some achievements and more possibilities, and a falling away of na'ive exuberance to which these instruments have been directed. To date, the political psychologist's
for the method. Abelson's (1968b, p. 346) closing statement is still apt: "The poten- principal tools have been scales employed in survey research, the focused interview,
tialities of the technique are indeed exciting, but it will take much more time to the psychobiography, and-to a much lesser extent--content analysis. But many
realize them than had been anticipated. Miracles are not to be expected; progress is." other techniques developed for the study of human behavior-participant observa-
tion, the laboratory and natural experiment, computer and human simulation, and
a variety of projective techniques--offer possibilities for increasing our understanding
of the intrapsychic basis of politics. Since research instruments shape the content as
well as the form of the data gathered (Knutson, 1972b), it has become increasingly
For their thoughtful critiques of the draft version of this chapter, I would like to
express public appreciation to Leopold Bellak, Fred I. Greenstein, William E. Henry, and
Paul H. Mussen.
414 Jeanne N. Knutson The New Frontier of Projective Techniques

important to widen the armamentarium of political psychology in order to check commonality behind the various tests of association and separately coined the terms
both our favorite hypotheses and our "established facts" against different types of projective techniques and projective methods to cover this form of research (with
data. The intellectual security which such multimethod research provides has been the historical priority belonging to Murray). As Frank's seminal monograph stated,
clearly recognized (Campbell and Fiske, 1959). "The dynamic conception of personality as a process of organizing experience and
Additionally, total reliance on personality scales for our knowledge of the structuralizing life space in a field leads to the problem of how we can reveal the
intrapsychic dimensions of political behavior is analogous to being limited to the way an individual personality organizes experience, in order to disclose or at least
stated preferences of a sample of voters in Bend, Oregon, in predicting the outcome gain insight into that individual's private world of meanings, significances, patterns,
of a national election. The employment of a variety of research methods increases and feelings" (p. 402).
both the range of human subjects on which inquiry can be feasibly focused and the Various other early studies uncovered, in perception and memory, systematic
type of information which may be gathered. Further, methods vary in their "discovery biases that could be related to personality factors. For example, the selective-percep-
power," with this power differential being a function of both the properties of the tion phenomenon was clearly illustrated in an early study of the development of
method and the use to which it is put. An experienced interviewer, for instance, can social attitudes in a southern community (Horowitz and Horowitz, 1938). Another
uncover sensitive areas unavailable to questionnaire research. At the same time, the study of that period (Seeleman, 1940-1911) found correlations ranging from .64 to
breadth of data provided by easily administered forced-choice survey questions can- .71 between a direct test of social attitudes and a recall of Negro and white photo-
not be approached by the more personalized depth methods. graphs. In addition, Edwards (1941) published evidence supporting the hypothesis
This chapter explores the utility of employing projective techniques in political that individuals tend to forget information which conflicts with existing frames of
research as an adjunct to the limited methods now in use. Since such use has been reference and to remember information which harmonizes. Thus, riot only through
rare, this subject is in many ways a new frontier. Like other frontiers, the explorer's, psychoanalysis but through the fields of perception and memory, psychologists were
investment is likely to be considerable and the rewards exciting but uncertain. Yet led to study the phenomenon of the selective ways in which an individual structures
the rich and varied literature on projective techniques can help us to assess the stimuli.
utility of expending energy in this direction by elucidating some basic issues involving The term projection has been able to cover this phenomenon because of
the meaning of "projective" in this sense, the import of the subject's response, and changes in the connotations of that word. According to Freudian theory, what occurs
the possible advantages which would outweigh the considerable investment of investi- in the mechanism of projection is that "internal perception is suppressed, and,
gator time. instead, its content, after undergoing a certain degree of distortion, enters conscious-
ness in the form of an external perception" (Lindzey, 1961, p, 30).
Historical Perspective
The essence of this conservative view is captured very well by Murray,
Undoubtedly, the precursor of the projective techniques available today is who suggests that an orthodox definition of projection implies ( a ) an
the technique of free association developed by Sigmund Freud. Through this method, actual misperception in which the individual believes something that is
it was early recognized that each individual has a "subjective or experiential side" manifestly false concerning another; (b) this misperception involves
attributing to the other person a tendency directed toward either the
(Bolgar, 1956), which is causally related to his overt behavior; furthermore, the best
perceiver or toward a third person; (c) the tendency is an important part
way to become privy to this inner world is to provide an unstructured situation of the perceiver's own personality; (d) the tendency is unacceptable to
which the person is free to structure in his idiosyncratic way. Although the free- the perceiver and he is unaware of its existence in his own makeup; (e)
association method remains the paradigm of such understanding, it requires con- the function of the process is to maintain self-esteem, or to escape from
siderable expertise and time; consequently, researchers have tried to develop other anxiety.
methods for tapping inner predispositions. Concurrently, the assumptions about
personality on which interpretations of responses are based have undergone con- However, projective techniques owe their name not to this psychoanalytic mechanism
siderable revision. of projection but rather to what Lindzey has called "generalized projectionm-
Since the late nineteenth century, psychologists have frequently employed namely, "a normal process whereby the individual's inner states or qualities influence
unstructured stimuli to test individual differences and to measure mental content. his perception and interpretation of the outer world."Thus, the term projection,
Word-association tests developed by Carl Jung, James Cattell, and Francis Galton outside of psychoanalytic discussions, is today assumed to include such intrapsychic
were soon followed by picture-association studies reported by H. L. Brittain, L. A. mechanisms as (1) broad processes of ascribing one's feelings and views to others,
Schwartz, D. J. Van Lennep, and A. Binet (Harrison, 1965; Lindzey, 1961). It was (2) ways in which personality needs distort perception, and (3) the mechanism of
not, however, until the advent of Hermann Rorschach's famous inkblot test in the classical projection rationalized by intellectual justification for the projection-a
1920s that, through the medium of psychoanalytic concepts, association to unstruc- phenomenon frequently employed in ethnocentrism (Murstein and Pryer, 1959; also
tured stimuli became connected to personality theory (Rabin, 1968). Shortly after- see Chapter Six above).
ward, L. K. Frank (1939) and Henry Murray (1938) independently discovered the Efforts to explore and measure the workings of this widened process termed
 Jeanne N. Knutson The New Frontier of Projective Techniques

projection led to the development of countless varieties of techniques. In addition fantasy mediated by ego processes. In Freudian terms, this fantasy is secondary-
to the Rorschach inkblots and such picture-association methods as Murray's Thema- rather than primary-process material. The growing understanding that the projective
tic Apperception Test, projective techniques include various devices involving mini- response is "mediated fantasy" is instructive because it epitomizes an important shift
mal cueing, such as auditory stimuli to which individuals give their interpretation, in meaning. From its earlier association with dreams, drug states, and other hyp-
incompleted sentences for which subjects supply the ending, story-completion proce- nogogic phenomena, fantasy has now come to refer to structured as well as unstruc-
dures, word-association tests, and drawing and drawing-completion tests. Shneidman tured imaginative processes. This broader usage reflects a differing view of the mean-
(1965, p. 500) summarizes the commonality of these techniques as "psychological ing of the projective response; hence, the continued application of "fantasy" to
measurement devices which are characterized (1) on the stimulus side by ambiguity projective techniques represents another case of new wine in old bottles.
in the stimulus, (2) on the response side by the multiplicity of responses permitted the The usefulness of projective responses is not seen as being reduced because they
subject in an open situation where the responses do not have a right or wrong are now understood to represent an individual's inner life seen in terms of his per-
character, and (3) in the interpretation aspect by the interest of the interpreter in sonalized coping activities and his own view of likely behavioral outcomes. On the
the unconscious or latent aspects of personality and in their amenability to holistic contrary, students of behavior have increasingly recognized such mediated psychic
personality anaiysis." needs as essential to understanding and predicting overt behavior. Indeed, "the
One might ask why, in sum, if projection is a basic and inescapable feature of manner in which a person copes with his-problems is the most revealing thing about
each person's dynamics, it is necessary to bother developing pictures, auditory stimuli, him. . . . Insofar as one can single out a particular flaw in current views of ego
elaborate stages with figures, word lists, and other devices. Why, in other words, defense, it is that writers on the subject have failed to mention the tremendous
does the researcher not simply ask each subject to tell five stories or give ten random importance of constructive strategies as a means of avoiding the vicissitudes that
statements? Murray (1943, p. 2) suggests that devices such as his TAT cards have make crippling defenses necessary" (Smith, Bruner, and White, 1956, pp. 282-283).
three advantages over collecting segments of subjects' thoughts which are not pro- Thus, it is today generally recognized that what has been labeled "cognitive
duced in response to specific stimuli: "It has been found (1) that pictures are control" (Gardner and others, 1959) intervenes between and mediates the impact
effective in stirring the imagination; (2) that they serve to force the subject to deal, of need on response. "The duty of cognitive control is to relate the functioning of
in his own way, with certain classical human situations; and finally (3) that the (a) the gratification of drives, (b) the modulation and delay of drive consumma-
advantages of using standard stimuli are here, as in other tests, considerable." tion, and (c) the appraisal of reality and reconciliation of environmental forces and
internal tension to each other within the person" (Murstein, 1963, p. 68). The
Projective Responses earlier view of a direct relationship between inner needs and projective responses
(Adorno and others, 1950) has generally been replaced by the understanding that
As stated previously, the intellectual tradition by which projective techniques the relationship between need and response is influenced by a number of other
were wedded to personality theory was psychoanalytic. Thus, when the intrapsychic intrapsychic factors. Hence, whereas early work (Sanford, 1936) demonstrated that
needs interpreted from Rorschach and TAT responses did not correspond to the intrapsychic needs influence a person's way of structuring a situation (both what he
respondent's observed behavior, a variety of p;ychoanalytic concepts-including perceives and what he ignores), other research began to illustrate that proje-~ b ' v e
re-
reaction formation, denial, repression, sublimation, and displacement-were em- sponse is influenced by such factors as the degree of drive arousal and of anxiety
ployed to account for the discrepancy. Concurrent with the psychoanalytically in- arousal. Further, as the concept of perceptual defense was clearly demonstrated,
fluenced interpretation of responses to projective techniques went (and with experimental evidence also indicated "that it is the acceptable, although perhaps
researchers in this tradition still goes) an emphasis on content (story themes and deprived, needs that are most frequently expressed in projective material" (Eriksen,
outcomes) rather than on the formal or stylistic features of the responses (Murstein, 1954, p. 139).
1963, p. 62). The traditional position that motives are directly expressed in projective
With the growing influence of ego psychology in the 1950sJhowever, the view (particularly TAT) responses-perhaps most ably defended by Atkinson (1961) and
of what projective responses mean began to change. Psychologists began to believe McClelhnd and his associates ( 1953)-has been borh challenged and illuminated
that intrapsychic needs may provide the basic stimuli for a person's responses but that by Lazarus (1966) in his "substitutive" theory, which states that "motives appear in
the type and direction of response are mediated by the functional capacity of the fantasy when they are not expressed behaviorally." Lazarus's theory actually supports
person's ego structure (such as ability to rationalize, degree of awareness of social the thesis that motives are directly expressed in projective responses since he provides
press, capacity to organize). In spite of Murray's emphasis on the similarities between theoretical rationale for cases where direct expression of motives does not occur. Thus,
TAT stories and other types of fantasy (specifically daydreaming), present projective his theory is consistent with Atkinson's (1961) statement that it is necessary to specify
theory has come to emphasize the differences rather than the similarities between the conditions when motives are expressed and, further, when they are related to
these verbal products. overt behavior.
While the element of fantasy (unguided free association without manifest Lrnarus (1966, p. 485) further suggests that TAT stories can vary on a con-
continuity) is undoubtedly present in projective responses, it is recognized to be tinuum from playful fantasy to problem-solving behavior necessary to gratify an
 Jeanne N. Knutson .The New Frontier of Projective Techniques

aroused motive-and, indeed, usually reflect both factors. "Depending on the mix,
F have been content rigidly to use one or two techniques to which they have
the contents will either directly or substitutively reflect motivations, positive corre- .
oftentimes become cultishly cathected. . . The normative aspirations of
lations with relevant behavior occurring in the case of the former, negative in the projective techniques can also be seen as a clear case of trying to placate
latter." He suggests that this mix is influenced by both personality and situational the other tribe's totem.
factors (and is true in any situation in which psychologists use behavior to infer
motives) : "The instrumental (goal-oriented) forms undoubtedly will increase rela- Concurrent with the interest in (and debate over) the utility of quantifying
tively under conditions that emphasize logical, coherent, socially meaningful stories . projective responses is the realization that situational or extrapsychic factors must be
that are consistent with a minimally ambiguous stimulus." Thus, needs may be considered when one assesses the responses of subjects. In addition to interviewer cues
expressed or not and may be expressed through fantasy or goal-oriented responses, ,(as mentioned by Lazarus, above), various studies of stimulus cards show that the
depending on the degree of need arousal, opportunities for its satisfaction, and the ;card itself "is by far the most important determinant of the content of a TAT
situational parameters relating to fantasy production. ,response7' (Murstein, 1963, p. 195). Based on this understanding, a number of
With the increasing realization that there are various determinants of how , ,,researchers (Harrison, 1965; Sigel and Hoffman, 1956; Feshbach, 1961) suggest that
intrapsychic needs influence projective responses and that the respondent may exer- -.projective devices which are custom-made to elicit the response area of interest are
cise a good deal of conscious control over his responses, there has come a correspond- .of value. As Feshbach (1961, p. 137) notes, however, the use of focused stimuli still
ing realignment of scoring techniques: "To get away from these easily censorable ,leaves the problems of inference to be solved: "It is by now evident that we should
responses, clinicians seem to be turning in increasing numbers to the use of formalistic - . not expect to find a simple, uniform relationship between 'covert' fantasy expression
variables which reflect how the story is told more than what the subject tells" (Mur- of a motive and 'overt' behavioral expressions. The proper question is not 'What is
stein, 1963, p. 364). In this vein, Riessman and Miller (1958) note that it is im- the relationship?' but rather 'Under what conditions would we expect to find a
portant to know how story length (as well as story theme) is affected by the subject's positive, inverse, or negligible correlation? "
social class. Indeed, problems of inference have been the focus of the most concerned
This distinction between content and structural variables has played an in- attention from projective theorists. Henry ( 1960, pp. 18-19) summarizes the issue:
creasingly important role in determining the meaning of projective responses. As
noted above, content variables refer to "the semantic meaning of the faqtasy response, We do assume that behavior bears some discoverable, if not direct, rela-
and, in practice, there is strong emphasis on attitudinal and motivational meaning" tion to underlying dynamics. And conversely, we do suppose that a partic-
(Kagan, 1961, p. 196) . Structural variables, however, deal with how the response ular pattern of inner dynamic relations has some greater tendency to be
is made rather than the meaning of the response and can refer to a variety of
.
related to one pattern of overt behavior than to another. . . The success
of predictions of overt behavior is, however, greatly limited by our very
processes. scanty knowledge of what psychological issues will manifest themselves in
which kind of behavior under what circumstances of correlative internal
First, a structural variable c.tn include the syntax of the subject's verbal dynamics and of external pressures and sanctions.
behavior and the ordering of language units in the verbal response . . .
the preferred use of specific language forms over others . . . the preferred For example, a study of lower- and middle-class boys (Mussen and Naylor, 1954)
perceptual organization of the stimulus . . . the coherence and logic of demonstrated that expression of aggression varies with both the degree of negative
the fantasy production . . . the tendency to "go beyond" what is given sanctions on aggression in the class group and the degree to which the person antici-
in the external stimulus. . . . One might also include the nonthematic
pates punishment for the expression of such aggression. Another study (Kagan,
behavior of the subject (e.g., his resistance or his subjective reaction to
the examiner) [Kagan, 1961, pp. 196-1 981. 1956) further illustrated that the expression of overt aggression varies in relation to
(1) the strength of the need, (2) the amount of anxiety accompanying the need, and
As a result of this emphasis on structural variables, a high degree of quantifi- (3) the kind of aggressive behavior; in general, the relationship between fantasy
cation now often replaces the earlier, predominantly qualitative, assessment methods. productions and overt aggression was closer when the fantasized aggressive behavior
This realignment has troubled a number of projective theorists. Atkinson (1961, was closely related to the overt criterion.
pp. 227-228), for example, suggests that examining isolated variables may obviate The levels of inference required and the resulting demands on the user of
the purpose for which projective measures were designed-namely, to study the whole projective techniques are well stated by Korner (1965, pp. 24-25) : "It is important
person, And Shneidman ( 1965, pp. 504-507) has incisively noted : that we are aware that our tests merely record behavior and that we can arrive at
clinical insights only through inference, which in turn requires a thorough familiarity
The worst indictment of projective techniques in America is not their lack with the principles of psychodynamics on the part of the interpreter." Korner goes
.
of rigor or validity-which . . can be viewed as an irrelevant demand on to observe that projective techniques are as yet of limited relevance because they
from the 'other point of view'-but rather that they themselves have not are slices of behavior and thus their utility rests on the inferences made about these
consistently pursued a global understanding of their human subjects but behavior segments. I t is here, however, that psychology, and particularly personality
420 Jeanne N. Knutson The New Frontier of Projective Techniques

theory, is presently lacking in knowledge (see Chapter Two above), and thus this carefully analyzed by Zimbardo (1964, p. 197), who criticizes direct measures of
deficit is felt equal!y by ai! other forms of behavioral measurement. As Korner notes, motivation employed after an experiment "because the subjects might not have been
successful prediction depends on the solution of two complex and perhaps unsolvable consciously aware of nor able to verbalize their true feelings, and the measure might
problems: evoke resistance in subjects for whom admitting to a state such as anxiety (or fear)
calls into question their masculinity." This viewpoint is supported by Campbell
The first one is the detection of all the innumerable variables that are at (1957a) and McClelland (1958) and, of course, is very meaningful in view of the
work in the process of an individual's reality adaptation to a need. I t is sensitive nature of much research in political psychology. Several interesting studies
probably not only the existence of these variables but also their interaction illustrate the value of employing projective measures of affect and motivation as a
which determines what form his adaptation will take. The other problem means of tapping areas in which anxiety is a factor. For example, Sarnoff and
is to,find the secret of ego synthesis, which probably consists of an organ- Zirnbardo ( 1961) used a projective postexperimental measure (subjects were shown
ismic process involving more than the sum of the variables at work and a photograph of a poker-faced person, who was supposedly about to participate in
which possibly is at the root of all the clinical discrepancies mentioned the same experiment, and asked to judge this person's reaction). They found a
before [p. 311.
significant correlation between subjects' projections of anxiety and the anxiety arousal
Thus, both projective behavior and reality behavior are valid indications of internal of experimental conditions to which they had been subjected, while a direct scale
predispositions, but between a person's needs and his behavior lie many intervening measure did not discriminate between conditions. In another well-known study, Clark
variables which need to be delimited. (1952) presented T A T cards to sexually stimulated subjects (1) in a classroom setting
I n sum, the meaning of projective responses has changed with the changing and (2) in a fraternity (where subjects were under the influence of alcohol). He
emphases in personality theory and in psychometrics. Projective responses are now found that the subjects used significantly more symbolic sexual imagery but less
seen as mediated and situationally influenced expressions of intrapsychic needs. manifest primary sexual imagery in their stories in the classroom, whereas in the
Although much research effort has been directed toward specifying what the mediat- fraternity their use of these categories was reversed.
ing and situational variables are, the problems of levels of inference and of multi- I n another interesting analysis of the relationship between projective and
determined causality have generally been unexplored. These serious issues, however, direct measures, Zimbardo and Formica (1963) found a correlation of .75 between
in many ways reflect the state of the art in psycholo~--or at least in personality direct and projective measures but discovered that subjects who had admitted auto-
measurement; thus, their existence does not provide an adequate reason to disallow biographically (prior to the study) to reserved natures in expressing emotions were
the utility of projective techniques in the study of political behavior. significantly more likely to express anxiety in projective measures than in the direct
measure. The authors suggest that the method of data gathering should be tailored
Projective Techniques and Other Methodologies to the expressive nature of the subjects, but they unfortunately provide no evidence
illustrating which measure was a better indicator of the actual anxiety felt by the
An area of particular interest to those using projective techniques in research subjects. It would Ee useful to replicate this study and to employ additionally a
is the relation between data gathered on projective measures and that acquired from physiological measure of anxiety.
more direct means. Of the number of studies focused on this area, the following are Other data suggest not only that different subjects may utilize projective
illustrative. devices in different ways but also that projective and- direct responses may have
Propper (1970)' replicating an earlier study by Davids and Pildner (1958), different meanings. One study (De Charms and others, 1955) measured the need for
~neasuredalienation by using a series of TAT cards and a sentence-completion test achievement in the standard way (through specially designed TAT cards) and in a
(the projective measures), as well as a self-rating scale and an affection questionnaire direct way (through answers to questions). The investigators found that two different
(the direct measures). Both studies found all intercorrelations to be significant at dimensions were being measured (which correlated +.23). Subjects who were con-
the .O1 level, "suggesting that Davids' objective and projective instruments are index- sciously achieving on the direct measure tended to be conformists, with a high valua-
ing the same dimensions of the alienation syndrome" (Propper, 1970, p. 43). tion of authority, a low valuation of unsuccessful people, and high scores on the F
In another study (Proshansky, 19431, ambiguous pictures of labor situations scale. Indirect achievers, on the other hand, tended to have internalized standards of
were intermingled with more typical TAT scenes as the projective measures. The excellence and to perform in a superior way on various tasks. It is not clear why
ratings from the subjects' descriptions correlated .77 and (in a second sample) .67 the investigators were surprised that two different dimensions appeared, since there
with a direct verbal scale of attitudes toward labor. Such correlations appear un- is a predictable psychodynamic difference between assertion and action in this area.
usually high. As Eriksen (1954, p. 440) notes in his summary of the literature: "In That is, it is one thing to state categorically that one conforms well to a major social
some experiments correlations of about .6 have been reported, [but] more typically value; it is 'quite a different thing to use one's abilities creatively in goal-oriented
the correlations have run about .4 or 3.'' behavior. I believe that this is an important point, because the often-found disparity
The area of the relationship between direct and projective scores has been between different measures purportedly assessing the same dimensions could fre-
422 Jeanne N. Knutson The New Frontier of Projective Techniques

quently be predicted from a careful analysis of the constraints inherent in the tions about actions that will take place under specified circumstances,
measures themselves (thus dispelling some unnecessary gloom surrounding the area and so forth [Greenstein and Tarrow, 1970, pp. 501-5021.
of psychometrics).
In a careful summary of indirect measures, Campbell (1950, pp. 30-31) notes I t appears to me that the above discussion raises an inconsequential, if not invalid,
that none of the many studies he reviewed illustrated that the indirect tests have distinction. Projective techniques, as Campbell (195713) has noted, vary in their
higher validity than the direct tests. However, the case for face validity of these degree of focus; what these investigators are describing are voluntary, free-response,
indirect measures is usually better than for direct measures. "In a number of the and either direct or indirect tests-and thus in the tradition of classical projective
disguised, structured tests, the distribution of scores and measures of internal con- techniques. Furthermore, projective techniques (like the so-called semiprojective
sistency demonstrates unequivocally that nonrandom, systematic errors, differences techniques) are also based on carefully adjusted ambi@ity; that is, although the
in perceptions, etc. exist, of which the respondents are presumably unaware. These stimulus rests on a large degree of psychological ambiguity, successful projective
systematic unconscious 'biases' are well worth study in their own right, and seem to lie techniques must present a fairly well-structured situation-or at least, as with TAT
close to the functional meaning of attitude." cards, the physical ambiguity must be minimal (Harrison, 1965). Finally, as we
In another thoughtful article, Campbell (195713, pp. 207-208) discusses the noted earlier, there are many different scoring interests, focusing on various levels
relation of projective techniques to other methods. He finds three main differences. and a wide range of psychic functioning and situationally determined behavior; the
The first distinction is between voluntary and objective tests. "In the voluntary test above-described scoring options fit well within traditional, nonclinical interests.
the respondent is given to understand that any answer is acceptable, and that there Projective techniques, then, can measure significant aspects of what direct
is no external criterion of correctness against which his answers will be evaluated." measures seek to capture, but most of the variance between these types of measures
The second distinction is between indirect and direct tests. "In the direct test, the is presently unaccounted for. (For a more detailed analysis of problems of validity,
respondent's understanding of the purpose of fie test and the psychologist's under- see Buros, 1965.) Perhaps the degree of missing variance can be significantly reduced
standing are in agreement." The third difference distinguishes free-response frok by a clearer understanding of the use which different personality types make of
structured tests. "The free-response format has the advantage of not suggesting different techniques of assessment. I t also appears necessary to define the target
answers or alternatives to the respondent, of not limiting the range of alternatives dimension more sharply, so that both the direct and indirect measures are similarly
available, nor of artificially expanding it through the suggestions provided in the focused and there is some intellectual confidence that they both measure the same
prepared alternatives." The classical projective measures are vo!untary, indirect, phenomenon. Finally, it should be underscored that projective techniques refer to
and free-response; projective techniques exist, however, for all permutations of these a wide variety of data-gathering devices which bear commonality because they are
dimensions except voluntary-direct-structured and two types that are inappropriate voluntary and usualIy indirect measures. Both a wide range of scoring focuses and
for personality measurement (objective-direct-free-response and objective-direct- methodologies and an equally vast array of structured ambiguity are at home within
structured). this commonality.
One final distinction should be made between projective techniques and other
methods. Two sets of investigators (Hanfmann and Getzels, 1955; Greenstein and Politically Relevant Uses
Tarrow, 1970) have raised a distinction between projective and what they call
"semiprojective" measures. These investigators use the term semiprojective to refer Projective techniques probably have proved most useful so far in the area of
to data-gathering techniques such as story-completion tests, "which are like fully ethnocentrism. Guggenheim (1969), for example, used projective techniques to focus
projective tests in that they involve fantasy elaboration on a stimulus, but which are on two previously separately documented relationships : ( 1) between being black
unlike them in two respects: and expressing lowered self-esteem and/or self-hatred, and ( 2 ) between cumulative,
chronic failure and an inability to realistically appraise task requirements, with
(1) The stimulus content is more structured and culturally patterned. It chronic failure presumably associated with being black in America. Connecting these
is ordinarily aspects of the stimulus, rather than the entire stimulus, that two issues is the view that level of aspiration is a function of self-esteem and that
are left ambiguous, and the ambiguities are carefully adjusted in order to self-esteem is a facet of self-concept. Guggenheim's study thus analyzed the interrela-
expose the respondent to specific issues with which the investigator is con- tionships of ethnic background, expectations for achievement, actual achievement,
cerned. Alternatively, the stimulus may be unambiguous but problematic, and self-esteem. His subjects (162 sixth-grade children from a Manhattan elementary
as in story completions that pose a dilemma in which a choice must be
school) were asked to "Draw a person like yourself" as a special version of Mach-
made between two equally desirable-or undesirable-options. In this
case, the ambiguity is in how to resolve the dilemma. (2) The resulting over's Draw-a-Person Test and were given a set of ten Semantic Differential Scales
data are interpreted at a surface (sociocultural) rather than a deep (psy- (Osgood, Suci, and Tannenbaum, 1957) related to the picture in order to assess
chodiagnostic) level. That is, the interpretive interest is in what orthodox self-esteem. This procedure was followed because "problems of social desirability and
projective testers ordinarily treat as chaff: values, cognitions, perceptual defensive responses which loom large when measuring self-esteem do not seem to be
sets, characteristic ways of perceiving typical social situations, expecta- adequately controlled in the phenomenological [self-rating] framework [usually em-
424 Jeanne N. Knutson The New Frontier of Projective Techniques 425

ployedl" (p. 69). On the basis of the semantic differential scales, 56 black and white in American society. (3) No Japanese people are assimilable in American society.
children were selected, differentiated on the basis of high or low self-esteem. This (4) Some Japanese people are not assimilable in American society. (5) No logical
sample was then given a specially designed achievement test which was administered conclusion can be dra.wn from the given statement." The control syllogism which
three times under different degrees of information regarding actual achievement. logically corresponded to the above stated: "If it is true that some pills are not poison
Guggenheim found that black children, as predicted, had significantly greater and if it is true that some poisons are not green, it is logical to conclude: (1) All
discrepancies between actual achievement and expectation-of achievement, but this pills are green. (2) Some pills are green. (3) No pills are green. (4) Some pills are
phenomenon was unrelated to degree of self-esteem. Information regarding actual not green. (5) No logical conclusion can be drawn from the given statement."
achievement reduced the discrepancy scores for the subjects as a whole. The amount While the direct measure did not offer this understanding, data from the
of discrepancy, however, was less important than the direction of the discrepancy syllogisms clearly indicated a pull toward tolerance (over the logical choice pattern
between actual and expected achievement: "Eighty-six percent of the Negro pupils, evidenced on the control syllogism). For example, whereas 85 percent of the sample
as compared to 58 percent of the white pupils, overestimated the number of problems (on the direct question) did not feel that "Japanese are inherently more cruel than
they thought they would get right." There was a similarly significant difference in white men" (and thus did not agree with popular propaganda statements of the
overestimation between high- and low-esteem children. with high self-esteem children times), the control syllogism choice was logically equivalent to the paired syllogism
also tending to predict higher scores than they achieved. Finally, while white children choice that "80 percent of the Japanese people are cruel"; on the emotion-laden
had significantly higher achievement scores than did black children, low-esteem white paired syllogism, however, the subjects tended to reject this ethnocentric statement
children had significantly higher expectation scores than did low-esteem black chil- in favor of "No logical conclusion can be drawn from the given statement." Thus
dren (and there was no significant difference between the high-esteem groups). This this use of projective techniques helped to separate the factors of socially desirability,
use of projective techniques thus suggests the surprising conclusion that while black popular responses, and intrapsychic proclivities.
children do exhibit generally lower school performance ( a phenomenon being at- To political psychologists, the best-known use of projective techniques in
tacked by federally funded programs to raise their self-esteem), black children do studying ethnocentrism is undoubtedly found in T h e Authoritarian Personality
not differ from white children in self-esteem. The performance-relevant personality (Adorno and others, 1950). (For a detailed discussion of this study and of the
difference may rather be defensive, overambitious, and unrealistic goal-setting be- development of its most visible contribution, the F scale, see Chapter Six in this
havior. volume. See Christie and Jahoda, 1954, for an analysis of validity problems in this
Several other studies fall in this area. Johnson (1949) successfully illustrated study.) Two further studies of this subject illustrate the utility of projective tech-
the development of Anglo-Spanish attitudes in the Southwest through specially niques in providing a clearer understanding of the psychodynamics of the prejudiced
designed TAT pictures. Seeleman (1940-1941), as mentioned previously, found high personality. In one study (Goldberg and Stern, 1952), a group of five hundred col-
correlations in two studies between a direct test of social attitudes and a recall of lege freshmen were given the E and F scales, with the highs and lows being selected
Negro and white photographs. Morgan (1945) investigated attitudes toward the for further study. These subjects were given several projective techniques, including
Japanese at the end of World War 11. By employing an unusual projective technique the Rorschach, the TAT, and a sentence-completion test. As in the original Berkeley
he attempted to counter the reduction in predictive value of public opinion polls study, it was found that ethnocentrics tended to repress or reject impulses, to hold
caused by "self-deceit" on issues which are highly emotion-laden. Using as subjects contradictory and distant parental images, and to show an absence of nurturant
170 students from psychology classes at Northwestern University, Morgan first capacities. In a second study (Sarnoff, 195I ) , one hundred Jewish male studects were
assessed their attitudes toward the Japanese by four items embedded in a fifty-item differentiated into highs and lows on an anti-Semitism scale from the Berkeley study
direct-attitude questionnaire. He then attempted to measure the veridical degree of and then given TAT cards and the Michigan Sentence Completion Test. Subjects
the questionnaire answers by giving respondents "two similar syllogisms, one con- high in ethnocentrism tended to have more negative and fewer positive attitudes
taining terms which have no personal value and the other involving personal con- toward parents and toward themselves, to be more frequently passive in the face of
victions or controversial issues; the differences in the patterns of response to the two interpersonal hostility, and to be less prone to retaliate against aggressors (that is,
syllogisms can be used as an index of the direction of the personal bias of the re- as noted in the previous study, to deny or repress impulses). Thus, data from these
spondents" (p. 27). (For evidence of the degree to which paired syllogisms can studies support the biographical and projective data gathered during the Berke!ey
measure personal bias, see his two earlier studies: Morgan and Morton, 1944, 1945.) research.
Morgan then employed the issues in the four direct-attitude questions in four The relationship between ethnocentrism and intolerance of ambiguity, a major
different syllogisms which were paired with four control syllogisms which used focus of the Berkeley authoritarianism work, received additional confirmation in a
emotionally indifferent terms. (These pairs were not given in juxtaposition.) For study by Block and Block (1951). These authors hypothesized that both ethno-
example, one direct question stated: "The Japanese are not assimilable in American centrism and intolerance of ambiguity are "subordinate manifestations" of a more
society." The corresponding syllogism stated: "If it is true that some Japanese people central personality dimension, which they labeled "ego control." (For a discussion
are not assimilable in American society, it is logical to conclude: (1) All Japanese of levels of intrapsychic analysis, see Chapter Two in this volume.) Using sixty-five
people are assimilable in American society. (2) Some Japanese people are assimilable males (homogeneous in social status and intelligence) from an elementary psychology
 The New Frontier of Projective Techniques 427
426 Jeanne N. Knutson

course, Block and Block first separated their subjects into overcontrollers, who "bind In general, this TAT response analysis gives an understanding of the psychodynamics
underlying the finding that-unlike the Dublin sample and the United States sample
their tensions excessively"; undercontrollers, who "do not bind their tensions suffi-
reported by Tapp and Kohlberg-the war-stressed Belfast sample does not evidence
ciently"; and appropriate controllers. This separation was made by judgments of
the age-related trend of moral development but generally remains at the lowest
stories told by the subjects to three Murray TAT cards.
level, suggesting that "the present conflict will effect a truncation in the development
Next, the experimenters made use of a different type of projective test: the
of moral judgment for the next generation of the Northern Irish."
well-known autokinetic phenomenon of the apparent movement of a pinpoint of
Turning from the dimensions of ethnocentrism, Dies (1968) reports a study
light focused on a wall of a darkened room. The subjects (in a hundred trials over
a forty-five-minute period) were asked to turnToff their light by button the instant dealing with the subject of efficacy, an area of continuing interest to political psy-
chologists. Specifically, Dies was concerned with Rotter's (1966) dimension of
they perceived it to move, and to write down the distance that it moved. Raters
internality-externality and employed the method--developed by Witkin and his
divided the subjects' scores on the basis of whether, over the series of trials, the
associates (1954)-of scoring TAT stories for story figures' ability to cope with
subject had established a norm or frame of reference for the degree of light move-
environmental and intrapersonal conflicts. The subjects (forty female psychiatric
ment. A statistically significant relationship (at the .02 level) was found between
nursing students) were given Rotter's internality-externality scale and two weeks
the establishment of a norm in the autokinetic experiment and a high score on the
later were asked to write stories to seven TAT cards. Their TAT scores allowed
Berkeley ethnocentrism scale. Further (at significance levels ranging between .05 and
correct classification of 80 percent of the subjects as to which end of the internality-
.01), overcontrollers were found to be highest in ethnocentrism, undercontrollers
were lowest, and appropriate controllers fell in between. Here the employment of a externality scale their score fell. (This correlation was unrelated to grades or to
mental health, as these factors were judged by experienced clinicians.) As noted
quite different method of measurement and a different point of view made an im-
in the previous section, there is a pressing need to understand why a direct measure
portant contribution by confirming the thesis of the Berkeley study, which had em-
picks up different scores for some respondents than does a projective measure; and
ployed another technique to measure intolerance of ambiguity.
thus deviant case analysis of the 20 percent whose scores did not agree would have
Also in this area is an ongoing study of the effects on moral development (in
been most helpful. This criticism is particularly important here since a secondary
Kohlberg's terms) of the stress and violence which children in Northern Ireland are
analysis of the TAT stories of the deviant scorers in terms of the subjects' scale
today experiencing (Fields, 1972). In a preliminary study in December 1971, the
responses might provide data leading to a refinement of our understanding of the
author interviewed twelve Catholic and twelve Protestant children from slum-
dimension of efficacy.
dwelling families in Belfast and an equal number (minus four Protestant children)
A second study which generally falls into this area of efficacy is reported by
in Dublin, employing a ten-card TAT and the Tapp questions on rules and laws
Mussen and Wyszynski (1952). The focus of this study is personality differences
(Tapp and Kohlberg, 1971). The TAT responses were analyzed in terms of Arnold's
between politically active and apathetic; for this purpose, the subjects selected were
(1962) method of story sequence analysis. Preliminary analysis indicates that the
156 University of Wisconsin undergraduates between the ages of seventeen and
children in Belfast
twenty-seven. In this study, the authors used a version of the projective question
delimited by Levinson (1950, pp. 545-548) as "an application of the general
are pessimistic and fatalistic. The people in their stories have little or no
control over their fate. They may choose to run away from c'troubles" principles of projective techniques to the questionnaire method and to the study of
but the trouble pursues them. People have incomprehensible motives for the dynamics of ideology. A projective question is an open-ended question which
destruction and children are quite helpless to contravert them. You may is answered in a few words or lines and which deals with unusual events or exper-
work hard to accomplish a task but more often your efforts end up in iences likely to have emotional significance for the individual. Care is taken to give
failure. Death and destruction are inevitable. When you try to correct an the question a 'homey,' even humorous, wording; also, an emphasis on the universal
injustice, you are hurt in the end. Ordinarily you try to tell others and get nature of certain emotional experiences (e.g., moods, embarrassment) may make the
them to help you, but they neither listen nor help. War is an everyday fact subject feel freer in giving an answer." Although this method is similar to other
of life and military action inevitably leads to grief and death. The prog- projective techniques, Levinson noted a number of advantages to its use in socio-
nosis, on the basis of motivation index scores, is quite dim. . . . The chil- psychological research: the items are readily comprehended; they can be answered
dren of Dublin, who share a similar cultural context but have not quickly ("eight items require only ten to fifteen minutes") ; they need no detailed
personally experienced war in their streets, demonstrate a more optimistic instructions; and, since they are admirably suited to questionnaire use, they provide
outlook. They have more concern with interpersonal relations and with a useful multimethod validation of dimensions measured in scale items.
the necessity for being good and "making amends." Their prognosis by In the Mussen-\Vyszynski (1952) study (at significance levels varying between
motivation index scores would appear to be more optimistic. The range of .05 and .lo), a number of meaningful differences appeared in the intrapsychic orien-
motivation index scores is parallel with those of children in the United tation of the active and the apathetic. Political apathy was seen as a fundamental
States and, in fact, somewhat higher than those of correspondingly lower
part of a basic passive orientation, identified by characteristics including "inability
class [Fields, 1972, p. 131.
428 Jeanne N. Knutson The New Frontier of Projective Techniques 429

to recognize personal responsibility or to examine-or cven accept-[one's] own party affiliation and the affiliation attributed to the photograph identified most posi-
emotions and feelings; vague, incomprehensible feelings of worry, insecurity, and tively by the subject, suggesting that a shared self-image may be an imporant factor in
threat; complete, unchallenging acceptance of constituted authority (social codes, a person's voting preferences. Interestingly, the working-class Conservatives usually
parents, religion) and conventional values" (p. 78). The politically active, on the identified positively with the photographs of the presumed Conservatives, but they
other hand, do not have feelings of insecurity and threat, and show "an emphasis on also tended to reject the presumed Conservatives as strongly as the two Labour
strivings for ego satisfactions, independence, maturity, and personal happiness. . . . groups did (selecting a presumed Conservative from photographs chosen for rejected
The sensitivity to others' feelings, emotions, and conflicts which is revealed by the identification). This finding suggests that working-class Conservatives have con-
politically active may also be interpreted as part of a generally active orientation, siderable ambivalence about their political orientation. Thus this study may have un-
since it may represent an outgoing response: an attempt to understand, and em- covered a fissure-in the form of a lack of personal identification with the
pathize with, others" (pp. 78-79). The investigators also note, however, that the Conservative image-in the bedrock of lower-class Conservatism, as well as evidence
positive desire to make social contributions must be accompanied by a lack of sense for the existence of a shared working-class identification which transcends party lines.
of threat for activity to be actualized and point out that authoritarianism (as Two other studies used a political cartoon to study attitudes. Fromme (1941)
measured by the F scale) does not differentiate between the active and the apathetic analyzed attitudes toward war, using as subjects thirty-five nonrandomly selected men.
student. Each subject was interviewed for four to seven hours and given (as a basis for the
Another research focus is the area of power, achievement, and affiliation interview probes) the SPSSI "Survey of Opinion on Methods of Preventing War."
motivation. First of all, Veroff and his associates (Veroff, 1961; Veroff and others, After the interview, in order to judge the stability of a subject's answers, the inter-
1960) used these dimensions in what, to my knowledge, is the only study in which viewer presented conflicting answers from an imaginary poll of experts. Additionally,
projective techniques were employed with a national sample (other than perhaps Fromme used five political cartoons, with the original captions removed; the subjects
some possibly projective questions). Although the interests of this nationwide study were asked to choose the most appropriate caption from among four. The subjects
of three basic motivations are generally unrelated to politics, a number of relation- were also given five TAT pictures, especially designed for the study. In spite of this
ships (for example, between the target dimensions and socioeconomic status) could wealth of data, the investigator offered n o definitive analysis of his materials but
be fruitfully reviewed by the political psychologist for heuristic purposes. did present suggestive evidence that "yes-no" responses (to the SPSSI Survey) about
In a directly political analysis of this dimension, Browning and Jacob (1964) half the time did not indicate the subject's true meaning.
gave the McClelland TAT series to a sample of fifty elected officials in two Louisiana Fillmore Sanford (1950b) reported a more carefully designed study which
parishes (67 percent of the elected officials) and, in an eastern city, to a random also used the cartoon as a projective technique. Sanford's study, which focused on
sample of twenty-three businessmen-politicians. They found that neither affiliation, leadership and authority, employed a random sample (N = 963) of Philadelphians.
power, nor achievement differentiated the politicians from the nonpoliticians. How- The immediate concern of the study was what people worry about. Fairly early in
ever, "politicians in high-potential positions scored much higher in both achievement the hour-long interview, the subjects were asked, "Do you think you worry more or
and power motivation than their matched sample (N = nine pairs)" (p. 85). Thus, less than most people?'' and then "What sort of thing do you worry about most?"
this controlled analysis suggests the existence of a relationship between role require- Later in the interview, the subjects were shown a cartoon which depicted a person
ments and personality type which their initial aggregate-level analysis missed. with a problem which the subject was asked to specify. The subjects were also asked
Several other political studies report imaginative uses of projective methods. when they sought advice and what kind of advice they needed.
Green and Stacey (1966) investigated the relationship between voting choice and
self-image. The investigators showed their subjects eight photographs of males and While 159 (16.5 percent) give no answers [to the cartoon queries], this
does not compare too unfavorably with the 13.2 percent who do not
asked the subjects to order the pictures along eight dimensions of personality (for
answer the direct question on worries or with the 13.8 percent who fail
instance, "Which of these men would you say was the most/least determined and to answer the question on the sort of advice needed. (For four of the six
confident? Now arrange the others in betweenn-p. 13). Each subject was then asked pictures used in the complete study, the percentage of "no answers" runs
which of the men he mostileast would like to be, "thus providing an index of around 4 percent.) In terms of the technical feasibility of the pictures,
preferred and rejected identification." Finally, the subjects were told that four of there is the additional factor that most people interviewed appear to
the men in the photographs were members of the Conservative Party and that four enjoy responding to the pictures. . . . They represent, apparently, a
were staunch Labour Party supporters, and they were asked to guess which was which. pleasant change of pace in a long interview [pp. 699-7011.
The random sample-consisting of eighty male voters who lived in London-was
composed of twenty middle-class Conservatives, twenty middle-class Labourites, Sanford alsp felt that the pictures elicited more specific and personal answers and
twenty working-class Conservatives, and twenty working-class Labourites. thus "gets past the psychological censor which often operates in the standard face-to-
Green and Stacey's data revealed th& each of the four subject groups (at a face interview" (p. 701 ) .
significance level of .05) had markedly different self-images, except for the two In the Sanford study, 201 people were retested one month later. The cartoon
working-class groups. The data also indicated a high relation between a subject's was as or more reliable than other measures; and, according to Sanford, the reliability
430 Jeanne N. Knutson The New Frontier of Projective Techniques 431
would have been higher "had the coding been based on categories of a more psy- obviated the causal evidence they sought, the investigators present impressive corre-
chological or more genotypical nature" rather than "relatively superficial ones, based lations between social character (such as exploitative or productive hoarding) and
on the clearly manifest content of- the answers" (pp. 701-703). Sanford's retest such behavioral indices as assumption of and functioning in family roles, degree of
analysis indicated that the pictures elicited basic attitudes rather than simply indicat- alcoholism, type of crops planted, and degree of economic success. Important here is
ing socially desirable responses. (The politically relevant findings of this study are the sensitive nature of the data derived from projective techniques and the value of
reported in Sanford, 1950a.) These findings raise the important issue in the use of employing such a technique in studying intrapsychic dynamics of a generally non-
projective techniques of the ways in which reliability is affected when coding moves literate population.
iron1 a simple response description (low reliability) to broader categories in terms Finally, three studies in political socialization have employed projective tech-
of themes or story imports (generally high reliability) to coding based on inferences niques as a necessary antidote to the almost total reliance here on data shaped by
about intrapsychic processes (reliability issue unclear) . survey methods. First is a study by Greenstein and Tarrow (1970) of children in
Another study employing a somewhat different technique in politically rele- America, France, and Great Britain. The preliminary report focuses on the method-
vant research is reported by Hanfmann and Getzels (1955), who gave a specially ology used (an episodes technique similar to the one employed by Hanfmann and
constructed "Episodes Test" to a sample of Soviet refugees (forty-one men and ten Getzels) and presents as data only a few illustrative case studies, so that the value
women) living in Munich plus a control sample of Americans. The episodes of their techniques of measurement remains to be assessed. Some of the episodes deal
(described above) consisted of a series of brief verbal descriptions of interpersonal indirectly with political themes (and are thus termed "parapolitical") ; others are
situations (five familial and five extrafamilial). The subject .was asked to describe directly political. An example of the former is as follows: "A group of children of
the probable development and outcome of the episode. The technique was labeled your -own age are playing. Some of them want to play one game. Others want
cr
semiprojective" because "it addresses itself to the level of realistic, social-directed another. There are not enough children to play both. Finish the story." An example
action, rather than to that of fantasy, and elicits material that is not too far removbd of a political episode: "One day the President (substitute Queen in England, Presi-
from the subject's conscious attitudes and from his manifest behavior" (p. 1 ) . dent of the Republic in France) was driving his car to a meeting. Because he was
Hanfmann and Getzels (1955) found that Americans were more likely to see late he was driving very fast. The police stop the car. Finish the story."
difficulties as internal; the Russians, as social and/or political. Additionally, the Greenstein and Tarrow (p. 505) suggest that their measures help to fill the
Russians expressed significantly more positive feelings toward the social group and "imperfect fit between inner perceptual experience and its representation in survey
did not feel as weak, helpless, or unconfident vis-&-vis the group as did the American data" and make this cogent comment: "No doubt many survey respondents express,
subjects. In short, "the Russians do not feel compelled, as the Americans are, to in Converse's (1970) term, 'nonattitudes' rather than actual attitudes when queried
defend their individual integrity against the group" (pp. 34-35). Correlated with by the pollster. But it would seem likely that in addition there are people who do
this, the Russians were less hero-oriented, but rather tended to express more spon- hold opinions, or at least reasonably stable 'outlooks,' and who are simply unable to
taneous empathy for all the actors. Of equal value are some distinctiois that did not make the connection between their own thought patterns and the formalized mold
occur: "The data of the episodes do not yield any material that would indicate a provided for them by the authors of survey questionnaires." The results of this study
greater incidence of authoritarian character traits in one or in the other national will certainly be of major interest to political psychologists.
group," nor were there any great differences in other "essential personality variables" A second use of projective techniques in pditical socialization research is
such as aggression and affection-thus negating national-character arguments (pp. found in reports of work by Adelson and O'Neil (1966), using 120 children taken
36-37). In many ways, this study illustrates the importance of employing projective equally from the fifth, seventh, ninth, and twelfth grades. Their interview schedule
techniques, for the responses give evidence not only of what attitude is held but also
was based on the following premise: "Imagine that a thousand men and women,
of why it is held (its meaning and importance to the person in terms of his own
dissatisfied with the way things are going in their country, decide to purchase and
needs and values). move to an island in the Pacific; once there, they must devise laws and modes of
Another study in which the psychic-social nexus receives focus is the lengthy
government." After this imaginary format was offered to and accepted by the sub-
psychoanalytically oriented study of a Mexican village reported by Fromm and
jects, the children were asked to discuss a number of hypothetical issues:
Maccoby (1970). Over a period of many years, a variety of research techniques
(participant observation, Rorschach tests, the TAT, in-depth questionnaires) probed
For example, the subject was asked to choose among several forms of
the character types of the villagers (in terms of Fromm's typdogy) and the relation- government and to argue the merits and difficulties of each. Proposed
ship between social character and behavior in a number of areas. The authors (p. 7) laws were suggested to him; he was asked to weigh their advantages and
stated their thesis as follows: "We believe that just as psychoanalysis studies the liabilities and answer arguments from opposing positions. The interview
character of an individual in terms of analyzing the underlying forces which in a leaned heavily on dilemma items, wherein traditional issues in political
structuralized form make up his character and motivate him to feel and think in theory are actualized in specific instances of political conflict, with the
certain ways, the character common to a whole group, social character, has the same subject asked to choose and justify a solution. The content of our inquiry
dynamic function and can be studied empirically." While their research design ranged widely to include, among others, the following topics: the scope
432 Jeanne N. Knutson The New Frontier of Projective Techniques 433

and limits of political authority, the reciprocal obligations of citizens and 7. Locus of Authority: internal or external, with man as actor or puppet:
state, utopian views of man and society, conceptions of law and justice, the ability of m a n to rationally shape his destiny
the nzture of the political process. 8. Group-Individual: the ability of a group to satisfy an individual's
needs (especially affection and esteem) : the group as deprivational
The investigators suggested a sound rationale for choosing this projective method of or nurturing of a n individual's relational needs
exploring the development of a sense of community: "Our pretesting had taught us
that direct questions on such large and solemn issues, though at times very useful, 9. Poverty: as due to human failure, social conditions, fate, etc.: social
tended to evoke simple incoherence from the cognitively unready, and schoolboy us. individual responsibility for those falling below norms and the
causality of social dependence
stock responses from the facile." Employment of this technique yielded a rich data
base from which it was possible to comprehend not only what the children understood 10. Systemic Loyalty: country as ego alien, deserving of higher regard
of the target concept at each developmental level but also both their misconceptions than self, etc. : the power of th,e system to claim allegiance
and the dynamics which served to unite these misconceptions. The nuances of shifts
from an egocentric to a sociocentric perspective are particularly well reflected in these A preliminary study used these cards1 to explore the politicization of young
data. children (fourteen subjects in each of the third and sixth grades). Analysis indicates
In addition to these two socialization studies, I am currently engaged in a that children at this age have well-defined views of themselves and others; on a pre-
research project with a similar focus. (See Knutson, 1974, for a preliminary report.) political level, they have a full range of ideological stances similar to the more
This work employs a specially developed political version of the TAT, designed to politicized viewpoints held by their elders. In this study, the children were also asked
correspond to empirical evidence (discussed earlier) that ( 1) focused cards tend to to tell "what you are thinking about and feeling" when shown ten politically relevant
be most predictive of target behavior; (2) cards that are psychologically ambiguous magazine pictures (for instance, of President Nixon, of a policeman, and of a
but physically structured have been most successful; and (3) cards need not depict prisoner in jail) and to define, if possible, eighteen political concepts (such as
figures similar to the subject's sex, culture, or class-indeed, the most useful cards democracy, laws, Republican). The combined projective responses suggest--contrary
are often set in other times or cultures in order-to relax the subject's defenses so to the predominant assumptions in socialization literature-that many children are
as to elicit psychically meaningful, veridical responses, although such responses would at best neutral and at times quite hostile to political objects; the data further indicate
be unacceptable to the subject if conscious and/or verbalized. that children have clearly defined and personally meanirigful beliefs about basic (as
The ten cards of what is called the Political Thematic Apperception Measure opposed to time-bound) political issues-such as how criminals should be viewed and
(PTAM) were developed to correspond to ten basic themes in politics and to elicit what role war should have. In sum, this assessment method allowed young children
(when present) five basic intrapsychic need-areas, as delimited by Maslow (physio- to discuss comprehensively subjects which (1) would be far too difficult for their
logical, safety or security, affection and belongingness, esteem and self-actualization). verbal ability if asked in questionnaire form, (2) would often lead to ' ' S C ~ O O ~ ~ O ~ "
The theme cards are : responses, parroting little-understood definitions and ideological segments, and (3)
would elicit a "positivity" effect if cues were given so that any socially "right"
1. Legalized Force (aggression) : the beneficial vs. deleterious effects, answer was available.
the identification of the subject with protagonist as "victim" or society
as "victimized," etc.: the necessity and role of legalized force Research Advantages
2. Human Nature: seen as basically cooperative, antagonistic in a
Hobbesian sense, etc.: the ability which humans possess t o solve To the typical-that is, skeptical-political psychologist, far from at ease with
h u m a n problems clinical methods and clinical analysis, the above literature review may be interesting
but may also raise a number of serious reservations. He may wonder, for instance,
3. Authority: useful and beneficent, threatening and alien, etc.: the
about the considerable time necessary for data gathering and evaluation, the develop-
relation which it is n.ecessary t o make with authority
ment of the requisite skills, the necessity of validating new instruments (such as the
4. Leader-Follower: the necessary relation between leader and group; PTAM) or new uses of old instruments (explicating the political relevance of
leader as object or subject of authority: the basis of the decision- standard Rorschach responses), and he may experience as well the conservative force
making process of inertia (both professional and personal). Therefore, the present section deals
5. Security in Society: the safety or danger of communal life, etc.: the with enumeration of the special advantages offered by the assessment techniques that
ability of m e n t o live communally and handle antisocial acts have come to light in the literature review. These advantages are in part overlapping;
6. Youth: seen as an alienated or integrated group, object of hostility, 1 Copies of these cards are available from the author to interested political psycholo-
curiosity, affection, etc.: the function and role of generational differ- gists who wish to use them for research purposes. Requests should be sent to the author at
ences The Wright Institute, 10837 Via Verona, Los Angeles, California 90024.
434 Jeanne N. Knutson
The New Frontier of Projective Techniques 435
their number, however, is suggestive of the potential of projective techniques for
own idiosyncratic feelings, meanings, and ideas" (Harrison, 1965, p. 563). Instead
political psychologists.
of merely giving evidence of "how much" of a trait a respondent possesses, the
Two of these advantages have been already discussed. First, as Greenstein
projective technique makes palpable the modus vivendi of the person in which any
and Tarrow (1970) note, projective techniques help to narrow the distance between
target trait operates (Murstein, 1961, p. 3) . Thus, to employ projective responses only
the researcher's determined view of the respondent's inner reality and that inner
to assess a single personality dimension obviates the valuable contribution which con-
reality as perceived and employed by the respondent. Thus, both the investigator
figurational analysis makes possible.
and his subject are assured that responses are rich in meaning and are the choice of
Also inherent in a study of the manner in which a person typically responds
the subject. Additionally, however, responses can later be classified and ordered along
is the advantage-a; suggested above--of a holistic approach to personality. As
more generally meaningful dimensions in an aggregate analysis. Second, as discussed
Atkinson (1961, p. 227) has commented, "When you go at projectives the way some
by Campbell (1950), the systematic biases which appear in projective responses
of us have, looking only at isolated motives (achievement, power, affiliation), you
provide a more adequate case for face validity than do the typical casually employed,
lose contact with the whole person." Projective data, however, allow interpretation
unvalidated personality scales.
on both levels. Not only can assessments be made similar to the trait measurements
Next-and intimately related to the above-projective techniques reduce and
of personality scales, but personality can also be seen as an open-ended process.
standardize the external constraints which inevitably shape the expression of intra-
An additional advantage of e~nployingprojective techniques in political-
psychic predispositions, so that predispositions stand out with greater clarity
psychological research is clearly the opportunity to delimit the conditions under which
(Fromme, 1941). As Greenstein and Tarrow (1970, pp. 498-499) note, "Clinicians
a need or trait may be expressed, as well as the manner (the intensity and direction)
have argued that the open-endedness of projectiQe procedures contributes to their
in which expression is likely to occur. Once a separate area of intrapsychic function-
diagnostic utility by ensuring that the categories a respondent uses are his own rather
ing has been delimited, it i; imperative that the political psychologist determine
than those built into the test." It is well understood today that behavior is the result of
under what conditions and in what manner that predisposition is likely to become
personality interacting with social and cultural norms and values and with situational
operative. As the requirement for political relevance circumscribes the employment
constraints. By the use of projective techniques, it is possible to determine and con-
of laboratory experimentation (Knutson, 1972b; see also Chapter Thirteen in this
trol the influence of situational constraints; intercultural research may then make it
volume), specially constructed projective measures provide an additional way to
possible to isolate the influence of cultural and social factors. Projective techniques
assess the situational imperatives which are likely to engage intrapsychic predisposi-
thus may facilitate our understanding of the determining influence of predispositions
tions. Thus, in the area of projective theory predictive validity may be a more im-
on behavior.
portant and valid goal than concurrent validity (Klopfer, 1968). As previously
In addition, projective techniques are of great heuristic value because they
discussed, prediction is a problem that has plagued personality research as a whole
provide rich opportunity to learn about what we continually attempt to measure
(Korner, 1965; see also Chapter Two above). Projective techniques suggest a fruitful
gropingly. This advantage was brilliantly illustrated in research on the authoritarian
avenue of increasing our predictive powers in this area.
personality (for instance, Frenkel-Brunswik and Sanford, 1945), where hypotheses
Further, projective techniques provide a wealth of data. Responses can be
concerning psychogenesis and psychodynamics were developed and refined from
scored to illustrate a variety of theoretical perspectives; they are not limited to the
projective data and then employed as a theoretical basis for scale development (see
immediate focus of the investigator but may be reanalyzed by others in terms of
Chapter Six above). Thus, projective techniques allow exploration of the dimensions
different or unthought-of perspectives (Block and Haan, 1971). If drawn from a
underlying the target behavior. In the Hanfmann-Getzels (1955) study, for example,
meaningful sample, such data thus .provide a "bank" of information and an avenue
the data not only illustrated that Russians and Americans differ along fairly obvious
of access which has been too seldom traveled into the wealth of material accumulated
dimensions of social and individual values, but also made it possible to ascertain the
for psychiatric purposes (see, for instance, Almond, 1954; Lasswell, 1930). The rich-
intrapsychic toll which such differences exact and the unitary nature of personality
ness of the data provided by projective techniques is of course related to its heuristic
from which such diversity stemmed. Thus, projective techniques allow us to under-
importance. As Greenstein and Tarrow (1970, p. 531) noted: "We have been
stand motivations which the respondent is unaware of or only barely so (Klopfer,
constantly impressed by the number of serendipitous, theoretically interesting themes
1968), and also those which the investigator, in all his wisdom, cannot previously
that emerge in our sample." (This idea is well illustrated in Veroff, 1961.)
know of or intuit.
Another major advantage in using projective techniques is that they provide
In addition, projective techniques can be scored in ways which focus on how
the respondent organizes intrapsychic imperatives while simultaneously organizing an opportunity to study motivations that may be clearly delimited but are socially
and integrating focused situational variables. In other words, projective techniques undesirable and/or ego-alien to the respondent. Thus, investigators are better
equipped to surmount the barrier of social desirability which is now understood to
provide a unique way of assessing the manner in which a person responds. Indeed,
shape the meaning of scale responses (as well as other types of human behavior). As
a basic rationale behind the use of this assessment method has been that "personality
reveals some of its organization and ways of organizing experience when the person noted by Morgan (1945, p. 219) : "The fact of the matter is that a survey is likely
to be most accurate when the data assembled are based on drab or settled issues and
is given more or less ambiguous stimuli to interpret and in the process projects his
is most untrustworthy when the question is one on which opinion is divided, when
436 Jeanne N. Knutson The New Frontier of Projective Tecl~niques

violent emotions are involved, or when mutual distrust prevails." On the other hand, graphic characteristics. Finally, projective techniques make it possible for a deeper
in giving projective responses, the subject "often allows himself a greater degee of understanding of the manner in which an individual organizes his inner and outer
freedom of expression because he is not openly telling about himself and giving his reality, as well as the probable direction and intensity of such organization.
own ideas about real people and how they act" (Aron, 1950, p. 489). The opportun-
ity to have a veridical view of inner predispositions is intimately tied to the fact that Overview
projective techniques are "less susceptible to faking. Even if an individual has some
psychological sophistication and is familiar with the general nature of a particular Projective techniques are no panacea, free of methodological problems and
instrument, such as the Rorschach or TAT, it is still unlikely that he can predict the research limitations. Like every other method of assessment, they offer certain unique
intricate ways in which his responses will be scored and interpreted. Moreover, the advantages as well as clear drawbacks. The above review suggests that they may
examinee soon becomes absorbed in the task and hence is less likely to resort to the complement in valuable ways more traditional measures which are currently employed
customary disguises and restraints of interpersonal comn~unication" (Anastasi, in and allow access into arenas generally closed to political research. In addition, I
Greenstein and Tarrow, 1970, p. 488). believe that their use -will promote the development of political psychology by their
Of further value in employing projective techniques is their facility of admin- imperious demands for explication of what we set out to measure. A projective
istration, particularly as a way of engaging the interest of those with little skill and/ response is an open-ended resource; what we choose to see its content and form as
or ease in accomplishing verbal tasks or achieving in testing situations. Riessman and revealing about its author is nonobvious. Thus, the use of projective techniques
Miller (1958), for example, have found that projective techniques are particularly requires especially rigorous attention to the validity of those levels of inference-
effective tools in research with lower socioeconomic groups. A number of other from the intrapsychic to political behavior-which we select to employ (Knutson,
investigators (for example, Green and Stacey, 1966; Greenstein and Tarrow, 1970) 1972b).
have spontaneously noted that their subjects enjoyed this research task. Related to Taken as a whole, the advantages discussed above should lead to greater
this advantage is another opportunity-namely, that projective techniques "will ,
specification of process assumptions, as we study intrapsychic dynamics under dif-
make possible the u:e of large and representative samples in testing personality ferent sets of constraints. Thus, the processes of bargaining and decision making, of
hypotheses heretofore based on and tested by the study of small numbers of people attitude formation, and of the development of political ideology are likely areas for
drawn from the 'captive' undergraduate population" (F. Sanford, 1950b, p. 709). the focus of future research. In addition, projective techniques can serve as a neces-
Another reason for employing projective techniques in political research sary antidote to past heavy reliance on survey research questionnaires and can help
touches on an area of particular concern to investigators in this area. Projective us assess standard results gained thereby. Projective responses thus make it possible
techniques make possible comparisons across cultural and age groups (Greenstein to enrich aggregate analysis by individual-level data and-as the above discussion of
and Tarrow, 1970) because it is possible to employ the same research stimuli with socialization studies suggests-to vitalize some simplistic assumptions (by, for exam-
subjects who vary widely in verbal fluency, mother tongue, social class, and age. ple, exploring the dynamics of extremism and conformity). Past use has also sug-
Thus, projective techniques provide a valuable asset to those interested in develop- gested that projective techniques present an interviewee with an enjoyable and
mental and cross-cultural research. welcome change from a stock set of questions and answers, as well as providing
Finally, projective techniques offer a way in which the more methodologically access to the meaning of his written records; hence, they may also be of value to
conscious investigator of today can validate his conclusions. As noted above, prqjective those concerned with the study of leadership. (For a valuable example of the use of
questions fit well within the constraints of survey research. A further step removed these techniques to analyze personal records, see Bellak, 1966.)
but still offering the same advantage are the numerous other methods of projective The view that projective techniques provide an unclouded window to the
measurement. When it is possible to achieve a high degree of accuracy in predicting unconscious has departed and has been replaced by some sobering caveats. With these
the response of a subject on a direct measure from his response on a projective caveats, however, the projective response has become more useful to the political
measure (or vice versa), the investigator can be a good deal more comfortable that psychologist. Gaining recognition is a view of projective responses as focused segments
he is tapping a psychically meaningful, stable response pattern and that he adequately of behavior, which are expressions of intrapsychic needs mediated by ego processes,
understands the meaning of that pattern. (For a further discussion of this point, see by (experimentally isolable) social and role conventions, and by cultural values.
Chapter Two in this volume.) Within this wider meaning, projective techniques hold forth the promise of enriching
Projective techniques, then, offer a number of advantages which urge their our scientific understanding by making us more deeply aware of the ways in which
inclusion in the research tools employed by political psychologists. While the advan- inner and outer reality are individually shaped into political behavior.
tages touch on a wide variety of areas, most compelling to me is the opportunity to
widen our data base. Projective responses can be coded and designed to elicit pre-
dispositions in sensitive areas where-apart from captive college samples-it is not
usually feasible to probe. Cross-cultural comparison and developmental studies are
aided as a similar stimulus can be used with samples which vary widely in demo-
 Political Psychology: A Pluralistic Universe 439

students of politics have been slow to construct bridges between political science and
psychology. That there now should be a Handbook df Political Psychology seems to
signal a major-and highly promising-change in intellectual climate. Each of the
preceding chapters in this handbook is an impressive accomplishment, not only for
the substantial amount of literature, much of it quite recent, that is summarized,
but also for the originality of the contributor's own synthesis. And just as the indi-
vidual chapters each go beyond their bibliographical building blocks, the collective
impression left by the full array of chapters considerably exceeds the sum of the parts.
In short, political psychology as an interdisciplinary endeavor seems finally to be

POLITICAL alive, wellJ and sufficiently developed not to recede with the next shift in academic
fashions.
There appears to be no way to "do justice" to each of the richly diverse con-
tributions to a volume of this sort, particularly if the concluding chapter is also to
PSYCHOLOGY: make its own statement. My tack, therefore, has been to compose an essay (in the
last three sections of this chapter) building on my own program over the years of
seeking to clarify investigative strategies for the study of political psychology (Green-

A PLURALISTK stein, 1969; Greenstein and Lerner, 1971). In that essay I seek simultaneously to
argue a series of propositions about the peculiar properties of political psychology
and to make at least some reference to each of'the fifteen chapters that precede
mine, pointing to connections among them and among other elements in the plural-
UNIVERSE istic universe of political psychology not reviewed in this volume. Although I draw
extensively on the formulations of several of the contributors, I have merely alluded
to other chapters of equal importance which did not lend themselves to my argument,
and have discussed the topic of several of the chapters rather than the specific points
made by the authors. What follows, therefore, is neither a summary nor an evalua-
Fred I. Greenstein tion of the other chapters, each of which needs to be approached on its own terms.

Institutional Connections Between Psychology and Political Science

The disciplines of political science and psychology have some common roots.
Both can be traced to Plato and the pre-Socratics. Both represent enduring pre-
occupations of mankind. Both have antecedents in philosophy, theology, and ethics
and are more and more difficult to distinguish from those endeavors the further one
looks back into intellectual history. Neither was present in universities as an officially
titled discipline before the nineteenth century,
"political psychology'' has two referents: But in spite of the elements of common heritage, there are significant diver-
the psychological components of human political behavior and the academic en-
gences as well. Political science departments evolved from law schools and depart-
deavor of applying psychological knowledge to the explanation of politics. Political
ments of history as well as drawing on philosophy and kindred fields. Some of the
psychology in the first sense is a permanent part of the human condition, funda-
most sizable tributaries leading into the endeavors summarized by the term psychology
mentally implicated in the extraordinary achievements of the species, but also in the have their origins in the 'natural, and especially the life, sciences. Furthermore,
chronic difficulties mankind faces in living in peace and managing material and psychology became well grounded and entrenched in its institutional settings and
human resources. The profound human need to come 'to grips with political psychol- traditions in the nineteenth century; political science has been a slowly developing
ogy in the first sense accounts for the need for systematic development of political
phenomenon of the twentieth century.
psychology in the second sense. There are far more psychologists than political scientists in the United States,
For reasons I shall attempt to suggest (in the first two sections of this chapter). probably because psychology-especially in the post-World War I1 years-has
438 become intricately and complexly intertwined in the society as an applied pursuit, as
well as being lodged in' the academy, whereas political science is largely an academic
440. Fred I. Greenstein Political Psychology: A Pluralistic Universe 441

endeavor. In 1947 the American Psychological and American Political Science applied specialty of personnel management and assessment on the part of civil ser-
Associations had approximately the same number of members; by 1957 the former vants and students of public administration. Thus, at that time, those political
was more than twice the size of the latter, although both had grown markedly scientists with a behavioral science bent were drawn toward sociology. (The 1961
(Behavioral and Social Sciences Survey, 1969, p. 23). I am informed by the APA interest in economics seems to have had a traditional nexus in the economic institu-
and APSA national offices that in 1972 the former had 2927 Fellows (members tions that impinge on government rather than representing an intellectual commit-
judged to have made distinguished contributions) and 23,870 Regular Members ment to the analytic tools and perspectives of modern economics. By the end of that
(members with Ph.D.s) ; the latter had only 12,402 members in the nearest equiva- decade, economics was still another strand in "behavioral" and "post-behavioral"
lent category to APA Regular Members and Fellows combined-individual (non- political science-for example, via the public choice literature.)
student) members. Another more fundamental sign of the greater tropism of political scientists
The growth of interest within political science in what is now called the towzrd socio!ogy than toward psychology is that political scientists and sociologists
behavioral sciences was presaged in the great University of Chicago political science have in fact worked together on similar problems in similar ways, whereas there are
department of the 1920s led by Charles E. Merriam, who in his New Aspects of far fewer evidences of such connections between political scientists and psychologists.
Politics (1925) explicitly called for a scientific political science that would draw on Thus political scientists and sdciologists eften publish in each other's journals; politi-
other disciplines, including psychology. At that time Merriam's student and col- cal scientists and psychologists rarely do. Within sociology a recognized subfield-
league, Harold D. Lasswell, became the Founding Father of political psychology as political socio1og)l-explicitly takes note of sociology's connections with political
an academic subdiscipline. I t was many years, however, until work informed by science. The American Psychological Association has divisions of industrial psychol-
disciplines other than law and the humanities began to appear with frequency in ogy, military psychology, esthetics, and even philosophical psychology, but none of
periodicals like T h e American Political Science Review, first in the late 1940s and political psychology.
then increasingly in the 1950s.
The behavioral science emphasis in political science, popularly labeled "be- '
Logical and Empirical Connections Between Psychology and Political Science
haviorism" or "behavioralism," with no intended killusion to the behaviorism of
Pavlov and Watson, was pronounced a success (and hence ready to be assimilated Modern political science has reached out more to sociology than to psychology
and transcended) by one of the movement's key figures, Robert A. Dahl, in his 1961 partly because of the stimulus properties of the academic discipline of psychology as
essay "The Behavioral Approach in Political Science: Epitaph for a Monument to it is commonly practiced, if not as it inevitably must be practiced. For many psy-
a Successful Protest." Yet it is interesting to look at the Biographical Directory of chologists it is a central if unarticulated major premise not to conceive of their
the American Political Science Association published in the same year (American science as one which attempts to explain concrete instances of behavior in particular
Political Science Association, 1961). The 1961 APSA directory gave members of the sociopolitical contexts. Rather, in the phrase of one psychologist, Richard Littman
association the opportunity to list the full range of cognate fields that interested (1961), psychology is often seen as a "socially indifferent" discipline.
them. Of the numerous "other disciplines" in which political scientists evinced an Littman's observations were made before the increasing preoccupation
interest, there were more references to history (471) and economics (279) than to throughout the 1960s and early 1970s in all the behavioral sciences with "relevance."
either of the two disciplines which, from the standpoint of a logic-of-the-behavioral- He used "indifference" in a technical sense to refer not to a callousness about practi-
sciences rationale set forth below, "oughtJ' to be the twin pillars of political science: cal utility but rather to a basic strategy of inquiry. In seeking universal principles of
sociology and psychology. Dahl's epitaph evidently was a bit premature. Between behavior, Littman suggested, the psychologist has traditionally stripped away ele-
sociology and psychology, moreover, the preference was clearly for the former, with ments that are specific-for example, to the behavior of a congressional committee
188 references to sociology and only 69 to psycho1ogy.l or at a political convention. "Psychologists do study and must study things and
Much of the interest in psychology that ,was expressed by APSA members in activities possessing social content. . . . I t is only that psychology has been a science
1961 seems to have reflected not preoccupation with psychological aspects of the that abstracts out of all these content-characterized behaviors the concepts which
political process but rather quite narrow and technical interest in the traditional form the jargons of its subdisciplines." This may thereby contribute to the develop-
ment of universal laws, but there is a negative consequence: "When colleagues in
1 The 1968 APSA directory (APSA, 1968) did list "political psychology" as one of
the categories among which members could choose in expressing their interests, but strangely other disciplines (mainly sociology, anthropology, political science, and economics)
enough did not include the more recognized category "political sociology." Reference to turn to psychology for help, they are disappointed and, indeed, often aggrieved.
"political psychology" was made by 462 APSA members, a roughly sixfold increase during What they begin to read with enthusiasm they put down with depression. What
a period when overall APSA membership doubled, but only about 3 percent of the associa- seemed promising turns out to be sterile, palpably trivial, or false and, in any case, a
tion's membership. The most recent directory (APSA, 1973) uses still another set of sub- waste of time'' (p. 235). When psychologists do pronounce on problems of politics
disciplinary categories, one which makes it impossible to estimate the total number of and society,2 Littman argues, their observations often are "na'ive" and "conventional"
members disposed to profess an interest in psychology. (There is a listing for a subset of
political psychology, "personality and motivation," which was chosen by 238 members of
the association.) 2 Those willing to comment on psychological aspects of politics usually have been
442 Fred I. Greenstein
Political Psychology: A Pluralistic Universe 443
because they "are ignorant of the historical dimensions of most social activity" and
psychology literature (including two contributors to this volume, Hyman and Yinger)
"do not see the complex interweaving of institutions and arrangements." Psycholo-
are sociologists. Psychologists, in fact, are frequently interested in groups and are
gists, he suggests, "tend to be like laymen when they confront social phenomena,"
sufficiently interested in the environment, the source of "stimuli," that for some psy-
particularly those that involve large-scale patterns because "the main areas of social
chologists, like B. F. Skinner, it is hard to see just where the psyche fits in at all.
activity are only the place where psychologists study interesting sorts of things, rather
As M. Brewster Smith remarks in his chapter in this volume, a Skinnerian emphasis
than being the focus of inquiry."
would incapacitate students of political behavior, since actors and actions are central
But, Littman continues, the reciprocal of psychologists' insensitivity to socio-
to their concerns. Most political actions of any analytic interest call for psychological
political realities is the proclivity of nonpsychologists in their analyses of politics and
as well as environmental explanations, for reasons suggested in Noam Chomsky's
society to "posit incorrect or weak laws about individual humans." Both sides of
(1959) well-known critique of Skinner's Verbal Behavior. Chomsky defends what
Littman's thesis are profoundly germane to the diverse literatures of political psy- Smith calls "a psychology that bets on the strategy of accounting for the flux of
chology and will be illustrated at various points ih this chapter. observed social behavior by abstracting analytically two classes of inferred, recon-
Even though the links between political science and sociology seem to have structed determinants: features of the situation of action and inferred dispositions or
been stronger to date than those between political science and psychology, the study properties of the behaving person." Referring to Skinner's reluctance to accept
of politics needs to be firmly grounded in bmoth disciplines. We can demonstrate this inferences about human dispositions as evidence for explaining behavior, Chomsky
statement on two levels: (1) by considering what appears to be the broad general
(p. 27) points out:
purposes and preoccupations of the three disciplines and (2) by looking closely at
the problem of explaining political or, for that matter, any other kind of patterned Insofar as independent neurophysiological evidence is not available, it is
human behavior. obvious that inferences concerning the structure of the organism are based
As disciplines, psychology and sociology have directly complementary and on observation of behavior and outside events. Nevertheless, one's estimate
equally far-reaching concerns. The territorial imperatives of these two disciplines of the relative importance of external factors and internal structures in
'
appear to be dual: first, in Inkeles' (1963, p. 319) phrase, psychology deals with the determination of behavior will have an important effect on the direc-
"the personal system," sociology with "the social system." Second, psychology is tion of research. . . . Putting it differently, anyone who sets himself the
concerned with those determinants of behavior that arise from within individuals, problem of analyzing the causation of behavior will (in the absence of
whereas sociology focuses on the effects of the environment, especially the human independent neurophysiological evidence) concern himself with the only
environment, on individuals' behavior. In each case the discipline is conceived of as data available, namely the record of inputs to the organism and the or-
a basic science, unbounded in the specific modes and contexts of human activity ganism's present response, and will try to describe the function specifying
with which it may be concerned. the response in terms of the history of inputs. . . . There are no possible
grounds for argument here. . . . The differences that arise between those
I t follows from the universal ambitions of sociology and psychology that all
who affirm and those who deny the importance of the specific "contribu-
institution-specific disciplines, the two most notable being political science and
tion of the organism" to learning and performance concern the particular
economics, are derivatives in a logical sense from the two basic disciplines. As character and complexity of this function, and the kinds of observations
Herbert Simon (1959, p. 253) puts it with reference to economics, any "verified and research necessary for arriving at a precise specification of it.
generalizations about human economic behavior must have a place in the more
general theories of human behavior to which psychology and sociology aspire." The Chomsky then goes on to comment: "If the oontribution of the organism is
reasons (from the standpoint of the present abstract reconstruction) for according complex, the only hope of predicting behavior even in a gross way will be through
the institution-specific fields independent disciplinary status is the amount of a very indirect program of research that begins by studying the detailed character of
specialized knowledge necessary to deal with their subject matter and the importance the behavior itself and the particular capacities of the organism involved" (emphasis
of that subject matter for the conduct of human affairs. supplied). Since in politics "the organism" (actor) is manifestly complex in its con-
We have already made it clear that the real-world denizens of academic tributions to political behavior, at least in some circumstances, Chomsky's account
departments do not fit neatly into analytic ideal types, in that actual political scientists of what is involved in explaining behavior leads to the same conclusion as our dis-
have drawn more on sociology than on psychology. I t is also the case that the disci- cussion of the logic of academic disciplines: Political studies must be jointly social
plinarians of sociology and psychology overlap enormously in what they actually do. and psychological.
Sociologists often study individuals and intrapsychic variables, even though they Often, as Betty Glad points out in this volume, the complex contributions of
frequently profess to draw the line at attitudes and to be not interested in something the personal properties of political actors to their behavior are remarked upon by
called "personality." Indeed, some of the most important contributors to the political political analysts only when that behavior deviates from what might "normally" be
expected of an actor in similar circumstances. And such is the reliance of political
not academic psychologists but psychoanalysts like Erich Fromm or culture-and-personality scientists on sociological premises that psychological explanatory notions are left
anthropologists like Margaret Mead. implicit or deliberately not introduced if a situational explanation can be found.
 Fred I. Greeiistein Political Psychology: A Pluralistic Universe 445

Commonly, psychological explanation seems to find its way into political citizens. There would be interest, however, if like the President's Commission on the
analysis as a self-conscious activity via the back door. When situational explanation Assassination of President John F. Kennedy (1964, pp. 669-740) we were attempting
does not seem satisfying, there is an impulse to turn to psychological explanations. to explain the behavior of Lee Harvey Oswald. Similarly, psychological analyses
The sequence is picturesquely illustrated by B. A. Farrell (1963, p. 11) : "When we would be not necessary in explanations of the great bulk of political leaders, but
are ordinarily puzzled by someone, for example, our new neighbor next door, what they would be in the cases of dramatically successful or unsuccessful leaders whose
puzzles us are the apparent inconsistencies in his life and general conduct. Thus, our achievements or failings could not readily be attributed to external circumstances.
new neighbor might say he is very interested in gardening, but he lets his large Yet, as Glad points out in Chapter Eleven in this volume, it is scarcely satisfactory
garden go to ruin; when his luggage came, it was seen to contain various pieces of to reserve psychological explanations for the extraordinary and to imply that normal
sporting equipment, but he appears to play no sport and hardly goes out at all." In and typical behavior is not psychologically interesting.
effect, we generate a prediction of how a person in this situation "should" behave. According to Glad, my own inventory of conditions requiring political analysts
Because our neighbor's behavior seems inconsistent with our prediction, we begin to to take account of psychological factors (Greenstein, 1967; 1969, ch. 2) is open to
seek further information about his inner dispositions. Our puzzlement is likely to precisely the criticism that it treats political psychology as a means of explaining
cease and our investigation is likely to come to a rest if we learn a bit more about deviance, rather than as an enterprise applicable to the full range of political be-
the individual's situation. Farrell's illustration concludes: "But when we discover h a ~ i o r The
. ~ discussion in question is a distillation of two assertions made over the
that his wife has just died, that gardening and sport were joint activities of theirs, years: assertions by critics of the use of psychological data to explain political
and that he is still too distressed by her death to take up the normal round again, we behavior--critics who nevertheless acknowledged that under some circumstances it is.
then feel we have solved the puzzle-that we now understand him." important to look at psychological variables; assertions by defenders of political-
This "situational" explanation, of course, contains many-we might even say psychological analysis, who went on to indicate circumstances under which it would
mainly-pychological elements. However, the explanatory impulse is to treat our not be profitable to seek psychological determinants of behavior. The qualifying
neighbor's perceptions and feelings as epiphenomena of his circumstances and to statements by both the critics and the defenders of political psychological analysis
focus only on his circumstances in our explanatory account of him. And if psycho- converge in a single set of contingent statements, the following pair of propositions
logical explanatory factors tend to be admitted through the back door, attention to being examples: Variations in the personal qualities of actors are more likely to
underlying personality structures and psychodynamics tends, as it were, to be ad- account for variations in their behavior to the degree that the actors are placed in
mitted only through the window, late at night. Where the actor's behavior seems ex- ambiguous (new, complex, or contradictory) situations. Variations in personal
plicable neither as a reflexive action to his environment nor in terms of his conscious qualities are more likely to account for behavioral variations to the degree that actors
orientations-for example, if it could be shown that the neighbor in Farrell's example are in situations where sanctions are not attached to alternative courses of action.
had never had a wife-then the tendency would be to turn to the tool kit of clinical Propositions of the foregoing sort should, however, merely be thought of as
psychology for an explanation. indications of the necessary conditions for analyzing psychological antecedents of
Ordinarily this rationale is implicit, but occasionally, as in Theodore Abel's behavior, rather than as indications of the suficient conditions. Furthermore, the
(1945, p. 459) critique of psychological explanations of Nazism, the assumption is entire emphasis on such contingencies (whether the situation is ambiguous, whether
explicitly stated: "In dealing with social problems, we should always bear in mind there are sanctions) results from the reliance of the behavioral sciences on a particular
the dictum of . . . Durkheim, namely, 'social facts must be interpreted by social intellectual convention for the explanation of behavior-namely, the examination of
facts.' This rule teaches focusing our attention first of all on historical processes, concomitant variation. Within the logic of variable analysis, the analyst's energies
on aspects of social structure, on group mores, and on sentiments for clues to causal are expended in seeking antecedent factors which covary with the phenomena being
factors. The logical rider to this rule is that only if we fail to obtain a satisfactory explained. But whenever the phenomenon of concern to the analyst does not vary-
explanation in terms of social processes should we go beyond the realm of social facts for example, in unambiguous or sanctioned situations where political actors with
and invoke the aid of psychology." differing psychological properties behave in the same way-the impulse is to seek
For another formulation which seems-by the very act of definition-explicitly no further antecedents of the behavior in question.
to equate psychological analysis with the analysis of deviance, we may note Linton's Analysis of concomitant variation is an attractive intellectual mode because
(1945, p. 26) use of the term personality: "In general, all the individuals who occi~py of the leverage it provides for testing hypotheses about causality. To be reminded
a given position in the structure of a particular society will respond to many situations of the difficulties of explaining invariant behavior, one need only consider the
in very much the same way. That any one individual of such a group manifests this
response proves nothing about his personality except that he has normal learning 3 Changes in the ordering of the propositions and in a number of other parts of the
ability. His personal predispositions will be revealed not by his culturally patterned discussion between my earlier version (Greenstein, 1967) and my revised version (Green-
responses but by his deviations from the culture pattern." By this reasoning there stein, 1969, ch. 2 ) were introduced in an only partly successful effort to deal with precisely
would be little interest in an account of the personal orientations toward their politi- the point raised by Glad-a point that was pressed on me by Michael Lerner, who not
cal leaders of the vast majority of politically quiescent or "cooperative" American coincidentally shares Glad's interest in intensive single-case analysis.
 Fred I. Greenstein Political Psychology: A Pluralistic Universe

tortured history of attempts to account for that virtual cultural universal, the incest 1959; for a discussion of the importance 01 studying "common forms of thinking
taboo. Yet this example also cuts in the other direction. The incest taboo evidently and behavior" rather than simply focusing on "individual differences and their
reflects fundamental ingredients in the human condition, and there appears to be no effects," see the conclusion to Browning's discussion, in this volume, of Bardach's
justification for failing to study fundamental phenomena because one's analytic work on political entrepreneurship and the Newell-Shaw-Simon models of problem
modes are better adapted to studying that which is less fundamental. solving.)
I t is instructive to shift from the logic of variable analysis to the logic of
behavior, returning to the familiar formula just reviewed (and discussed in this Problems in the Study of Political Psychology
volume by Davies and Knutson as well as Smith) that behavior is a function of the
situation impinging on the person and the person's dispositions. We immediately The following remarks, stimulated by the fifteen previous chapters, are de-
recognize that all situational stimuli are mediated by actors with psychological signed to illustrate four general propositions about the nature of "real-world" political
properties. Much human behavior-including the great bulk of "conforming" and psychology and therefore about the intellectual requirements of the academic sub-
especially habitual behavior-is not even evoked by immediate situational stimuli. discipline of the same name. These are propositions which can be simply stated, but
And even within the logic of variable analysis, what is treated as a variable depends which have implications that run deep and that are very often unappreciated. Taken
upon the interests of the investigator, so that behavior viewed as invariant from one together they help to explain the slowness to take root of academic political psychol-
standpoint may vary in interesting ways from another. Further, for some purposes ogy and the many resistances political scientists have shown to accepting and drawing
the entire logic of variable analysis becomes strained, and it becomes more comfort- on the existing political psychology literature. The propositions:
able to take a case-by-case approach. Political analysts need to be interested in single 1. The connections between psychological phenomena and political behavior
(especially well-placed) individuals and in case studies in general (Eckstein, forth: and processes are complex and often indirect.
coming) ; and such interests tend to reduce the sway of variable analysis and to 2. The complexity and indirectness of connection are such that political
increase the desirability of studying conformity as well as deviance. psychologists need to invest considerable effort in conceptualizing their investigations
The notion that we need accounts of conforming as well as of nonconforming and framing their research in terms of explicit theories that take account of the
behavior is congenial to a school of sociological and anthropological thought which diverse types of predispositional and environmental antecedents of individual and
concerns itself with the psychological prerequisites of social systems. Writers on this collective behavior.
topic have been concerned with personality and role requirements. One way we can 3. The previous points art not meant as a counsel of vagueness ("It's all
think of any social or political system, they point out, is in terms of the sets of actions so complicated! So what can we say?"), but instead lead to a highly specific admoni-
that must be performed in order to maintain it in existence. What roles must be filled tion: "Search for contingent relationships-for interactive rather than direct effects."
in the system, and what are their requirements? People must be motivated to fill A successful search may entail finding structural effects in which, for example, a
these roles, whether by the satisfaction they receive from the role performance, by particular psychological factor has different properties and consequences in different
other gratifications they are accorded, or because of sanctions inflicted for perform- sociopolitical settings. I t also may be dependent on identifying psychological typol-
ance or nonperformance. Even where role performance depends -upon sanctions, ogies which differentiate among political actors in terms of the way they will respond
psychological understanding of why people respond to particular rewards and punish- to similar environmental stimuli. (Knutson quotes Allport's [1937, p. 3251 delightful
ments is important. For, as Spiro (1961, p. 102) notes, "Unless the members of a analogy: "The same heat that melts the butter hardens the egg.")
society have certain personality drives which can be reduced by acquiring positive, 4. Finally, I repeatedly revert to an issue not discussed in any detail in my
and avoiding negative, sanctions, it is unlikely that these sanctions would serve as own earlier writings on this topic, but very commonly raised by those political scien-
techniques of social control. . . . Social sanctions serve as techniques of social control tists whose experiences have provided them with detailed inside exposure to the
because they function as motivational variables." intricacies of government and politics and to the specialized norms of the political
According to this way of thinking, the task of the political psychologist would process. Most efforts at the formal application of the diverse theories and method-
be to understand the psychological underpinnings of the commonplace and normal in ologies of psychology (and psychiatry) to politics seem to the close political observer
political behavior, no less than the obscure and unexpected. This would be a compre- (and the political actor) to be insensitive to "political realities," including their
hensive undertaking and presumably would involve "the psychological 'mapping' or psychological components. By the same token, the common-sense psychological poli-
'census' of major subgroups and total societal populations" (Inkeles, 1963, p. 383). tical explanations by nonpsychologists tend to lack rigor and theoretical grounding.
Included would be analyses of the various roles and personnel in the political system, In short, there is a gap in the extant literature between political psychology and
asking questions such as: What kinds of people are selected to fill each role? How do political psychology.
the personal characteristics of the role incumbents affect their role performance? Davies' introductory chapter, "Origins and Dimensions of Political Psychol-
How does filling a role affect the role incumbent? Moreover, such analysis need not ogy," very substantially amplifies the points made above about the slowness with
assume that role requirements are static, or even that they cannot be substantially which political psychology has emerged from the broad matrix of psychological in-
altered by the role incumbents themselves. (On role and personality, see Levinson, quiry, the nature of that matrix, and the buzzing, blooming diversity of the political-
448 Fred I. Greenstein Political Psychology: A Pluralistic Universe

psychological inquiry to date. The three chapters that then follow under the heading shortcomings of personality theories that focus only on Freud's earlier concerns, the
"Basic Psychological Constructs"-Knutson on personality, Smith on attitude, and unconscious and the ego defenses. As Smith puts it, "a less restrictive conception of
Lane on belief-each parallel Davies' introduction in seeking to connect political persorlality is also possible, and . . . has much to recommend it" Nevertheless,
psychology with certain general problems of psychological inquiry. because frameworks like Smith's and Katz's for examining the motivational basis of
In the latter part of his chapter, Smith presents an up-to-date summary political orientations make it possible to consider both the inner depth and the
exposition of his exceptionally useful "map" for the analysis of personality and outer periphery of psychic dispositions, they are peculiarly useful for those who are
politics. This map, which is valuably complemented by Knutson's similar formulation less interested in the intradisciplinary issues within psychology than in being alert to
in her chapter on personality, serves as a convenient way to visualize the connections all of the facets of individual character that may affect political behavior.
among the diverse antecedents of human behavior: situational stimuli; opinions and The main part of Smith's chapter consists of a detailed account and evaluation
attitudes; the patterns of personality predispositions that underlie attitudes and of the several proliferating psychological literatures on attitudes and the conditions
opinions; personal socializing experiences that determine psychological predispositions; under which they arise, persist, and change. Smith draws on an impressive inventory
and overarching societal and political institutions, past and present, that provide the (derived by Zimbardo and Ebbeson from Karlins and Abelson) of forty propositions
wider context of individual political and social development of behavior. and subpropositions about the conditions under which attitudes are subject to in-
The panel numbered I11 in Smith's Figure 1 provides a key to the "functional fluence. Yet, as Smith points out, in spite of the vast effort and ingenuity that have
approach" to the study of attitudes-the approach that Smith and another con- gone into the study of attitudes, the number of nonobvious findings in the literature
tributor to this volume, Katz (1960), have been instrumental in developing. This is not astounding, and the less obvious findings are subject to interesting controversies
approach, which stresses that political and other social orientations may serve differ- and failures of replication. Moreover, Smith reminds us, there are substantial dis-
ent functions in the psychic economies of different individuals, encourages flexible crepancies, which Hovland noted a number of years ago, between experimental
multivariate conceptualization of political thought and behavior, in that it recognizes findings about opinion change and observations about the effects of influence efforts
the complexity and diversity of human motivation. Assumptions of the sort that in actual field contexts. These discrepancies, which probably help account for the
underlie the Smith and Katz formulations sensitize one to recognize a particularly continuing failure of psychologists studying attitudes and students of politics studying
politically important pair of complementary complexities : ( 1) Similar patterns of voting and public opinion to draw in detail on each other's work, provide a first
political orientation and behavior can perform different psychological functions for illustration of the final two propositions enunciated at the start of the section-that
different individuals (for instance, one politician may, in Smith's terminology, be connections between psychology and politics are likely to be complexly interactive
basing his political behavior on object-appraisal needs, another on ego-defensive (and therefore not easily replicated from one context to the next) and that "pure"
needs). ( 2 ) Essentially the same pattern of motivational needs may exhibit itself in psychological theories may often be inapplicable to "applied" political realities.
different behavior patterns (as in left- and right-wing authoritarianism). Knutson's treatment of the study of personality and politics contrasts usefully
Smith notes that his map and the functional approach more generally have with the extended chain of commentary that I presented on this topic (Greenstein,
had greater influence in political science than in his own discipline. This may be 1969). She delves in considerable detail into the substance of diverse personality
appropriate. Psychologists are able to practice a division of labor that leads some theories and the concrete findings of personality research, whereas I attempted to ask
investigators to focus on affect and the emotions, some on social conformity and non- in a general way how questions about personality and politics can be clearly phrased,
cor,formity, and some on cognitive patterns (the three general functions that opinions what kinds of questions typically arise in personality-and-politics inquiry, and what
may serve for the personality) ; but political scientists, precisely because theirs is an might be necessary in establishing a systematic program of inquiry around the diverse
applied specialty and they are interested in concrete behavior, need to consider the types of questions or analytic tasks.
whole panoply of psychic possibilities. Knutson's discussion of the differences between trait personality theories and
Since political scientists are inevitably forced to grapple with the complexity holistic theories does not describe an unbridgeable gulf, since examination of the
of real-world political behavior and its many bewildering complications, they are pattern of presenting traits of an individual or type in the light of some appropriate
lilrely to find it difficult to inform their inquiries in a fruitful way with psychologies personality theory will ordinarily lead to more holistic formulations about overall
that see humanity in terms of a restricted range of traits. Hall and Lindzey (1957) personality dynamics and about the antecedents of these dynamics. (Compare Green-
in their summary of diverse personality theories remark that psychoanalytic psychol- 1969, ch. 3.) Nevertheless, Knutson does point to a genuine distinction to
ogy enables one to see the social actor as "a full-bodied individual living partly in a which students of political psychology need to be sensitive. In what, if any, kind of
world of reality and partly in a world of make-believe, beset by conflicts and inner formal personality theory should political analysts seek to ground their work? As
contradictions, yet capable of rational thought and action, moved by forces of which Hall and Lindzey (1957, p. 557) wisely suggest, there is more to be said for immers-
he has little knowledge and by aspirations which are beyond his reach, by turn ing oneself in a theory one finds congenial and then using it skeptically and eclecti-
confused and clear-headed, frustrated and satisfied, hopeful and despairing, selfish cally to inform one's work than in seeking to synthesize a master theory. My own
and in short, a complex human beingm (p. 72). predilection, following from the observation made above about the merits of Smith's
Both Smith and Katz, in their chapters in this volume, cogently point out the rather vision of the dimensions of psychic phenomena, is for a liolistic
 Fred I. Greenstein Political Psychology: A Pluralistic Universe 451

rather than a trait approach. Trait psychologies are of use for certain things that 1969)-apparently because most of the shapers of the political socialization literature
students of political psychology do, notably correlational analyses. But if one seeks have been interested in learning about the advent of political orientations with an
to examine the actual political behavior of actors in context, it becomes necessary ultimate view to contributing to the understanding d how political systems function
to think of personal qualities as more than disaggregated bundles of traits. The rather than with an intrinsic interest in human development. Consequently, they
political scientist resembles the clinician in that each needs to make a prediction have tended to pay attention to psychological structures such as party identification-
about this patient, or this political actor, in this concrete reality context. (Deliberately which have great political consequences but do not play a sufficient part in human
simplified diagnoses can sometimes be analytically suggestive, however.) See Payne development to interest students of development in general-and to ignore more
(1972) and Payne and Woshinsky (1972) for an interesting effort to apply a single- "fundamental" aspects of development. (For an expansion of this point, see Green-
trait motivational typology to political analysis. stein, 1970a.)
Lane's chapter on political belief in this volume goes over many of the same Following Niemi's chapter, the two remaining chapters in the section on
issues as Smith's psychology-grounded chapter on political attitudes; but for Lane, "Forming and Maintaining Stable OrientationsJ' deal with two of the most extensive
the political scientist, the adjective political is at least as important as the noun. The of the several literatures in which a psychological or social-psychological typology is
mosaic of propositions listed by Smith was derived from the innumerable studies by applied to political behavior-the authoritarian and the anomie-alienation literatures.
psychologists who (precisely as Littman indicates) are indifferent whether their Of the two, and of typologies in general, the authoritarian construct is by far the
subject matter is explicitly political. Psychologists have studied opinions in mani- most widespread psychological classification scheme to have been applied to political
festly political contexts, but they seem as eager to look at attitudes toward tooth psychology. Christie and Cook (1958) were able to list 260 bibliographical references
brushing (Janis and Feshbach, 1953) or even toward the eating of grasshoppers relating to the authoritarian personality through 1956, only six years after publication
(Smith, 1961). What emerges from Lane's analysis is an account of belief in which of T h e Authoritarian Personality. This inventory preceded the point at which the
both content of belief and the sociopolitical context in which belief arises are im- analysis of response sets in F scale questionnaires became a booming source of articles
portant. Lane's contention that belief systems are necessary to enable political and in psychology journals. A 1967 literature review required a short monograph (Kirscht
other social actors to thread their ways through life reaches back to the still im- and Dillehay). In spite of the many critiques of the authoritarian construct (perhaps
pressive discussion by Walter Lippmann (1922) of how "pictures inside our heads of the most easily accessible and well rounded of them being the chapter on authori-
the world beyond . . . reach" mediate human encounters with reality. tarianism in Roger Brown's [I9651 social psychology textbook) interest in authoritar-
Lane also draws extensively on the sociology-of-knowledge literature, again ian character and personality endures. In 1972 numerous articles (for example,
illustrating the way in which the requirements of ~oliticalanalysis seem often to lead Gabennesch, Simpson, Herzon, Kohn, Wright, Thompson and Michel, and Rogh-
political scientists, even psychologically oriented political scientists like Lane (see mann and Sodeur) using the term authoritarian and addressing themselves directly
Lane, 195913, 1962, 1972), to rely more on sociology than on psychology. In a sense, to issues involved in this literature continued to appear. Furthermore, the issue of
Mannheim's (1949) classic work on ideology and utopia is directly cognate for authoritarianism is one which from the beginning extended itself well beyond the
sociology to the once-thriving industry of attitude research within psychology. domain of formal government politics into studies of authoritarian child-rearing
Niemi, in his succinct summary of the political socialization literature that has practices, religious habits, approaches to the perception and treatment of mental
burgeoned forth so extraordinarily since the introduction of that term by Hyman illness, and so forth.
(1959), reports on the notable expansion of knowledge about the events occurring Sanford, whose chapter is informed by his research with Adorno, Frenkel-
in panel I1 of Smith's map--namely, the social environment as a context for the Brunswik, and Levinson in the team that produced The Authoritarian Personality
development of personality and the acquisition of attitudes. Not least in this accom- (1950), notes that the phenomenon of psychological authoritarianism was sufficiently
~lishmenthas been the major national survey of American high school seniors with evident in the daily life of the Germany of the 1920s and 1930s to lead the group
which Niemi has been associated (see, for example, Jennings and Niemi, 1968b). of scholars in the Institute for Social Research to emigrate in advance of active Nazi
For the most part, political socialization research has dealt with the anteced- persecution. I t seems to me that at base the interest in authoritarianism, despite the
ents of what Almond and Verba (1963) call "subject" and "citizen" orientations: critiques, may persist because once sensitized to the notion that some individuals
nonleaders' dispositions vis-A-vis the political order and their own participation have reciprocal needs to dominate "inferiors" and to kowtow to c'superiors," we all
therein. There are unfortunately still few bridges between the socialization and the recognize instances of this type of individual in our everyday environments.
political recruitment literatures; although there is much fragmentary information In Personality and Politics I quote Fielding's assertion (in 1747) that "slaves
about the in-role socialization experiences of political leaders, that information is and flatterers . . . exact the same taxes on all below them which they themselves
not yet systematically summarized. In addition, the issues about the fundamental pay to all above them." Another striking example of the long-standing recognition in
properties of belief and personality dealt with in the Smith, Knutson, and Lane the conventional wisdom of the culture of phenomena described in T h e Authoritarian
chapters obviously have important preadult antecedents, and these have not been Personality is Samuel Butler's description of the Victorian clergyman Theobald
dealt with in the political socialization literature. Nor have most students of political Pontifex in The W a y of All Flesh (published in 1903). Sanford in this volume de-
socialization drawn explicitly on developmental psychology (but see Merelman, scribes one of the defining traits of the authoritarian syndrome, "authoritarian
452 Fred I. Greenstein Political Psychology: A Pluralistic Universe 453
aggression," as the ego-defensive strategy of one who "has identified himself with 1972 presidential campaign, which included such ingredients of sociomoral polariza-
the ingroup authorities of his childhood and found in the tendency to punish wrong- tion as a defense of "Gay Liberation" by a homosexual delegate to the Democratic
doing in others a safe and fairly well-sanctioned outlet for his aggression." Butler National Convention and the defection of many middle-class ethnic and working-
presents an episode in which Theobald, already in "a bad temper," is offended at class voters from the traditional Democratic coalition, at least partly in reaction to
his young son's pronunciation of "come" as "tum." Butler describes Theobald's effort life-style issues. Hyman and Sheatsley convincingly showed in their i~ifluential1954
to instruct his young son to give up the baby-talk pronunciation of the word. critique of The Authoritarian Personality that many of the values and attitudes
associated with the authoritarian "symptomatology" were not necessarily the outward
"Ernest," said Theobald, from the armchair in front of the fire, manifestations of the psychodynamics posited by the Berkeley investigators (that is,
where he was sitting with his hands folded before him, "don't you think it reaction formations to repressed antiauthority impulses), but instead were conven-
would be very nice if you were to say 'come' like other people, instead of tionally learned working-class cognitions. But, as Christie (1954) argued, it still
'tum' ?" seemed likely that some individuals who exhibit authoritarian presenting patterns,
"I do say tum," replied Ernest, meaning that he had said "come"
perhaps especially those of middle-class backgrounds, did so because of ego-defensive
... psychodynamic antecedents. An interesting possibility, in view of the evidence of a
Theobald noticed that he was being contradicted in a moment . . .
"NO, Ernest, you don't," he said, "you say nothing of the kind, you migration of rigid child-rearing practices "downward" in the social structure
say 'turn,' not 'come.' Now say 'come' after me, as I do." (Bronfenbrenner, 1958) and in view of the affluent backgrounds of the first wave
"Tum," said Ernest, at once, "is that better?" I have no doubt he of radical student activists in the 1960s (Hacks, 1967), is that new-style working-
thought it was, but it was not. class authoritarianism may have a stronger ego-defensive component than old-style
"Now, Ernest, you are not taking pains, you are not trying as you working-class authoritarianism.
ought to do . . . Smith's map, with its built-in assumption that outward behavioral patterns
The boy remained silent for a few seconds and then said "tum" may perform different inner functions for different individuals, is tailor-made for
again. clarifying the requirements of establishing whether and to what degree authoritarian
I laughed, but Theobald turned to me impatiently and said, patterns have a basis in learned cognitions (Smith's "object appraisal" category) or
"Please do not laugh, Overton, it will make the boy think it does not mat- in ego-defensive needs. Only a handful of the countless studies of authoritarianism-
ter, and it matters a great deal;" then turning to Ernest he said, "Now,
a notable example being the ingenious experimental study of authoritarianism and
Ernest, I will give you one more chance, and if you don't say 'come,' I
repression by Kogan (1956)-provided solid evidence of the original explanation
shall know that you are self-willed and naughty."
He looked very angry, and a shade came over Ernest's face, like that authoritarianism is in essence an outward manifestation of the play of mecha-
that which comes upon the face of a puppy when it is being scolded with- nisms of ego defense. But the cognitivists-for example, Selznick and Steinberg
out understanding why. The child saw well what was coming now, was (1969), on whose work Sanford comments so generously-seem to me often to err
frightened, and, of course, said "tum" once more. in the opposite direction from the original Berkeley group; that is, they fail to con-
"Very well, Ernest," said his father, catching him angrily by the ceive of the possibility that both kinds of authoritarianism-ego-defensive as well as
shoulder. "I have done my best to save you, but if you will have it so, you cognitive-may exist and that the initial conceptualization of research and develop-
will," and he lugged the little wretch, crying by anticipation, out of the ment of instrumentation should take this into account. Furthermore, it is not just in
room. A few minutes more and we could hear screams coming from the the authoritarianism literature that mechanical application of a psychodiagnostic
dining-room, and knew that poor Ernest was being beaten. instrument to members of diverse social groups appears to have misclassified the
conventional cognitive patterns of particular social groups as psychodynamic sympto-
Butler is worth quoting at this length both to illustrate the social reality of the matology. Gynther (1972) argues that this is precisely what occurs when the MMPI
constellation of traits described in the authoritarian construct and for the period- scores of American blacks are interpreted in terms of white norms.
piece tone of its specific details. The latter helps illustrate Sanford's interesting Yinger's chapter on anomie and alienation deals with the array of typologies
observation that, since the pioneering study he and his Berkeley associates conducted within sociology that may well earn the-most-tangled-empirical-literature award,
in the early post-World War I1 years, there does appear to have been genuine which would no doubt be earned by the authoritarianizm literature within psychology.
characterological change in society. At the same time, it would be premature to There have been a. very large number of empirical ventures, psychometric scales, and
compose an obituary for the original typology, as can be seen from Rupert Wilkin- statistical data-reduction procedures, in which responses to questionnaire items de-
son's recent (1972) use of the typology for a broad-gauged survey of a remarkably signed to measure the Durkheim- and Marx-derived concepts of anomie and aliena-
diverse array of political trends. tion were intercorrelated and sometimes also used in conjunction with items from
Sanford's references to such features of the classical authoritarian syndrome scales designed to measure authoritarianism. Yet th:s literature remains far from
as "rigid adherence" to "middle-class values'' and "moral indignation with respect having reached a "final" empirical closure.
to the behavior of other people" seem remarkably undated in the context of the Both the anomie-alienation literature and the authoritarianism literature have
 Political Psychology: A Pluralistic Universe 455
454 Fred I. Greenstein
placed, role-defining leaders-it seems appropriate to think of the individual as a
been complexly extended, controversial, and plagued by the formula "failure to
separate entity acting on the polity.
replicate" because of a pair of complementary problems. The first of these problems
In dealing with the motivational underpinnings of leadership styles and
is insufficient awarenesj. that relationships between psychological dispositions and
behavior, one can usefully return for still another application of the left side of the
political behavior ought i n principle to be highly variable, depending upon further
personality and personal dispositions panel (111) in Smith's map, which, like Katz's
aspects of the social and psychological context. Yinger has been a continuing spokes-
own seminal work, suggests that attitudes and other political behavior may to varying
man for the sort of contextual social-psychological analysis, which is also called for degrees serve diverse motives. A number of motivational classifications are possible.
by Smith's mapping formulation and in Knutson's personality chapter. While Yinger Smith's, which resembles Freud's tripartite conception of psychic structure, dis-
does not systematically summarize his own general conceptualization of the overall tinguishes the rational-cognitive ("object appraisal"), the social ("mediation of self-
strategies appropriate for studying psychology and sociopolitical structure in the other relationships"), and the emotions-management ("ego defense and externaliza-
present volume (for his earlier statements, see Yinger, 1963, 1965), his view about tion") needs of individuals.
the need for systematic simultaneous attention of psychological and sociological Each of Smith's functions can be linked with a leadership type. His category
factors informs his treatment of the checkered bodies of literature he reviews and "mediation of self-other relationships" (response to social influences) provides a use-
helps to clarify. As he comments, "There is not a single empirical investigation of ful lever for understanding the political behavior of many-perhaps the great bulk
anomie and deviant behavior . . . that has systematically included simultaneous of all-democratic leaders. Many politicians are likely to be preselected and then
aspects of collectivity, subgroup, and individual attributes in relation to deviant further socialized by their roles into a psychological stance in which sensitivity to
behavior." Eliminate the word deviant and substitute political behavior and we have others and their demands provides a principal motivational engine for the leader's
a somewhat, but fortunately not completely, accurate characterization of the general continuing activities. This sort of leader, who resembles Riesman's (1950) other-
state of systematic attention to the interaction of the social and the psychological directed type, is likely to come forth with such formulas as the following assertion,
in political behavior research. quoted by V. 0. Key (1958, p. 497n) : "I have outlined my views without equivoca-
The second problem is that the anomie-alienation literature and the authori- tion and those views will be my continued views unless the people at meetings and
tarianism literature have suffered from shortcomings in instrumentation. As Sanford through cards and letters give evidence of the voters' opposition." To the degree that
suggests, there has been an inordinate reliance on standard, fixed-choice attitude a political or other actor's behavior is motivated by conformity needs, that behavior,
and personality scale measures (whether or not corrected for response set) to the as Goldhamer (1950, p. 353) has pointed out, is likely to "have a somewhat fortui-
exclusion of other measurement devices, such as controlled clinical procedures, tous character in relation to [other aspects of] the personality and be dependent
projective devices of the sort discussed by Knutson in this volume, and unobtrusive largely on attendant situational factors."
field observation procedures. Too often, investigators have "operationalized" phe- Such an other-directed leadership style resembles one of the leadership types
nomena of the greatest complexity via crude measures of the independent variables; in Barber's interesting schema of leadership styles developed in his study of Connecti-
they have failed to recognize that a multitude of dispositional and environmental cut state legislators (Barber, 1965) and expanded on in his typological analysis of
factors may contribute to a single item of behavior, that dfferent types of actors will twentieth-century American presidents (Barber, 1972a). The style in question is
respond differently to similar environmental influences, and that even slightly differ- called "the spectator" in Barber's first book and the "passive-positive" in his 1972
ent environmental contexts may yield quite different patterns of behavior. book. Among American presidents, William Howard Taft and Warren G. Harding
For a good example of a study that breaks out of the standard psychometric appear to fit this category. Barber's classification schema is based on a dichotomous
mode, the reader should examine Smith's (1965) use of Q-sort techniques to code cross-classification of politicians in terms of two variables-their level of political
psychiatric interviews in a study of authoritarianism. Of the many studies over the activity (active-passive) and the general emotional tone of their stance toward life
years in which personality inventories and attitude scales are intercorrelated, perhaps (positive-negative) .
the most carefully designed and sensitively analyzed is the body of data on American But leaders are not all chameleonlike passive-positives. Indeed, as Lewis
citizens and their leaders collected by Herbert McClosky in the late 1950s and Dexter (1969, pp. 159-160) has noted in a masterful treatment of the social psy-
subsequently reported by McClosky and his associates (McClosky, 1958; McClosky chology of the relationships among representatives and their constituents, even when
and Schaar, 1965; McClosky, 1967a; DiPalma and McClosky, 1970; Sniderman and a politician seeks to mirror his constituency, the shapelessness of public demands may
Citrin, 1971). make it impossible for him not to lead. Independence on the part of leaders also is
Katz's chapter on patterns of leadership is the first of three chapters that often imposed by features of the interpersonal networks that surround them, as when
appear under the heading "The Nexus of Individual and Polity." Literally, of course, a congressman who consciously seeks to subordinate himself to constituency opinion
there is no "nexus" between individuals and polities in that the latter are composed encounters in the course of attempting to do so the very constituency activists who
of the former. For this reason, my own predilection is to borrow the language used take their own cues from the congressman himself. Thus, even the political leader
to connect microeconomic and macroeconomic phenomena and to speak of "aggrega- with conformist social needs in fact finds politics a kind of Rorschach.
tion" (Greenstein, 1969). Nevertheless, in the case of leaders--especially pivotally
456 Fred I. Greenstein Political Psychology: A Pluralistic Universe 457
Smith's two other functional categories also can serve as psychological modal- This criticism of efforts to find the personality of the political leader is in no
ities around which distinctive leadership styles cluster, meshing suggestively with two way inconsistent with Lasswell's (1930) famous formula asserting that "Politicd
other of the Barber categories. Smith's "ego defense and externalization" category Man" displaces private motives on public objects, rationalizing them in the public
seems to converge with Barber's "advertiser" or "active-negative" type (Wilson, interest. Lasswell's "Political Man" refers both to more and to less than all actors in
Hoover, Lyndon Johnson, Nixon), presidents who Barber feels were prone to "act what society conventionally labels the political arena; he has in mind power seekers,
out" inner tensions in their behavior in office. And Smith's "object appraisal" cate- excluding politicians not oriented to power and including power seekers in the
gory parallels Barber's "active-positive" type (Franklin Roosevelt, Truman, Ken- church, business, and other areas. From his earliest work on personality and politics
nedy), who takes a generally rational approach to appraising and coping with the to the present, it has been a central Lasswellian tenet that the tatterdemalion array
objects he encounters in the political arena. Barber's fourth type, "the reluctant" or of individuals who happen to become inducted into the diverse array of roles con-
"passive-negative," is perhaps less likely than the others to reflect distinctive person- stituting the formal political process are unlikely to have common psychologica1
ality needs in that this type of individual (for instance, Coolidge or Eisenhower) is properties. Therefore, Lasswell has argued, the political psychologist needs to identify
typically someone who dislikes politics but is drafted to political office because of his
functionally distinctive political roles (regardless of their formal institutional nexus)
esteem in the c o m m ~ n i t y . ~
and then to seek distinctive psychological concomitants (Lasswell, 1930, ch. 4;
Barber's T h e Lawmakers (1965) concludes (pp. 219-227) with an important
Lasswell, 1948, ch. 1 ) . One of the few studies to take this admonition seriously is
theoretical attack on the general problem of whether political leaders as a class have the fascinating analysis of politicians in diverse functional roles and political settings
distinctive personality constellations and, if so, what kind. Barber's formulation is an by Browning and Jacob (1964). (For the empirical delineation of a political per-
exemplary instance of contextual treatment of social and psychological variables. The
sonality type reminiscent of Lasswell's Political Man, see Browning, 1968b.)
political career, he points out, is negatively valued by most Americans. Politics is a
Barber's treatment of presidential personalities and political styles provides a
"late-entry" occupation, as well. Thus, the individual who shifts from political acti- valuable extension of Richard Neustadt's penetrating analysis of the role require-
vism in the general population to more or lejs ful!-time political activity needs both ments of the modern American presidency, discussed by Katz in his survey of
to justify to himself a career change and to explain to others his entry into a some- "patterns of leadership." Katz is a psychologist who hews more closely to analysis of
what disreputable activity. These external social pressures increase the motivational how psychological processes exhibit themselves in specific institutional contexts than
threshold for political participation, Barber argues. Only individuals with excep-
the archetypically "socially indifferent" psychologist referred to by Richard Littman.
tionally powerful needs to do so will take the necessary. step. This motivation n a y
(See his editorial in Volume 1, Number 1, of the Journal of Social a n d Personality
come from compensatory needs to propitiate the psychic deficiencies which, in Barber's Psychology [1965], urging more psychological research that takes realistic account
opinion, commonly motivate his advertisers, spectators, and in some cases his
of social contexts.) Katz's sensitivity to specific institutional configurations and their
reluctants; or it may come from the need for self-fulfillment on the part of individuals
implications for political roles leads him to summarize sympathetically Neustadt's
with special emotional and cognitive strengths. (Compare the application of Maslow's analysis of how Franklin Roosevelt deliberately snarled the lines of authority among
crself-actualization'' formulation to politics by Davies, 1963, and Knutson, 1972a.)
his subordinates in order to maintain presidential control of policy making. A less
Thus, Barber's analysis explicitly assumes the diversity of motivation of different
politically astute psychologist might well merely attribute such a presidential leader-
types of political actors and strongly suggests that efforts such as DiRenzo's (1967)
ship practice to "poor administration."
application of a Rokeach dogmatism scale to Italian legislators, with a view to
In spite of Katz's political perspicacity, a comparison of Icatz's references to
"distinguishing a general personality type that constitutes the professional politician"
the American presidency and Neustadt's discussion (which, while not stated in a
(p. 6 ) , are doomed to failure. (For an expansion of this point, see Greenstein,
formalized manner, is susceptible to more or less formal summary) provides a further
1970b.) The same criticism applies to other efforts (for example, Hennessy, 1959;
illustration of proposition 4 at the beginning of this section, dealing with the dif-
Schwartz, 1969) to establish a unimodal political-leadership personality type.
ferences between psychologists' and political scientists' perspectives on politics.
Neustadt discusses many more aspects of the political psychology of presidents that
4 Barber's work on presidential character is discussed in this volume by both Knutson
and Katz. I agree with them about the desirability of further conceptional clarification and are presidency-specific than does Katz; and Katz is far more likely than Neustadt
operationalization of the variables that define his typology. For example, the two dichoto- to deal with the psychology of leadership as a general phenomenon.
mized variables of affect and activity might be treated as continuous variables, and the Apart from practicing political science as a vocation, through his service with
affect variable might be so conceived as to take account of ambivalence and mood swings. Truman and Kennedy, Neustadt has had extensive participant-observer experience
Knutson and Katz seem to me to underestimate the degree to which Barber's formulations with the tribal mores along the Potomac. This experience informs his delineation of
have already received some quantitative grounding in the correlational data he presents in
The Lawmakers and the carefully ordered case-material presentations in both The Law- I
a presidential role (or rather a range of role possibilities) which is in many respects
makers and The Presidential Character. It should be noted that Barber redefined his affect sui generis and therefore only partially susceptible to analysis in terms of generalized
variable from "affect toward role" to "general outlook on life" in his presidential study, organization and leadership theories. The modern president has enormous formal
presumably because all individuals who make their way through the recruitment process powers of the sort that Neustadt metaphorically describes as "clerical." But formal
to the presidency are at least somewhat positive toward that role. powers do not automatically entail leadership in the sense of a distinctive presidential
458 Fred I. Greenstein Political Psychology: A Pluralistic Universe 459

impact on his administration. Furthermore, while the president can i n principle act ship plays ego-defensive functions. The enigmatic Abraham Lincoln would also have
unilaterally to tell his ccsubordinates" what to do, in most instances peremptory at least intermittently fit into this category.
command (as practiced, for example, when Truman dismissed MacArthur) is so Katz's concern with leadership, especially as amplified here to take account
politically costly as to be only a last resort. of leaders' styles and personalities, is closely paralleled by Glad's chapter on psycho-
The president's command power within the government is limited, not only biography, which I shall therefore refer to here rather than in connection with the
because of the constitutional separation of powers but also because the fragmentation other four essays in the methodological section. There is a well-worn debate within
of the American political system is inconsistent with organizing even the administra- psychology about the problem, in Gordon Allport's (1962) phrase, of "the general
tive branch (much less the entire government) on a classically pyramidal basis. The and the unique in psychological science." For some psychologists-more so for clini-
model of hierarchy applies better to some aspects of the president's jurisdiction than cians than for the seekers of general laws referred to by Littman-personality analysis
to others, but in general it is useful to think of the president as being surrounded is not satisfactory unless it is capable of application to specific individuals in all of
by a series of semiautonomous individuals and entities, among whom there are many their personal distinctiveness. For others, the individual is useful at best as a source
reciprocal interdependencies, rather than as the head of any "subordinates" other of illustrative material in presentations of data designed to support universal generali-
than those on his official staff. Both the norms of national politics and the realities zations. Whatever may be appropriate for psychology, the political analysis cannot
of the reciprocities make the president a bargainer and a persuader, to the degree content itself with the strategy of ignoring the concrete particularity of the individual
that his own self-definition of his role leads him to want to accomplish ends for actor. The analyst of the modern, institutionalized American presidency, for example,
which the cooperation of others is necessary-that is, to the degree that he has any has only six cases at his disposal-Franklin Roosevelt, Truman, Eisenhower, Kennedy,
goals other than that of being inactive. Johnson, and Nixon.
An ideal-typical Neustadt president-one who is seeking to leave a mark on Even though some political analysts may seek to transcend individual cases,
public policy and has the professionalism to know what a president can and cannot the society's needs for understanding are such that other political analysts (whatever
do in the American political context-behaves in ways which, if duplicated in the their training and formal connections) are bound to look at the interaction of role
daily behavior of an individual not responding to similar situational imperatives, and personality in these five individuals. This need, of course, goes well beyond the
would seem to reflect excessive, perhaps even pathological, vanity and opportunism. American presidency, as can be seen by the intense critical interest with which new
First, he exhibits a hypertrophy of a certain kind of rather coolly aloof empathy, psychological interpretations of historically pivotal figures are received-for example,
leading him constantly to ask of other political actors, "What does he want?" "How Lacy Baldwin Smith (1971) on Henry VIII and Langer's belatedly published (1972)
can I arrange matters so .it is in his interest to want to do what I want him to do?" psychodiagnosis of Hitler (conducted for the American government during World
Second, such a president cultivates a kind of controlled narcissism, recognizing the War 11), as well as many other of the psychobiographies referred to in Glad's
great instrumental importance of having a formidable "professional reputation"- chapter.
that is, a reputation among his significant others in the Washington policy-making Stating it generally, analyses of the psychological properties of individual
community as someone who systematically and consistently is prepared to reward his political actors are necessary to the degree that two conditions obtain. First, the
allies and punish his adversaries to advance his policy goals. Finally, the president actor is lodged i n a role which leaves some room for the role incumbent's personal
who practices Neustadt's how-to-succeed-in-the-White-House precepts will, as Katz qualities to contribute to uariance i n role performance. Some roles are sufficiently
notes, often deliberately set his associates at counterpurposes (even if he does not open-ended and encumbered as to resemble the Rorschach inkblot. Such roles in-
actually play them off against one another in Franklin Roosevelt fashion) and will evitably exhibit the personal qualities of role incumbents. Other roles are so hedged
in general comport himself in ways that maximize his own decision-controlling by external pressures-a limiting case might be the galley slave-that individual role
options. (On the social psychology of presidents' relationships with their advisers, incumbents are largely substitutable for one another. (See the discussion of "actor
see George, 1971, and Janis, 1972.) dispensability" in Greenstein, 1969, ch. 2, as well as the remarks in this chapter on
From the standpoint of Barber's political-leadership characterology, a Neu- necessary versus sufficient conditions for using psychological data.)
stadtian president cannot by definition be one of Barber's two passive types. One Among contemporary leadership roles, the official positions of leaders in
wants, of course, to assume that an effective president "ought" to fall into Barber's totalitarian nations and in developing nations in which the political forces impinging
rational, problem-solving-oriented, "active-positive" category. But, as Barber (1972b) upon government are loosely institutionalized probably leave more room for personal
pointed out in his pre-1972 election attempt to assess Nixon and McGovern in terms idiosyncrasy to come into play than do the roles of leaders of industrialized democ-
of his typology, the active-positive individual-probably as a result of his own relative racies with well-institutionalized interest groups. Leadership roles in the new nations
lack of anxiety-tends to lack an empathic "resonance" with the fears and hopes appear to have some of the same malleability that was evident in the premodern
of others. And, in fact, two of the three twentieth-century presidents who presided West. Norman Cantor (1967, p. 136) says of the kingship in Angevin England:
over the most substantial policy innovations-Wilson and Johnson (but not the third, "Royal government was always heavily dependent on the king's personality, and
Roosevelt)-fall into Barber's category of the active-negative leader for whom leader- without an intelligent and energetic monarch on the throne the effectiveness of royal
460 Fred I. Greenstein Political Psychology: A Pluralistic Universe 461

administration and law was bound to be diminished. On the other hand, a king of man and Bloom's valuable discussion of "Assumptive Frameworks in International
great qualities could have a profound impact on the expansion of royal authority." Politics." By "assumptive frameworks" they mean the complex of underlying assump-
Similarly, in contemporary preindustrial settings, the personality of a Sukarno, an tions in terms of which "national decision makers, opinion leaders, or involved
Nkruma, or a Nasser appears to leave a vivid imprint on role performance. A striking citizens . . . formulate specific opinions and arrive at decisions on issues of interna-
recent illustration of the lability of political roles in the new nations is provided by tional politics." Thus, their discussion is in part about the kind of basic psychological
the behavior of General Idi Amin of Uganda, who expelled his nation's 50,000 issue discussed by Smith and Lane in their chapters on attitudes and beliefs, rather
Asian residents after reporting that God had instructed him to do so in a dream and than about the mixture of psychological and structural concerns that would be in-
whom one journalist (Munnion, 1972) described as a "heavyweight boxer, flyweight volved in linking data on individuals aggregatively into full-fledged accounts of
philosopher, and seemingly punch-drunk president [who has] transformed life in his institutional functioning. The Kelman-Bloom discussion does differ from those
. . . state into a nightmare of terror for inhabitants of all hues." of Smith and Lane, however, in its emphasis on decision makers and the influential
Tucker (1965) comments on the profound effect the personality of a totali- minority outside of government rather than on the full population (mass orientations
tarian leader can have on both the leadership style in a political system and on the toward foreign policy are notoriously weakly grounded and labile) and in a further
overall performance of the system, not as in the Third World case because of the respect: Kelman and Bloom place a particular emphasis on assurnptive frameworks
vacuum of forces surrounding the leader, but rather because the bureaucratic struc- which they feel are inadequate in that they "impose unnecessary limits on the range
ture of the totalitarian state serves "as a conduit of the dictatorial psychology." If, of alternatives considered in the formulation and execution of policy."
on the other hand, we look at the institutionalization of leadership in industrialized While the treatment of assumptive frameworks that makes up the major
democracies, it may seem at first glance as if the "role requirements" virtually totally part of Kelman and Bloom's chapter is not addressed to locating assumptive frame-
mold the behavior of leaders. We note, for example, that various American presidents works in their sociopolitical contexts, Kelman and Bloom make clear, in the literature
-for instance, Trllman (Rogow and Lasswell, 1963) and Nixon-have behaved in summary with which their chapter begins, that they reject the emphasis of those
the White House in ways not generally expected on the basis of their previous careers, writers on international conflict, from William James (1910) on, who assume that
and that politicians of different ideological persuasions tend to converge in their the causes of war are exclusively psychological. Echoing the other critics of the war-is-
political programs in a two-party system (Downs, 1957, ch. 8 ) . Nevertheless, even in-the-minds-of-men school of thought, such as Waltz (1959), Kelman and Bloom
leaders of developed democracies, whose roles force somewhat similar behavior upon comment that they hope to see developed "not a complete and self-contained social-
them, may vary rather strikingly in the conduct of their duties. One need only com- psychological theory of international relations, but a general theory of political
pare the styles in office of American presidents to recognize the many aspects of behavior at the system level in which social-psychological factors play a part, once
personal variation possible even within well-institutionalized roles (Hargrove, 1966; the points in the process at which they are applicable have been properly identified."
Barber, 1972a) .
Reference to the American president, who used to be described in a multiply To focus attention on the individual actor does not imply that such an
ambiguous locution as "the most powerful democratic leader in the world," brings analysis can provide a total picture of international political processes.
us to the second factor which makes it incumbent upon political analysts to attend Certainly, every participant in both the national and international systems,
to an individual actor's psychological properties. T h e greater the oonsequences of the at whatever level he makes his contribution, is heavily influenced by the
actor's behavior, the more necessary it is to study that leader intensively. Even if only political, economic, and social realities that surround him, and acts within
nuances of personal variation are evident from role incumbent to role incumbent, the constraints and procedures of institutional structures. Moreover, once
the choices faced by political leaders are often balanced so delicately that the added an individual actor has developed a point of view, made a decision, or
personal feather-and it may be more than a feather--can be decisive. The modern taken a politically relevant action, he has contributed only one of a large
American president is in some respects (especially where congressional approval is number and variety of inputs out of which a political outcome is shaped.
necessary) extraordinarily encumbered. But he also is a human lynch pin in a There is no substitute for analysis of the structural mechanisms and sys-
thermonuclear-weapons equilibrium-which, if disturbed, could wipe out much of temic processes that select among the myriad of individual inputs and
the earth's population in a matter of hours. Short of the thermonuclear-weapons- mold them into final policy decisions.
launching powers of the president-which seem fortunately to be unusable-are his
capacities for exercising only imperfectly restrained powers for prosecuting unde- Both the foregoing passage by Kelman and Bloom and Kelman's other writ-
clared limited wars. For a somewhat loosely argued version of a thesis that can be ings (for instance, Kelman, 1965, pp. 565-607; Kelman, 1970) underline the point
defended more rigorously on Lyndon Johnson's personal responsibilities for the that psychological data are necessary (as the sociologizers fail to appreciate) but
Vietnam escalation, see Wicker (1968). not su.fficient (as the psychologizers fail to appreciate) to understand international
This reminder of the impact of American presidential behavior on the sta- politics. Yet, in a way that parallels the differences between Katz's and Neustadt's
bility of the international political arena provides an appropriate transition to Kel- treatments of the presidency, the Kelman-Bloom discussion does not exhibit the same
462 Fred I. Greenstein Political Psychology: A Pluralistic Universe 463

sense of the specific norms, style of behavior, and perceived contingencies of interna- motely resembles the sociopolitical conditions that minimize the use of violence
tional political actors that is to be found in some of the writings in what Kelman within the so-called stable nations.
and Bloom refer to as "the traditional discipline of international relations." Such concerns led Osgood some years ago (1955) to point to what he con-
Kelman and Bloom make no claims about the distribution, of their several sidered three fundamentally incorrect premises behind much of the psychological
presumptive frameworks in populations of leaders, but instead focus on why the literature on conflict resolution-that the "normal" state of relations among nations
frameworks appear to be inaccurate or dysfunctional and what alternative frame- is harmonious rather than competitive (on this point Osgood's argument can be
works might replace them. Nevertheless, by specifying possible cognitive and affective assimilated to the broader school of thought that emphasizes the functions of social
sets that govern political behavior, their formulation naturally connects with empiri- conflict), that it is possible usefully to discuss international tensions q u a tensions
cal efforts to observe, delineate, and classify frames of reference in actual popula- "abstracted from the actual political circumstances in which they occur," and that
tions; for example, the "folk taxonomy" studies of cognitive anthropologists such as "conflicts of national interest and power" are "merely the superficial symptoms of
Tyler (1969) and the attempts to characterize political leaders' operational codes some sociopsychological malady and not . . . anything fundamentally inherent in
and conceptions of the rules of the game by political scientists such as Leites (1959) the existing conditions of international society." Kelman's own writings (1965, 1970,
and George ( 1969). and the passage with Bloom quoted above) constitute an important source of re-
Furthermore, given the importance of well-placed individual actors in political orientation of international political psychology in the direction of political realism.
behavior, one inevitably is led to compare categorizations of types of assumptive Yet there do continue to be persistent differences in emphasis in the way that psy-
frameworks, like those presented by Kelman and Bloom, with the actual frameworks chological aspects of international politics are treated by writers who come to these
of individual decision makers, so far as these can be ascertained, recognizing that issues with different disciplinary backgrounds-differences that are at once a source
abstract types of framework will inevitably seem "thin" when compared with the of creative intellectual tension and an evidence of a still imperfect scholarly synthesis.
variegated thought and behavior of a complex, sophisticated individual. If one, for Political scientists such as Key and Munger (1959) sometimes phrase the
example, compares Landau's (1972) detailed reconstruction of Henry Kissinger's desideratum of attending to the distinctive norms and situational demands of the
thought and behavior with the Kelman-Bloom frameworks, at least one consistency political arena in terms of the desirability of studying "politics" rather than "psy-
and one inconsistency are evident. Kissinger's balance-of-power frame of reference chology" or "sociology." This is an unfortunate usage: the psychology (sociology) of
is consistent with the Kelman-Bloom assumptive framework that treats the nation- politics is just as psychological (sociological) as any aspect of the parent discipline,
state as supreme (it would be hard to see how a national representative could escape but behind the shaky semantics is a point of fundamental importance about the need
to take account of what is distinctive about political contexts and dispositions. In
this framework), but Kissinger does not appear to have viewed the Vietnam conflict
this connection, Burton's experimental efforts to encourage conflict resolution by
in aggressor-defender terms, whatever the rhetoric of the president he served.
Greek and Turkish Cypriots and Doob's similar experiments with Africans of different
Therefore, as Kelman and Bloom further develop their interesting analysis of
nationalities, as fascinating as they are, are likely to arouse less enthusiasm among
assumptive frameworks, it is to be hoped that they will establish links with the
political scientists than among psychologists for precisely the reasons Kelman and
literature on actual leaders' frameworks, whether collective or individual. Correct
Bloom recognize as accounting for their efficacy in the experimental context: these
description of existing assumptive frameworks has some bearing on the task of pre-
procedures bring "together representatives of nations or national (ethnic) com-
scribing alternative frameworks and persuading decision makers to adopt them. One
munities involved in active conflict, for face-to-face communication in a relatively
of the points in the inventory of propositions about attitude change presented by
isolated setting, free from governmental and diplomatic protocol."
Smith is that "a communicator's effectiveness is increased if he initially expresses From the standpoint of the concerns of political scientists, it may be more to
some views that are also held by his audience." the point to develop experimental settings with demand characteristics like those in
Those political scientists who, like Neustadt, specialize in analyzing politics the interesting experimental attempt by Herman and Herman (1967) to simulate
in roughly the same terms that politicians experience politics have been among the the outbreak of World War I. The Hermans explicitly encouraged their experimental
critics of efforts to apply notions derived from psychology to the resolution and subjects to assume a psychological mode that is ubiquitous in political life and
elimination of international conflict. One such writer is the sometime State Depart- certain other aspects of social life, but rarely studied by psychologists-that of the
ment consultant-political scientist Robert Osgood (not to be confused with the agent or representative. A number of years ago Kelman (1965, p. 597) pointed to
psychologist Charles Osgood). The very title of Osgood's book, Ideals and Self- the desirability of incorporating "into the laboratory situation the significant condi-
Interest in America's Foreign Relations (1953), in itself suggests a set of ubiquitous tions of the international situation," and more recently Michel and Dillehay (1969)
concerns of the politician that s ~ m epsychologists desc?ibing and prescribing for have sought to study reference behavior of representatives. The Herman and Herman
international politics (especially those of the minds-of-men school) have been wont also is interesting in its imaginative effort to "match" the personalities of
to ignore, or to treat in a wholly pejorative manner. These are concerns with national the subjects with those of the principal great-power leaders and their
power in a world at least partly composed of adversary relationships, genuine con- agents in 1914. Here again the problem of realism arises, however. As commendable
flicts of interest, and the absence of an international social order which even re- as this matching effort was, there is a breathtaking inferential leap between 1960's
 Fred I. Greenstein Political Psychology: A Pluralistic Universe 465

American high school students who have been classified on the basis of California to deny that, say, the nature of endocrine functioning or cortical structures in
Psychological Inventory and Semantic Differential personality tests and induced to humankind may be consequential for conflict within and between nations, but rather
play a hypothetical simulation game and the actual behavior in 1914 of Nicholas 11, to insist on explicit, self-c~nsciousexamination of the way different "levels" of
Wilhelm 11, Berthelot, Grey, and their compatriots. phenomena are interconnected.
If Kelman and Bloom, in focusing on the psychological phenomenon of A quite striking and suggestive research report by Peter Bourne ( 1971), former
assumptive frameworks, are not explicitly concerned in their chapter with aggregat- chief of the Neuropsychiatry Section of the U.S. Army Medical Team in Vietnam,
ing micro-level data to account for macro-level patterns, Davies surely does have uses both physiochemical and psychological data in an analysis that is consistent
this interest in the second of his chapters in this handbook. The topics referred to in with the monitory themes of this chapter. Using medical and logistical technology
his title-aggression, violence, revolution, and war-run the gamut from the indi- which was not available as recently as the Korean war, Bourne measured adrenal
vidual to the intersocietal levels. Furthermore, Davies is alone among the contributors function via levels of 17-hydro-oxycorticosteroid in urine, thus obtaining a physical
in emphasizing that the individual psyche is rooted in nontrivial ways in its somatic measure of chronic inner stress. He studied members of a helicopter medical aid
base, drawing on some of the work (especially with infrahuman subjects) that takes team and a Special Forces group on hazardous duty in Vietnam. There were regular
account of the physiochemical antecedents of behavior. psychiatric interviews and physiochemical measurements during episodes of relative
One source of difficulty with the challenging literatures discussed by Davies quiet and during periods of intense combat or imminent threat of combat. Previous
is the diversity of the phenomena subsumed by the terms aggression and violence. research had established normal' 17-hydro-oxycorticosteroid levels and normal altera-
Even ruling out what Davies excludes in his initial definitional discussion, an extra- tions of level in individuals subjected to stress (race drivers, runners, patients prior
ordinarily varied range of somatic and motivational antecedents ought in principle to to open-heart surgery).
be responsible for such diverse phenomena as crimes of passion, premeditated murder, Bourne's findings exhibit exactly the sort of complexly interactive pattern
violence ensuing from racial conflict, so-called official violence such as capital pun- 1 have been suggesting "ought" to be found in careful, theoretically informed analyses
ishment and the use of physical force by police, wartime violence by coolly profes- of the connections between psychological and physiological variables with behavior.
sional soldiers, the violence of passionate zealots, and so forth. I t is only a semantic Contrary to the assumption that military activity is likely to have its motivating basis
accident that singular terms apply to such varied phenomena. In this area, as in that jn states of emotional excitation, the helicopter medical crew members (all of whom
of establishing the psychological qualities of politicians, it would seem necessary to were enlisted men) showed abnormally low levels of adrenal secretion. These levels
identify functionally discrete types of violence and aggression in order to identify did not vary as the intensity of combat exposure increased, although they did vary
reasonably stable and distinctive antecedents. under situations of personal stress, such as severe accidental injury. The helicopter
There is a second difficulty with the challenging bodies of literature discussed crewmen appeared able to ward off severe anxiety by concentrating on the mechani-
by Davies, especially when they are viewed in relation to one another. This is the cal tasks at hand and by a series of psychological defenses of a magical-fatalistic sort.
problem of "reductionism," which Davies frequently alludes to, and its obverse, the The Special Forces combat team exhibited a somewhat different pattern and
task of systematically aggregating individual-level data to account for system-level certainly one that justifies still another use on my part of the overworked term
phenomena. Expanding on my own summary of a chain of linkage (Greenstein, "complex." Members of this team, which was in enemy territory and "under con-
1969, ch. 5 ) to take account of issues discussed by Davies, we need to recognize stant threat of attack by an overwhelmingly superior force," also showed lower mean
that the relationships are not necessarily simple and direct along the following causal secretion levels than normal populations. But the secretion levels of officers were
chain: (1) genetic and acquired physiochemical dispositions, (2) childhood en- significantly higher than those of the enlisted men, and this difference was wholly
vironmental influences, (3) adult personality, (4) adult sociopolitical orientations, consistent with the psychiatric evidence that the officers underwent severe stress as
(5) individual behavior, and (6) collective sociopolitical outcomes. Thus : "Similar a result of the uncertainty and responsibility of leadership, whereas the enlisted men
somatic resources (1) and similar childhood environmental experiences (2) can "bound" anxiety via bravado, magical thinking, and a concentration of the routine
produce different personalities, if only because of the many complex ways in which mechanical aspects of their tasks. Furthermore, at the point of imminent enemy
(1) and (2) can interact. attack on the camp, officers' secretion levels rose dramatically and those of enlisted
We have already noted that similar personalities (3) can accommodate differ- rnen declined. Bourne (1971, p. 287) advances the following speculation at the con-
ent (if not the full possible range of) sociopolitical orientations ( 4 ) . clusion of his paper:
Sociopolitical orientations (4) such as those discussed by Smith and Lane do
not in themselves lead to predictable behavior patterns because behavior also is a At first glance, it might appear that these findings in humans contradict
function of situational influences (Panel I V of Smith's map). the extensive data on the physiological aspects of aggressive behavior in
Finally, individual actions (5) are far from simply additive in accounting for animals. However, closer examination indicates that, for many reasons,
collective phenomena (6) : millions of actions by peasants may count for less than warfare among nations cannot be equated either psychologically or phys-
the behavior of a Lenin." iologically with aggressive behavior in animals. First, it is highly institu-
The point of enumerating all these junctures and potential disjunctions is not tionalized, with the individual soldier having virtually no control over and
466 Fred I. Greenstein Political Psychology: A Pluralistic Universe 467

little emotional investment in the decisions made by his leaders. He finds includes in his design situational as well as dispositional variables, thus correcting for
himself compelled to follow their wishes, and his concern becomes one of one major shortcoming in the survey literature. Surveys have produced mountains
surviving in a socially acceptable way rather than being in a state of of psychological data, but much less in the way of a capacity to predict behavior,
personal aggressive arousal. Second, warfare, particularly for the pilot
because of the absence in most surveys of evidence about the environmental antece-
dropping bombs or even the foot soldier using a gun, has become a mech-
anistic act which in most instances is quite depersonalized. Much of the dents of behavior.
time the soldier feels he is merely doing a job and experiences little sense Both McConahay's discussion of experiments and Browning's balanced, lucid
of animosity or aggressivity. Third, human psychic processes enable man treatment of simulation point to the advantages of the methodologies of deliberate
to divorce himself emotionally from events that are threatening or aggres- and radical simplification for the clarification of key theoretical issues about causal
sive in a way which presumably animals are incapable of doing. relationships. The seeming paradox in the three complementary methodologies of
surveying, experimentation, and simulation is that as capacity for precise, controlled
Bourne's conclusion fits remarkably well with the carefully reasoned recent treatment of data increases, capacity to deal realistically and convincingly with
discussion and data presentation by Ray (1972), who argues against the widespread "recognizably important issues" often seems to decrease. And when the techniques for
assumption among some political psychologists that "psychopathologicalJJor at least precise simplification are tinkered with to accommodate to the nonsimplified nature
aggressive attitudes are the necessary antecedents of "militarismJ'; both commentators of reality, we are in danger of winding up with the worst of all worlds, a model so
deny that a single dispositional state can account for military behavior. In this complex that it does not allow clear interpretation and is not even an empirical slice
respect, it is also interesting to note Dicks' (1972) conclusion, on the basis of his of "the real world," an outcome discussed in Browning's interesting comments about
psychiatric interviews with Nazi S.S. killers, that conformity needs rather than "Bonini's paradox." In spite of the problems of realism inherent in simulation and
psychiatric disorders of the sort that produce aggressive symptomatology motivated experimentation, the gain in clarity about causal relationships and control over data
their behavior. makes it highly desirable that they continue to be employed and perfected.
I now turn to an overly brief set of remarks on four of the five chapters in Hyman, McConahay, Browning, and &utson (in her valuable review of
the "Methods of Inquiry" section. (Glad's chapter has already been discussed.) In projective testing procedures in terms of their application to political analysis) are
reprinting his methodological papers, Samuel Stouffer (1962) commented that the all commendably innocent of methodological imperialism. They recognize the im-
one most sought after by students requesting reprints was "Some Observations on portance of multiple, complementary techniques of inquiry, even within an investiga-
Study Design" (pp. 290-299). This essay of Stouffer's might usefully be read as a tive mode. The general notion that no finding is adequately documented without "a
bridge to the methodological chapter in this volume by McConahay on experiments triangulation of measurement processes" (Webb and others, 1966, p. 3) is well
and by Hyman on surveys, in that it elegantly shows how the full experimental recognized, if not always acted on, within psychology (Campbell and Fiske, 1959).
design of experimental group, control group, and before-after measurement is both I t needs to be better recognized by political scientists. Thus, we may hope to see
a research design ideal and the basis of the explanatory logic in the many investiga- more studies with mixed instrumentation and research design. An issue of sufficient
tions that for one reason or another fall short of experimental design, notably surveys. importance and complexity might well call for surveys using projective as well as
Hyman's ingenious invocation of the "most ingenious paradoxJ' from T h e conventional nonprojective instruments, which are paralleled by coordinate experi-
Pirates of Penzance is by way of recognition that, in spite of the supremacy in theory ments, simulations, life histories, and field and archival observations. More often, of
of experimental design, it is nonexperimental survey research with all of its causal course, it is the more or less free marketplace of individual scholarly enrepreneurship
ambiguities that has rendered the bulk of important, nonobvious, statistically that produces such diversity, rather than a single master research design.
grounded contributions to political psychology. My own resolution of this paradox
will by now be painfully predictable. Surveys take account of how orientations and Literatures Not Covered in This Volume
behavior exhibit themselves in the actual settings of the "real world." If, as I have
been arguing, the concrete contexts of politics are highly consequential and cannot The publishing dates of the successive handbooks of social psychology were
be "stripped awayJJor abstracted without grievous loss, then it will not be surprising 1935, 1954, and 1965, and the magnitude of the project increased dramatically
if we find political analysts turning to the survey in most instances--or occasionally with each edition (Murchison, 1935, volume; Lindzey, 1954, 2 volumes; Lindzey
to the field experiment for such specialized problems as the study of how to enhance and Aronson, 1968-1969, 5 volumes). Clearly each earlier handbook failed to cover
voter turnout. I t also will not be surprising to find that "conflicting results [are] some of the topics in the following handbook, either because the topic was not ripe
derived from experimental and field studies" (Hovland, 1959). for review or because the appropriate reviewer was unavailable.
McConahay's chapter on experimentation, which because of the paucity of This volume lacks literature summaries on some important topics for the un-
bona fide political experiments focuses perforce on methodological and design issues, assailable reason that the relevant literatures have not yet been produced. Inventories
begins with an extraordinarily interesting illustration, from the literature. on male of potential but presently nonexistent literatures, like censuses of unborn populations,
and female sexuality, of one way in which experiments can be more interesting probably serve largely as projective tests of the compiler of the inventory. My only
than surveys. As McConahay's example makes clear, the experimenter automatically suggestion about future needs is a broad one, following from much of what I have
468 Fred I. Greenstein Political Psychology: A Pluralistic Universe 469

said: there is a great deal of important but unsystematic psychological content in attitude scale-see Oppenheim (1966). Parallel to the chapters in this volume on
the many, many writings within political science (and overlapping the work of surveys, experimentation, and simulation, there are available sources that can
journalists). Integration of the political and the psychological in political psychology introduce the student of political psychology to two other fundamental resources used.
is undoubtedly one of the major tasks ahead. by political scientists to gather data, particularly data on how institutions operate:
Of those political psychology literatures that presently exist, but which it was field studies (McCall and Simmons, 1969) and archival and library studies (Vose,
not convenient to review in this volume, several are worth noting, following the main forthcoming).
headings of the table of contents:
1. Smith notes that for purposes of expository clarity he has omitted from The Literatures of Political Psychology: Let Many Flowers Bloom
his map reference to a critically important class of phenomena that intervene
between stable personal predispositions and situational stimuli: perceptions. Perhaps Having taken the unconventional tack of summarizing my conclusions earlier
the single most interesting (but difficult to achieve at the present stage of knowledge) in this chapter, there remains only one final point to make-that it is misleading to
addition to the section of this handbook on basic psychological constructs would speak of political psychology in the singular. It is chimerical to imagine that a "truly
have been a review of what is presently known about political perception and the comprehensive" handbook will ever be poxsible, or even desirable. As Donald Camp-
related issues of attention and selective exposure to political stimuli (see, for example, bell (1969b) has pointed out in his proposal for a "fish-scale" approach to the unity
Freedman and Sears, 1965). The importance of systematically understanding the of knowledge, it is impossible for any single work to encompass all of the diverse
interaction between the goals of political actors and their perceptions of their en- data, methodologies, and perspectives that might be applied to the issues it seeks to
vironments is a theme running through Browning's chapter in this volume. cover. But some integration will occur if inquiries are so organized that one investiga-
2. Under the heading "Forming and Maintaining Stable Orientations," this I tion or specialty partially overlaps another, after the fashion of the successive over-
lappings of the scales of a fish. As reassuring as Campbell's image is in an intellectual
handbook contains Niemi's excellent review of the literature on the childhood
antecedents of adult citizen and subject behavior, but there are no reviews of the universe in which one can never "adequately cover the literature," the fish-scale
better-developed literature on the adult aspects of these phenomena. Fortunately, there metaphor does not go quite far enough. Even if "knowledge" did have bounded
is an excellent and highly comprehensive recent review article by Sears (1969) and contours and we did know the shape of the fish and therefore could supply it with all
literature reviews on voting by Converse and political participation by Verba and of its scales, we would still have to take account of the inevitability and desirability
Nie appear in still another handbook: T h e Handbook of Political Science (1974). that diverse viewpoints be expressed about each set of issues. In short, we need not
Reviews are also available elsewhere of what might be called the old and new only the overlapping scales but also alternative renderings of each of the scales.
comparative political psychology literature, the former being the traditional culture- As a result of the widespread intellectual ferment within political psyc~ology
I
and-personality writings (Inkeles and Levinson, 1954, 1969) ; the latter, the political- that this volume exhibits we can expect substantial expansion of the universe of
culture literature (Pye and Verba, 1965). In addition, the time may be ripe for political psychology. The cumulating logic of inquiry points to this. So does the need
reviews of typological literatures other than those discussed by Sanford and Yinger- of our species to come to grips with its own condition-before we are too late. But
for example, dogmatism (Rokeach, 1960) and Machiavellianism (Christie and Geis, it is neither inevitable nor necessarily desirable that the future handbooks of political
1970), or for a reexamination of the numerous other studies correlating personality psychology will reveal a universe that is less pluralistic than the stimulating array of
traits with political orientations and behavior, or for a summary of the numerous theory, conceptualization, and evidence to be found in the present volume.
scattered studies of the objective and subjective characteristics of different populations
of political leaders and activists (for example, Salisbury, 1966; Conway and Feigert,
1968) .
3. The section in this volume on "The Nexus of Individual and Polity" could
at some point (the present time is probably premature) be expanded to review the
range of particular types of political contexts (such as legislatures and courts) and
the,more or less distinctive psychological demands they may make on political actors.
In addition, there are the many challenging, if not invariably clear, broad psycho-
p~~losophical characterizations of polity and society by writers such as Brown (1959),
Fromm ( 1941) , and Marcuse ( 1959), which could usefully be reviewed in terms of
their empirical and normative implications.
4. Of the present methodological chapters, only Knutson's, on the under-
utilized but promising approach of projective techniques, deals with instrumentation.
For a discussion of the more conventional approach to eliciting political-psychologi-
cal observations from populations of individuals-namely, the questionnaire and
 REFERENCES

ABCARIAN, G., AND STANAGE, S. "Alienation and the Radical Right." Journal of Politics, 1965,
27, 776-796.
ABEL, T. "IS a Psychiatric Interpretation of the G,erman Enigma Necessary?" American
Sociological Review, 1945, 10, 457-464.
ABELSON, H., COHEN, R., HEATON, E., AND SLIDER, C. "Public Attitudes Toward an Experience
with Erotic Materials." Technical Reports of the Commission on Obscenity and
Pornography. Vol. 6. Washington, D.C. : U.S. Government Printing Office, 1970.
ABELSON, R. P. '(Mathematical Models of the Distribution of Attitudes Under Controversy."
In N. Frederiksen and H. Gulliksen (Eds.), Contributions to Mathematical Psy-
chology. New York: Holt, 1964.
ABELSON, R. P. "Psychological Implication." In R. P. Abelson, E. Aronson, W. J. McGuire,
T. M. Newcomb, M. J. Rosenberg, and P. H. Tannenbaum (Eds.), Theories of
Cognitive Consistency: A Sourcebook. Chicago: Rand McNally, 1968a.
ABELSON, R. P. "Simulation of Social Behavior." In G. Lindzey and E. Aronson (Eds.), The
Handbook of Social Psychology. (Rev. Ed.) Vol. 2. Reading, Mass.: Addison-Wesley,
1968b.
ABELSON, R. P. "The Ideology Machine." Paper prepared for annual meeting of American
Political Science Association, Chicago, 1971.
ABELSON, R. P. "The Structure of Belief Systems." In K. Colby and R. Schank (Eds.),
Computer Simulation of Thought and Language. San Francisco: Freeman, 1973.
ABELSON, R. P., ARONSON, E., MCGUIRE, W. J., NEWCOMB, T. M., ROSENBERG, M. J., AND
TANNENBAUM, P. H. (Eds.) Theories of Cognitive Consistency: A Sourcebook.
Chicago: Rand McNally, 1968.
 7-

References References
ABELSON, R. P., AND BERNSTEIN, A. "A Computer Simulation Model of Community Refer-
ALLPORT, G. W. "Attitudes." In C. Murchison (Ed.), A Handbook of Social Psychology.
endum Controversies." Public Opinion Quarterly, 1963, 27, 93-122. Worcester, Mass.: Clark University Press, 1935. Pp. 798-844. Reprinted in G. W.
ABELSON, R. P., AND CARROLL, J. D. "Computer Simulation of Individual Belief Systems."
Allport, T h e Nature of Personality. Reading, Mass.: Addison-Wesley, 1950.
American Behavioral Scientist, 1965, 8, 24-30. ALLPORT, G. W. Personality: A Psychological Interpretation. New York: Holt, 1937.
ABERBACH, J. D. "Alienation and Political Behavior." American Political Science Review,
ALLPORT, G. W. "Personality: A Problem for Science or a Problem for Art?" Review of
1969,63,86-99. Psychology, 1938, 1, 488-502.
ABERBACH, J. D., AND WALKER, J. L. "Political Trust and Racial Ideology." American Political
ALLPORT, G. W. T h e Use of Personal Documents in Psychological Science, Bulletin 49. New
Science Review, 1970, 64, 1199-1219. York: Social Science Research Council, 1942.
ABERLE, D. F. "The Psychosocial Analysis of a Hopi Life History." I n Y. A. Cohen (Ed.),
ALLPORT, G. W. "The General and the Unique in Psychological Science." Journal of Per-
Social Structure and Personality: A Casebook. New York: Holt, 1961.
sonality, 1962, 30, 405-422.
ABRAHAM, K. "Amenhotep (Ikhnaton) : A Psychoanalytic Contribution to the Understanding
ALLPORT, G. w., BRUNER, J. s., AND JANDORF, E. M. "Personality Under Social Catastrophe:
of His Personality and the Monotheistic Cult of Anton." Imago, 1912, 1, 334-360.
Ninety Life Histories of the Nazi Revolution." Character and Personality, 1941, 10,
Translated in Psychoanalytic Quarterly, 1935, 4, 537-569.
1-22.
ABRAMSON, P. "Political Efficacy and Political Trust Among Black School-Children: Two
ALLPORT, G. w., AND ODBERT, H. "Trait-Names: A Psycho-Lexical Study." Psychological
Explanations." Journal of Politics, 1972, 34, 1243-1275.
Monographs, 1936, 47, whole No. 21 1.
ABRAMSON, P., AND INGLEHART, R. "The Development of Systematic Support in Four Western
ALMOND, G. A. T h e Appeals of Communism. Princeton: Princeton University Press, 1954.
Democracies." Comparative Political Studies, 1970, 2, 419-442. ALMOND, G. A. T h e American People and Foreign Policy. New York: Harcourt, 1950. New
ADELSON, J. "A Study of Minority Group Authoritarianism." Journal of Abnormal and
York: Praeger, 1960a.
Social Psychology, 1953, 48, 477-485.
ADELSON, J. "The Political Imagination of the Young Adolescent." Daedalus, 1971, 100, 1 ALMOND,G. A. "Introduction: A Functional Approach to Comparative Politics." I n G.
Almond and J. Coleman (Eds.), T h e Politics of the Developing Areas. Princeton:
1013-1050.
Princeton University Press, 1960b.
ADELSON, J., AND O'NEIL, R. "Growth of Political Ideas in Adolescence: The Sense of Com-
ALMOND, G. A. "A Comparative Study of Interest Groups and the Political Process." In H.
munity." Journal of Personality and Social Psychology, 1966, 4, 295-306.
Eckstein and D. E. Apter (Eds.), Comparative Politics. New York: Free Press, 1963.
ADELSON, J., AND O'NEIL,R. "Growth of the Idea of Law in Adolescence." Developmental
ALMOND, G. A., AND VERBA, S. T h e Civic Culture: Political Attitudes and Democracy in Five
Psychology, 1969, 1, 327-332. Nations. Princeton, N.J.: Princeton University Press, 1963. Boston: Little, Brown,
ADELSON, J., AND SULLNAN, P. "Ethnocentrism and Misanthropy." American Psychologist,
1965.
1952, 7, 330. American Political Science Association. Biographical Directory. (4th Ed.) Washington, D.C.,
ADLER, A. Understanding Human Nature. New York: Fawcett World Library, 1969. OPigi-
1961, 1968.
nally published 1927. American Political Science Association. Biographical Directory. (6th Ed.) Washington, D.C.,
ADLER,N., AND HARRINGTON, C. T h e Learning of Political Behavior. Glenview, Ill.: Scott,
1973.
Foresman, 1970. ANDERSON, B., ZELDITCH, M., JR., TAKAGI, P., AND WHITESIDE, D. "On Conservative Attitudes."
ADORNO, T. "Anti-Semitism and Fascist Propaganda." I n E. Simmel (Ed.), Anti-Semitism:
Acta Sociologica, 1965, 8, 189-204.
A Social Disease. New York: International Universities Press, 1946.
ANGELL,R. C. "Preferences for Moral Norms in Three Problem Areas." American Journal
ADORNO, T., FRENKEL-BRUNSWIK, E., LEVINSON, D., AND SANFORD, N. T h e Authoritarian
of Sociology, 1962, 67, 650-660.
Personality. New York: Harper, 1950. With a preface by Max Horkheimer.
APSLER,R. "Effects of the Draft Lottery and a Laboratory Analogue on Attitudes." Journal
AGGER, R. E., GOLDSTEIN, M., AND PEARL, S. "Political Cynicism: Measurement and Meaning."
of Personality and Social Psychology, 1973.
Journal of Politics, 1961, 23, 477-506.
ARDREY, R. African Genesis. New York: Dell, 1961.
AGGER, R., AND OSTROM, V. "The Political Structure of a Small Community." Public Opinion
ARDREY, R. T h e Territorial Imperative. New York: Atheneum, 1966.
Quarterly, 1956, 20, 89-102. I
ARENDT, H. T h e Origins of Totalitarianism. New York: Harcourt, 1951.
ALGER,C. F. "Comparison of Intranational and International Politics." American Political
ARNOLD, M. Story Sequence Analysis. New York: Columbia University Press, 1962.
Science Review, 1963, 57,406-419.
ARON,B. "The Thematic Apperception Test in the Study of Prejudiced and Unprejudiced
ALKER,H. A. "IS Personality Situationally Specific or Intrapsychically Consistent?" Journal
Individuals." I n T. Adorno, E. Frenkel-Brunswik, D. Levinson, and R. N. Sanford
of Personality, 1972, 40, 1-16.
(Eds.), T h e Authoritarian Personality. New York: Harper, 1950.
ALLAND, A., JR. T h e Human Imperative. New York: Columbia University Press, 1972.
ARONOFF, J. Psychological Needs and Cultural Systems. Princeton, N.J.: Van Nostrand,
ALLINSMITH, w., AND ALLINSMITH, B. "Religious Affiliation and Politico-Economic Attitude."
1967.
Public Opinion Quarterly, 1948, 12, 377-389.
ARONOFF, J. "Psychological Needs as a Determinant in the Formation of Economic Systems:
ALLISON, G. T. Essence of Decision. Boston: Little, Brown, 1971.
A Confirmation." Human Relations, 1970, 23, 123-138.
ALLPORT, F. Social Psychology. Boston: Houghton Mifflin, 1924.
ARONSON, E., AND CARLSMITH, J. M. "Experimentation in Social Psychology." I n G. Lindzey
ALLPORT, .G. W. "The Composition of Political Attitudes." American Journal of Sociology,
and E. Aronson (Eds.), T h e Handbook of Social Psychology. (Rev. Ed.) Vol. 2.
1930,35,220-228.
Reading, Mass. : Addison-Wesley, 1968.
474 References References 475

ASCH,S. E. Social Psychology. Englewood Cliffs, N.J. : Prentice-Hall, 1952. BELL,D. T h e End of Ideology: O n the Exhaustion of Political Ideas in the Fifties. New
ATKINSON, J. "Discussion of Paper by R. Lazarus." In J. Kagan and C. Lesser (Eds.), York: Free Press, 1960.
Contemporary Issues in Thematic Apperceptive Methods. Springfield, Ill.: Thomas, BELL, D. (Ed.) T h e Radical Right. Garden City, N.Y.: Doubleday, 1963.

1961. BELL, W. "Anomie, Social Isolation, and Class Structure." Sociometry, 1957, 20, 105-116.
AXELROD, J., FREEDMAN, M., HATCH,w., KATZ,J., AND SANFORD, N. Search for Relevance. BELL,w., HILL,R., AND WRIGHT,C. Public Leadership. San Francisco: Chandler, 1961.
San Francisco : Jossey-Bass, 1969. BELLAK, L. ''Somerset Maugham: A Thematic Analysis of Ten Short Stories." In R. White
AZRAEL, J. "Soviet Union." I n J. Coleman (Ed.), Education and Political Development. (Ed.), T h e Study of Lives. Chicago: Aldine-Atherton, 1966.
Princeton : Princeton University Press, 1965. BEM, D. "Self-Perception: An Alternative Interpretation of Cognitive Dissonance Phenomena."
BAIN, R. "Spencer's Love for George Eliot." Psychoanalytic Review, 1927, 14, 37-55. Psychological Review, 1967, 74, 183-200.
BALDWIN, A. L. ''Personal Structure Analysis: A Statistical Method for Investigating the BENEDICT, R. Patterns of Culture. Boston: Houghton Mifflin, 1934.
Single Personality." Journal of Abnormal and Social Psychology, 1942, 37, 163-183. BERADT, C. T h e Third Reich of Dreams. Chicago: University of Chicago Press, 1968.
BALES, R. F. "Task Roles and Social Roles in Problem-Solving Groups." I n E. Maccoby, BERELSON, B., LAZARSFELD, P., AND MC PHEE,W. Voting: A Study of Opinion Formation in
T.. M. Newcomb, and E. L. Hartley (Eds.), Readings in Social Psychology. New a Presidential Election. Chicago: University of Chicago Press, 1954.
York: Holt, 1958. BERGER, A. s., GAGNON, J. H., AND SIMON,W. "Pornography: High School and College
BANFIELD, E. T h e Moral Basis of a Backward Society. New York: Free Press, 1958. Years." Technical Reports of the Commission on Obscenity and Pornography. Vol. 9.
BARAKAT, H. "Alienation: A Process of Encounter Between Utopia and Reality." British Washington, D.C. : U.S. Government Printing Office, 1970.
Journal of Sociology, 1969, 20, 1-10. BERKOWITZ, L. Aggression: A Social Psychological Analysis. New York: McGraw-Hill, 1962.
BARBER, J. D. T h e Lawmakers: Recruitment and Adaptation to Legislative Life. New Haven, BERKOWITZ, L. (Ed.) T h e Roots of Aggression: A Re-examination of the Frustration-Aggres-
Conn. : Yale University Press, 1965. :ion Hypothesis. New York: Atherton, 1969.
BARBER, J. D. Power in Committees: A n Experiment in the Governmental Process. Chicago: I BERKOWITZ, L. "The Study of Urban Violence: Some Implications of Laboratory Studies
Rand McNally, 1966. of Frustration and Aggression." American Behavioral Scientist, March-April 1971,
BARBER, J. D. T h e Presidential Character: Predicting Performance in the White House. pp. 14-17.
Englewood Cliffs, N.J. : Prentice-Hall. 1972a. BETTELHEIM, B. T h e Informed Heart. New York: Free Press, 1960.
BARBER, J. D. "The Question of Presidential Character." Saturday Review, 23 September, BETTELHEIM, B. T h e Children of the Dream. New York: Avon, 1969.
1972b, pp. 62-66. BILLY, J.-S. "Le Probleme de la FinalitC des SocietCs Politiques et les Explications Ideo-
BARDACH, E. T h e Skill Factor in Politics: Repealing the Mental Commitment Laws in
logiques." International Social Science Bulletin, 1953, 5, 51-74.
California. Berkeley: University of California Press, 1972. BLACK, M. Critical Thinking. Englewood Cliffs, N.J. : Prentice-Hall, 1946.
BARKER, E. "Authoritarianism of the Political Right, Center and Left." Journal of Social BLALOCK, H. M. Social Statistics. New York: McGraw-Hill, 1960.
Issues, 1963, 19, 63-74. BLAU, P. Exchange and Power in Social Life. New York: Wiley, 1964.
BARNARD, C. I. T h e Functions of the Executive. Cambridge, Mass.: Harvard University ~ , M., AND SCOTT, W. R. Formal Organizations. San Francisco: Chandler, 1962.
B L A P.
Press, 1938. BLAUNER, R. Alienation and Freedom: T h e Factory Worker and His Industry. Chicago:
BASS, B. "Authoritarianism or Acquiescence." Journal of Abnormal and Social Psychology, University of Chicago Press, 1964.
1955, 51, 616-623. BLITZER,C. A n Immortal Commonwealth: T h e Political Thought of James Harrington.
BAUMOL, W. J., AND QUANDT, R. E. "Rules of Thumb and Optimally Imperfect Decisions."
New Haven : Yale University Press, 1960.
American Economic Review, 1964,54, 23-46. BLOCK,JACK. "Personality Measurement: Overview." I n D. Sills (Ed.), International
BAUMRIND, D. "Some Thoughts on Ethics of Research: After Reading Milgram's 'Behavioral
Encyclopedia of the Social Sciences. New York: Macmillan, 1968.
Study of Obedience.' " American Psychologist, 1964, 19, 421423.
BLOCK, JEANNE. "Generational Continuity and Discontinuity in the Understanding of
BAUMRIND, D. "Child Care Practices Anteceding 'Three Patterns of Preschool Behavior."
Societal Rejection." Journal of Personality and Social Psychology, 1972, 22, 333-345.
Genetic Psychology Monographs, 1967, 75,43-88. I
BLOCK, J. AND BLOCK, J. "An Investigation of the Relationship Between Intolerance of
BAY,C. T h e Structure of Freedom. Stanford, Calif.: Stanford University Press, 1958.
BECK,P. "Youth and the Politics of Realignment." In R. Niemi (Ed.), New Views of Ambiguity and Ethnocentrism." Journal of Personality, 1951, 19, 303-311.
BLOCK, j., AND HAAN, N. Lives Through Time. Berkeley: Bancroft Books, 1971.
Children and Politics. San Francisco: Tossev-Bass.
- ,
forthcomin~.1974. --
BLOCK, J. HAAN,N., AND SMITH, M. B. "Spcialization Correlates of Student Activism."
-2 - -

BECK,S. J. "The Science of Personality: Nomothetic or Idiographic?" Psychological Review,

1953, 60, 353-359. Journal of Social Issues, 1969, 25, 143-177.
BLONDEL, J. Voters, Parties, and Leaders. Hammondsworth, England: Penguin Books, 1963.
BECKER,H. P. "Normative Reactions to Normlessness." American Sociological Review,
1960, 25, 803-810. BLOOM, B. Stability and Change in Human Characteristics. New York: Wiley, 1965.

BECKER, H. S. ''Social Observation and Social Class Studies." In D. Sills (Ed.), International BLUMENTHAL, M. D., AND OTHERS. Justifying Violence: Attitudes of American Men. Ann
Encyclopedia of the Social Sciences. New York: Macmillan, 1968. Arbor: University of Michigan Institute for Social Research, 1972.
Behavioral and Social Sciences Survey Committee. T h e Behavioral and Social Sciences: BLYDENBURGH, J. C. 'A Controlled Experiment to Measure the Effects of Personal Contact
Outlook and Needs. Englewood Cliffs, N.J.: Prentice-Hall, 1969. Campaigning." Midwest Journal of Political Science, 1971, 15, 365-381.
476 References References '

BODER, D. P. "The Adjective-Verb Quotient: A Contribution to the Psychology of Language." puter Simulation." I n W. D. Coplin (Ed.), Simulation in the Study of Politics.
Psychological Record, 1940, 3, 310-343. Chicago: Markham, 1968a.
BOLGAR, H. "A Re-evaluation of Projective Theory." I n B. Klopfer (Ed.), Advances in BROWNING, R. P. "The Interaction of Personality and Political System in Decisions to Run
Rorschach Technique. Vol. 2. New York: World Book Co., 1956. for Office: Some Data and a Simulation Technique." Journal of Social Issues, 1968b,
BONINI, C. P. Simulation of Information and Decision Systems in the Firm. Englewood 24, 93-109.
Cliffs, N.J. : Prentice-Hall, 1963. BROWNING, R. P. Businessmen in Politics: Motivation and Circumstances in the Rise to
BOOCOCK, S. s., AND SCHILD, E. O. (Eds.) Simulation Games in Learning. Beverly Hills, Calif.: Power. (Doctoral dissertation, Yale University) Ann Arbor, Mich.: University
Sage Publications, 1968. Microfilms, 1969. No. 69-12, 908.
BOUDON, R. "Reflexions sur la Logique des Modtles SimulCs." European Journal of Sociology, BROWNING, R. P. ''Innovative and Non-innovative Decision Processes in Government Bud-
1965, 6, 3-20. geting." I n I. Sharkansky (Ed.), Policy Analysis in Political Science. Chicago:
BOURNE, P. G. "Altered Adrenal Function in Two Combat Situations in Viet Nam." I n B. E. Markham, 1970.
Eleftheriou and J. P. Scott (Eds.), T h e Physiology of Aggression and Defeat. New BROWNING, R., AND JACOB, H. "Power Motivation and the Political Personality." Public
York: Plenum, 1971. Opinion Quarterly, 1964, 28, 75-90.
BOWLBY, J. Attachment. New York: Basic Books, 1969. BROWNLEE, K. A. Statistical Theory and Methodology in Science and Engineering. New
BOYD, E. "Sex in Biography." Harper's, 1932, 165, 752-759. York: Wiley, 1960.
BRECHT,A. Political Theory: T h e Foundations of Twentieth Century Political Thought. BRUNER, J. s., GOODNOW, J. J., AND AUSTIN, G. A. A Study of Thinking. New York: Wiley,
Princeton : Princeton University Press, 1959. 1956.
BREED, w., AND KTSANES,T. "Pluralistic Ignorance in the Process of Opinion Formation." BRYAN,J., AND TEST, M. "Models and Helping: Naturalistic Studies of Aiding Behavior."
Public Opinion Quarterly, 1961, 25,382-392. Journal of Personality and Social Psychology, 1967, 6, 400407.
I
BREMER, S. A. National and International Systems: A Computer Simulation. (Doctoral dis- BUDNER,S. "Intolerance of Ambiguity as a Personality Variable." Journal of Personality,
sertation, Michigan State University) Ann Arbor, Mich.: University Microfilms, 1962, 30, 29-50.
1971. NO. 71-18, 175. BULMUGH, B. "Alienation in the Ghetto." American Journal of Sociology, 1967, 72, 469478.
BRICKMAN, P., SHAVER,P., AND ARCHIBALD, P. "American Tactics and American Goals in BURKE, E. Reflections on the Revolution in France. Chicago: Reg~ery, 1955. Originally
Vietnam as Perceived by Social Scientists." I n W. Isard (Ed.), Vietnam: Issues and published 1790.
Alternatives. Cambridge, Mass.: Schenkman, 1969. BUROS,o. K. (Ed.) T h e Sixth Mental Measurements Yearbook. Highland Park, N.J.:
BRIDGMAW, P. W. T h e Logic of Modern Physics. New York: Macmillan, 1928. Gryphon Press, 1965.
BRIM, JR., o., AND WHEELER, S. Socialization After Childhood: T w o Essays. New York: BURTON, J. W. Conflict and Communication: T h e Use of Controlled Communication in
Wiley, 1966. International Relations. London: Macmillan, 1969.
BRINTON, C. T h e Anatomy of Revolution. Englewood Cliffs, N.J.: Prentice-Hall, 1938. New BURTON, J . W. World Society. Cambridge: Cambridge University Press, 1972.
York: Vintage, 1965. BURTON, R., AND WHITING,J. "The Absent Father and Cross Sex Identity." Merrill-Palmer
BRODIE, B. A. "A Psychoanalytic Interpretation of Woodrow Wilson." World Politics, 1957, Quarterly, 1961, 7, 85-95.
9,413422. BUSHMAN,R. L. "On the Uses of Psychology: Conflict and Conciliation in Benjamin
BRODIE, F. M. Thaddeus Stevens. New York: Norton, 1966. Franklin." History and Theory, 1966, 5, 225-240.
BRONFENBRENNER, U. "Socialization and Social Class Through Time and Space." I n E. E. BUTLER, D., AND STOKES, D. Political Change in Britain. New York: St. Martin's, 1969.
Maccoby, T. M. Newcomb, and E. L. Hartley (Eds.), Readings in Social Psychology. BW, D. 'cDimensions of Social Conflict in Latin America." In L. H. Masotti and D. R.
(3rd Ed.) New York: Holt, 1958. Bowen (Eds.), Riots and Rebellion: Civil Violence in the Urban Community.
BRONFENBRENNER, U. TWOWorlds of Childhood. New York: Russell Sage, 1970. Beverly Hills, Calif. : Sage, 1968.
BROWN,D. M. T h e White Umbrella: Indian Political Thought from Manu to Gandhi. CAHALAN, D. "Correlates of Respondent Accuracy in the Denver Validity Survey." Public
Berkeley: University of California Press, 1959. I
Opinion Quarterly, 1968, 32, 607-621.
BROWN,D., AND BRSTRYN, D. "College Environment, Personality and Social Ideology of CAMPBELL, A., CONVERSE, P., MILLER, w., AND STOKES, D. T h e American Voter. New York:
Three Ethnic Groups." Journal of Social Psychology, 1956, 44, 279-288. Wiley, 1960.
BROWN, D., FREEDMAN, M., SANFORD, N., AND WEBSTER, H. "Personality Development During CAMPBELL, A., CONVERSE, P., MILLER, w., AND STOKES,D. Elections and the Political Order.
the College Years." Journal of Social Issues, 1956, 12(41). New York: Wiley, 1966.
BROWN,J. F. Psychology and the Social Order. New York: McGraw-Hill, 1936. CAMPBELL, A., GURIN, G., AND MILLER, W. T h e Voter Decides. New York: Harper, 1954.
BROWN, N. O. Life Against Death. Middletown: Wesleyan University Press, 1959. CAMPBELL, A., AND SCHUMAN, H. "Racial Attitudes in Fifteen American Cities." 111Supple-
BROWN, R. Social Psychology. New York: Free Press, 1965. mental Studies for the National Advisory Committee on Civil Disorders. Washington,
BROWN, S. R. ccConsistencyand the Persistence of Ideology: Some Experimental Results." D.C.: U.S. Government Printing Office, July 1968, pp. 11-67.
Public Opinion Quarterly, 1970, 34, 60-68. CAMPBELL, D. T. "The Indirect Assessment of Social Attitudes." Psychological Bulletin,
BROWN,S. R. "Experimental Design and the Structuring of Theory." Experimental Study 1950, 47, 15-38.
of Politics, 1971, 1, 1-41. CAMPBELL, D. T. "Factors Relevant to the Validity of Experiments in Social Settings."
BROWNING, R. P. "Hypotheses About Political Recruitment: A Partially Data-Based Com- Psychological Bulletin, 1957a, 54, 297-312.
 References References

CAMPBELL, D. T. "A Typology of Tests, Projective and Otherwise." Journal of Consulting CHOMSKY, N. "Review of Verbal Behavior by B. F. Skinner." Language, 1959, 35, 2658.
Psychology, 195713, 21, 207-210. CHRISTIANSEN, B. "The Scientific Status of Psychoanalytic Clinical Evidence." Inquiry,
CAMPBELL, D. T. "Recommendations for APA Test Standards Regarding Construct, Trait, or 1964, 7, 47-49.
Discriminant Validity." American Psychologist, 1960, 15, 546-553. CHRISTIE,R. "Authoritarianism Re-examined." I n R. Christie and M. Jahoda (Eds.),
CAMPBELL, D. T. "Social Attitudes and Other Acquired Behavioral Dispositions." In S. Koch Studies in the Scope and Method of "The Authoritarian Personality." New York:
(Ed.), Psychology: A Study of a Science. Vol. 6. New York: McGraw-Hill, 1963. Free Press, 1954.
Pp. 94-172. CHRISTIE, R. "Eysenck's Treatment of the Personality of Communists." bsychological Bulletin,
CAMPBELL, D. T. "Quasi-Experimental Design." In D. L. Sills (Ed.), International Encyclo- 1956, 53,411-430.
pedia of the Social Sciences. Vol. 5. New York: Macmillan, 1968. CHRISTIE, R., AND COOK,P. "A Guide to Published Literature Relating to the Authoritarian
CAMPBELL, D. T. "Reforms as Experiments." American Psychologist, 1969a, 24, 409429. Personality Through 1956." Journal of Psychology, 1958, 45, 171-199.
CAMPBELL, D. T. "Ethnocentrism of Disciplines and the Fish-Scale Model of Omniscience." CHRISTIE, R., AND GEIS, F. L. (Eds.) Studies in Machiavellianism. 'New York: Academic Press,
In M. Sherif and C. W. Sherif (Eds.), Interdisciplinary Relationships in the Social 1970.
Sciences. Chicago: Aldine, 196913. CHRISTIE,R., HAVEL, J., AND SEIDENBERG, B. "IS the F Scale Irreversible?" Journal of Ab-
CAMPBELL, D. T., AND FISKE, D. W. "Convergent and Discriminant Validation by the Multi- normal and Social Psychology, 1958, 56, 143-159.
trait-Multimethod Matrix." Psychological Bulletin, 1959, 56, 81-105. CHRISTIE, R., AND JAHODA, M. (Eds.) Studies in the Scope and Method of "The Authoritarian
CAMPBELL, D. T., AND MC CANDLESS, B. "Ethnocentrism, Xenophobia and Personality." Hu- Personality". New York: Free Press, 1954.
man Relations, 1951, 4, 186-192. CLARK, L. P. Lincoln: A Psycho-biography. New York: Scribner's, 1933.
CAMPBELL, D. T., AND ROSS,H. L. "The Connecticut Crackdown on Speeding: Time-Series CLARK, R. "The Projective Measurement of Experimentally Induced Levels of Sexual Mmti-
Data in Quasi-Experimental Analysis." Law and Society Review, 1968, 3 ( 1 ) , 33-53. vation." Journal of Experimental Psychology, 1952, 44, 391-399.
CAMPBELL, D. T., AND STANLEY, J. C. "Experimental and Quasi-Experimental Designs for CLARKE, J. "Family Structure and Political Socialization Among Urban Black Children:
Research on Teaching." In N. L. Gage (Ed.), Handbook of Research on Teaching. Some Observations on Father Absence." Midwest Journal of Political Science, 1973.
Chicago : Rand McNally, 1963. CLARKSON, G. P. E. "Decision Making in Small Groups: A Simulation Study." Behavioral
CANNING, R. R., AND BAKER,J. M. "Effect of the Group on Authoritarian and Non-Authori- Science, 1968, 13, 288-305.
tarian Persons." American Journal of Sociology, 1959, 64,579-581. CLARKSON, G . P. E., AND SIMON, H. A. LcSimulationof Individual and Group Behavior."
CANTOR, N. F. The English: A History of Politics and Society to 1760. New York: Simon
American Economic Review, 1960, 50,920-932.
and Schuster, 1967. CLAUSEN, A. R. L c R e ~ p ~Validity:
n~e Vote Report." Public Opinion Quarterly, 1968, 32,
CANTRIL, H. T h e Politics of Despair. New York: Basic Books, 1958. 588-606.
CANTRIL, H. T h e Pattern of Human Concerns. New Brunswick, N.J.: Rutgers University
CLINARD, M. B. (Ed.) Anomie and Deviant Behavior: A Discussion and Critique. New York:
Press, 1965. Free Press, 1964.
CARLSMITH, J. M., COLLINS, B. E., AND HELMREIGH, R. L. "Studies on Forced Compliance. I:
CLOWARD, R. A., AND OHLIN, L. E. Delinquency and Opportunity. New York: Free Press,
The Effect of Pressure for Compliance on Attitude Change Produced by Face-to-Face 1960.
Role-Playing and Anonymous Essay Writing." Journal of Personality and Social CNUDDE, c., AND NEUBAUER, D. Empirical Democratic Theory. Chicago: Markham, 1969.
Psychology, 1966, 4, 1-13. UOE,R. M. "Conflict, Interference and Aggression: Computer Simulation of a Social Process."
CARR, L. G. "The Srole Items and Acquiescence." American Sociological Review, 1971, 36,
Behavioral Science, 1964, 9, 186-196.
287-293. 9
COHEN, A. K. "The Sociology of the Deviant Act: h o m i e Theory and Beyond." American
CASH,W. J. T h e Mind of the South. New York: Knopf, 194L
Sociological Review, 1965, 30, 5-14.
CASPARY, W. "The 'Mood Theory': A Study of Public Opinion and Foreign Policy." Ameri-
GOHEN,K. J., AND CYERT,R. M. "Simulation of Organizational Behavior." In J. G. March
can Political Science Review, 1970, 64, 536-547.
(Ed.), Handbook o f Organizations. Chicago: Rand McNally, 1965.
CATTELL, R, "New Concepts for Measuring Leadership in Terms of Group Syntality." Human
COHN, T. S. "The Relation of the F Scale to a Response to Answer Positively." American
Relations, 1951, 4, 161-184.
CATTELL, R. Personality and Motivation: Structure and Measurement. Yonkers-on-Hudson,
Psychiatrist, 1953, 8, 335. (Abstract)
N.Y. : World Book, 1957. COHN, T., AND CARSCH, H. L'Admini~tration of the F Scale to a Sample of Germans." Journal
CENTERS, R. The Psychology of Social Classes. Princeton: Princeton University Press, 1949.
of Abnormal and Social Psychology, 1954, 49, 471.
CHEIN,I. T h e Science of Behavior and the Image of Man. New York: Basic Books, 1972. COKER, F. W. Recent Political Thought. New York: Appleton-Century, 1934.

CHERRYHOLMES, C. H., AND SHAPIRO, M. J. Representatives and Roll Calls: A Computer COI.BY,K. M. "Computer Simulation of Neurotic Processes." In R. W. Stacy and B, D.
Simulation of Voting in the Eighty-Eighth Congress. Indianapolis: Bobbs-Merrill, Waxman (Eds.), Computers in Biomedical Research. Vol. 1. New York: Academic
1969. Press, 1965.
CHICKERING, A. Education and Identity. San Francisco: Jossey-Bass, 1969. COLBY, K. M. "A Programmable Theory of Cognition and Affect in Individual Personal
CHILD,I. L., STORM,T., AND VEROFF,J. "Achievement Themes in Folk Tales Related to Belief Systems." In R. P. Abelson, E. Aronson, W. J. McGuire, T. M. Newcomb,
Socialization Practice." In J. W. Atkinson (Ed.), Motives in Fantasy, Action and M. J. Rosenberg, and P. H. Tannenbaum (Eds.), Theories of Cognitive Consistency:
Society. Princeton: Van Nostrand, 1958. A Source Book. Chicago: Rand McNally, 1968.
480 References References 481

COLBY, K. M., AND SCHANK,R. (Eds.) Computer Simulation of Thought and Language. San COOPERSMITH, S. T h e Antecedents of Self-Esteem. San Francisco: Freeman, 1967.
Francisco : Freeman, 1973. CORNBLETH, C. "Political Socialization and the Social Studies: Political Beliefs of Mexican-
COLEMAN, J. S. "Relational Analysis: The Study of Social Organization with Survey American Youth." Paper presented at meeting of American Educational Research
Methods." Human Organization, 1958, 17, 28-36. Association, New York, February 1971.
COLEMAN, J. S. T h e Adolescent Society. New York: Free Press, 1961. CORNFIELD, J., AND TUKEY, J. W. "Average Values of Mean Squares in Factorials." Annals
COLEMAN, J. S. " C ~ l l e ~ t Decisions."
i~e Sociological Inquiry, 1964a, 34, 166-181. of Mathematical Statistics, 1956, 27, 907-949.
COLEMAN, J. S. c'Impli~ation~ of the Findings on Alienation." American Journal of Sociology, COSER, L. T h e Functions of Social Conpict. New York: Free Press, 1956.

1964b, 70, 76-78. COUCH,A., AND KENISTON,K. "Yeasayers and Naysayers: Agreeing Response Set as a
COLEMAN, J. S. "Mathematical Models and Computer Simulation." I n R. E. L. Faris (Ed.), I Personality Variable." Journal of Abnormal and Social Psychology, 1960, 60, 151-174.
Handbook of Modern Sociology. Chicago: Rand McNally, 1964c. cox, D. R. "The Use of a Concomitant Variable in Selecting an Experimental Design. Bio-
COLEMAN, J. S. (Ed.) Education and Political Development. Princeton: Princeton Uni- metrika, 1957, 44, 150-158.
versity Press, 1965a. cox, D. R. Planning of Experiments. New York: Wiley, 1958.
COLEMAN, J. S. "The Use of Electronic Computers in the Study of Social Organizations." CRAWFORD, T. J., AND NADITCH, M. "Relative Deprivation, Powerlessness, and Militancy: The
European Journal of Sociology, 1965b, 6,89-107. Psychology of Social Protest." Psychiatry, 1970, 33, 208-223.
COLEMAN, J. S. Equality of Educational Opportunity. Washington: United States Office of CRECINE, J. P. "A Computer Simulation Model of Municipal Budgeting." Management
Education, 1966a. Science, 1967, 13, 786815.
COLEMAN, J. S. T h e Game of Democracy. Washington, D.C.: National 4-H Club Founda- I CRECINE, J. P. Governmental Problem-Solving, A Computer Simulation of Municipal Budget-
tion, 1966b. ing. Chicago: Rand McNally, 1069.
COLES,R. "The Limits of Psychiatry." Progressive, May 1967, pp. 32-34. CRITTENDEN, J. "Aging and Party Affiliation." Public Opinion Quarterly, 1962, 26, 648-657.
Commission on Obscenity and Pornography. Report of the Commission on Obscenity and CRONBACH, L. Educational Psychology. New York: Harcourt, 1963.
Pornography. New York: Bantam, 1970. CRONBACH, L., AND MEEHL,P. "Construct Validity in Psychological Tests." Psychological
COMSTOCK, C. "Swing Voters: A Comparison of Attitudes on Leadership Style and on the Bulletin, 1955, 52, 281-302.
War in Viet Nam, Expressed by -Undecided Voters and by Supporters of McGovem CROSSMAN, R. (Ed.) T h e God That Failed. New York: Harper, 1949.
and Nixon." Berkeley, Calif.: Wright Institute, 1972. GROTTY, W. J. "Party Effort and Its Impact on the Vote." American Political Science
COMSTOCK, c., AND DUCKLES, R. "The Assessment of Destructiveness in Personality: Working Review, 1971, 65, 439-450.
Papers." Berkeley, Calif.: Wright Institute, 1972. CROW, W. J. "A Study of Strategic Doctrines Using the Inter-Nation Simulation." Journal
CONNELL, R. "Research and Thought on the Development of Political Beliefs in Children: A of ConfEict Resolution, 1963, 7, 580-589.
Critical Survey." Paper presented at meeting of Australasian Political Studies Associa- CRUTCHFIELD, R. "Character and Conformity." American Psychologist, 1955, 10, 191-198.
tion, Sydney, August 1969. CUMMINGS, E., AND HENRY,W. Growing Old: T h e Process of Disengagement. New York:
CONVERSE, P. E. "The Nature of Belief Systems in Mass Publics." I n D. E. Apter (Ed.), Basic Books, 1961.
Ideology and Discontent. New York: Free Press, 1964. CUTLER,N. "Generation, Maturation and Party Affiliation: A Cohort Analysis." Public
CONVERSE, P. E. "Of Time and Partisan Stability." Comparative Political Studies, 1969, 2, Opinion Quarterly, 1970, 33,583-588.
139-171. CUTLER, S. J. "Membership in Voluntary Associations and the Theory of Mass Society."

CONVERSE, P. E. "Attitudes and Non-Attitudes: Continuation of a Dialogue." I n E. Tufte Unpublished Doctoral dissertation, University of Michigan, 1969.
(Ed.), T h e Quantitative Analysis of Social Problems. Reading, Mass.: Addison- CUTRIGHT, P. "Measuring the Impact of Local Party Activity on the General Election
Wesley, 1970. Vote." Public Opinion Quarterly, 1963, 27, 372-386.
CONVERSE, P. E. "Non-voting Among Young Adults in the United States." In W. Crotty, D. CUTRIGHT, P., AND ROSSI, P. H. "Grass Roots Politicians and the Vote." American Sociologi-
Freeman, and D. Gatlin (Eds.), Political Parties and Political Behavior. Boston: cal Review, 1958, 23, 171-179.
Allyn and Bacon, 1971. CYERT,R. M., AND MARCH,J. G. A Behavioral Theory of the Firm. Englewood Cliffs, N.J.:
CONVERSE, P. E. "Public Opinion and Voting Behavior." I n F. I. Greenstein and N. W. Prentice-Hall, 1963.
Polsby (Eds.), T h e Handbook of Political Science. Vol. 6. Reading, Mass.: Addison- DAHL, R. A. "The Behavioral Approach in Political Science: Epitaph for a Monument to a
Wesley, forthcoming. Successful Protest." American Political Science Review, 1961, 55, 763-772.
CONVERSE, P. E., AND DUPEUX, G. c r P ~ l i t i ~ i ~ a of
t i othe
n Electorate in France and the United DAVIDS, A. "Comparison of Three Methods of Personality Assessment: Direct, Indirect and
States." Public Opinion Quarterly, 1962, 26, 1-23. Projective." Journal of Personality, 1955, 23, 432-440.
CONVERSE, P. E., AND SCHUMANN, H. "Silent Majorities and the Vietnam War." Scientific DAVIDS, A., AND PILDNER, H. ' c C ~ m p a r i ~of~ Direct
n and Projective Methods of Personality
American, 1970, 222, 17-25. Assessment Under Different Conditions of Motivation." Psychological Monographs,
CONWAY, M. M., AND FEIGERT, F. B. "Motivation, Incentive Systems, and the Political Party I 1958, 72, whole No. 464.
Organization." American Political Science Review, 1968, 62, 1159-1 173. DAVIES, A. F. Private Politics: A Study of Five Political Outlooks. Melbourne: Melbourne
COOMBS, S. L., FRIED, M., AND ROBINOVITZ, S. H. "An Approach to Election Simulation Through University Press, 1966.
Modular Systems." I n W. D. Coplin (Ed.), Simulation in the Study of Politics. DAVIES,A. F. "Criteria for the Political Life History." Historical Studies of Australia and
Chicago: Markham, 1968. New Zealand, 1967,13(49), 76-85.
482 References References 483

DAVIES, A. "The Child's Discovery of Nationality." Australian and New Zealand Journal of Political Learning. Technical Report No. 63. Madison: Wisconsin Research and
Sociology, 1968, 4, 107-125. Development Center for Cognitive Learning, 1968.
DAVIES, J. C. "The Political Implications of Psychoanalytic and Academic Psychology." Un- DENNIS, J., AND MC CRONE, D. "Preadult Development of Political Party Identification in
published doctoral dissertation, University of California, Berkeley, 1952. Western Democracies." Comparative Political Studies, 1970, 3, 243-263.
DAVIES, J. C. "Toward a Theory of Revolution." American Sociological Review, 1962, 27, DERIVERA, J. T h e Psychological Dimension of Foreign Policy. Columbus, Ohio: Merrill,
5-19. 1968.
DAVIES, J. C. Human Nature in Politics. New York: Wiley, 1963. DEUTSCH, K. W. T h e Nerves of Government. New York: Free Press, 1963.
DAVIES, J. C. "The Family's Role in Political Socialization." Annals of the American DEUTSCH, M. "A Theory of Cooperation and Competition." Human Relations, 1949, 2, 129-
Academy of Political and Social Science, 1965, 361, 10-19 (September). 152.
DAVIES, J. C. "The J-Curve of Rising and Declining Satisfactions as a Cause of Some Great DEUTSCH, M. "Trust, Trustworthiness, and the F Scale." Journal of Abnormal and Social
Revolutions and a Contained Rebellion." I n H. D. Graham and T. R. Gurr (Eds.), Psychology, 1960, 61, 138-141.
T h e History of Violence in America. New York: Bantam, 1969. DEUTSCH, M., AND KRAUSS, R. M. "The Effect of Threat on Interpersonal Bargaining."
DAVIES, J. C. "Violence and Aggression: Innate or Not?" Western Political Quarterly, 1970, Journal of Abnormal and Social Psychology, 1960, 61, 181-189.
23, 611-623. DEUTSCHER, I. Stalin: A Political Biography. New York: Oxford University Press, 1949.
DAVIES, J. C. (Ed.) When Men Revolt and Why. New York: Free Press, 1971. DEVEREAUX, G . ''Charismatic Leadership and Crisis." Psychoanalysis and the Social Sciences,
DAVIES, J. C. "The Need Hierarchy and Political Theory." Unpublished manuscript, 1972. 1955, 4, 145-157.
DAVOL, S. H., AND REIMANS, G. "The Role of Anomie as a Psychological Concept." Journal DE VOTO, B. "The Skeptical Biographer." Harper's, 1933,166, 181-192.
of Individual Psychology, 1959, 15, 215-225. DEXTER, L. A. T h e Sociology and Politics of Congress. Chicago: Rand McNally, 1969.
DAWSON, P. A. "The Structural Nature of Attitudes: An Experimental Investigation of Atti- DEXTER, L. A. Elite and Specialized Interviewing. Evanston, Ill.: Northwestern University
tudes Toward the War in Viet Nam." Experimental Study of Politics, 1971, 1, 61-86. Press, 1970.
DAWSON, R. E., AND PREWITT, K. Political Socialization. Boston: Little, Brown, 1969. DICKS, H. V. "The Authoritarian Personality: A Critical Appreciation." Human Relations,

DAY, R. H., AND TINNEY,E. H. "HOWto Co-operate in Business Without Really Trying: A 1951, 4, 203-211.
Learning Model of Decentralized Decision Making." Journal of Political Economy, DICKS, H. V. Licensed Mass Murder: A Socio-Psychological Study of Some S.S. Killers. New

1968, 76, 583-600. York: Basic Books, 1972.

DIES, R. "Development of a Projective Measure of Perceived Locus of Control." Journal of
DEAN, D. G. "Alienafion: Its Meaning and Measurements." American Sociological Review,
1961, 26, 753-758. Projective Techniques and Personality Assessment, 1968, 32,487-490.
DIETZEL, c., AND ABELES, N. "Thematic Drive Expression and Self-Esteem." Journal of
DE CHARMS, R. Personal Causation: T h e Internal Affective Determinants of Behavior. New
York: Academic Press, 1968. Personality Assessment, 1971, 35, 442-447.
DI PALMA, G., AND MC CLOSICY, H. "Personality and Conformity: The Learning of Political
DE CHARMS, R., MORRISON, H., REITMAN, w., AND MC CLELLAND, D. "Behavioral Correlates of
Attitudes." American Political Science Review, 1970, 64, 1054-1073.
Directly and Indirectly Measured Achievement Motivation." In D. McClelland
DI RENZO, G. J. Personality, Power and Politics: A Social Psychological Analysis of the Italian
(Ed.), Studies in Motivation. New York: Appleton-Century, 1955.
Deputy and His Parliamentary System. Notre Dame, Ind.: University of Notre Dame
DE FLEUR,M., AND DE FLEUR,L. "The Relative Contribution of Television as a Learning
Press, 1967.
Source for Children's Occupational Knowledge." American Sociological Review,
DOLLARD, J. Criteria for the Life History. New Haven: Yale University Press, 1935.
1967,32, 777-789. DOLLARD, J., DOOB,L. w., MILLER,N. E., MOWRER, O. H., AND SEARS,R. R. Frustration and
DEGRAZIA, S. "Mahatma Gandhi: The Son of His Mother." Political Quarterly, 1948a,
Aggression. New Haven: Yale University Press, 1939.
19, 336-348. DOLLARD, J., AND MOWRER, O. H. ''A Method of Measuring Tension in Written Documents."
DE GRAZIA, S. T h e Political Community: A Study of Anomi'e. Chicago: University of Chicago
Journal of Abnormal and Social Psychology, 1947, 42, 3-32.
Press, 1948b. DONALD, D. H. "Between History and Psychology: Reflections on Psychobiography." Paper
DE LAMATER, J., KATZ,D., AND KELMAN, H. C. "On the Nature of National Involvement: A
presented at meeting of American Psychiatric Association, Dallas, May 1972.
Preliminary Study." Journal of Conflict Resolution, 1969, 13, 320-357.
noon, L. W. "The Behavior of Attitudes." Psychological Review, 1947, 54, 135-156.
DELGADO, J. M. R. Physical Control of the Mind. New York: Harper, 1969.
DOOB, L. W. (Ed.) Resolving Conflict in Africa: T h e Fermeda Workshop. New Haven: Yale
DENNIS,J. Political Learning in Childhood and Adolescence: A Study of Fifth, Eighth, and University Press, 1970.
Eleventh Grades in Milwaukee, Wisconsin. Technical Report No. 98. Madison: DOOB, L. W. "The Impact of the Fermeda Workshop on the Conflicts in the Horn of Africa."
University of Wisconsin Research and Development Center for Cognitive Learning, International Journal of Group Tensions, 1971, 1,91-101.
1969. DOOLEY, L. "Psychoanalytic Studies .of Genius." American Journal of Psychology, 1916, 27,
DENNIS,J. A Survey and Bibliography of Contemporary Research on Political Learning and 363416.
Socialization. Occasional Paper No. 8. Madison: University of Wisconsin Research DOUVAN, E., AND WALKER, A. M. "The Sense of Effectiveness in Public Affairs." Psychological
and Developnlent Center for Cognitive Learning, 1967; updated version, 1971. Monographs, 1956, 70 (no. 22), 1-19.
DENNIS,J. Socialization to Politics: A Reader. New York: Wiley, 1973. DOWNS, A. A n Economic Theory of Democracy. New York: Harper, 1957.
DENNIS, J., BILLINGSLEY, K., AND THORSON, S. A Pilot Experiment in Early Childhood DRIVER, M. J. "Conceptual Structure and Group Processes in an Inter-Nation Simulation.
 References References 485

ELDERSVELD, S. J., AND DODGE, R. W. "Personal Contact or Mail Propaganda? An Experiment

Part One: The Perception of Simulated Nations." Educational Testing Service
Research Bulletin, 1962, RB 62-15. in Voting Turnout and Attitude Change." In D. Katz, D. Cartwright, S. EldersvekI,
DRIVER, M. J. A Structural Analysis of Aggression, Stress and Personality in an Inter-Nation
and A. M. Lee (Eds.), Public Opinion and Propaganda. New York: Dryden, 1954.
ELMS, A., AND MILGRAM, S. ccPers~nality Characteristics Associated with Obedience and
Simulation. Institute Paper No. 97. Lafayette, Ind.: Institute for Research in the
Eehavioral, Economic and Management Sciences, Purdue University, 1965. Defiance Toward Authoritative Command." Journal of Experimental Research in
DRUCKMAN, D. "Ethnocentrism in the Inter-Nation Simulation." Journal of Conflict Resolu-
Personality, 1966, 1, 282-289.
EMSHOFF, J. R. "A Computer Simulation Model of the Prisoner's Dilemma." Behavioral
tion, 1968, 12, 45-68.
DUBERMAN, M. "Politics and the Man," New York Times, December 10, 1967.
Science, 1970, 15, 304-3 17.
ENDLEMAN, R. "Review of T h e University Crisis Reader." Contemporary Sociology, 1972,
DULLES,A. W. "A Foreign-Affairs Scholar Views the Real Woodrow Wilson." Look, 13
December 1966. 1, 3-9.
ENGLE, B., AND FRENCH, T. M. "Some Psychodynamic Reflections upon the Life of Solon."
DURBIN, E. F. M., AND BOWLBY, J. Personal Aggressiveness and War. London: Kegan Paul,
1939. Psychoanalytic Quarterly, 1951, 20, 253-273.
ERBE, W. "Social Involvement and Political Activity: A Replication and Elaboration."
DURKHEIM, E. T h e Division of Labor in Society. New York: Free Press, 1933. Originally
published 1893. American Sociological Review, 1964, 29, 198-215.
DURKHEIM, E. Suicide. New York : Free Press, 1951. Originally published 1897.
ERIKSEN,C. "Needs in Perception and Projective Techniques." Journal of Projective Tech-
DUTTON, J. M., AND STARBUCK, w. H. Computer Simulation of Human Behavior. New York: niques, 1954, 18, 435440.
Wiley, 1971. ERIKSON, E. H. "Hitler's Imagery and German Youth." Psychiatry, 1942, 5, 475493.

DYE, T. R., AND ZEIGLER, L. H. T h e Irony of Democracy. Belmont, Calif.: Wadsworth, 1970. ERIKSON, E. H. Childhood and Society. (2nd ed.) New York: Norton, 1963. Originally
DYMOND, R. "Can Clinicians Predict Individual Behavior?" Journal of Personality, 1954, 61, published 1950.
339-342. ERIKSON, E. H. "The Problem of Ego Identity." Journal of the American Psychoanalytic
DYSON, J. w., AND SCIOLI, F. P. "An Experimental Test of Consistency Theory." Exferimental Association, 1956, 4, 58-121.
Study of Politics, 1971, 1, 118-138. ERIKSON, E. H. Young Man Luther: A Study in Psychoanalysis and History. New York:

EASTON, D. A Systems Analysis of Political Life. New York: Wiley, 1965. Norton, 1958.
EASTON, D., AND DENNIS, J. "The Child's Acquisition of Regime Norms." American Political ERIKSON, E. H. "The Strange Case of Freud, Bullitt, and Woodrow Wilson: I (A Dubious
Science Review, 1967, 61, 25-38. Collaboration) ." New York Review of Books, 9 February 1967.
EASTON,D., AND DENNIS, J. Children in the Political System. New York: McGraw-Hill, ERIKSON, E. H. "On the Nature of Psycho-Historical Evidence: In Search of Gandhi."
1969. Daedalus, 1968, 97, 695-730.
EASTON,D., AND HESS, R. D. "The Child's Political World." Midwest Journal of Political ERIKSON, E. H. Gandhi's Truth. New York: Norton, 1969.

Science, 1962, 6, 229-246. ERNST, M. L., AND LOTH,D. Report on the American Communist. New York: Holt, 1952.
ECKHARDT, w., AND LENTZ,T. F. "Factors of War/Peace Attitudes." Peace Research Reviews, ETZIONI,A. T h e Active Society. New York: Free Press, 1968.
1967, I ( 5 ) . ETZIONI,A. c ' S ~ ~ i a l - P ~ y ~ h o l oAspects
g i ~ a l of International Relations." In G. Lindzey and E.
ECKSTEIN, H. "Case-Study and Theory in Political Science." In F. I. Greenstein and N. W. . .. . Mass.
Aronson (Eds.), T h e Handbook of Social Psychology. (2nd Ed.) Vol. 5. Reading,
: Addison-Wesley, 1969.
Polsby (Eds.), T h e Handbook of Political Science. Reading, Mass.: Addison-Wesley,
forthcoming. EULAU, H., BUCHANAN, w., FERGUSON, L. c., AND WAHLKE, J. C. "Career Perspectives of
EDEL, L. "The Biographer and Psycho-Analysis." International Journal of Psycho-Analysis, American State Legislators." In D. Marvick (Ed.), Political Decision Makers. New
1961, 42, 45%466. York: Free Press, 1961.
EDINGER, L. J. "Political Science and Political Biography: Reflections on the Study of EULAU, H., AND SCHNEIDER, P. "Dimensions of Political Involvement." Public Opinion
Leadership." Journal of Politics, 1964, 26, 423-439, 648-676. Qua~terly,1956, 20, 128-142.
EDINGER, L. J. Kurt Schumacher: A Study in Personality and Political Behavior. Stanford, EVAN,W. "Cohort Analysis of Survey Data: A Procedure for Studying Long Term Opinion
Calif. : Stanford University Press, 1965. Change." Public Opinion Quarterly, 1959; 23, 63-72.
EDWARDS, A. L. "Political Frames of Reference as a Factor lnfluencing Recognition." Journal EYSENCK, H. J. T h e Psychology of Politics. London: Routledge and Keagan Paul, 1954a.
of Abnormal and Social Psychology, 1941, 36, 34-50.
- -
EYSENCK, H. J. "The Science of Personality: Nomothetic!" Psychological Review, 1954b,
EDWARDS, A. L. Experimental Design in Psychological Research. (Rev. Ed.) New York: Holt,
61, 339-342.
1960.
FAGEN,R. "Cuba: The Political Content of Adult Education." In R. Sigel (Ed.), Learning
EDWARDS, L. P. T h e Natural History of Revolution. Chicago: University of Chicago Press,
About Politics. New York: Random House, 1970.
1927.
FALK, J. L. L ' I ~Distinguishing
~ ~ e ~ Idiographic from Nomothetic Approaches to Personality
EINSTEIN,A., AND FREUD,S. W h y War? Paris: International Institute of Intellectual Co-
Theory." Psychological Review, 1956, 63, 53-62.
operation, 1932.
FANON, F. T h e Wretched of the Earth. New York: Grove Press, 1968.
ELDERSVELD, S. J. ''Experimental Propaganda Techniques and Voting Behavior." American
FARBER, M. L. "The Problem of National Character: A Methodological Analysis." Joul'nal of
Political Science Review, 1956, 50, 154-165.
ELDERSVELD, S. J. Political Parties. Chicago: Rand McNally, 1964.
Psychology, 1950, 30, 307-316.
486 References References 487

FARRELL,B. A. Introduction. Sigmund Freud, Leonardo Da Vinci and a Memory of His FRANCIS, W. L. "Political Process Simulations." I n H. Guetzkow, P. Kotler, and R. L.
Childhood. Harmondsworth, Middlesex, England: Penguin Books, 1963. Schultz (Eds.), Simulation in Social and Administrative Science. Englewood Cliffs,
FEARING, F. ''Psychological Studies of Historical Personalities." Psychological Bulletin, 1927, N.J. : Prentice-Hall, 1972.
24, 521-539. FRANK, J. D. Persuasion and Healing: A Comparative Study of Psychotherapy. Baltimore:
FEIERABEND, I. K., AND FEIERABEND, R. L. c ' A g g r e ~ ~Behaviors
i~e Within Polities, 1948-1962. Johns Hopkiis Press, 1961.
A Cross-National Study." Journal of Conflict Resolution, 1966, 10, 249-271. FRANK, L. K. "Projective Methods for the Study of Personality." Journal of Psychology,
FEIERABEND, I. K., AND FEIERABEND, R. L. "Systematic Conditions of Political Aggression: 1939, 8, 389413.
An Application of Frustration-Aggression Theory." I n I. K. Feierabend, R. L. Feiera- FREE, L. "Gauging Thresholds of Frustration." I n J. C. Davies (Ed.), When Men Revolt and
bend, and T. R. Gurr (Eds.), Anger, Violence, and Politics. Englewood Cliffs, N.J.: Why. New York: Free Press, 1971.
Prentice-Hall, 1972. FREEDMAN, J. L., AND SEARS,D. O. "Selective Exposure." I n L. Berkowitz (Ed.), Aduances
FEIERABEND, I. K., FEIERABEND, R. L., AND GURR, T. R. (Eds.) Anger, Violence, and Politics. in Experimental Social Psychology. Vol. 2. New York: Academic Press, 1965.
Englewood Cliffs, N.J. : Prentice-Hall, 1972. FREEDMAN, R., HAWLEY, A. H., LANDECKER, W. s., LENSKI, G. E., AND MINER, H. M. Principles
FEIERABEND, I. K., FEIERABEND, R. L., AND NESVOLD, B. A. "Social Change and Political of Sociology. (Rev. Ed.) New York: Holt, 1956.
Violence: Cross-National Patterns." I n H. D. Graham and T. R. Gurr (Eds.), T h e FREEMAN, L. ''TWOProblems in Computer Simulation in the Social and Behavioral Sciences."
History of Violence in America. New York: Bantam Books, 1969. Social Science Information, 1971, 10, 103-112.
FELDMAN, K., AND NEWCOMB, T. T h e Impact of College on Students. San Francisco: Jossey- FRENCH,J. R. P., AND RAVEN,B. H. "The Bases of Social Power." I n D. Cartwright (Ed.),
Bass, 1969. Studies in Social Power. Ann Arbor: University of Michigan Press, 1959.
FELDT,L. S. "A Comparison of the Precision of Three Experimental Designs Employing a FRENKEL-BRUNSWIK, E. "Further Explorations by a Contributor." I n R. Christie and M.
Concomitant Variable." Psychometrika, 1958, 23, 335-353. Jahoda (Eds.), Studies in the Scope and Method of "The Authoritarian Personality."
FENDRICH, J. M., AND AXELSON, L. J. "Marital Status and Political Alienation Among Black New York: Free Press, 1954.
Veterans." American Journal of Sociology, 1971, 77, 245-261. FRENKEL-BRUNSWIK, E., AND SANFORD, R. N. "Some Personality Factors in Anti-Semitism."
FERGUSON, G. A. Statistical Analysis in Psychology and Education. New York: McGraw-Hill, Journal of Psychology, 1945, 20, 271-291.
1959. FREUD,S. Leonard0 da Vinci: A Psycho-Sexual Study of an Infantile Reminiscence. New
FESHBACH, S. "The Influence of Drive Arousal and Conflict upon Fantasy Behavior." In York: Dodd, Mead, 1932. Originally published 1910.
J. Kagan and G. Lesser (Eds.), Contemporary Issues in Thematic Apperceptive FREUD,S. Group Psychology and the Analysis of the Ego. London: Hogarth, 1922.
Methods. Springfield, Ill.: Thomas, 1961. FREUD, S. Ciuilization and Its Discontents. London: Hogarth, 1930.
FESTINGER,L. A Theory of Cognitive Dissonance. Stanford, Calif.: Stanford University FREUD, S. New Introductory Lectures in Psycho-Analysis. London: Hogarth, 1933,.
Press, 1957. FREUD, s., AND BULLITT, W. C. Thomas Woodrow Wilson: A Psychological Study. Boston:
FESTINGER, L., AND CARLSMITH, J. M. "Cognitive Consequences of Forced Compliance." Houghton Mifflin, 1967.
Journal of Abnormal and Social Psychology, 1959,58,203-210. FREY, F. "Surveying Peasant Attitudes in Turkey." Public Opinion Quarterly, 1963, 27,
FIEDLER, F. E. "The Contingency Model: A Theory of Leadership Effectiveness." I n H. 335-355.
Proshansky and B. Seidenberg (Eds.), Basic Studies in Social Psychology. New York: FRIEDRICH, C. J., AND CHAPMAN, J. (Eds.) Nomos V I : Justice. New York: Atherton, 1963.
Holt, 1965. FRIJDA, N. H. c ' P r ~ b l e mof~ Computer Simulation." Behavioral Science, 1967, 12, 59-67.
FIEDLER, F. E. A Theory of Leadership Effectiveness. New York: McGraw-Hill, 1967. FROMM,E. Escape from Freedom. New York: Holt, 1941.
FIELD,J., AND ANDERSON, R. "Ideology in the Public's Conceptualization of the 1964 Elec- FROMM,E. Man for Himself. New York: Holt, 1947.
tion." Public Opinion Quarterly, 1969, 33, 380-398. FROMM,E. T h e Sane Society. Greenwich: Fawcett Publications, 1955.
FIELDS, R. "Conversations with Children Under Siege." Unpublished manuscript, 1972. FROMM, E., AND MACCOBY, M. Social Character in a Mexican Village: A Sociopsychoanalytic
FINIFTER, A. W. "Dimensions of Political Alienation." American Political Science Review, Study. Englewood Cliffs, N.J. : Prentice-Hall, 1970.
1970, 64, 389-410. FROMME, A. "On the Use of Certain Qualitative Methods of Attitude Research: A Study
FISHBEIN, M. "An Investigation of the Relationship Between Beliefs About an Object and of Opinions on the Methods of Preventing War." Journal of Social Psychology,
the Attitude Toward That Object." Human Relations, 1963, 16, 233-240. 1941, 13,429-459.
FISHBEIN,M. "Attitude and the Prediction of Behavior." I n M. Fishbein (Ed.), Readings in FUESS,C. M. "Debunkery and Biography." Atlantic, 1933, 151, 347-356.
Attitude Theory and Measurement. New York: Wiley, 1967. GABENNESCH, H. "Authoritarianism as World View." American Journal of Sociology, 1972,
FISHER, R. A. Statistical Methods for Research Workers. London: Oliver and Boyd, 1925. 77, 857-875.
FISHER, R. A. T h e Design of Experiments. London: Oliver and Boyd, 1935. GAGE,N., AND CHATTERGEE, B. "The Psychological Meaning of Acquiescence Set: Further
FISHER,R. J. ''Third Party Consultation: A Method for the Study and Resolution of Con- Evidence." Journal of Abnormal and Social Psychology, 1960, 60, 280-283.
flict." Journal of Conflict Resolution, 1972, 16, 67-94. GALTUNG, J. "Social Position, Party Identification and Foreign Policy Orientation: A Nor-
FISHMAN, J. A. Language and Nationalism. Rowley, Mass.: Newbury, 1972. wegian Case Study." I n J. N. Rosenau (Ed.), Domestic Sources of Foreign Policy.
FLACKS, R. "The Liberated Generation: An Exploration of the Roots of Student Protest." New York: Free Press, 1967.
Journal of Social Issues, 1967, 23, 52-75. GALTUNG, J. "Foreign Policy Opinion as a Function of Social Position." In J. N. Rosenau
488 References References 489
/
(Ed.), International Politics and Foreign Policy. (2nd Ed.) New York: Free Press, GOLDBERG, s., AND STERN, G. "The Authoritarian Personality and Education." American
1969. 1
Psychologist, 1952, 7, 372-375.
GAMSON, W. A. Power and Discontent. Homewood, Ill.: Dorsey Press, 1968. GOLDHAIVIER, H. c ' P ~ b l iOpinion
~ and Personality." American Journal of Sociology, 1950,
GAMSON, W. A., AND MODIGLIANI, A. Untangling the Cold War: A Strategy for Testing Rival 55,346-354.
Theories. Boston: Little, Brown, 1971. GORDEW, R. L. Interviewing: Strategy, Techniques, and Tactics. Homewood, Ill.: Dorsey Press,
GARCIA, c. "An Inquiry into the Development of Political System Values Among Mexican- 1969.
American Children." Unpublished manuscript, University of New Mexico, 1972. GORDON, L. "A Comment on Political Orientation and Riot Participation." American
GARDNER, R., HOLZMAN, P., KLEIN,G., LINTON,H., AND SPENCE, D. "Cognitive Control, A Sociological Review, 1972, 37, 379.
Study of Individual Consistencies in Cognitive Behavior." Psychological Issues, 1959, GOSNELL, H. F. Getting Out the Vote. Chicago: University of Chicago Press, 1927.
1, No. 4. GOTTERSFELD, H., AND DOZIER, G. "Changes in Feelings of Powerlessness in a Community
GARRATY, J. A. "The Interrelations of Psychology and Biography." Psychological Bulletin, Action Program." Psychological Reports, 1966, 10, 978.
1954,51,569-582. GOTTFRIED, A. BOSSCermak of Chicago. Seattle: University of Washington Press, 1962.
GARRATY, J. A. T h e Nature of Biography. New York: Knopf, 1957. GOTTSCHALK, L., KLUCKHOHN, c., AND ANGELL, R. T h e Use of Personal Documents in
GEORGE, A. L. "Power as a Compensatory Value for Political Leaders." Journal of Social History, Anthropology and Sociology. New York: Social Science Research Council,
Issues, 1968, 24, 2949. n.d.
GEORGE, A. L. "The 'Operational Code': A Neglected Approach to the Study of Political GOUGH, H. G. California Psychological Inventory. Palo Alto, Calif.: Consulting Psychologists
Leaders and Decision-Making." International Studies Quarterly, 1969, 13, 190-222. Press, 1956.
GEORGE, A. L. "Some Uses of Dynamic Psychology in Political Biography: Case Materials GOUGH,H., AND SANFORD, N. "Rigidity as a Psychological Variable." Mimeographed. Institute
on Woodrow Wilson." I n F. I. Greenstein and M. Lemer (Eds.), A Source Book for of Personality Assessment and Research, University of California, Berkeley, 1952.
the Study of Personality and Politics. Chicago: Markham, 1971. GRAHAM, H. D., AND GURR, T. R. (Eds.) T h e History of Violence in America. Report to the
GEORGE, A. L. "The Case for Multiple Advocacy in Making Foreign Policy." American National Violence Commission. New York: Bantam, 1969.
Political Science Review, 1972, 66, 751-785. GRANBERG, D. "War Expectancy and the Evaluation of a Specific War." Journal of ConfEict
GEORGE, A. L., AND GEORGE, j. L. Woodrow Wilson and Colonel House: A Personality Study. Resolution, 1969, 13, 546549.
New York : Dover, 1956. GREEN, R., AND STACEY, B. "A Flexible Projective Technique Applied to the Measurement of
GERTZOG, I. N. "The Electoral Consequences of a Local Party Organization's Registration the Self-Image of Voters." Journal of Projective Techniques and Personality Assess-
Campaign: The San Diego Experiment." Polity, 1970, 3, 247-264. ment, 1966, 30, 12-15.
GERWIN, D. Budgeting Public Funds. Madison: University of Wisconsin Press, 1969. GREENACRE, P. Swift and Carroll. New York: International Universities Press, 1955.
GERKE,O. Political Theories of the Middle Age. Trans. F. W. Maitland. Boston: Beacon, GREENBERG, E. "Children and the Political Community: A Comparison Across Racial Lines:'
1958. Canadian Journal of Political Science, 1969,2,471-492.
GILBERT, D., AND LEVINSON, D. "Role Performance, Ideology and Personality in Mental GREENBERG, E. "Black Children and the Political System." Public Opinion Quarterly, 1970a,
Hospital Aides." I n M. Greenblatt and others (Eds.), T h e Patient and the Mental 34,333-345.
Hospital. New York: Free Press, 1957. GREENBERG, E. "Children and Government: A Comparison Across Racial Lines." Midwest
GILBERT, G. M. T h e Psychology of Dictatorship. New York: Ronald Press, 1950. Journal of Political Science, 1970b, 14,249-275.
GILSON,c., AND ABELSON, R. P. "The Subjective Use of Inductive Evidence." Journal of GREENBERG, E. (Ed.) Political Socialization. New York: Atherton, 1970c.
Personality and Social Psychology, 1965, 2, 301-310. GREENBLATT, M., LEVINSON, D. j., AND WILLIAMS, R. H. (Eds.) T h e Patient and the Mental
GLAD, B. Charles Evans Hughes and the Illusions of Innocence. Urbana: University of Hospital. New York: Free Press, 1957.
Illinois Press, 1966. GREENSTEIN, F. I. "The Benevolent Leader: Children's Images of Political Authority."
GLAD, B. "Review of C. Beradt, T h e Third Reich of Dreams." American Political Science American Political Science Review, 1960, 54,934-943.
Review, 1968, 62, 545. GREENSTEIN, F. I. Children and Politics. New Haven: Yale University Press, 1965.
GLAD, B. "The Significance of Personality for Role Performance as Chairman of the Senate GREENSTEIN, F. I. "The Impact of Personality on Politics: An Attempt to Clear Away the
Foreign Relations Committee: A Comparison of Borah and Fulbright." Paper Underbrush." American Political Science Review, 1967, 61, 629-641.
presented at annual meeting of American Political Science Association, New York, GREENSTEIN, F. I. "Harold D. Lasswell's Concept of Democratic Character." Journal of
1969. Politics, 1968a, 30, 69G709.
GLENN,N. "Aging, Disengagement and Opinionation." Public Opinion Quarterly, 1969, GREENSTEIN, F. I. "Socialization: Political Socialization." I n D. Sills (Ed.), International
33, 17-33. Encyclopedia of the Social Sciences. New York: Macmillan, 1968b.
GLENN,N. "Problems of Comparability and Trend Studies with Opinion Poll Data." Public GREENSTEIN, F. I. Personality and Politics. Chicago: Markham, 1969.
Opinion Quarterly, 1970,34, 82-91. GREENSTEIN, F. I. "A Note on the Ambiguity of 'Political Socialization': Definitions, Criti-
GLENN,N., AND GRIMES,M. "Aging, Voting and Political Interest." American Sociology cisms and Strategies of Inquiry." Journal of Politics, 1970a, 32, 969-977.
Review, 1968, 33, 563-575. GREENSTEIN, F. I. "Review of Personality, Power and Politics." Political Science Quarterly,
GLOCK, C. (Ed.) Survey Research in the Social Sciences. New York: Russell Sage, 1967. 1970b, 85,365-368.
490 References References 491

GREENSTEIN, F. I. "The Study of Personality and Politics: Overall Considerations." In F. I. GYNTHER, M. D. "White Norms and Black MMPIs: A Prescription for Discrimination."

Greenstein and M. Lerner (Eds.), A Source Book for the Study of Personality and Psychology Bulletin, 1972, 78, 38f3-402.
Politics. Chicago: Markham, 1971. HAAS,E. B. Beyond the Nation-State: Functionalism and International Organization. Stan-
GREENSTEIN, F. I., AND LERNER, M. A Source Book for the Study of Personality and Politics. ford, Calif.: Stanford University Press, 1964.
Chicago: Markham, 1971. HAIRE, M. "Projective Techniques in Marketing Research." Journal of Marketing, 1950,

GREENSTEIN, F. I., AND TARROW, S. "The Study of French Political Socialization." World 649-656.
Politics, 1969, 22, 95-137. HALL, C . s., AND LINDZEY, G. Theories of Personality. New York: Wiley, 1957.

GREENSTEIN, F. I., AND TARROW, S. "Political Orientations of Children: The Use of a Semi- HALL, E. "Will Success Spoil B. F. Skinner?" An Extensive Interview. Psychology Today,

Projective Technique in Three Nations." Sage Professional Papers in Comparative - 1972, 6, 65ff.
Politics 01-0009, 1, 479-558. Beverly Hills: Sage, 1970. HAMBLIN, R. L. "The Dynamics of Racial Discrimination." Social Problems, 1962, 10, 103-

GREGG, L. w., AND SIMON,H. A. "Process Models and Stochastic Theories of Simple Concept 121.
Formation." Journal of Mathematical Psychology, 1967, 4, 246-276. HAMILTON, R. "A Research Note on the Mass Support for 'Tough' Military Initiatives."

GRONSETH, E. "The Impact of Father Absence in Sailor Families on the Personality Structure American Sociological Review, 1968, 33,439-445.
and Social Adjustment of Adult Sailor Sons." In N. Anderson (Ed.), Studies of the HANFMANN, E., AND GETZELS, J. "Interpersonal Attitudes of Former Soviet Citizens, as
Family. Vol. 2, Part 1. Gottingen: Vanderhoeck and Ruprecht, 1957. Studied by a Semi-projective Method." Psychological Monographs, 1955, 389, whole
GUEDALLA, P. "The Method of Biography." Journal of the Royal Society of Arts, July 1939, issue.
pp. 925-936. HANNA, J. F. L'Information-The~reti~ Techniques for Evaluating Simulation Models." I n
GUETZKOW, H. Multiple Loyalties. Princeton: Princeton University Press, 1955. J. M. Dutton and W. H. Starbuck (Eds.), Computer Simulation of ~ u m a Behavior.
n
GUETZKOW, H. (Ed.) Simulation in Social Scz'ence: Readings. Englewood Cliffs, N.J.: New York: Wiley, 1971.
Prentice-Hall, 1962. HARGROVE, E. C. Presidential Leadership: Personality and Political Style. New York:

GUETZKOW, H. "Some Correspondences Between Simulations and 'Realities' in International Macmillan, 1966.
Relations." I n M. Kaplan (Ed.), New Approaches to International Relations. New HARLOW, H. F. "Mice, Monkeys, Men and Motives." Psychological Review, 1953, 60, 23-32.
York: St. Martin's, 1968. HARLOW, H. F., AND HARLOW, M. K. "Social Deprivation in Monkeys." Scientific American,
GUETZKOW, H. "Simulations in the Consolidation and Utilization of Knowledge About 1962, 207, 136-146.
International Relations." In D. G. Pruitt and R. C. Snyder (Eds.), Theory and HARLOW, H. F., AND SUOMI,S. J. "Induced Psychopathology in Monkeys." Engineering and
Research on the Causes of War. Englewood Cliffs, N.J.: Prentice-Hall, 1969. Science (Pasadena: California Institute of Technology), 1970, 33, 8-14.
GUETZKOW, H., ALGER,C. F., BRODY, R. A., NOEL,R. c., AND SNYDER, R. C. Simulation in In- HARLOW, H. F., AND ZIMMERMAN, R. R. "The Development of Affectional Response in Infant
ternational Relations: Developments for Research and Teaching. Englewood Cliffs, Monkeys." Proceedings of the American Philosophical Society, 1958, 102, 501-509.
N. J. : Prentice-Hall, 1963. HARNED,L. "Authoritarian Attitudes and Party Activity." Public Opinion Quarterly, 1961,
GUETZKOW, H., AND GYR,J. "An Analysis of Conflict in Decision-Making Groups." Human 25, 393-399.
Relations, 1954, 7, 367-381. HARRIS, D., GOUGH, H., AND MARTIN, W. E. "Children's Ethnic Attitudes: 11. Relationship to
GUETZKOW, H., KOTLER, P., AND SCHULTZ, R. L. (Eds.) Simulation in Social and Administra- Parental Beliefs Concerning Child Training." Child Development, 1950, 21, 169-181.
tive Science. Englewood Cliffs, N.J.: Prentice-Hall, 1972. HARRISON, R. "Thematic Apperceptive Methods." In B. Wolman (Ed.), Handbook of
GUGGENHEIM, F. "Self-Esteem and Achievement Expectations for White and Negro Children." Clinical Psychology. New York: McGraw-Hill, 1965.
Journal of Projective Techniques and Personal Assessment, 1969, 33, 63-71. HARTMAN, D. A. "The Psychological Point of View in History: Some Phases of the Slavery
GUILFORD, J. Personality. New York: McGraw-Hill, 1959.
Struggle." Journal of Abnormal Psychology, 1922-1923, 17, 261-273.
GULLAHORN, J., AND GULLAHORN, J. E. "A Computer Model of Elementary Social Behavior."
HARTMANN, G. W. "A Field Experiment on the Comparative Effectiveness of 'Emotional' and
Behavioral Science, 1963, 8, 354-362.
'Rational' Political Leaflets in Determining Election Results." Journal of Abnormal
GULLAHORN, J., AND GULLAHORN, J. E. "Computer Simulation of Role Conflict Resolution."
and Social Psychology, 1935, 31, 99-1 14.
I n J. M. Dutton and W. H. Starbuck (Eds.), Computer Simulation of Human
HARVEY, O. J., HUNT, D. E., AND SCHRODER, H. M. Conceptual Systems and Personality
Behavior. New York: Wiley, 1971.
Organization. New York: Wiley, 1961.
GURIN,G., VEROFF,J., AND FELD,S. Americans View Their Mental Health: A Nationwide
Interview Study. New York: Basic Books, 1959. HATHAWAY, s., AND MC KINLEY,J. C . Manual for the Minnesota Multiphasic Personality
GURIN, P. "Motivation and Aspirations of Southern Negro College Youth." American Journal
Inventory. (Rev. Ed.) New York: Psychological Corporation, 1951.
HAWLEY, W. D., MC CONAHAY, J. B., MC CONAHAY, S. F., NELSON, K., AND GRUBER, J. E. A n
of Sociology, 1970, 75, 607-631.
GURIN, P., GURIN, G., LAO, R. c., AND BEATTIE, M. "Internal-External Control in the Motiva- Evaluation of High School in the Community. New Haven: Educational Research
tional Dynamics of Negro Youth." Journal of Social Issues, 1969, 25, 29-53. Services, 1972.
HAY, J., AND NICOLAY, J. G . Abraham Lincoln. Chicago: University of Chicago Press, 1966.
GURR, T. R. W h y Men Rebel. Princeton: Princeton University Press, 1970.
GUSFIELD, J . R. "Mass Society and Extremist Politics." American Sociological Review, 1962,
Originally published 1886.
27, 19-30. HAYS,W. L. Statistics for Psychologists. New York: Holt, 1963.
GUYTON, A. C. Textbook of Medical Physiology. (3rd Ed.) Philadelphia: Saunders, 1966.
HAYTHORN, w., COUCH, A., HAEFNER, D., LANGHAM, P., AND CARTER, L. "The Behavior of
492 References References 493

Authoritarian and Equalitarian Personalities in Groups." Human Relations, 1956a, HOLLANDER, E. P., AND JULIAN, J. P. "Studies in Leader Legitimacy, Influence and Innova-
9, 57-74. tion." I n L. Berkowitz (Ed.), Advances in Experimental Social Psychology. Vol. 5.
HAYTHORN, w., COUCH,A., HAEFNER, D., LANGHAM, P., AND CARTER, L. "The Effects of New York: Academic Press, 1970.
Varying Combinations of Authoritarianism and Equalitarianism in Leaders and HOLLINGSHEAD, A. Elmtown's Youth. New York: Wiley, 1949.
Followers." Journal of Abnormal and Social Psychology, 1956b, 53, 210-219. HOLST,E. VON, AND ST. PAUL, U. VON. "Electrically Controlled Behavior." Scientific Amer-
HEIDER, F. T h e Psychology of Interpersonal Relations. New York: Wiley, 1958. ican, March 1962, 206, 50-59.
HEIST,P. "Diversity in College Students' Characteristics." Journal of Educational Psychology, HOLSTI, O. R. "Comparative LoperationalCodes' of Recent U.S. Secretaries of State: John
1960, 33, 279-291. Foster Dulles." Paper presented at meeting of American Political Science Association,
HEMPEL,C . G. Aspects of Scientific Explanations. New York: Free Press, 1965. New York, September 1969.
HENNESSY, B. "Politicals and Apoliticals: Some Measurements of Personality Traits." Mid- HOLSTI, O. R., BRODY, R. A., AND NORTH, R. C. "The Management of International Crisis:
west Journal of Political Science, 1959,3, 336355. Affect and Action in American-Soviet Relations." I n D. G. Pruitt and R. C. Snyder
HENRY,W. "The Thematic Apperception Test." In A. Carr and others (Eds.), T h e (Eds.), Theory and Research on the Causes of War. Englewood Cliffs, N.J.: Prentice-
Prediction of Overt Behavior Through the Use of Projective Techniques. Springfield, Hall, 1969. Pp. 62-79.
Ill.: Thomas, 1960. HOLT, R. R. "The Nature of TAT Stories as Cognitive Products: A Psychoanalytic Approach."
HERMAN, S. N. Israelis and Jews: T h e Continuity of an Identity. New York: Random House, In J. Kagan and G. Lesser (Eds.), Contemporary Issues in Thematic Apperceptive
1970. Methods. Springfield, Ill.: Thomas, 1961.
HERMANN, C. F. "Validation Problems in Games and Simulations with Special Reference to HOLT, R. R. Unpublished working paper. Research Center for Mental Health, New York
Models of International Politics." Behavioral Science, 1967, 12, 216231. University, 1972.
HERMANN, C. F. Crises in Foreign Policy: A Simulation Analysis. Indianapolis: Bobbs- HOLTZMAN, W. "Personality Measurement: Personality Inventories." I n D. Sills (Ed.),
Merrill, 1969. International Encyclopedia of the Social Sciences. New York: Macmillan, 1968.
HERMANN, C. F., AND HERMANN, M. G. "An Attempt to Simulate the Outbreak of World HOLZBERG, J. D. "The Clinical and Scientific Methods: Synthesis or Antithesis?" Journal of
War I." American Political Science Review, 1967, 61, 400-416. Projectiue Techniques, 1957, 21, 227-242.
HERMANN, M. G. "Testing a Model of Psychological Stress." Journal of Personality, 1966, HOLZBERG, J. D. "The Clinical and Scientific Methods: Synthesis or Antithesis?" Journal of
34, 381-396. Personality, 1962, 30, 377403.
HERO,A., JR. c L L i b e r a l i ~ m - C ~ n ~ e ~Revisited:
a t i ~ m Foreign vs. Domestic Federal Policies, HOMANS, G. T h e Human Group. New York: Harcourt, 1950.
1937-1967." Public Opinion Quarterly, 1969, 33, 399-408. HOOK, S. T h e Hero in History. Boston: Beacon, 1943.
HERRING, E. P. T h e Politics of Democracy. New York: Holt, 1940. HORKHEIMER, M. (Ed.) Studien iiber Autoritat und Familie. Paris: Alcan, 1966.
HERSEY,J. T h e Algiers Motel Incident. New York: Bantam, 1968. HORNEY, K. T h e Neurotic Personality of Our Time. New York: Norton, 1937.
HERZON, F. D. "A Review of Acquiescence Response Set in the California F Scale." Social
HORNEY, K. Neurosis and Human Growth. New York: Norton, 1950.
Science Quarterly, June 1972, pp. 6678. HOROWITZ, E., AND HOROWITZ, R. 'cDe~elopment of Social Attitudes in Children." Sociometry,
HESS, R. "The Socialization of Attitudes Toward Political Authority: Some Cross National
1938,1,301-338.
Comparisons." International Social Science Journal, 1963, 15, 542-559.
HOROWITZ, I. L., AND LIEBOWITZ, M. "Social Deviance and Political Marginality: Toward a
HESS, R., AND EASTON, D. "The Child's Image of the President." Public Opinion Quarterly,
Redefinition of the Relation Between Sociology and Politics." Social Problems, 1968,
1960, 24, 632-644.
15, 280-296.
HESS, R., AND TORNEY, J. T h e Development of Political Attitudes in Children. Chicago:
HORST, P. "The Prediction of Personal Adjustment and Individual Cases." In P. F. Lazarsfeld
Aldine, 1967.
and M. Rosenberg (Eds.), T h e Language of Social Research: A Reader in the
HIMELHOCH, J. "Tolerance and Personality Needs: A Study of the Liberalization of Ethnic
Attitudes Among Minority Group College Students." American Sociological Review, Methodology of Social Research. New York: Free Press, 1955.
HORTON, J. "The Dehumanization of Anomie and Alienation: A Problem in the Ideology of
1950,15,79-88.
HIRSCH,H. Poverty and Politicization. New York: Free Press, 1971. Sociology." British Journal of Sociology, 1964, 15, 283-300.
HITSCHMANN, E. "Some Psycho-Analytic Aspects of Biography." International Journal of HORTON, J. E., AND THOMPSON, W. E. L c P ~ ~ e r l e and
~ ~Political
n e ~ ~ Negativism: A Study of
Psycho-Analysis, 1956, 37, 265-269. Defeated Local Referendums." American Journal of Sociology, 1962, 67, 485-493.
HOBBES, T. Leviathan. Oxford: Clarendon, 1946. Originally published-1651. HOVLAND, C. I. "Reconciling Conflicting Results Derived from Experimental and Survey
HOFSTADTER, R. T h e Paranoid Style in American Politics and Other Essays. New York: Studies of Attitude Change." American Psychologist, 1959, 14, 8-17.
Knopf, 1965. HOVLAND, C. I., AND JANIS, I. L. Personality and Persuasibility. New Haven: Yale University
HOFSTADTER, R. "The Strange Case of Freud, Bullitt, and Woodrow Wilson: 11." New York Press, 1959.
Review of Books, 9 February 1967, pp. 3-8. HOVLAND, C . I., JANIS, I. L., AND KELLEY, H. H. Communication and Persuasion. New Haven:
HOLLANDER, E. rcAuthoritariani~m and Leadership Choice in a Military Setting." American Yale University Press, 1953.
Psychologist, 1953, 8, 368-369. HOVLAND, C. I., LUMSDAINE, A. A., AND SHEFFIELD, F. D. Experiments in Mass Communica-
HOLLANDER, E. P. "Conformity, Status and Idiosyncrasy Credit." Psychological Review, 1958, tion: Studies in Social Psychology in World W a r I I . Vol. 3. Princeton, N.J.: Princeton
65(2), 117-127. University Press, 1949.
494 References References 495

HOWARD, J. w., JR. "Judicial Biography and the Behavioral Persuasion." Paper Presented at JAHODA, G. "The Development of Children's Ideas About Country and Nationality. Part I :
meeting of American Political Science Association, New York, September 1969. The Conceptual Framework." British Journal of Educational Psychology, 1963a, 33,
HOWARD, J., AND SOMERS, R. "Resisting Institutional Evil from Within." I n N. Sanford and 47-60.
C. Comstock (Eds.), Sanctions for Evil. San Francisco: Jossey-Bass, 1971. JAHODA, G. "The Development of Children's Ideas About Country and Nationality. Part 11:
HOWREY, P., AND KELEJIAN, H. H. "Simulation Versus Analytical Solutions." I n T. H. Naylor National Symbols and Themes." British Journal of Educational Psychology, 1963b,
(Ed.), T h e Design of Computer Simulation Experiments. Durham, N.C.: Duke 33, 143-153.
University Press, 1969. JAHODA, G. "Children's Concepts of Nationality: A Critical Study of Piaget's Stages." Child
HUGHES,H. S. History as an Art and as a Science. New York: Harper, 1964. Development, 1964, 35, 1081-1092.
HUNT, J. "Traditional Personality Theory in the Light of Recent Evidence." American JAKOBOVITS, L. A. "Evaluational Reactions to Erotic Literature." Psychological Reports,
Scientist, 1965, 53, 80-96. 1965, 16, 985-994.
HUNT, w., CRANE, w., AND WAHLKE,J. "Interviewing Political Elites in Cross-Cultural JAMES,W. T h e Principles of Psychology. New York: Dover, 1950. Originally published-1890.
Comparative Research." American Journal of Sociology, 1964, 70, 59-68. JAMES, W. "The Moral Equivalent of War." International Conciliation, 1910 (27), 3-20.
HUNTINGTON, S. P. "Conservatism as an Ideology." American Political Science Review, 1957, JANIS, I. L. ccDeci~ional Conflicts: A Theoretical Analysis." Journal of Conflict Resolution,
51, 454-473. 1959, 3, 6-27.
HYMAN,H. H. "Community Background in Public Opinion R e ~ e a r c h Journal.~ of Abnormal JANIS, I. L. "Groupthink Among Policy Makers." I n N. Sanford and C. Comstock (Eds.),
and Social Psychology, 1945, 40, 411-413. Sanctions for Evil. San Francisco: Jossey-Bass, 1971.
HYMAN, H. H. Survey Design and Analysis. New York: Free Press, 1955. JANIS, I. L. Victims of Groupthink. Boston: Houghton Mifflin, 1972.
HYMAN, H. H. Political Socialization. New York: Free Press, 1959. JANIS, I. L., AND FESHBACH, S. "Effects of Fear-Arousing Communications." Journal of
HYMAN,H. H. Secondary Analysis of Sample Surveys: Principles, Procedures, and Poten- Abnormal and Social Psychology, 1953, 48, 78-92.
tialities. New York: Wiley, 1972a. JANIS, I. L., AND SMITH, M. B. "Effects of Education and Persuasion on National and
HYMAN,H. H. "Strategies in Comparative Survey Research." I n R. Smith (Ed.), Social International Images." I n H. C. Kelman (Ed.), International Behavior: A Social
Science Methods. New York: Free Press, 1972b. Psychological Analysis. New York: Holt, 1965.
HYMAN, H. H., COBB, W. J., FELDMAN, J. J., HART, C. w., AND STEMBER, C. H. Interviewing JANOWITZ, M., AND MARVICK, D. "Authoritarianism and Political Behavior." Public Opinion
in Social Research. Chicago: University of Chicago Press, 1954. Quarterly, 1953, 17, 185-201.
HYMAN,H. H., LEVINE, G., AND WRIGHT, C. "Studying Expert Informants by Survey JAROS,D. "Children's Orientations Toward the President: Some Additional Theoretical
Methods: A Cross-National Inquiry." public Opinion Quarterly, 1967, 31, 9-26. Considerations and Data." Journal of Politics, 1967, 29, 368-387.
HYMAN, H. H., AND SHEATSLEY, P. B. "Some Reasons Why Information Campaigns Fail." JAROS, D. Socialization to Politics. New York: Praeger, 1972.
Public Opinion Quarterly, 1947, 11, 412323. JAROS,D., HIRSCH,H., AND FLERON,F., JR. "The Malevolent Leader: Political Socialization
HYMAN, H. H., AND SHEATSLEY, P. B. "Trends in Public Opinion on Civil Liberties." Journal in an American Subculture." American Political Science Review, 1968, 62, 564-575.
of Social Issues, 1953, 9, 6-16. JENNINGS,H. H. Leadership and Isolation: A Study of Personality in Interpersonal Rela-
HYMAN,H. H., AND SHEATSLEY, P. B. "The Authoritarian Personality-A Methodological tions. New York: Longmans, 1943.
Critique." I n R. Christie and M. Jahoda (Eds.), Studies in the Scope and Method JENNINGS,M., AND NIEMI, R. ccPatternsof Political Learning." Haruard Educational Review,
of "The Authoritarian Personality." New York: Free Press, 1954. 1968a, 38,443467.
HYMAN, H. H., AND SHEATSLEY, P. B. "Attitudes Toward Desegregation." Scientific American, JENNINGS, M., AND NIEMI,R. "The Transmission of Political Values from Parent to Child."
1956, 195, 35-39. American Political Science Review, 1968b, 62, 169-184.
HYMAN, H. H., AND SINGER, E. (Eds.) Readings in Reference Group Theory and Research. JENNINGS,M., AND NIEMI,R. Families, Schools and Political Learning. Princeton, N.J.:
New York: Macmillan, 1968. Princeton University Press, forthcoming, 1974.
INBAR, M., AND STOLL, C. Simulation and Gaming in Social Science. New York: Free Press, JESSOR,R., GRAVES, T., HANSON, R. c., AND JESSOR,S. L. Society, Personality and Deviant
1971. Behavior. New York: Holt, 1968.
INGLEHART, R. "Changing Value Priorities and European Integration." Journal of Common JOHNSON, A. "Tendencies of Recent American Biography." Yale Review, 1912, 1, 390403.
Market Studies, 1971a, 10, 1-36. JOHNSON,C. "An Experimental Analysis of the Origin and Development of Racial Attitudes
INGLEHART, R. "The Silent Revolution in Europe: Intergenerational Change in Post- with Special Emphasis on the Role of Bilingualism." Unpublished doctoral disserta-
Industrial Societies." American Political Science Review, 1971b, 65, 991-1017. tion, University of Colorado, 1949.
INKELES, A. "Sociology and Psychology." I n S. Koch (Ed.), Psychology: A Study of a JOHNSON, G. W. Randolph of Roanoke. New York: Minton, Beach, 1929.
Science. New York: McGraw-Hill, 1963, Pp. 317-387. JONES, E. "The Case of Louis Bonaparte, King of Holland." Journal of Abnormal Psychology,
INKELES,A., AND LEVINSON, D. J. "National Character: The Study of Modal Personality 1913, p. 289.
and Sociocultural Systems." In G. Lindzey and E. Aronson (Eds.), T h e Handbook JONES,E. T h e Life and Work of Sigmund Freud. New York: Basic Books, 1953, 1955,
of Social Psychology. (2nd Ed.) Reading, Mass.: Addison-Wesley, 1969. Originally 1957. 3 Vols.
published 1954. JONES,H. M. "Methods in Contemporary Biography." English Journal, 1932, 21, 113-122.
INSKO, C. A. Theories of Attitude Change. New York: Appleton-Century, 1967. JONES,M., BAYLEY, N., MAC FARLANE, J., AND HONZIK,M. T h e Course of Human Develop-
ISRAEL, F. L. Nevada's Key Pittman. Lincoln: University of Nebraska Press, 1963. ment. Waltham, Mass.: Xerox College Publishing, 1971.
496 References References 497
JOSEPHSON, E., AND JOSEPHSON, M. (Eds.) Man Alone: Alienation in Modern Society. New KELMAN, H. C. "Compliance, Identification and Internalization: Three Processes of Attitude
York: Dell, 1962. Change." Journal of Conflict Resolution, 1958, 2, 51-60.
JOSEPHSON, M. "Historians and Mythmakers." Virginia Quarterly Review, 1940, 16, 92-109. KELMAN, H. C. (Ed.) International Behavior: A Social-Psychological Analysis. New York:
KAGAN, J. "The Measurement of Overt Aggression from Fantasy." Journal of Abnormal and Holt, 1965.
Social Psychology, 1956, 52, 390-393. KELMAN, H. C. "Education for the Concept of a Global Society." Social Education, 1968a,
KAGAN,J. "Stylistic Variables in Fantasy Behavior." In J. Kagan and G. Lesser (Eds.), 32, 661-666.
Contemporary Issues in Thematic Apperceptive Methods. Springfield, Ill. Thomas, KELMAN, H. C. "International Relations: Psychological Aspects." I n D. L. Sills (Ed.),
1961. International Encyclopedia of the Social Sciences. Vol. 8. New York: Macmillan,
KAGAN,J. "Motives and Development." Journal of Personality and Social Psychology, 1972, 196813.
22, 51-66. KELMAN, H. C. A T i m e to Speak: O n Human Values and Social Research. San Francisco:
KAGAN, J., AND MOSS, H. Birth to Maturity. New York: Wiley, 1962. Jossey-Bass, 1968c.
KAMIN,L. J. "Ethnic and Party Affiliations of Candidates as Determinants of Voting." KELMAN, H. C. "Patterns of Personal Involvement in the National System: A Social-
Canadian Journal of Psychology, 1958,12, 205-212. Psychological Analysis of Political Legitimacy." I n J. N. Rosenau (Ed.), International
KANOUSE, D. E., AND ABELSON, R. P. "Language Variables Affecting the Persuasiveness of Politics and Foreign Policy. (2nd Ed.) New York: Free Press, 1969.
Simple Communications." Journal of Personality and Social Psychology, 1967, 7, 158- KELMAN,H. C. "The Role of the Individuai in International Relations: Some Conceptual
163. and Methodological Considerations." Journal of International Affairs, 1970, 24, 1-17.
KAPLAN, M. A. "Review of Klineberg's T h e Human Dimension in International Relations." KELMAN, H. C. "Language as an Aid and Barrier to Involvement in the National System."
American Political Science Review, 1964, 58, 682-683. I n J. Rubin and B. H. Jernudd (Eds.), Can Language Be Planned? Sociolinguistic
KARDINER, A., AND OTHERS. T h e Psychological Frontiers of Society. New York: Columbia Theory and Practice for Developing Nations. Honolulu: University Press of Hawaii,
University Press, 1945. 1971.
KARIEL, H. S. "The Political Relevance of Behavioral and Existential Psychology." American KELMAN, H. C. "The Problem-Solving Workshop in Conflict Resolution." In R. L. Men-itt
Political Science Review, 1967, 61, 334342. (Ed.), Communication in International Politics. Urbana: University of Illinois Press,
KARLINS, M., AND ABELSON, H. I. Persuasion: HOW Opinions and Attitudes Are Changed. 1972.
(2nd Ed.) New York: Springer, 1970. KELMAN, H., AND BARCLAY, J. "The F Scale as a Measure of Breadth of Perspective." Journal
KATZ, D. "The Functional Approach to the Study of Attitudes." Public Opinion Quarterly, of Abnormal and Social Psychology, 1963, 67, 608-615.
1960, 24, 163-204. KELMAN, H. c., AND EZEKIEL, R. s., with the collaboration of R. B. Kelman. Cross-National
KATZ,D. "Editorial." Journal of Personality and Social Psychology, 1965, 1, 1-2. Encounters: T h e Personal Impact of an Exchange Program for Broadcasters. San
KATZ, D. "The Practice and Potential of Survey Methods in Psychological Research." I n Francisco: Jossey-Bass, 1970.
C. Glock (Ed.), Survey Research in the Social Sciences. New York: Russell Sage, KELMAN,H. c., AND LAWRENCE, L. H. "Assignment of Responsibility in the Case of Lt.
1967. Calley: Preliminary Report on a National Survey." Journal of Social Issues, 1972,
KATZ, D., AND ELDERSVELD, S. "The Impact of Local Party Activity upon the Electorate." 28(1), 177-212.
Public Opinion Quarterly, 1961, 25, 1-24. KENISTON,K. T h e Uncommitted: Alienated Youth in American Society. New York: Har-
KATZ,D., AND KAHN, R. L. T h e Social Psychology of Organizations. New York: Wiley, 1966. court, Brace and World, 1965.
KATZ, D., MC CLINTOCK, c., AND SARNOFF, I. "Ego-Defense and Attitude Change." Human KENNAN,G. F. "It's History, But Is It Literature?" New York Times Book Review, April
Relations, 1956, 9, 2745. 26, 1959.
KATZ,D., MC CLINTOCK, c., AND SARNOFF, I. "The Measurement of Ego Defense as Related KENYON, S. "The Development of Political Cynicism Among Negro and White Adolescents."
to Attitude Change." Journal of Personality, 1957, 25, 465474. Paper presented at meeting of American Political Science Association. New York,
KATZ, D., AND SCHANCK, R. Social Psychology. New York: Wiley, 1938. September 1969.
KATZ, D., AND STOTLAND, E. ''A Preliminary Statement to a Theory of Attitude Structure and KERR,E. H. "Changing Attitudes Through International Participation: European Parlia-
Change." I n S. Koch (Ed.), Psychology: A Study of a Science. Vol. 3. New York: mentarians and Integration." International Organization, 1973, 27, 45-83.
McGraw-Hill, 1959. KEY,V. o., JR. Politics, Parties and Pressure Groups. 4th ed. New York: Crowell, 1958.
KATZ,E., AND LAZARSFELD, P. F. Personal InFuence: T h e Part Played by People in the Flow KEY, V. o., JR. Public Opinion and American Democracy. New York: Knopf, 1961.
of Mass Communication. New York: Free Press, 1955. KEY,V. o., JR. T h e Responsible Electorate: Rationality in Presidential Voting. Cambridge:
KATZ, J. (Ed.) N o T i m e for Youth. San Francisco: Jossey-Bass, 1968. Harvard University Press, 1966.
KELLEY,H. H. "TWOFunctions of Reference Groups." In G. E. Swanson, T. M. Newconb, KEY,V. o., JR. AND MUNGER, F. "Social Determinism and Electoral Decision: The Case of
and E. L. Hartley (Eds.), Readings in Social Psychology. 2nd Ed. New York: Holt, Indiana." I n E. Burdick and A. J. Brodbeck (Eds.), American Voting Behavior.
1952. New York: Free Press, 1959.
KELLEY,IT. H. "Attribution Theory in Social Psychology." In D. Levine (Ed.), Nebraska KIELL, N. Psychological Studies of Famous Americans. New York: Twayne, 1964.
Symposium on Motivation. Lincoln: University of Nebraska Press, 1967. RIESLER,C. A., COLLINS,B. E., AND MILLER,N. Attitude Change: A Critical Analysis of
KELLY, E. "Consistency of the Adult Personality." American Psychologist, 1955, 10, 659-681. Theoretical Approaches. New York: Wiley, 1969.
498 References I
References 499

KILLIAN,L. M. "Leadership in the Desegregation Crisis: An Institutional Analysis." I n M. KOHLBERG, L. "Deve!opment of Moral Character and Moral Ideology." I n M. L. Hoffman

Sherif (Ed.), Intergroup Relations and Leadership. New York: Wiley, 1962. and L. W. Hoffman (Eds.), Review of Childhood Development. Vol. 1. New York:
KILLIAN,L. M., AND SMITH,C. U. "Negro Protest Leaders in a Southern Community." Social Russell Sage, 1964.
KOHLBERG, L. "Moral Development." I n D. Sills (Ed.), International Encyclopedia of the
Forces, 1960, 38, 253-257.
KINSEY,A. c., POMEROY, W. B., AND MARTIN, C. E. Sexual Behavior in the Human Male. Social Sciences. Vol. 10. New York: Macmillan, 1968.
KOHLBERG, L. Stages in the Development of Moral Thought and Action. New York: Holt,
Philadelphia: Saunders, 1948.
KINSEY,A. c., POMEROY, W. B., MARTIN, C. E., AND GEBHARD, P. H. Sexual Behavior in the
1971.
Human Female. Philadelphia: Saunders, 1953. KOHN,P. M. "The Authoritarianism-Rebellion Scale: A Balanced F Scale with Left-Wing
KINTZ,B., DELPRATO, D., METTE, D., PERSONS, c., AND SCHAPPE, R. "The Experimenter Effect."
Reversals." Sociometry, 1972, 35, 176-189.
KOMORITA, S . S. "Cooperative Choice in a Prisoner's Dilemma Game." Journal of Personality
Psychological Bulletin, 1965, 63, 223-232.
KIRBY,D. "A Counter-Culture Explanation of Student Activism." Social Problems, 1971, and Social Psychology, 1965,2, 741-745.
KORNER, A. "Theoretical Considerations Concerning the Scope and Limitations of Projective
19, 203-216.
KIRSCHT,J. P., AND DILLEHAY, R. C. Dimensions of Authoritarianism: A Review of Research
Techniques." I n B. Murstein (Ed.), Handbook of Projective Techniques. New York:
and Theory. Lexington: University of Kentucky Press, 1967. Basic Books, 1965.
KORNHAUSER, W. T h e Politics of Mass Society. New York: Free Press, 1959.
KISH,L. Survey Sampling. New York: Wiley, 1965.
KRAMER, G. H. "The Effects of Precinct-Level Canvassing on Voting ~ehavior." Public
KLECKA, W. "Applying Political Generations to the Study of Political Behavior: A Cohort
Analysis." Public Opinion Quarterly, 1971, 35, 358-373. Opinion Quarterly, 1970, 34, 560-572.
KRAUT, R. E., AND MC CONAHAY, J . B. "An Experimental Study of the Effects of 'Public
KLINEBERG, O. Tensions Affecting International Understanding. New York: Social Science
Research Council, 1950. Opinion Polling' and Alienation Reduction upon Turnout in Primary Elections."
KLINEBERG, O. T h e Human Dimension in International Relations. New York: Holt, 1964.
Paper presented at annual meeting of American Political Science Association, Chicago,
KLINEBERG, O. "Prejudice: The Concept." I n D. Sills (Ed.), International Encyclopedia of
1971.
the Social Sciences. New York: Macmillan, 1968. KRAUT, R. E., AND MC CONAHAY, J. B. "An Experimental Study of the Effects of Being Inter-

KLINEBERG, o., AND ZAVALLONI, M. Nationalism and Tribalism Among African Students.
viewed upon Voting in Subsequent Primary Elections." Unpublished paper, Yale
Paris: Mouton, 1969. University, 1972.
KLOPFER, W. "The Metamorphosis of Projective Methods." Journal of Projective Techniques
KRECH,D. "The Chemistry of Learning." Saturday Review, 20 January 1968, pp. 48-68.
KROEBER, A. L. Configurations of Culture Growth. Berkeley: University of California Press,
and Personality Assessment, 1968, 32, 402-404.
KLUCKHOHN, c., AND MURRAY, H. A. "Personality Formation: The Determinants." I n C.
1944.
KROEBER, A. L., AND PARSONS, T. "The Concepts of Culture and of Social System." American
Kluckhohn, and H. A. Murray (Eds.), Personality in Nature, Society and Culture.
(2nd Ed.) New York: Knopf, 1953. Sociological Review, 1958, 23, 582-583.
KROEBER, T. "The Coping Functions of the Ego Mechanisms." I n R. White (Ed.), T h e
KLUCKHOHN, F. R., AND STRODTBECK, F. L. Variations in Value Orientation. New York:
Harper, 1961. Study of Lives. New York: Atherton, 1966.
KUBIE, L. S. Neurotic Distortion of the Creative Process. New York: Noonday Press, 1961.
KNUTSON, J. "Personality Correlates of Political Belief: Left, Right and Center." Chicago:
American Political Science Association, 1971. LAMARE, J. W. "University Education in American Government: An Experimental Approach

KNUTSON, J. T h e Human Basis of the Polity: A Psychological Study of Political Men.

to a Growing Problem." Experimental Study of Politics, 1971, 1, 122-148.
LAMBERT, W. E., AND KLINEBERG, O. Children's Views of Foreign Peoples: A Cross-National
Chicago: Aldine-Atherton, 1972a.
KNUTSON,J . "Some Problems of Assessment in Political Psychology." Washington, D.C.: Study. New York: Appleton-Century, 1967.
American Political Science Association, 197213. LAMPRECHT, F., EIcHELMAN, B., THOA, N. B., WILLIAMS, R. B., AND KOPIN, I. J. "Rat Fighting

KNUTSON,J . "The Political Relevance of Self-Actualization." I n A. Wilcox (Ed.), Public Behavior: Serum Dopamine-B-Hydroxylase and Hypothalamic Tyrosine Hydroxylase."
Opinion and Political Attitudes. New York: Wiley, 1973. Science, 1972, 177, 1214-1215.
LANDAU, D. Kissinger: T h e Uses of Power. Boston: Houghton Mifflin, 1972.
KNUTSON, J . "Individual Differences in Pre-political Ideology : The Basis of Political Learn-
ing." I n R. Niemi (Ed.), New Views of Children and Politics. San Francisco: Jossey- LANE,R. E. "P01iti~a1Character and Political Analysis." Psychiatry, 1953, 16, 387-398.
Bass, forthcoming, 1974. LANE,R. E. "Political Personality and Electoral Choice." American Political Science Review,
KOCH,S. "Epilogue." I n S. Koch (Ed.), Psychology: A Study of Science. Vol. 3. New York: 1955, 49, 173-190.
LANE, R. E. "Fathers and Sons: Foundations of Political Belief." American Sociological
McGraw-Hill, 1959.
KOESTLER, A. T h e Ghost in the Machine. New York: Macmillan, 1967. Review, 1959a, 24, 502-5 11.
LANE, R. E. Political Life: W h y People Get Involved in Politics. New York: Free Press,
KOESTLER, A. "Man-One of Evolution's Mistakes?" New York Times Magazine, 19 October
1969, pp. 28ff. 1959b.
LANE, R. E. T h e Liberties of Wit: Humanism, Criticism and the Civic Mind. New Haven:
KOGAN, N. "Authoritarianism and Repression." Journal of Abnormal and Social Psychology,
1956, 53, 34-37. Yale University Press, 1961.
LANE, R. E. Political Ideology: W h y the American Common Man Believes What He Does.
KOGAN,N., AND WALLACH, M. Risk Taking: A Study in Cognition and Personality. New
York: Holt, 1964. New York: Free Press, 1962.
 500 References References 501

LANE,R. E. "The Decline of Politics and Ideology in a Knowledgeable Society." American LEGGETT, J. C. Class, Race and Labor: Working-Class Consciousness in Detroit. New York:
Sociological Review, 1966, 31, 649-662. Oxford University Press, 1968.
LANE, R. E. "Personality, Political: The Study of Political Personality." I n D. Sills (Ed.), LEIGHTON, A. E. T h e Governing of Men. Princeton: Princeton University Press, 1945.
International Encyclopedia of the Social Sciences. New York: Macmillan, 1968. LEITES,N. O n the Game of Politics in France. Stanford, Calif.: Stanford University Press,
LANE,R. E. Political Thinking and Consciousness. Chicago: Markham, 1969. 1959.
LANE, R. E. Political Man. New York: Free Press, 1972. LERNER, D. T h e Passing of Traditional Society. New York: Free Press, 1958.
LANE,R. E., AND SEARS, D. O. Public Opinion. Englewood Cliffs, N.J.: Prentice-Hall, 1964. LEVENS, H. "Organizational Affiliation and Powerlessness: A Case Study of the Welfare
LANGER, W. C. T h e Mind of Adolf Hitler: T h e Secret Wartime Report. New York: Basic Poor." Social Problems, 1968, 16, 18-32.
Books, 1972. LEVENTHAL, H., JACOBS,R., AND KUDIRKA, N. ccA~thoritariani~m, Ideology and Political
LANGER, W. L. "The Next Assignment." American Historical Review, 1958, 63, 283-304. Candidate Choice." Journal of Abnormal and Social Psychology, 1964, 69, 539-549.
LANGTON, K. "Peer Group and School and the Political Socialization Process." American LEVIN, M. "Social Climates and Political Socialization." Public Opinion Quarterly, 1961,
Political Science Review, 1967, 61, 751-758. 25, 596606.
LANGTON, K. Political Socialization. New York: Oxford University Press, 1969. LE VINE,R. "Political Socialization and Culture Change." I n C. Geertz (Ed.), Old Societies
LANGTON, K., AND JENNINGS, M. c c P ~ l i t i ~Socialization
al and the High School Civics Cur- and New States. New York: Free Press, 1963.
riculum in the United States." American Political Science Review, 1968, 62, 852-867. LEVINSON, D. J. "Projective Questions in the Study of Personality and Ideology." I n T.
LAPONCE, J. A., AND SMOKER, P. (Eds.) Experimentation and Simulation in Political Science. Adorno, E. Frenkel-Brunswik, D. Levinson, and R. N. Sanford (Eds.), T h e Author-
Toronto: University of Toronto Press, 1972. itarian Personality. New York: Wiley, 1950.
LASKI,H. Communism. New York: Holt, 1927. LEVINSON, D. J. "Authoritarian Personality and Foreign Policy." Journal of Conflict Resolu-
LASSWELL, H. D. Psychopathology and Politics. Chicago: University of Chicago Press, 1930. tion, 1957, 1, 37-47.
LASSWELL, H. D. World Politics and Personal Insecurity. New York: McGraw-Hill, 1935. LEVINSON, D. J. "The Relevance of Personality for Political Participation." Public Opinion
New York: Free Press, 1965. Quarterly, 1958, 22, 3-10.
LASSWELL, H. D. "Agitation." In Encyclopedia of the Social Sciences. Vol. 1. New York: LEVINSON, D. J. c c R ~ l ePersonality,
, and Social Structure in the Organizational Setting."
Macmillan, 1936a. Journal of Abnormal and Social Psychology, 1959,58, 170-180.
LASSWELL, H. D. Politics: W h o Gets What, When, How. New York: Whittlesey House, LEVINSON, D. J . "Conservati~m and Radicalism." I n D. Sills (Ed.), International Encyclo-
1936b. pedia of the Social Sciences. New York: Macmillan, 1968.
LASSWELL, H. D. Power and Personality. New York: Viking, 1948. LEVINSON, D. J., AND HUFFMAN, P. "Traditional Family Ideology and Its Relation to Per-
LASSWELL, H. D. T h e Political Writings of Harold D. Lasswell. New Ycrk: Free Press, 1951. sonality." Journal of Personality, 1955, 23, 251-273.
LASSWELL, H. D. "The Selective Effect of Personality on Political Participation." I n R. LEVINSON, D. J., AND SANFORD, N. ''A Scale for the Measurement of Anti-Semitism." T h e
Christie and M. Jahoda (Eds.), Studies in the Scope and Method of "The Author- Journal of Psychology, 1944, 17, 339-370.
itarian Personality." New York: Free Press, 1954. LEVY, D. M. "Anti-Nazis: Criteria of Differentiation." Psychiatry, 1948, 11, 125-167.

LASSWELL, H. D. "The Political Science of Science." American Political Science Review, LEVY, L. H. Conceptions of Personality. New York: Random House, 1970.

1956, 50,961-979. LEWIN, K. Field Theory in Social Science. New York: Harper, 1951.

LASSWELL, H. D. Psychopathology and Politics: A New Edition with Afterthoughts by the

LEWIS,F. "What is Alienation? The Career of a Concept." New Politics, 1962, 1, 116134.
LEWIS, O. Tepoztlan, Village in Mexico. New York: Holt, 1960.
Author. New York: Viking, 1960.
LIEBSCHUTZ, s., AND NIEMI, R. ccP~litical Socialization Among Black Children: The Develop-
.LASSWELL, H. D. "A Note on 'Types' of Po!itical Personality: Nuclear, Correlational,
ment of Attitudes and the Impact of Curriculum and Teachers." I n R. Niemi (Ed.),
Developmental." Journal of Social Issues, 1968, 24, 81-91.
New Views of Children and Politics. San Francisco: Jossey-Bass, forthcoming, 1974.
LASSWELL, H. D., AND KAPLAN, A. Power and Society: A Framework for Political Inquiry.
LIFTON, R. J. Thought Reform and the Psychology of Totalism: A Study of "Brainwashing"
New Haven: Yale University Press, 1950.
in China. New York: Norton, 1961.
LASSWELL, H. D., AND LERNER, D. (Eds.) World Revolutionary Elites. Cambridge: M I T
LIFTON,R. J. Death in Life. Survivors of Hiroshima. New York: Random House, 1968.
Press, 1965. LIKERT, R. T h e Human Organization: Its Management and Value. New York: McGraw-Hill,
LATANE, B., AND DARLEY, J. M. T h e Unresponsive Bystande;: W h y Doesn't He Help? New
1967.
York: Appleton-Century, 1970. LINDLEY, D. v., AND MILLER, J. C. P. Cambridge Elementary Statistical Tables. Cambridge
LAZARSFELD, P. F., BERELSON, B., AND GAUDET, H. T h e People's Choice. New York: Duell, University Press, 1958.
1944. New York: Columbia University Press, 1948. LINDQUIST, E. F. Design and Analysis of Experiments i n Psychology and Education. Boston:
LAZARUS, R. "A Substitutive-Defensive Conception of Apperceptive Fantasy." I n J. Kagan Houghton M a i n , 1953.
and G. Lesser (Eds.), Contemporary Issues in Thematic Apperceptive Methods. LINDZEY, G. (Ed.) Handbook of Social Psychology. Reading, Mass.: Addison-Wesley, 1954.
Springfield, Ill. : Thomas, 1961. 2 Vols.
LAZARUS, R. Personality and Adjustment. Englewood Cliffs, N.J.: Prentice-Hall, 1963. LINDZEY, G. T h e Assessment of Human Motives. New York: Holt, 1958.
LAZARUS, R. "Story Telling and the Measurement of Motivation." Journal of Consulting LINDZEY, G. Projective Techniques and Cross-Cultural Research. New York: Appleton-
Psychology, 1966, 30, 483487. Century, 1961.
502 References References 503

LINDZEY, G., AND ARONSON, E. (Eds.) T h e Handbook of Social Psychology. Reading, Mass.: LYND, R. s., AND LYND, H. M. Middletown: A Study in Contemporary American Culture.
Addison-Wesley, 1968-1969. 5 Vols. (2nd Ed.). New York: Harcourt, 1929.
LINTON, R. T h e Cultural Background of Personality. New York: Appleton-Century, 1945. LYND,R. s., AND LYND,H. M. Middletown in Transition. New York: Harcourt, 1937.
LIPPMANN, W. Public Opinion. New York: Harcourt, 1922. LYNN, D., AND SAWREY, W. "Effects of Father Absence on Norwegian Boys and Girls."
LIPSET,S. M. "Some Social Requisites of Democracy." American Political Science Review, Journal of Abnormal and Social Psychology, 1959, 19, 258-262.
1959, 53, 69-105. LYONS, S. "The Political Socialization of Ghetto Children. Journal of Politics, 1970, 32,
LIPSET,S. M. Political Man, T h e Social Bases of Politics. New York: Doubleday, 1960. 288-304.
LIPSET,S. M. T h e First N e w Nation: T h e United States in Historical and Comparative MC CALL,G. J., AND SIMMONS, J. L. (Eds.) Issues in Participant Observation: A T e x t and
Perspective. New York: Basic Books, 1963. Reader. Reading, Mass. : Addison-Wesley, 1969.
LIPSET, S. M. "Three Decades of the Radical Right: Coughlinites, McCarthyites, .and MC CLELLAND, D. "Methods of Measuring Human Motivation." I n J. Atkinson (Ed.), Motives
Birchers." I n D. Bell (Ed.), T h e Radical Right. Garden City, N.Y.: Doubleday, in Fantasy, Action and Society. Princeton: Van Nostrand, 1958.
1964. MC CLELLAND, D., ATKINSON, J., CLARK, R., AND LOWELL, E. T h e Achievement Motive. New
LIPSET,S. M., AND RAAB, E. T h e Politics of Unreason. New York: Harper, 1970. York: Appleton-Century, 1953.
LIPSET,S. M., TROW,M., AND COLEMAN, J. Union Democracy. New York: Free Press, 1956. MC CLELLAND, D. T h e Achieving Society. New York: Free Press, 1967.
LIPSITZ,L. "Work Life and Political Attitudes: A Study of Manual Workers." American MC CLOSKY, H. "Conservatism and Personality." American Political Science Review, 1958,
Political Science Review, 1964, 58, 951-962. 52, 2745.
LIPSITZ,L. "Working-Class Authoritarianism : A Reevaluation." American Sociological Re- MC CLOSKY, H. "Personality and Attitude Correlates of Foreign Policy Orientation." I n J. N.
view, 1965, 30, 103-109. I Rosenau (Ed.), Domestic Sources of Foreign Policy. New York: Free Press, 1967a.
LIPSKY, M. "Protest as a Political Resource." American Political Science Review, 1968, 62, Pp. 51-109.
1144-1 158. MC CLOSKY, H. "~urveyResearch in Political Science." In C. Glock (Ed.), Survey Research
LITT, E. "Civic Education, Community Norms and Political Indoctrination." American i n the Social Sciences. New York: Russell Sage, 1967b. Pp. 63-143.
Sociological Review, 1963, 28, 69-75. MC CLOSKY, H., AND DAHLGREN, H. "Primary Group Influence on Party Loyalty." American
LITTMAN, R. A. "Psychology: The Socially Indifferent Science." American Psychologist, 1961, Political Science Review, 1959, 53, 757-776.
16, 232-236. MC CLOSKY, H., AND SCHAAR, J. H. "Psychological Dimensions of Anorny." American Socio-
LOEHLIN, J. C. Computer Models of Personality. New York: Random House, 1968. logical Review, 1965, 30, 14-40.
LOEVINGER, J . "The Meaning and Measurement of Ego Development." American Psychol-
MACCOBY, E., MATTHEWS, R., AND MORTON, A. ''Youth and Political Change." Public Opinion
ogist, 1966, 21, 195-206. Quarterly, 1954, 18, 23-29.
LOEVINGER, J., AND WESSLER, R. Measuring Ego Development. Vol. 1. Construction and Use
of a Sentence Completion Test. San Francisco: Jossey-Bass, 1970.
I MACCOBY, M. "Emotional Attitudes and Political Choices." Politics and Society, 1972, 2, 21 1.
MC CONAUGHY, J . B. "Certain Personality Factors of State Legislators in South Carolina."
LORENZ, K. O n Aggression. New York: Bantam, 1967. Originally published 1963 as Der
American Political Science Review, 1950, 44, 897-903.
Sogenannte Bose: Zur Naturgeschichte der Aggression.
MC DILL,E. L., AND RIDLEY, J. C. "Status, Anomia, Political Alienation, and Political Partici-
LORINSKAS, R. A., HAWKTNS, B. w., AND EDWARDS, S. "The Persistence of Ethnic Voting in
pation." American Journal of Sociology, 1962, 68, 205-213.
Rural and Urban Areas: Results from the Controlled Election Method." Social
MC DOUGALL, W. Introduction to Social Psychology. (14th Ed.) Boston: Luce, 1921. Origi-
Science Quarterly, 1969, 49, 871-899.
LOVEJOY, A. T h e Great Chain of Being. Cambridge, Mass.: Harvard University Press, 1936.
nally published 1908.
LOWE,C. M., AND DAMANKOS, F. J. "Psychological and Sociological Dimensions of Anomie MC FARLAND, A. S. Power and Leadership in Pluralist Systems. Stanford, Calif.: Stanford

in a Psychiatric Population." Journal of Social Psychology, 1968, 74, 65-74. University Press, 1969.
LOWE,F., AND MCCORMACK, T. "A Study of the Influence of Formal and Informal Leaders MC GEE, H. "Measurement of Authoritarianism and Its Relation to Teachers' Classroom
in an Election Campaign." Public Opinion Quarterly, 1956, 20, 651-662. Behavior." Unpublished doctoral dissertation. University of California, Berkeley,
LOWENSTEIN, K. "Ideologies and Their Institutions: The Problem of Their Circulation." 1954.
Western Political Quarterly, 1953a, 6, 689-706. MC GUIGAN, F. J. "The Experimenter: A Neglected Stimulus Object." Psychological Bulletin,
LOWENSTEIN, K. (Rapporteur) "Report of the Second International Congress of Political 1963, 60,421-428.
Science, 1952." International Social Science Bulletin, 1953b, 5. MC GUIRE,W. J. "A Syllogistic Analysis of Cognitive Relationships." In M. J. Rosenberg,
LOWENTHAL, L., AND GUTERMAN, N. Prophets o f Deceit. New York: Harper, 1949. C. I. Hovland, W. J. McGuire, R. P. Abelson, and J. W. Brehm (Eds.), Attitude
LUTTERMAN, K. G., AND MIDDLETON, R. "Authoritarianism, Anomia, and Prejudice." Social Organization and Change. New Haven: Yale University Press, 1960.
Forces, 1970, 48, 485-492. MC GUIRE,W. J . "Inducing Resistance to Persuasion." I n L. Berkowit? (Ed.), Advances in
LYLE, W. H., AND LEVITT, E. E. "Permis~ivene~~, Authoritarianism, and Parental Discipline of E,xperimental Social Psychology. Vol. 1. New York: Academic ~ r b s 1964.
,
Grade School Children." Journal of Abnormal and Social Psychology, 1955, 51, 42-46. MC GUIRE, W. J. "The Nature of Attitudes and Attitude Change." I n G. Lindzey and E.
LYND, H. M. ''Must Psychology Aid Reaction?" I n Toward Discovery. New York: Sarah Aronson (Eds.), T h e Handbook of Social Psychology. (Rev. Ed.) Vol. 3. Reading,
Lawrence College, 1965. Mass. : Addison-Wesley, 1968.
504 References References 505

MCILLWAIN, C. H. T h e High Court of Parliament and Its Supremacy. New Haven: Yale MARX, K., AND ENGELS, F. T h e Manifesto of the Communist Party. New York: International
University Press, 1910. Publishers, 1968. Originally published 1848.
MAC NER,R. Society, A Textbook of Sociology. New York: Holt, 1937. MASLOW, A. H. "The Dynamics of Psychological Security-Insecurity." Character and Per-
MAC NER,R. T h e W e b of Government. New York: Macmillan, 1948. sonality, 1941-1942, 10, 331-344.
MAC KINNON, W. J., AND CENTERS,R. "Authoritarianism and Internationalism." Public MASLOW, A. H. "Authoritarian Character Structure." Journal of Social Psychology, 1943a,
Opinion Quarterly, 1956, 20, 621-630. 18,401-411.
MC PHAIL, C. ''Civil Disorder Participation: A Critical Examination of Recent Research." MASLOW, A. H. "Dynamics of Personality Organization: 11." Psychological Review, November
American Sociological Review, 1971, 36, 1058-1073. 1943b, 50, 370-396.
MC PHEE,W. N. "Note on a Campaign Simulator." I n W. N. McPhee (Ed.), Formal Theories MASLOW, A. H. "A Theory of Human Motivation." Psychological Review, 1943c, 50, 370-396.
of Mass Behavior. New York: Free Press, 1963. MASLOW, A. H. Motivation and Personality. New York: Harper, 1954.

MC PHEE,W. N., AND FERGUSON, J. "Political Immunization." I n W. N. McPhee and W. A. MASLOW, A. H. "Power Relationships and Patterns of Personal Development." I n A. Korn-
Glaser (Eds.), Public Opinion and Congressional Elections. New York: Free Press, hauser (Ed.), Problems o f Power in Amsrican Democracy. Detroit: Wayne State
1962. University Press, 1957.
MC PHEE, W. N., FERGUSON, J., AND SMITH, R. B. "A Model for Simulating Voting Systems." MATTHEWS, D., AND PROTHRO, J. Negroes and the N e w Southern Politics. New York:
I n J. M. Dutton and W. H. Starbuck (Eds.), Computer Simulation of Human Harcourt, 1966.
Behavior. New York: Wiley, 1971. MATZA,D. "Subterranean Traditions of Youth." Annals of the American Academy of
MC PHEE, W. N., AND SMITH,R. B. "A Model for Analyzing Voting Systems." I n W. N. Political and Social Science, 1961, 338, 102-118.
McPhee and W. A. Glaser (Eds.), Public Opinion and Congressional Elections. New MAY, H. F. Protestant Churches and Industrial America. New York: Harper, 1949.
York: Free Press, 1962. MAY, M. A. A Social Psychology o f War and Peace. New Haven: Yale University Press,
MAC RAE, J., AND SMOKER, P. "A Vietnam Simulation." Journal of Peace Research, 1967, 1943.
4, 1-24. MAY,R. Power and Innocence: A Search for the Sources of Violence. New York: Norton,
MC WHINNEY, W. H. "Simulating the Communication Network Experiments." Behavioral 1972.
Science, 1962, 9, 80-84. MAZLISH, B. (Ed.) Psychoanalysis and History. Englewood Cliffs, N.J.: Prentice-Hall, 1963.
MAIER,N. R. F. Frustration: T h e Study of Behavior Without a Goal. New York: McGraw- MAZLISH,B. I n Search of Nixon: A Psychohirtorical Inquiry. New York: Basic Books, 1972.
Hill, 1949. MEAD,M. Soviet Attitudes Toward Authority. New York: McGraw-Hill, 1951.
MALONE, D. "Biography and History." In J. R. Strayer (Ed.), T h e Interpretation of History. MEIER, D. L., AND BELL,W. "Anomia and Differential Access to the Achievement of Life
Princeton: Princeton University Press, 1943. Goals." American Sociological Review, 1959, 24, 189-202.
MANN, F. C. "Toward an Understanding of the Leadership Role in Formal Organizations." MERELMAN, R. cLTheDevelopment of Political Ideology: A Framework for the Analysis of
In R. Dubin, G. Homans, and D. Miller (Eds.), Leadership and Productivity. San Political Socialization." American Political Science Review, 1969, 63, 750-767.
Francisco: Chandler, 1964. MERELMAN, R. "The Development of Policy Thinking in Adolescence." American Political
MANN,M. "The Social Cohesion of Liberal Democracy." American Sociological Review, Science Review, 1971a, 65, 1033-1047.
1970, 35, 423439. MERELMAN, R. Political Socialization and Educational Climates. New York: Holt, 1971b.
MANNHEIM, K. Ideology and Utopia. Trans. L. Wirth and E. Shils. New York: Harcourt,
MERRIAM, C. E. N e w Aspects of Politics. Chicago: University of Chicago Press, 1925. (Revised
1949. 1931; reprinted revision, with a foreword by Barry D. Karl, published 1970.)
MARCH,J . G., AND SIMON,H. A., with H. Guetzkow. Organizations. New York: Wiley, 1958.
MERRIAM, C. E. American Political Ideas, 1865-1917. New York: Macmillan, 1929.
MARCUSE, H. Eras and Civilization. Boston: Beacon Press 1959. Rev. Ed., 1966.
MERRIAM, C. E. T h e Making of Citizens: A Comparative Study of Methods of Civic Training.
MARK,V. H., AND ERVIN,F. R. Violence and the Brain. New York: Harper, 1970.
MARTIN, W. c., AND BENGTON, V. L. "Alienation and Age: A Context Specific Approach."
Chicago: University of Chicago Press, 1931.
MERRIAM, C. E. Political Power. New York: Collier, 1964. Originally published 1934.
Unpublished manuscript, 1971.
MERTON, R. K. Social Theory and Social Structure. (Rev. Ed.) New York: Free Press, 1957.
MARTIN,J. G., AND WESTIE, F. R. "The Tolerant Personality." American Sociological Review,
1959, 24, 521-528. (Enlarged edition 1968.)
MICHEL, J. B., AND DILLEHAY, R. C. "Reference Behavior Theory and the Elected Representa-
MARVICK, D., AND BAYES, J. "Domains and Universes: Problems in Concerted Use of Multiple
Data Files for Social Science Inquiries." I n M. Dogan and S. Rokkan (Eds.), tive." Western Political Quarterly, 1969,22, 759-773.
Quantitative Ecological Analysis in the Social Sciences. Cambridge: M I T Press, MICHELS, R..Political Parties. New York: Free Press, 1915. Reprinting, 1949.
1969. MIDDLETON, R. "Alienation, Race and Education." American Sociological Review, 1963, 28,

MARX,K. Early Writings. Trans. T. B. Bottomore. New York: McGraw-Hill, 1964. Origi- 973-977.
nally published 1844. MIDDLETON, R., AND PUTNEY,S. "Political Expression of Adolescent Rebellion." American
MARX, K. "Wage Labor and Capital." I n K. Marx and F. Engels, Selected Works i n T w o Journal of Sociology, 1963, 68,527-535.
Volumes. Vol. 1. Moscow, Foreign Languages Publishing House, 1849. See also MIDLARSKY, M., AND TANTER, R. "Toward a Theory of Political Instability in Latin America."
R. C. Tucker (Ed.), T h e Marx-Engels Reader. New York: Norton, 1972. Journal of Peace Research, 1967,3, 209-227.
506 References References 507
MILBRATH, L. W. Political Participation: How and W h y Do People Get Involved in Politics? MURCHISON, C. (Ed.) A Handbook of Social Psychology. Worcester, Mass.: Clark University
Chicago: Rand McNally, 1965. Press, 1935.
MILBRATH, L. w., AND KLEIN,W. W. "Personality Correlates of Political Participation." Acta MURPHY, G., AND LIKERT, R. Public Opinion and the Individual. New York: Harper, 1938.
Sociologica, 1962, 6, 53-66. MURPHY, G., MURPHY, L. B., AND NEWCOMB, T. M. Experimental Social Psychology. (Rev.
MILGRAM, S. "Behavioral Study of Obedience." Journal of Abnormal and Social Psychology, Ed.) New York: Harper, 1937.
1963, 67, 371-378. Reprinted in W. G. Bennis and others (Eds.), Interpersonal MURRAY, H. A. Explorations in Personality. New York: Oxford University Press, 1938.
Dynamics: Essays and Readings in Human Interaction. Homewood, Ill.: Dorsey Press, MURRAY, H. A. "In Nomine Diaboli." New England Quarterly, 1941, 24, 435452.
1964. MURRAY, H. A. Thematic Apperception Test Manual. Cambridge: Harvard University
MILGRAM, S. "Some Conditions of Obedience and Disobedience to Authority." Human Rela- Press, 1943.
tions, 1965, 18, 57-76. M U R R AH.~ , A. "Personality: Components of an Evolving Personological System." I n D.
MILLER,C. R., AND BUTLER, E. W. "Anomia and Eunomia: A Methodological Evaluation of Sills (Ed.), International Encyclopedia of the Social Sciences. New York: Macmillan,
Srole's Anomia Scale." American Sociological Review, 1966, 31, 400406. 1968.
MILLER, D. R., AND SWANSON, G. E. T h e Changing American Parent. New York: Wiley, 1958. MURRAY, H. A., AND KLUCKHOHN, C. "Outline of a Conception of Personality." In C. Kluck-
MILLER, D. R., AND SWANSON, G. E. Inner Conflict and Defense. New York: Holt, 1960. hohn and H. A. Murray (Eds.), Personality in Nature, Society and Culture. New
MILLER,J. G. "Living Systems: The Organization." Behavioral Science, 1972, 17, 1-182. York: Knopf, 1949.
MILLER,S. M., AND RIESSMAN, F. "Working Class Authoritarianism: A Critique of Lipset." MURSTEIN, B. "The Role of the Stimulus in the Manifestation of Fantasy." In J. Kagan and
British Journal of Sociology, 1961, 12, 263-276. G. Lesser (Eds.), Contemporary Issues in Thematic Apperceptive Methods. Spring-
MILLER, W. "One Party Politics and the Voter." American Political Science Review, 1956, field, Ill.: Thomas, 1961.
50, 707-725. MURSTEIN, B. Theory and Research in Projective Techniques. New York: Wiley, 1963.
MILLETT, K. Sexual Politics. Garden City, N.Y.: Doubleday, 1970. MURSTEIN, B. Handbook of Projective Techniques. New York: Basic Books, 1965.
MISCHEL, W. Personality and Assessment. New York: Wiley, 1968. MURSTEIN, B., AND PRYER,R. "The Concept of Projection: Review." Psychological ~ u l l e t i n ,
MISIAK,H., AND SEXTON, V. History of Psychology. New York: Grune and Stratton, 1966. 1959, 56, 353-374.
MITRANY, D. A Working Peace System. Chicago: Quadrangle Books, 1966. MUSSEN, P., AND NAYLOR, H. "The Relationship Between Overt and Fantasy Aggression."
MIZRUCHI, E. H. "Social Structure and Anomia in a Small City." American Sociological Journal of Abnormal and Social Psychology, 1954, 49, 235-240.
Review, 1960, 25, 645-654. MUSSEN, P., AND WYSZYNSKI, A. "Personality and Political Participation." Human Rela-
MIZRUCHI, E. H. Success and Opportunity: A Study of Anomie. New York: Free Press, 1964. tions, 1952, 5, 65-82.
MODIGLIANI, A. "Hawks and Doves, Isolationism and Political Distrust: An Analysis of National Opinion Research Center. Opinion News, August 1948.
Public Opinion on Military Policy." American Political Science Review, 1972, 66, NAYLOR, T. H. (Ed.) T h e Design of Computer Simulation Experiments. Durham, N.C.: Duke
960-978. University Press, 1969.
MONYPENNY, W. F., AND BUCKLE, G. E. T h e Life of Benjamin Disraeli. New York: Mac- NAYLOR, T. H. (Ed.) Computer Simulation Experiments with Models of Economic Systems.
millan, 1913-1920. 6 Vols. New York: Wiley, 1971.
MOORE, C. G. "Simulation of Organizational Decision Making: A Survey." I n W. D. Coplin NAYLOR, T. H., BALINTFY, J. L., BURDICK, D. s., AND CHU, K. Computer Simulation Tech-
(Ed.), Simulation in the Study of Politics. Chicago: Markham, 1968. niques. New York: Wiley, 1966.
MORGAN, J. "Attitudes of Students Toward the Japanese." Journal of Social Psychology, NEAL,A. G., AND RETTIG, S. "Dimensions of Alienation Among Manual and Non-manual
1945, 21, 219-222. Workers." American Sociological Review, 1963, 28, 599-608.
MORGAN, J., AND MORTON, J. "The Distortion of Syllogistic Reasoning Produced by Personal NEAL, A. G., AND RETTIG, S. "On the Multidimensionality of Alienation." American Socio-
Conviction." Journal of Social Psychology, 1944, 20, 39-59. logical Review, 1967, 32, 54-64.
MORGAN, J., AND MORTON, J. "Distorted Reasoning as an Index of Public Opinion." School NEAL,A. G., AND SEEMAN, M. "Organizations and Powerlessness: A Test of the Mediation
and Society, 1945, 57, 333-335. Hypothesis." American Sociological Review, 1964, 29, 216-226.
MORLEY, J . T h e Life of William Ewart Gladstone. New York: Macmillan, 1903. 3 Vols. NELSON, J. I. "Anomie: Comparisons Between the Old and New Middle Class." American
MORROW, C. "Aggression Against Whom?" Unpublished doctoral dissertation, Yale Uni- Journal of Sociology, 1968, 74, 184-192.
versity, 1972. NETTLER,G. "A Measure of Alienation." American Sociological Review, 1957,. 22,. 670-677.
MOYER, K. E. T h e Physiology of Hostility. Chicago: Markham, 1971. NETTLER, G. "A Further Comment on 'Anomy.' " American Sociological Review, 1965, 30,
MUELLER, J. "Presidential Popularity from Truman to Johnson." American Political Science
762-763.
Review, 1970, 64, 18-34. NETTLER, G., AND HUFFMAN, J. ''Political Opinion and Personal Security." Sociometry,
MUELLER, J. "Trends in Popular Support for the Wars in Korea and Vietnam." American
1957,20,51-66.
Political Science Review, 1971, 65, 358-375. NEURINGER, C. "The Relationship Between Authoritarianisni, Rigidity and Anxiety." Journal
MUMFORD, L. "The Task of Modern Biography." English Journal, 1934, 23, 1-9. of General Psychology, 1964, 71, 169-175.
MUNNION, C. "If Idi Amin of Uganda Is a Madman, He's a Ruthless and Cunning One." NEUSTADT, R. E. Presidential Power: T h e Politics of Leadership. New York: Wiley, 1960.
New York Times Magazine, 12 November 1972. NEWCOMB, T. Personality and Social Change. New York: Holt, 1943.
508 References I References 509
OGBURN, W. F. Social Change with Respect to Culture and Original Nature. New York:
NEWCOMB, T. M. "Some Patterned Consequences of Membership in a College Community."
Huebsch, 1922.
I n T. M. Newcomb and E. L. Hartley (Eds.), Readings in Social Psychology. New
OLDS, J. "Pleasure Centers of the Brain." Scientific American, October 1956, 195, 106-116.
York: Holt, 1947.
OLDS, J. "Self-Stimulation of the Brain." Science, 1958, 127, 315-324.
NEWCOMB, T., KOENIG, K., FLACKS, R., AND WARWICK, D. Persistence and Change. New York: OLDS, M. E., AND OLDS, J. "Approach-Escape Interactions in Rat Brain." American Journal
Wiley, 1967.
of Physiology, 1962, 203, 803-810.
NEWELL,A. Human Problem Solving. Englewood Cliffs, N.J.: Prentice-Hall, 1972. OLSEN, M. E. "Alienation and Political Opinions." Public Opinion Quarterly, 1965, 29, 200-
NEWELL,A., SHAW, J. c., AND SIMON,H. A. "Elements of a Theory of Human Problem
212.
Solving." Psychological Review, 1958, 65, 151-166.
OLSEN, M. E. "Perceived Legitimacy of Social Protest Actions." Social Problems, 1968, 15,
NEWELL, A., AND SIMON, H. A. "The Logic Theory Machine: A Complex Information Process-
297-310.
ing System." I R E Transactions on Information Theory, 1956, It-2, 61-79.
OLSEN, M. E. "TWOCategories of Political Alienation." Social Forces, 1969, 47, 288-299.
NEWELL, A., AND SIMON, H. A. Human Problem Solving: A New Approach to the Psychology O'NEIL, W. M., AND LEVINSON, D. "A Factorial Exploration of Authoritarianism and Some of
of Human Problem Solving Based on Detailed Analyses of Human Behavior and
Its Ideological Concomitants." Journal of Personality, 1954, 22, 449463.
Basic Studies of the Information Processing Systems. Englewood Cliffs, N.J.: Prentice- OPPENHEIM, A. N. Questionnaire Design and Attitude Measurement. New York: Basic Books,
Hall, 1972.
1966.
NEWELL, A., TONGE, F. M., FEIGENBAUM, E. A., GREEN, B. F., KELLY, H. A., AND MEALY, G. OPPENHEIMER, F. T h e State. Indianapolis: Bobbs-Merrill, 1914.
Infornzation Processing Language V Manual. (2nd Ed.) Englewood Cliffs N.J.: ORBELL, J. M., DAWES, R. M., AND COLLINS, N. J. "Grass Roots Enthusiasm and the Primary
Prentice-Hall, 1964. Vote." Western Political Quarterly, 1972, 25, 249-259.
NICOLSON, H. T h e Development of English Biography. London: Hogarth, 1927. ORNE, M. T. "On the Social Psychology of the Psychological Experiment with Particular
NIE, N., POWELL,G., JR., AND PREWITT,K. "Social Structure and Political Participation: Reference to Demand Characteristics and Their Implications." American Psychologist,
Developmental Relationships, I." American Political Science Review, 1969a, 63, 361- 1962,17, 776-783.
378. ORUM, A. (Ed.) T h e Seeds of Politics: Youth and Politics in America. Englewood Cliffs,
NIE, N., POWELL,G., JR., AND PREWITT, K. "Social Structure and Political Participation: N.J.: Prentice-Hall, 1972.
Developmental Relationships, 11." American Political Science Review, 1969b, 63, OSGOOD, C. E. A n Alternative to War or Surrender. Urbana: University of Illinois Press,
808-832. 1962.
NIEMI,R. "Review of R. Hess and J. Torney, T h e Development of Political Attitudes in OSGOOD, c. E. "Calculated De-escalation as a Strategy.'' I n D. G. Pruitt and R. C. Snyder
Children." Contemporary Psychology, 1969, 14, 497-498. (Eds.), Theory and Research on the Causes of War. Englewood Cliffs, N.J.: Prectice-
NIEMI,R. L ' C ~ l l e ~ t iInformation
ng About the Family: A Problem in Survey Methodology." Hall, 1969.
I n J. Dennis (Ed.), Socialization to Politics: A Reader. New York: Wiley, 1973. OSGOOD, C. E., SUCI,G. J., AND TANNENBAUM, P. H. T h e Measurement of Meaning. Urbana:
NISKANEN, W. Bureaucracy and Representative Government. Chicago: Aldine-Atherton, University of Illinois Press, 1957.
1971. OSGOOD, R. E. Ideals and Self-Interest in America's Foreign Relations. Chicago: University
NNEKAWA, A. M. "Factors Associated with Authoritarianism in Japan." Unpublished doctoral of Chicago Press, 1953.
dissertation, New York University, 1960. OSGOOD, R. E. "A Critique of Clinical Psychological Approaches to the Explanation of
NOELLE,E., AND NEUMANN, E. T h e Germans: Public Opinion Polls 1947-1966. Allensbach International Conflict." Social Problems, 1955, 2, 176-180.
and Bonn: Verlag fiir Demoskopie, 1967. OSTROGORSKI, M. Democracy and the Organization of Political Parties. New York: Macmillan,
NOELLE-NEUMANN, E. "Wanted: Rules for Wording Structured Questionnaires." Public 1902.
Opinion Quarterly, 1970, 34, 191-201. PADIOLEAU, J.-G. "Note sur les simulations en Sociologie Politique." Revue Fran~aise de

NOGEE,P., AND LEVIN, M. "Some Determinants of Political Attitudes Among College Voters." Sociologie, 1969, 10, 201-2 10.
Public Opinion Quarterly, 1958, 22, 449463. PAIGE,G. D. T h e Korean Decision. New York: Free Press, 1968.
NORMAN, W. "Toward an Adequate Taxonomy of Personality Attributes: Replicated Factor
PAMMETT,J. "The Development of Political Orientations in Canadian School Children."
Structures in Peer Nomination Personality Ratings." Journal of Abnormal and Social Canadian Journal of Political Science, 1971, 4, 132-141.
Psychology, 1963, 66, 574-583. PARROTT, G., AND BROWN, L. "Political Bias in the Rokeach Dogmatism Scale." Psychological
NORTH, R. c., BRODY, R. A., AND HOLSTI, O. R. "Some Empirical Data on the Conflict Spiral."
Reports, 1972, 30, 805-806.
PARRY, H., AND CROSSLEY, H. "Validity of Responses to Survey Questions." Public Opinion
Peace Research Society (International) Papers, 1964, 1, 1-14.
Quarterly, 1950, 14, 61-80.
NOVE, A. Economic Rationality and Soviet Politics, or Was Stalin Really Necessary? New
PARSONS, T. T h e Social System. New York: Free Press, 1950.
York: Praeger, 1964.
PARSONS, T. Structure and Process in Modern Societies. New York: Free Press, 1960.
OCHS, S. Elements of Neurophysiology. New York: Wiley, 1965.
PARSONS,T., AND SHILS,E. A. (Eds.) Toward a General Theory of Action. New York:
OFSHE, R., AND OFSHE, S. L. "Choice Behavior in Coalition Games." Behavioral Science,
Harper, 1951.
1970a, 15,337-349. PASSINI,F., AND NORMAN, W. "A Universal Conception of Personality Structui-e?" Journal
OFSHE,R., AND OFSHE,S. L. Utility and Choice in Social Interaction. Englewood Cliffs, of Personality and Social Psychology, 1966, 4, 44-49.
N.J. : Prentice-Hall, 1970b.
 References References

PATCHEN, M. The American Public's View of U.S. Policy Toward China. New York: Council PINNER, F. "Parental Overprotection and Political Distrust." Annals of the American
on Foreign Relations, 1964. Academy of Political and Social Science, 1965, 361,58-70.
PATRICK, J . "The Impact of an Experimental Course 'American Political Behavior' on the POOL, I. DE s., ABELSON, R. P., AND POPKIN,S. Candidates, Issues and Strategies: A Computer
Knowledge, Skills and Attitudes of Secondary School Students." Social Education, I.' Simulation of the 1960 and 1964 Presidential Elections. (2nd Ed.) Cambridge: M I T
1972, 36,-168-179. .,.
s t ,
Press, 1965.
' '
PAVLOV, I. P. Conditioned RefEexes. New York: Dover, 1960. Originally published 1927. POPPER, K. R. The Logic of Scientific Discovery. New York: Basic Books, 1959.
PAYNE,J . C. Incentive Theory and Political Process: Motivation and Leadership in the POWELL, C. A. "Simulation: The Anatomy of a Fad. A C r i t i ~ u eand a Suggestion with
Dominican Republic. Lexington, Mass.: Heath, 1972. Respect to Its Use in the Study of International Conflict." Acta Politica, 1969, 4,
PAYNE,J. L., AND WOSHINSKY, O. H. "Incentives for Political Participation." World Politics, 299-330.
1972, 24, 518-546. POWELL, E. H. The Design of Discord: Studies of Anomie. New York: Oxford University
PAYNE,R. "The Effect of Preshock on Pain-Elicited Aggression." Paper presenied at meeting Press, 1970.
of Western Psychological Association, San Diego, March 1968. President's Commission on the Assassination of President John F. Kennedy. Report. Wash-
PEABODY, D. "Attitude Content and Agreement Set in Scales of Authoritarianism, Dogmatism, ington, D.C.: U.S. Government Printing Office, 1964.
,. Anti-Semitism, and Economic Conservatism." Journal of Abnormal and Social , PREWITT,K., EULAU, H., AND ZISK, B. "Political Socialization and Political Roles." Public
Psychology, 1961, 63, 1-12. Opinion Quarterly, 1966-1967, 30, 569-582.
PEAK,H. Obseruations on the Characteristics and Distribution of German Nazis. Washington, PREMTT,K., AND OKELLO-OCULI, J. "Political Socialization and Political Education in the
D.C.: American Psychological Association, 1945. New Nations." In R. Sigel (Ed.), Learning About Politics. New York: Random
PEAK,H. "Problems of Objective Observation." In L. Festinger and D. Katz (Eds.), Re- House, 1970.
search Methods in the Behavioral Sciences. New York: Dryden, 1953. PRICE, D. E., AND LUPFER, M. "Volunteers for Gore: The Impact of a Precinct-Level Canvass
PEAR,T. H. (Ed.) Psychological Factors of Peace and War. London: Hutchinson, 1950. in Three Tennessee Cities." Journal of Politics, 1973.
PEARLIN, L. I. "Alienation from Work: A Study of Nursing Personnel." American Sociological PROCTOR, C. H. "Comment on Carr's Srole Items." American Sociological Reuiew, 1971, 3 6
.
A: Review, 1962,27,314-326.
PELZ,D. C. "Leadership Within a Hierarchical Organization." Journal of Social Issues, 1951.
I " 1107-1108. I 1 . .
PROPPER, M. "Direct and Projective Assessment of Alienation Among Affluent Adolescent
7,49-55. Males." Journal of Projective Techniques and Personality Assessment, 1970, 34, 41-44.
PERRY, S. E. "Notes on the Role of the National: A Social-Psychological Concept for the PROSHANSKY, H. "A Projective Method for the Study of Attitudes." Journal of Abnormal
$1, .
Study of International Relations." Journal of ConfEict Resolution, 1957, 1, 346-363. and Social Psychology, 1943, 38, 383-395.
PETTEE, G. S. The Process of Revolution. New York: Harper, 1938. PROTHRO, E. T., AND MELIKIAN, L. "The California Public Opinion Scale in an Authoritarian
PETTIGREW, T. "Personality and Sociocultural Factors in Intergroup Attitudes: A Cross- I Culture." Public Opinion Quarterly, 1953, 17, 353-362.
National Comparison." Journal of Conflict Resolution, 1958, 2, 29-42. PRUITT, D. G. "Definition of the Situation as a Determinant of International Action." In
PETTIGREW, T. F. "Social Evaluation Theory: Convergences and Applications." In D. Levine H. C. Kelman (Ed.), International Behavior: A Social-Psychological Analysis. New
(Ed.), Nebraska Symposium on Motivation. Lincoln: University of Nebraska Press, York: Holt, 1965.
1967. PRUITT,D. G., AND GAHAGAN, J. P. "Campus Crisis: The Search for Power." Unpublished
PIAGET, J. The Moral Judgment of the Child. New York: Free Press, 1965. Originally manuscript, State University of New York at Buffalo, 1972.
published 1932. PRUITT, D. G., AND SNYDER, R. C. (Eds.) Theory and Research on the Causes of War. Engle-
PIAGET,J., AND WEIL, A. "The Development in Children of the Idea of the Homeland and of wood Cliffs, N. J. : Prentice-Hall, 1969.
a Relation with Other Countries." Interna,tional Social Science Bulletin, 1951, 3, PUTNAM, R. "Studying Elite Culture: The Case of 'Ideology.' " American Political Science
561-578. Review, 1971, 65, 651-681.
PIERCE, E. L. Memoir and Letters of Charles Sumner. London: Sampson Low, 1878-1893. PYE, L. W. "Personal Identity and Political Ideology." In D. Marvick (Ed.), Political.De-
4 Vols. cision Makers. New York: Free Press, 1961. i- ?.i.-.
sI :

PILIAVIN, I., RODIN, J., AND PILIAVIN, J. A. "Good Samaritanism: An Underground Phenom- PYE, L. W. Politics, Personality and Nation Building. New Haven: Yale University Press,
enon?" Journal of Personality and Social Psychology, 1969, 13, 289-299. 1962.
PILISUK,M., AND RAPOPORT, A. "A Non-Zero Sum Game Model of Some Disarmament Prob- PYE, L. w., AND VERBA, S. (Eds.) Political Culture and Political Development. Princeton:
lems." Peace Research Society (International) Papers, 1964, 1, 57-58. Princeton University Press, 1965. : , 1 ., l i .
-. .
PILISUK, M., AND SKOLNICK, P. ''Inducing Trust: A Test of the Osgood Proposal." Journal of QUAY, W. B., BENNETT, E. L., ROSENZWEIG, M..R.,'AND KRECH, D. "Effects of ~s'olationand
Personality and Social Psychology, 1968,8, 121-133. 1 Environmental Complexity on Brain and Pineal Organ." Physiology and Behavior,
PIN, E. "Hypotheses Relative i la Desaffection Religieuse dans les Classes Inferieures." Social 1969,4,489-494. '.

Compass, 1962, 9, 5 15-537. RABIN, A. Projective Techniques in Personality Assessment. New York: Springer. 1968.
A - , --
PINE,F., AND LEVINSON, D. J. "TWOPatterns of Ideology, Role Concepion, and Personality RAINIO, K. "Social Interaction as a Stochastic Learning Process." European Journal of
Among Mental Hospital Aides." In M. Greenblatt, D. J. Levinson, and R. H. Williams Sociology, 1965, 6, 68-88.
(Eds.), The Patient and the Mental Hospital. New York: Free Press, 1957. RAINIO, K. "A Study on Socio-metric Group Structure: An Application of a
512 References
References 513

Gaming and Simulation." I n D. G. Pruitt and R. C. Snyder (Ecis.), Theory and Re-
Theory of Social Interaction." I n J. Berger, M. Zelditch, Jr., and B. Anderson (Eds.),
search on the Causes of War. Englewood Cliffs, N.J.: Prentice-Hall, 1969.
Sociological Theories in Progress. Boston: Houghton Mifflin, 1966.
RANSFORD, H. E. ' L I ~ ~ l a t i oPowerlessness
n, and Violence: A Study of Attitudes and Par- ROBINSON, J. P., AND HEFNER, R. "Multidimensional Differences in Public and Academic Per-

ticipation in the Watts Riot." American Journal of Sociology, 1968, 73, 581-591. ceptions of Nations." Journal of Personality and Social Psychology, 1967, 7, 251-259.
ROBINSON, J. P. AND HEFNER,R. "Perceptual Maps of the World." Public Opinion Quarterlyj
RAPOPORT, A. "Conflict Reso!urion in the Light of Game Theory and Beyond." I n P. Swingle
(Ed.), T h e Structure of Conflict. New York: Academic Press, 1970. 1968,32, 273-280.
ROBINSON, J. P., RUSK,J., AND HEAD,K. Measures of Political Attitudes, Ann Arbor, Mich.:
RAPOPORT, A., AND CHAMMAH, A. M. Prisoner's Dilemma: A Study in Conjlict and Coopera-
tion. Ann Arbor: University of Michigan Press, 1965. Survey Research Center, 1968.
ROBINSON, J. P., AND SHAVER, P. Measures of Social Psychological Attitudes. Ann Arbor,
RAPOPORT, D. "Principles Underlying Projective Techniques." Character and Personality,
1942,10,213-219. Mich. : Survey Research Center, 1969.
ROBINSON, W. "The Motivational Structure of Political Participation." American Sociological
RASER, J. R. "Personal Characteristics of Political Decision-Makers: A Literature Review."
Peace Research Society (International) Papers, 1966, 5, 161-181. Review, 1952, 17, 151-156.
ROETHLISBERGER, F. J., AND DICKSON, W. J. Management and the Worker. Cambridge: Har-
RASER, J. R. Simulation and Society: A n Exploration of Scientific Gaming. Boston: Allyn and
Bacon, 1969. vard University Press, 1939.
RGGERS, E. Diffusion of Innovations. New York: Free Press, 1962.
RASER, J. R., CAMPBELL, D. T., AND CHADWICK, R. W. ''Gaming and Simulation for Develop-
ROOERS, H., JR., AND TAYLOR, G. "The Policeman as an Agent of Regime Legitimation." Mid-
ing Theory Relevant to International Relations." General Systems, 1970, 15, 183-204.
RASER, J . R., AND CROW, W. J. "A Simulation Study of Deterrence Strategies." In D. G.
west Journal of Political Science, 1971, 15, 72-86.
ROGHMANN, K., AND SODEUR, W. "The Impact of Military Service on Authoritarian Attitudes:
Pmitt and R. C. Snyder (Eds.), Theory and Research on the Causes of War. Engle-
wood Cliffs, N.J. : Prentice-Hall, 1969. Evidence from West Germany." American Journal of Sociology, 1972, 78, 418-433.
ROGOW, A. A. James Forrestal: A Study of Personality, Politics and Policy. New York:
RAY, J. J. "Militarism, Authoritarianism, Neuroticism and Antisocial Behavior." Journal of
Conflict Resolution, 1972, 16, 3 19-340. Macmillan, 1963.
R O ~ W A., A. 'rRe~iewof V. Wolfenstein, T h e Revolutionary Personality." American Political
REBACK, G. L. "The Effects of Precinct-Level Voter Contact Activities on Voting Behavior."
Experimental Study of Politics, 1971, 1, 65-98. Science Review, 1968, 62,604-666,
ROGOW, A. A., AND LASSWELL, H. D. Power, Corruption and Rectitude. Englewood Cliffs,
REED, J., JR. T h e Enduring South. Lexington, Mass.: Heath, 1972.
REICH, C. T h e Greening of America. New York: Random House, 1970.
N.J. : Prentice-Hall, 1963.
REICH, W. T h e Mass Psychology of Fascism. (3rd Ed.) New York: Orgone Press, 1946.
ROHTER, I. S. "The Genesis of Political Radicalism: The Case of the Radical Right." I n S.

Originally published 1933. Sigel (Ed.), Learning About Politics: A Reader in Political Socialization. New York:
RENSHON, S. T h e Psychological Origins of Political Eficacy: T h e Need for Personal Control.
Random House, 1970.
ROIG, c., AND BILLON-GRAND, F. La Sociolisation Politique des Enfants: Contribution d Pe'tude
Washington, D.C.: American Political Science Association, 1972.
de la Formation des Attitudes Politiques en France. Paris: Librairie Armand Colin,
RICHARDS, I. A. Practical Criticism. New York: Harcourt, 1954.
1968.
RICHARDSON, L. F. Arms and Insecurity. Chiczgo: Quadrangle Books, 1960.
ROKEACH, M. "The Nature and Meaning of Dogmatism." Psychological Review, 1954, 61,
RIESMAN, D. T h e Lonely Crowd. New Haven: Yale University Press, 1950.
194-204.
RIESMAN, D., AND GLAZER, N. Faces in the Crowd: Individual Studies in Character and ROKEACH, M. "Political and Religious Dogmatism: An Alternative to the Authoritarian
Politics. New Haven: Yale University Press, 1952. Personality." Psychological Monographs, 1956, 70, 1 4 3 .
RIESSMAN, F., AND MILLER,S. "Social Class and Projective Tests." Journal of Projective ROKEACH, M. T h e Open and Closed Mind: Investigations into the Nature of Belief Systems
Techniques, 1958, 22, 432-439. and Personality Systems. New York: Basic Books, 1960.
RING, K. ccExperimental Social Psychology: Some Sober Questions About Some Frivolous ROKEACH, M. Beliefs, Attitudes, and Values: A Theory of Organization and Change. San
Values." Journal of Experimental Social Psychology, 1967, 3, 113-123. Francisco : Jossey-Bass, 1968.
ROAZEN,P. Freud: Political and Social Thought. New York: Knopf, 1968. I ROPER, E. YOU and Your Leaders: Their Actions and Your Reactions, 1936-1956. New York:
ROBERTS, A. H., AND JESSOR, R. "Authoritarianism, Punitiveness and Perceived Social Status." Morrow, 1957.
Journal of Abnormal and Social Psychology, 1958,56,311-314. RORER, L. C. "The Great Response-Style Myth." Psychological Bulletin, 1965, 63, 129-156.
ROBERTS, A. H., AND ROKEACH, M. "Anomie, Authoritarianism and Prejudice: A Replication." ROSENAU, J. N. National Leadership and Foreign Policy: A Case Study in the Mobilization
American Journal of Sociology, 1956, 61,355-358. of Public Support. Princeton: Princeton University Press, 1963.
ROBINSON, C. J. "Malcolm Little as a Charismatic Leader." Paper presented at meeting of ROSENAU, J. N. "Pre-Theories and Theories of Foreign Policy." I n R. B. Farrell (Ed.),

American Political Science Association, Los Angeles, September 1970. Approaches to Comparative and International Politics. Evanston, Ill:. Northwestern
ROBINSON, J. A., ANDERSON, L., HERMANN, M., AND SNYDER,R. "Teaching with Interna- University Press, 1966.
tion Simulation and Case Studies." American Political Science Review, 1966, 63, ROSENAU, J. N. (Ed.) Domestic Sources of Foreign Policy. New York: Free Press, 1967.

53-65. ROSENBERG, M. "Some Determinants of Political Apathy." Public Opinion Quarterly, 1954,

ROBINSON, J. A., HERMANN, C. F., AND HERMANN, M. G. "Search Under Crisis in Political 18,349-366.
 References I References 515
ROSENBERG, M. "Misanthropy and Political Ideology." American Sociological Review, 1956, SANFORD, F. Authoritarianism and Leadership. Philadelphia: Stephenson Brothers, 1950a.
21, 69G695. SANFORD, F. "The Use of a Projective Device in Attitude Surveying." Public Opinion
ROSENBERG, M. "Misanthropy and Attitudes Toward International Affairs." Journal of Con- Quarterly, 1950b, 14, 697-709.
flict Resolution, 1957, 1, 340-345. SANFORD, N. "The Effect of Abstinence from Food upon Imaginal Processes: A Preliminary
ROSENBERG, M. T h e Logic of Survey Analysis. New York: Basic Books, 1968. Experiment." Journal of Psychology, 1936, 2, 129-136.
ROSENBERG, M. J. "An Analysis of Affective-Cognitive Consistency." I n M. J. Rosenberg, SANFORD, N. "Individual and Social Change in a Community Under Pressure: The Oath
C. I. Hovland, W. J. McGuire, R. P. Abelson, and J. W. Brehm (Eds.), Attitude Controversy." Journal of Social Issues, 1953, 9, 25-42.
Organization and Change: An Analysis of Consistency Among Attitude Components. SANFORD, N. "Recent Developments in Connection with the Investigation of the Authoritarian
New Haven: Yale University Press, 1960. Personality." Sociological Review (New Series, British), 1954, 2, 11-33.
ROSENBERG, M. J. "Images in Relation to the Policy Process: American Public Opinion on SANFORD, N. "The Approach of the Authoritarian Personality." I n J. L. McCary (Ed.),
Cold-War Issues." I n H. C. Kelman (Ed.), International Behavior: A Social- Psychology of Personality. New York: Logos Press, 1956.
Psychological Analysis. New York: Holt, 1965a. SANFORD, N. "Impact of a Women's College upon Its Students." In A. Traxler (Ed.), Long-
ROSENBERG, M. J . "When Dissonance Fails: On Eliminating Evaluation Apprehension from Range Planning for Education. Washington, D.C.: American Council on Education,
Attitude Measurement." Journal of Personality and Social Psychology, 1965b, 1, 2842. 1957.
ROSENBERG, M . J. "Attitude Change and Foreign Policy in the Cold War Era." I n J. N. SANFORD, N. "The Professor Looks at the Student." In R. Cooper (Ed.), T h e T w o Ends of
Rosenau (Ed.), Domestic Sources of Foreign Policy. New York: Free Press, 1967. Pp. the Log. Minneapolis: University of Minnesota Press, 1958.
111-159. SANFORD, N. "Knowledge of Students Through the Social Studies." In N. Brown (Ed.),
ROSENBERG, M. J., AND ABELSON, R. P. "An Analysis of Cognitive Balancing." In M. J.
Spotlight on the College Student. Washington, D.C.: American Council on Education,
Rosenberg, C. I. Hovland, W. J. McGuire, R. P. Abelson, and J. W. Brehm (Eds.), 1959a.
Attitude Organization and Change: A n Analysis of Consistency Among Attitude Com- SANFORD, N. "Motivation of High Achievers." I n 0. D. David (Ed.), The Education of
ponents. New Haven: Yale University Press, 1960. Women: Signs fur the Future. Washington, D.C.: American Council on Education,
ROSENBERG, M. J., VERBA, s., AND CONVERSE, P. E. Vietnam and the Silent Majority. T h e
1959b.
Dove's Guide. New York: Harper, 1970.
SANFORD, N. "The Development of Maturity of Personality in College." I n T. R. McConnell
ROSENTHAL, H. "Election Simulation." European Journal of Sociology, 1965, 6, 21-42.
(Ed.), Selection and Educational Differentiation. Berkeley, Calif.: Field Service
ROSENTHAL, H. "Voting and Coalition Models in Election Simulations." I n W. D. Coplin
Center and Center for the Study of Higher Education, University of California, 1960a.
(Ed.), Simulation i n the Study of Politics. Chicago: Markham, 1968.
SANFORD, N. "Theories of Higher Education and the Experimental College." I n S. Harris
ROSENTHAL, R. Experimenter Effects in Behavioral Research. New York: Appleton-Century,
(Ed.), Higher Education in the United States. Cambridge: Harvard University
1966.
Press, 1960b.
ROSENTHAL, R. "Interpersonal Expectations: Effects of the Experimenter's Hypothesis." In
SANFORD, N. "Recent Research on the American College Student." In N. Brown (Ed.),
R. Rosenthal and R. L. Rosnow (Eds.), Artifact in Behavioral Research. New York:
Academic Press, 1969. Orientation to College Learning. Washington, D.C.: American Council on Education,
ROSENTHAL, R., AND ROSNOW,R. L. (Eds.) Artifact i n Behavioral Research. New York: 1961.
SANFORD, N. "Ends and Means in Higher Education." I n K. Smith (Ed.), Current Issztes
Academic Press, 1969.
ROSSITER, C. T h e Political Thought of the American Revolution. New York: Harcourt, 1953.
in Higher Education. Washington, D.C.: National Education Association, 1962a.
ROSTOW, W. W. T h e Process of Economic Growth. New York: Norton, 1952. SANFORD, N. T h e American College: A Psychological and Social Interpretation of the Higher

ROSZAK, T. T h e Making of a Counter Culture. Garden City, N.Y.: Doubleday, 1969. Learning. New York: Wiley, 1962b.
ROTTER, J. B. "Generalized Expectancies for Internal Versus External Control of Reinforce- SANFORD, N. "The Freeing and Acting Out of Impulse in Late Adolescence: Evidence from
ment." Psychological Monographs, 1966, 80 (no. 1), 1-28. Two Cases." I n R. White (Ed.), T h e Study of Lives: Essays on Personality in Honor
RUGGIERO, G. DE. T h e History of European Liberalism. Trans. R. G. Collingwood, London: of Henry A . Murray. New York: Atherton, 1963.
Oxford University Press, 1927. SANFORD, N. (Ed.) College and Character: A Briefer Version of "The American College."
RUSHING, W. A. "Class, Culture and Social Stmcture and Anomie." American Journal of New York: Wiley, 1964a.
Sociology, 1971, 76, 857-872. SANFORD, N. "Ego Processes in Learning." In N. M. Lambert and others (Eds.), T h e Pro-
RUSSET,B. M. "Inequality and Instability: The Relation of Land Tenure to Politics." World tection and Promotion of Mental Health in Schools. Public Health Service Publication
Politics, 1964, 16, 442-454. 1226, Mental Health Monograph 5. Washington, D.C.: U.S. Department of Health,
RUSTOW, D. A. "The Study of Leadership." Daedalus, 1968, 97. Education and Welfare, 1964b.
KUTHERFORD, B. "Psychopathology, Decision-Making and Political Involvement." Journal of SANFORD, N. ' L P ~ y ~ h o l o g iand
~ a l Developmental Aspects of the Adolescent Years as They
Conflict Resolution, 1966, 10, 387-407. Apply to the Use of Alcohol." I n H. B. Bruyn (Ed.), Alcohol and College Youth.
SALISBURY, H. T h e 900 Days. New York: Harper, 1969. Berkeley, Calif.: American College Health Association, 1966a.
SALISBURY, R. H. "The Urban Party Organization Member." Public Opinion Quarterly, 1966, SANFORD, N. Self and Society: Social Change and Individual Development. New York:
29,550-564. Atherton, 1966b.
516 References References 517

SANFORD, N. Where Colleges Fail: A Study of the Student as a Person. San Francisco: SCHLESINGER, A. M. JR., T h e Age of Roosevelt. Vol. 2. T h e Coming of the New Deal. Boston:

Jossey-Bass, 1967. Houghton Mifflin, 1959.

SCHLESINGER, J. A. Ambition in Politicr: Political Careers in the United States. Chicago:
SANFORD, N. "Education for Individual Development." American Journal of Orthopsychiafry,
Rand McNally, 1966.
1968a, 30,858-865.
ScHMmL, F. "Psychoanalysis and History." Psychoanalytic Quarterly, 1962, 31, 532-548.
SANFORD, N. "Personality: The Field." I n D. Sills (Ed.), International Encyclopedia of the
Social Sciences. New York: Macmillan, 1968b. ScHNEmER, J. "The Cultural Situation as a Condition for the Achievement of Fame."
American Sociological Review, 1937, 2, 480491.
SANFORD, N. "The University and the Life of the Student: The Next 100 Years." In J.
SCHOENBERGER, R. A. "Conservatism, Personality and Political Extremism." American Poli-
Walsh (Ed.), T h e University in a Developing World Society. South Bend, Ind.:
tical Science Review, 1968, 62, 868-877.
University of Notre Dame Press, 1968c.
SCHRODER, H. M., DRIVER, M. J., AND STREUFERT, S. Human Information Processing. New
SANFORD, N. "Research with Students as Action and Education." American Psychologist,
York: Holt, 1967.
1969,24,544-546.
SCHUMAN, H. "The Random Probe." American Sociological Review, 1966, 31, 218-222.
SANFORD, N. "The Campus Crisis in Authority." Educational Record, 1970a, pp. 112-115.
SCHWARTZ, D. a. rcToward a Theory of Political Recruitment." Western Political Quarterly,
SANFORD, N. Issues in Personality Theory. San Francisco: Jossey-Bass, 1970b.
1969,22,552-571.
SANFORD, N. "Whatever Happened to -4ction Research?" Journal of Social Issues, 1970c, 26,
SCHWARTZ, D. C. "A Theory of Revolutionary Behavior." In J. C. Davies (Ed.), When Merz
3-23.
Reuolt and Wlzy. New York: Free Press, 1971.
SANFORD, N. "Academic Culture and the Teacher's Development." Soundings, 1971-1972,
ScHwmTz, M. Public Opinion and Canadian Identity. Berkeley: University of California
357-371.
Press, 1967.
SANFORD, N. "The New Values and Faculty Response." In E. McGrath (Ed.), Prospects for
SCOTT, W. A. "Attitude Measurement." In G. Lindzey and E. Aronson (Eds.), T h e Handbook
Renewal. San Francisco : Jossey-Bass, 1972.
of Social Psychology. (Rev. Ed.) Vol. 2. Reading, Mass.: Addison-Wesley, 1968.
SANFORD, N., ADKINS,M., MILLER, R. B., AND COBB, E. "Physique, Personality, and Scholar-
SCOTT, W. R. "Field Methods in the Study of Organizations." In J. G. March (Ed.), Hand-
ship." Monographs of the Society for Research in Child Development, 1943, 8.
book of Organizations. Chicago: Rand McNally, 1965.
SANFORD, N., AND COIVISTOCK, C. (Eds.) Sanctions for Evil. San Francisco: Jossey-Bass, 1971.
SEARS, D. O. "Review of R. Hess and J. Torney, T h e Development of Political Attitudes in
SANFORD, N., WEBSTER, H., AND FREEDMAN, M. "Impulse Expression as a Variable of Person-
Children." Harvard Educational Review, 1968, 3(?, 571-578.
ality." Psychological Monographs, 1957, 71.
SARBIN, T. R., AND ADLER, N. c'Self-Rec~n~titution Processes: A Preliminary Report." Psy- SEARS,D. O. "Political Behavior." In G. Lindzey anc! E. Aronson (Eds.), T h e Handbook of
choanalytic Review, 1970-1971, 57, 599-616. Social Psychology. (2nd Ed.) Vol. 5. Reading, Mass.: Addison-Wesley, 1969.
SEARS, D. o., AND M C cONAHAY, J. B. T h e Politics of violence. Boston: Houghton Mifflin,
SARNOFF, I. "Identification with the Aggressor: Some Persona!ity Correlates of Anti-Semitism
Among Jews." American Psychologist, 1951, 6, 3 19-320. 1973.
SARNOFF, I. c L P ~ y ~ h o a n a ITheory
y t i ~ and Socia! Attitudes." Public Opinion Quarterly, 1960, SEELEMAN, V. "The Influence of Attitude upon the Remembering of Pictorial Material."

24,25 1-279. Archives of Psychology of N.Y., 1940-1941,36, 258.

SARNOFF,I. Personality Dynamics and Development. New York: Wiley, 1962. SEEMAN,M. "On the Meaning of Alienation." American Sociological Review, 1959, 24,
SARNOFF, I., AND KATZ,D. "The Motivational Bases of Attitude Change.'~ournalof Abnor-
783-791.
SEEMAN, M. "Alienation and Social Learning in a Reformatory." American Journal of
mal and Social Psychology, 1954, 49, 115-124.
SARNOFF, I., AND ZIMBARDO, P. "Anxiety, Fear and Social Affiliation." Journal of Abnormal
Sociology, 1963, 69, 27C-283.
and Social Psychology, 1961, 62,356-363. SEEMAN, M. "Alienation, Membership and Political Knowledge: A Comparative Study."

SAWYER, J., AND GUETZKOW, H. "Bargaining and Negotiation in International Relations." I n Public Opinion Quarterly, 1966, 30, 353-367.
H. C. Kelnlan (Ed.), International Behavior: A Social-Psychological Analysis. New SEEMAN, M. "Alienation and Knowledge: A Comparative Study of Powerlessness and Learn-

York: Holt, 1965. ing." Sociometry, 1967a, 30, 105-123.

SCHACHT, R. Alienation. Garden City, N.Y.: Doubleday, 1971. With an introductory essay by SEEMAN, M. "On the Personal Consequences of Alienation in Work." American Sociological
Walter Kaufman. Review, 1967b, 32,273-285.
SCHAULAND, H., NAYLOR, T. H., AND KORNBERG, A. "ELECTION: A Campaign simulator." SEEMAN, M. "The Urban Alienations: Some Dubious Theses from Mam to Marcuse." Jozlr-
Paper prepared for annual meeting of American Political Science Association, nal of Personality and Social Psychology, 1971, 19, 135-143.
Chicago, 1971. SEEMAN, M. "The Signals of '68: Alienation in Pre-Crisis France." American Sociological
SCHEIN, E. H., SCHNEIER, I., AND BARKER, C. H. Coercive Persuasion: A Socio-Psychological Review, 1973,37,385402.
Analysis of "Brainwashing" of American Civilian. Prisoners by the Chinese Com- SELZNICK, G. J., AND STEINBERG, S. T h e Tenacity of Prejudice: Anti-Semitism in Con-
munists. New York : Norton, 1961. temporary America. New York: Harper, 1369.
SCHELLING, T. c., AND HALPERIN, M. H. "Pre-emptive, Premeditated and Accidental War." SERENO, R. T h e Rulers. New York: Praeger, 1962.
I n D. G. Pruitt and R. C. Snyder (Eds.), Theory and Research on the Causes of SHAFFER,W. R. Computer Simulations of Voting Behavior. New York: Oxford University
War. Englewood Cliffs, N.J. : Prentice-Hall, 1969. Press, 1972.
SCHILDER, P. "The Analysis of Ideologies as a Psycho-Therapeutic Method Especially in
SHAND, A. F. T h e Foundations of Character. London: Macmillan, 1914.
Group Treatment." American Journal of Psychiatry, 1936, 93, 601-617.
 References
References 519
SHAPIRO, M. J. "Rational Political Man: A Synthesis of Economic and Social-Psychological
SINGER, J. D. "Inter-Nation Influence: A Formal Model." American Political Science Re-
Perspectives." American Political Science Review, 1969, 63, 1106-1119.
view, 1963, 57, 420430.
SHAW,R. E. "Cognition, Simulation and the Problem of Complexity." Journal of Structural
SINGER,J. D. "Data-Making in Intemational Relations." Behavioral Science, 1965, 10,
Learning, 1971, 2, 31-44.
68-80.
SHERIF,C. w., SHERIF,M., AND NEBERGALL, R. E. Attitude and Attitude Change: T h e Social
SINGER, J. D. "Man and World Politics: The Psycho-Cultural Interface." Journal of Social
Judgment Approach. Philadelphia : Saunders, 1965.
Issues, 1968, 24, 127-156.
SHERIF, M. T h e Psychology of Social Norms. New York: Harper, 1936.
SINGER,J. D. "The Correlates of War Project: Interim Report and Rationale." World
SHERIF, M. "Group Influences upon the Formation of Norms and Attitudes." In T. M. New-
Politics, 1972, 24, 243-270.
comb and E. L. Hartley (Eds.), Readings in Social Psychology. New York: Holt,
SKINNER, B. F. Walden Two. New York. Macmillan, 1948.
1947.
SKINNER, B. F. Verbal Behavior. New York: Appleton-Century, 1957.
SHERIF, M., AND CANTRIL, H. T h e Psychology of Ego-Involvements. New York: Wiley, 1947.
SKINNER, B. F. Beyond Freedom and Dignity. New York: Knopf, 1971.
SHERIF,M., AND HOVLAND, C. I. Social Judgment: Assimilation and Contrast Egects in Com-
munication and Attitude Change. New Haven: Yale University Press, 1961. SMELSER,N. J., AND W. T. (Eds.) Personality and Social Systems. New York: Wilev. 1965.
, z

SHNEIDMAN, E. Thematic Test Analysis. New York: Grune and Stratton, 1951. SMITH, B. F. Adolf Hitler. Stanford, Calif.: Hoover Institution on War, Revolution and
SHNErnMAN, E. "Projective Techniques." I n B. Wolman (Ed.), Handbook of Clinical Peace, 1967.
Psychology. New York: McGraw-Hill, 1965. SMITH, E. E. "The Power of Dissonance Techniques to Change Attitudes." Public Opinion
SHURE, G. H., MEEKER, R. J., AND HANSFORD, E. A. "The Effectiveness of Pacifist Strategies in Quarterly, 1961, 25, 626-639.
Bargaining Games." Journal of Conflict Resolution, 1965, 9, 106-1 17. SMITH, G. H. ''Liberalism and Level of Information." Journal of Education and Psychology,
SIEGEL, A. E., AND SIEGEL, S. "Reference Groups, Membership Groups and Attitude Change." 1948a, 39,65-8 1.
Journal of Abnormal and Social Psychology, 1957,55, 360-364. SMITH, G. H. "The Relation of 'Enlightenment' to Liberal-Conservative Opinions." Journal
SIGEL, I., AND HOFFMAN, M. "The Predictive Potential of Projective Tests for Nonclinical of Social Psychology, 1948b, 28, 3-17.
Populations." Journal of Projective Techniques, 1956, 20, 261-264. SMITH,L. B. Henry V I I I : T h e Mask of Royalty. London: Jonathan Cape, 1971.
SIGEL,R. "Assumptions About the Learning of Political Values." Annals of the American SMITH,M. B. "Review of T h e Authoritarian Personality." Journal of Abnormal and Social
Academy of Political and Social Science, 1965, 361, 1-9. Psychology, 1950, 45, 775-779.
SIGEL, R. Learning About Politics. New York: Random House, 1970.
SMITH, M. B. "Opinions, Personality and Political Behavior." American Political Science
SIGUSCH, v., SCHMIDT, G., REINFELD, R., AND WEIDERMANN-SUTOR, I. Psychosexual Stimulation:
Review, 1958,52, 1-17.
Sex Differences." Journal of Sex Research, 1970, 6, 10-24.
SMITH, M. B. "An Analysis of Two Measures of 'Authoritarianism' Among Peace Corps
SILBERSTEIN, F. B., AND SEEMAN, M. "Social Mobility and Prejudice." American Journal of
Teachers." Journal of Personality, 1965, 33,513-535.
Sociology, 1959, 65, 258-264.
SMITH, M. B. "Attitude Change." I n D. Sills (Ed.), International Encyclopedia of the Social
SIMMONS,J. "Liberalism, Alienation and Personal Disturbance." Sociology and Social Re-
search, 1965, 49, 456464. Sciences. Vol. 1. New York: Macmillan, 1968a.
SMITH, M. B. "Competence and Socialization." I n J. Clausen (Ed.), Socialization and
SIMMONS, J. L. "Some Intercorrelations Among Alienation Measures." Social Forces, 1966,
44, 370-372. Society. Boston: Little Brown, 1968b.
SIMON,H. A. Administrative Behavior: A Study of Decision-Making Processes in Administra- SMITH, M. B. "A Map for the Analysis of Personality and Politics." Journal of Social Issues,
tive Organization. New York: Macmillan, 1947. 1968~,24, 15-28.
SIMON,H. A. "Theories of Decision-Making in Economics and Behavorial Science." American SMITH, M. B. "Personality in Politics. A Conceptual Map, with Application to the Problem
Economic Review, 1959,40,253-283. of Political Rationality." In 0. Garceau (Ed.), Political Research and Political
SIMON, H. A., SMITHBURG, D. w., AND THOMPSON, V. A. Public Administration. New York: Theory: Essays in Honor of V . 0. Key, Jr. Cambridge: Harvard University Press,
Knopf, 1950. 1968d.
SIMON, J., AND RESCHER, P. ''Cause and Counterfactual." Philosophy of Science, 1963, 33, SMITH, M. B. "Allport, Murray, and Lewin on Personality Theory: Notes on a Confrontation."
323-340. Journal of the History of the Behavioral Sciences, 1971a, 7, 353-362.
SIMPSON,E. L. Democracy's Stepchildren. San Francisco: Jossey-Bass, 1971. SMITH, M. B. "Review of Fromm and Maccoby's Social Character in a Mexican Village."
SIMPSON,G. E., AND YINGER, J. M. Racial and Cultural Minorities. (4th Ed.) New York: Contemporary ~ s ~ c h o l 1971b,
o ~ ~ ,1.6, 635.
Harper, 1972. SMITH, M. B. "ISExperimental Social Psychology Advancing?" Review of L. Berkowitz (Ed.),
SIMPSON,M. c'A~thoritariani~m and Education: A Comparative Approach." Sociometry, Advances in Experimental Social Psychology, Vols. 1-5. Journal of Experimental
1972,35,223-234. Social Psychology, 1972,8, 8696.
SIMPSON,R. L., AND MILLER, H. M. "Social Status and Anomia." Social Problems, 1963, 10, SMITH, M. B. "ISPsychology Relevant to New Priorities?" American Psychologist, 1973.
256264. SMITH,M. B., BRUNER, J., AND WHITE, R. Opinions and Personality. New York: Wiley, 1956.
SINGER,J. D. "The Level-of-Analysis Problem in Intemational Relations." In K. Knorr and SMOKER, P. Analyses of Conflict Behaviours in an International Processes Simqlation and an
S. Verba (Eds.), T h e Interngtional System. Princeton: Princeton University Press,
International System 1955-60. Evanston, Ill.: International Processes Simulation
1961. Project, Northwestern University, 1968a.
520 References References, 521

SMOKER,P. International Processes Simulation: A Man-Computer Model. Evanston, 111.: STOUFFER, S. A. Social Research to Test Idens: Selected Writings of Samuel A. Stouter. New
International Processes Simulation Project, Northwestern University, 196813. York: Free Press, 1962.
STOUFFER, S. A., AND OTHERS. T h e American Soldier. Studies in Social Psychology in World
SMOKER, P. "Social Research for Social Anticipation." American Behavioral Scientist, 1969,
12, 7-13. War I I . Vol. 1: Adjustment in Army Life. Vol. 2: Combat and Its Aftermath.
SMOKER, P. "International Processes Simulation: An Evaluation." Peace Research Reviews,
Princeton : Princeton University Press, 1949.
STRACHEY, L. Eminent Victorians. New York: Capricorn Ed., 1963. Originally published 1918.
1970, 4.
STRAUSS, H. J. "Revolutionary Types: Russia, 1905." Journal oi Conflict Resolution, 1973, 17.
SMOKER, P. ccIntemational Relations Simulations." In H. Guetzkow, P. Kotler, and K. L.
Schultz (Eds.), Simulation in Social and Administratizte Science. Englewood Cliffs, STRAUSS, L. T h e Political Philosophy of Hobbes. Trans. E. M. Sinclair. Chicago: University

N.J. : Prentice-Hall, 1972. of Chicago Press, 1952.

STREUFERT, S., CLARDY, M., DRIVER, M., KARLINS, M., SCHRODER, H. M., A N D SUEDFELD, P.
SNIDERMAN, P. M., AND CITRIN, J. "Psychological Sources of Political Belief: Self-Esteem and
Isolationist Attitudes." American Political Science Review, 1971, 65, 401-417. 'rA Tactical Game for the Analysis of Complex Decision Making in Individuals and
SNOW, EDGAR, Red Star over China. New York: Grove Press, 1968. Originally published 1938.
Groups." Psychological Reports, 1965, 17, 723-729.
STREUFEXT, s., AND STREUFERT, S. "Effects of Conceptual Structure, Failure and Success on
SNYDER, R. .c., BRUCK, H. w., AND SAPIN, B. (Eds.) Foreign Policy Decision-Making. New
York: Free Press, 1962. Attribution of Causality and Interpersonal Attitudes." Joi~rnal of Persomality and
SOKOL, R. "Power Orientation and McCarthyism." American Journal of Sociology, 1968,
Social Psychology, 1969, 2, 138-147.
73,443452. STREUFERT, s., STREUFERT, s., AND CASTORE, C. "Complexity, Increasing Failure and Decision

SOMIT,A., TANENHAUS, J., WILKE,w., AND COOLEY, R. "The Effect of the Introductory
Making." Journal of Experimental Research in Personality, 1969, 3, 293-300.
Political Science Course on Student Attitudes Toward Personal Political Participa- STREUFERT, s., SUEDFELD, P., AND DRIVER, M. "Conceptual Structure, Information Search and
tion." American Political Science Review, 1958, 52, 1129-1 132. Information Utilization." Journal of Personality and Social Psychology, 1965, 2,
SOREL, G. Reflections on Violence. New York: Collier, 1961. Originally published 1908. 736-740.
SOROKIN, P. T h e Sociology of Revolution. Philadelphia: Lippincott, 1925.
STREUNING, E. L., AND RICHARDSON, A. H. "A Factor Analytic Exploration of the Alienarion,

SPIRO, M. "Social Systems, Personality, and Functional Analysis." In B. Kaplan (Ed.),

Anomia and Authoritarianism Domain." American Sociological Review, 1965, 30,
Studying Personality Cross-Culturally. New York: Harper, 1961. 768-776.
SPITZ, R. A. "The Role of Ecological Factors in Emotional Development in Infancy." Child SULLIVAN, H. S. "Conceptions of Modern Psychiatry." Psychiatry, 1940, 3, 1-117.

Development, 1949,20, 145-155. SUTTON,F. x., HARRIS, S. E., KAYSEN,c., AND TOBIN, J. T h e American Business Creed.
SPITZ,R. A. A Genetic Field Theory of Ego Formation: Its Implications for Pathology. New Cambridge : Harvard University Press, 1956.
York: International Universities Press, 1959. SWINGLE, P. "Dangerous Games." I n P. Swingle (Ed.), T h e Structure of Conflict. New

SROLE, L. "Social Integration and Certain Corollaries: An Exploratory Study." American York: Academic Press, 1970a.
Sociological Review, 1956, 21, 709-716. SWINGLE, P. (Ed.) T h e Structure of Conflict. New York: Academic Press, 1970b.
SROLE,L. "A Comment on 'Anomy.' " American Sociological Review, 1965, 30, 757-762. SYKES,G. (Ed.) Alienation: T h e Cultural Climate of Modern Man. New York: Braziller,
STAGNER, R. "Fascist Attitudes: Their Determining Conditions." Journal of Social Psychology, 1964. 2 Vols.
1936, 7,438454. TAJFEL, H. "The Formation of National Attitudes: A Social-Psychological Perspective." In
STANLEY, J. C. "Fixed, Random and Mixed Models in the Analysis of Variance as Special M. Sherif and C. W. Sherif (Eds.), Interdisciplinary Relationships in the Social
Cases of Finite Model 111." Psychological Reports, 1956,2, 369. Sciences. Chicago : Aldine, 1969.
STANTON, A. H., AND PERRY,S. E. (Eds.) Personality and Political Crisis. New York: Free TANNENBAUM, A. J. (Issue Ed.) "Alienated Youth." Journal of Social Issues, 1969, 25.
Press, 1951. TANSEY, P. J., AND UNWIN, D. Simulation and Gaming ir. Education. London: Methuen, 1969.
STAPEL, J. "The Convivial Respondent." Public Opinion Quarterly, 1947-1948, 11, 524-529.
TANTER, R. "International War and Domestic Turmoil: Some Contemporary Evidence." In
STARK, S. "Toward a Psychology of Charisma: I. The Innovation Viewpoint of Robert
H. D. Graham and T. R. Gurr (Eds.), T h e History of Violence in America. New
Tucker." Psychological Reports, 1968, 23, 1163-1 166.
York: Bantam, 1969.
STARK, S. "Toward a Psychology of Charisma: 11. The Pathology Viewpoint of James C.
TANTER, R., AND MIDLARSKY, M. 'A Theory of Revolution.'' Journal of Conflict Resolution,
Davies." Psychological Reports, 1969, 24, 88-90.
1967, 11, 264-280.
STARK,W. T h e Sociology of Religion. New York: Fordham University Press, 1967. 3 Vols.
TAPP,J., AND KOHLBERG, L. "Developing Senses of Law and Legal Justice." Journal of Social
STEVENSON, H., AND STUART,E. "A Developmental Study of Racial Awareness in Young
Issues, 1971, 27, 65-92.
Children." Child Development, 1958, 29, 399-409.
TAPPER, E., AND BUTLER, R. "Continuity and Change in Adolescent Political Party Prefer-
STOKES, D. E. "Some Dynamic Elements of Contests for the Presidency." American Political
Science Review, 1966, 60, 19-28. ences." Political Studies, 1970, 18, 390-394.
TAVISS, I. "Changes in the Form of Alienation: The 1900's Versus 1950's." American Socio-
STOKES, D. E., CAMPBELL, A., AND MILLER, W. E. c c C ~ m p ~ n eofn t Electoral
~ Decision."
American Political Science Review, 1958,52, 367-387. logical Review, 1969, 34, 46-57.
STOKES,D. E., AND MILLER, W. E. "Party Government and the Saliency of Congress." Public TAWNEY, R. H. Religion and the Rise of Capitalism. Gloucester, Mass.: Smith, 1963.

Opinion Quarterly, 1962, 26, 531-546. Originally published 1926.

STOUFFER, S. A. Communism, Conformity and Civil Liberties. New York: Doubleday, 1955. TAWNEY, R. H. Equality. New York: Harcourt, 1929.
 References References 523
TAYLOR, J. "A Personality Scale of Manifest Anxiety." Journal of Abnormal and Social UNDERWOOD, B. J. Psychological Research, New York: Appleton-Century, 1957.
Psychology, 1953, 48, 285-290. VAILLANCOURT, P. "The Stability of Children's Political Orientations: A Panel Study." Paper
TEMPLETON, F. "Alienation and Political Participation: Some Research Findings." Public presented at meeting of American Political Science Association, Los Angeles, Septem-
Opinion Quarterly, 1966, 30, 249-261. ber 1970.
TERHUNE, K. W. "The Effects of Personality in Cooperation and Conflict." In P. Swingle VALEN, H., AND KATZ,D. Political Parties in Norway. Oslo: University of Oslo Press, 1964.
(Ed.), T h e Structure of Conflict. New York: Academic Press, 1970. VAUGHN, G., AND WHITE, K. LLC~nformity and Authoritarianism Re-examined." Journal of
TERHUNE,K. w., AND FIRESTONE, J. M. Psychological Studies in Social Interaction and ! Personality and Social Psychology, 1966, 3, 363-366.
Motives (SIAM). Phase 2: Group Motives in an International Relations Game. CAL VERBA, S. Small Groups and Political Behavior. Princeton: Princeton University Press, 1961.
Report No. VX-2018-G-2. Buffalo, N.Y.: Cornell Aeronautical Laboratory, 1967. I
VERBA, s., BRODY, R. A., PARKER, E. B., NIE, N. H., POLSBY,N. w., EKMAN, P., AND BLACK, G. S.
TESLER,L., ENEA,H., AND COLBY, K. M. "A Directed Graph Representation for Computer "Public Opinion and the War in Vietnam." American Political Science Review, 1967,
Simulation of Belief Systems." Mathematical Biosciences, 1968, 2, 1940. 61,317-333.
THIBAUT,J. w., AND RIECKEN, H. W. "Authoritarianism, Status and the Communication of VERBA, s., AND NIE, N. Participation in America. New York: Harper, 1972.
Aggression." Human Relations, 1955, 8, 95-120. VERBA, s., AND NIE, N. "Citizenship and Civil Milieu." In F. I. Greenstein and N. W. Polsby
THOMAS, L. "Political Attitude Congruence Between Politically Active Parents and College- (Eds.), T h e Handbook of Political Science. Reading, Massachusetts: Addison-Wesley,
Age Children: An Inquiry into Family Political Socialization." Journal of Marriage 1974.
and the Family. 1971,33,375-386. VEROFF, J. "The Use of Thematic Apperception to Assess Motivation in a Nationwide Inter-
THOMAS,W. I. "The Unadjusted Girl." Excerpted in M. Janowitz (Ed.), W . I . Thomas on view Study." Psychological Monographs, 1960, 74(12).
Social Organization and Social Personality. Chicago: University of Chicago Press, VEROFF,J. "Thematic Apperception in al~ationwideSample Survey." I n J. Kagan and G.
1966. Pp. 117-139. Originally published 1923. 1 Lesser (Eds.), Contemporary Issues in Thematic Apperceptive Methods. Springfield,
THOMAS, W. I., AND ZNANIECKI, F. T h e Polish Peasant in Europe and America. Vol. 1. Ill.: Thomas, 1961.
Boston: Badger, 1918. MCARY, J. "Word Association and Opinion Research." Public Opinion Quarterly, 1948, 12,
THOMPSON, R. c., AND MICHEL,J. B. "h;leasuring Authoritarianism: A Comparison of the 81-98.
F and D Scales." Journal of Personality, 1972, 40, 180-190. VOSE,C. E. " ~ a n u s c r i ~Collections
t and Reference Books as Data for Political Science." In
THOMPSON, W. E., AND HORTON, J. E. ' r P ~ l i t i ~Alienation
al as a Force in Political Action." F. I. Greenstein and N. W. Polsby (Eds.), T h e Handbook of Political Science. Vol.
Social Forces, 1960, 38, 190-195. 2. Reading, Mass. : Addison-Wesley, forthcoming.
THORNDIKE, R., AND GALLUP, G. "Verbal Intelligence of the American Adult." Journal of WALLACH, M. "Commentary: Active-Analytical vs. Passive-Global Cognitive Functioning."
Genetic Psychology, 1944, 30, 75-85. I n S. Messick and J. Ross (Eds.), Measurement in Personality and Cognition. New
THURSTONE, L. L. "Attitudes Can Be Measured." American Journal of Sociology, 1928, 33, York: McGraw-Hill, 1962.
529-554. I WALLAS, G. Human Nature in Politics. New York: Knopf, 1921. Originally published 1908.
TILLER,P. "Father Absence and Personality Development of Children in Singular Families." WALTER, J., AND STINNETT, N. "Parent-Child Relationships: A Decade Review of Research."
Vol. 2, Part 2. In N. Anderson (Ed.), Studies of the Family. Gottingen: Vanderhoeck Journal of Marriage and the Family. 1971, 33, 70-111.
and Ruprecht, 1957. WALTZ, K. N. Man, the State, and War. New York: Columbia University Press, 1959.
TOCQUEVILLE, A. DE. T h e Old Regime and the French Revolution. New York: Harper, 1856.
WARNER, L. G., AND DEFLEUR, M. L. "Attitude as an Interactional Concept: Social Constraint
TOLLEY,H., JR. "Socialization to War: A Questionnaire Study of Children's Attitudes and
and Social Distance as Intervening Variables Between Attitudes and Actions." Ameri-
Knowledge." Paper presented at meeting of American Political Science ~ssociation,
can Sociological Review, 1969, 34, 153-169.
Washington, D.C., September 1972.
WARNER, W. L. Democracy in Jonesville. New York: Harper, 1949.
TOLMAN, E. C. Drives Toward War. New York: Appleton-Century, 1942.
WATSON, J. B. Behaviorism. Chicago: University of Chicago Press, 1961. Originally pub-
TOZZER, A. M. "Biography and Biology." American Anthropologist, 1933, 35, 418-432.
TROTSKY, L. T h e History of the Russian Revolution. New York: Simon and Schuster, 1932. lished 1924.
TRUEBLOOD, C. K. "Sainte-Beuve and the Psychology of Personality." Character and Person- WEBB, E. J., CAMPBELL, D. T., SCHWARTZ, R. D., AND SECHREST, L. B. Unobtrusive Measures:

ality, 1939, 8, 120-143. Nonreactive Research in the Social Sciences. Chicago: Rand McNally, 1966.
TUCKER,R. C. "The Dictator and Totalitarianism." World Politics, 1965, 17, 555-583. WEBER, M. T h e Protestant Ethic and the Spirit of Capitalism. New York: Scribner's 1958.
TURK, H. "Comparative Urban Studies in Interorganizational Relations." Sociological In- Originally published 1904-1905.
quiry, 1969, 39, 108-1 10. WEBER, M. T h e Theory of Social and Economic Organization. Trans. M. Henderson and T.
TURK, H. "Interorgani~ational Networks in Urban Society: Initial Perspectives and Com- Parsons. New York: Oxford University Press, 1947. Originally published 1922.
parative Research." American Sociological Review, 1970, 35, 1-19. WEBSTER, H., FREEDMAN, M., AND HEIST,P. "Personality Changes in College Students." In
TURNER, R. H. "The Public Perception of Protest." American Sociological Review, 1969, N. Sanford (Ed.), T h e American College. New York: Wiley, 1962.
34,815-830. WEBSTER, H., SANFORD, N., AND FREEDMAN, M. 'rA New Instrument for Studying Authori-
TURNER, R. H., AND KILLIAN, L. M. Collective Behavior. Englewood Cliffs, N.J.: Prentice- tarianism in Personality." Journal of Psychology, 1955, 40, 73-84.
Hall, 1957. WEDGE, B. c c K h r u ~ h ~ at h e ~a Distance-A Study of Public Personality." Transaction,
TYLER, S. A. (Ed.) Cognitive Anthropology. New York: Holt, 1969, October 1968.
524 References
1 References 525

WOODWARD, I., AND ROPER, E. "Political Activity of American Citizens." American Political
WEICK, K. E. "Laboratory Experimentation with Organizations." I n J. G. March (Ed.),
Science Review, 1950,44, 872-885.
Handbook of Organizations. Chicago: Rand McNally, 1965.
WOODWARD, w. E. George Washington: T h e Image and the Man. New York: Boni and
WEINSTEIN, E. A. "Denial of Presidential Disability: A Case Study of Woodrow Wilson."
Livenvright, 1926.
Psychiatry, 1967, 30, 376-390.
WRIGHT, c., AND CANTOR, M. "The Opinion Seeker and Avoider: Steps Beyond the Opinion
WELDON, T. D. T h e Vocabulary of Politics. London: Penguin, 1953.
Leader Concept." Pacific Sociological Review, 1967, 10, 3 3 4 3 .
WESTBY, D. L., AND BRAUNGART, R. G. "The Alienation of Generations and Status Politics:
I WRIGHT, J. D. "The Working Class, Authoritarianism, and the War in Vietnam." Social
Alternative Explanations of Student Political Activism." I n R. Sigel (Ed.), Learning
Problems, 1972, 20, 134-150.
About Politics: A Reader in Political Socialization. New York: Random House, 1970.
I WRIGHT,Q. A Study of War. (Abridged Ed.) Chicago: University of Chicago Press, 1964.
WHITE, R, K. Nobody Wanted War: Misperception in Vietnam and Other Wars. Garden
WRIGHT, Q. "The Study of War." I n D. Sills (Ed.), International Encyclopedia of the Social
City, N.Y. : Doubleday, 1968.
Sciences. Vol. 16. New York: Macmillan, 1968.
WICKER, T. JFK and LBJ: T h e Influence of Personality upon Politics. New York: William
WYATT,F. "Review of P. Greenacre, Swift and Carroll: A Psychoanalytic Study of T w o
Noirow, 1968.
Lives." Contemporary Psychology, 1956, 1, 105-107.
WILKINSON, R. T h e Broken Rebel: A Study in Culture, Politics and Authoritarian Character.
WYLIE, L. Village in the Vaucluse. New York: Harper, 1964.
New York: Harper, 1972.
YINGER, J. M. "Research Implications of a Field View of Personality." American Journal of
WILLCOX, W. B. Portrait of a General. New York: Knopf, 1964.
Sociology, 1963, 68, 58C592.
WILLIAMS, R., AND WRIGHT, C. "Opinion Organization in a Heterogeneous Adult Population."
YINGER, J. M. Toward a Field Theory of Behanior: Personality and Social Structure. New
Journal of Abnormal and Social Psychology, 1955,51, 559-564.
York: McGraw-Hill, 1965.
WILLNER,A. R. Charismatic Political Leadership: A Theory. Research Monograph 32.
I YINGER, J. M. T h e Scientific Study of Religion. New York: Macmillan, 1970.
Princeton: Center of International Studies, Princeton University, 1968. YOUNG, o. R. T h e Intermediaries: Third Parties in International Crises. Princeton: Princeton
WILLNER, A. R. "Quasi-Revolutionary Charismatic Political Leaders and Periods of Crisis."
University Press, 1967.
Paper presented at meeting of American Political Science Association, New York, YOUNG, P. V. Scientific Social Surveys and Research. (4th. Ed.) Englewood Cliffs, N.J.:
September 1969. Prentice-Hall, 1966.
WILSON, J. Q. "The Strategy of Protest: Problems of Negro Civic Action." Journal of Con-
ZALD,M. N., AND SCHLIEWEN, R. aLEthn~-Metl~odology and Simulation of Organizational
flict Resolution, 1961, 3, 291-303. Decision Making." I n W. D. Coplin (Ed.), Simulation in the Study of Politics.
WILSON, J. Q. "Review of Gottfried's Boss Cermak of Chicago." American Journal of Chicago: Markham, 1968.
Sociology, 1962, 68, 375. ZELDITCH, M., JR., AND EVAN, W. M. "Simulated Bureaucracies: A Methodological Analysis."
WILSON, J. Q. "Violence, Pornography and Social Science." Public Interest, 1971, 22, 45-61. I n H. Guetzkow (Ed.), Simulation in Social Science: Readings. Englewood Cliffs,
WILSON, R. A. "Anomie in the Ghetto: A Study of Neighborhood Type, Race and Anomie." N.J. : Prentice-Hall, 1962.
I
American Journal of Sociology, 1971, 77, 68-88. ZELIGS, M. Friendship and Fratricide, A n Analysis of Whittaker Chambers and Alger Hiss.
WILSON, W. Constitutional Government in the United States. New York: Columbia Univer- New York: Viking, 1966.
sity Press, 1908. ZELLMAN, G., AND SEARS, D. 'cChildhood Origins of Tolerance for Dissent." Journal of Social
WINER, B. J. Statistical Principles in Experimental Design. (2nd Ed.) New York: McGraw- Issues, 1971, 27, 109-136.
Hill, 1971. ZIBLATT, D. "High School Extracurricular Activities and Political Socialization." Annals of
WIRTH, L. c r C ~ n ~ eand n s ~Mass
~ Communication." American Sociological Review, 1948, the American Academy of Political and Social Science, 1965, 361, 20-31.
13, 1-15. ZIMBARDO, P. "Relationship Between Projective and Direct Measures of Fear Arousal." Jour-
MSH, M., DEUTSCH, M., AND BIENER,L. "Differences in Conceptual Structures of Nations: nal oj Abnormal and Social Psychology, 1964, 68, 196199.
An Exploratory Study." Journal of Personality and Social Psychology, 1970, ZIMBARDO, P. "The Human Choice: Individuation, Reason and Order Versus Deindividuation,
16,361-373. I
Impulse and Chaos." I n D. Levine (Ed.), Nebraska Symposium on Motivation. Lin-
WITKIN, H. A., DYK,R. B., FATERSON, H. F., GOODENOUGH, O. R., AND KARP, S. A. Psychological' coln: University of Nebraska Press, 1969.
Diferentiation: Studies of Development. New York: Wiley, 1962. EMBARDO, P., AND EBSESEN, E. B. Influencing Attitudes and Changing Behavior. Reading,

WITKIN,H., LEWIS,H., HERTZMAN, M., MACHOVER, K., MEISSNER, P., AND WAPNER,S. Per- Mass.: Addison-Wesley, 1969.
sonality Through Perception. New York: Harper, 1954. ZIMBARDO, P., AND FORMICA, R. "Emotional Comparison and Self-Esteem as Determinants of

WOLFENSTEIN, E. V. T h e Revolutionary Personality: Lenin, Trotsky and Gandhi. Princeton: Affiliation." Journal of Personality, 1963, 31, 141-162.
Princeton University Press, 1967.
WOLFENSTEIN, E. V. Personality and Politics. Belrnont, Calif.: Dickenson, 1969.
WOLFERS, A. "The Actors in International Politics." In W. T. R. Fox (Ed.), Theoretical
Aspects of International Relations. Notre Dame, Ind.: University of Notre Dame
Press, 1959.
WOLFINGER, R. E. "The Influence of Precinct Work on Voting Behavior." Public Opinion
Quarterly, 1963, 27, 387-398.
 Index

BAYES, J., 330, 504 BROWN, L., 52,509

BAYLEY, N.. 39, 495 BROWN, R., 14111, 153n, 155, 451, 476
BROWN, S. R., 356, 370, 476
BROWNING, R. P., 41, 46, 229, 383-412, 428,
447, 457, 467, 476477

INDEX BROWNLEE, K. A., 359, 477

BRSTRYN,
BRUCK,
D., 155, 476
H. w., 266, 520
BRUNER, J . s., 75, 78, 95, 96, 106, 298, 307,
312, 313, 417, 473, 477, 519
BRYAN, J., 378, 477
BUCHANAN, w., 229, 485
BUCKLE, G. E., 297, 506
BUDNER, s., 168, 477
BULU~T, W. c., 299, 316, 487
BULLOUGH, B., 182, 199, 477
BURDICK, D. s., 386, 410, 507
BUROS, o. K., 423, 477.
Author BURTON, J. w., 268, 289-290, 293, 463, 477
BURTON, R., 129, 477
ABCARIAN, G., 200, 471 APSLER, R., 380, 473 BUSHMAN, R. L., 314n, 319, 477
ABEL,T., 444, 471 ARCHIBALD, P., 264, 476 BUTLER, D., 129, 134, 477
ABELSON, H. I., 64, 357, 358, 449, 471, 496 ARDREY, R., 239, 240-241, 259, 473 BUTLER, (E. w., 182, 506
ABELSON, R. P., 60, 62, 63, 69, 71, 360, 383, ARENDT, H., 302, 473 BUTLER, R., 129,521
391, 393, 394-395, 396-397, 407, 410, ARNOLD, M., 426-427, 473
412, 471472, 488, 496, 511, 514 ARON, B., 143, 436, 473
ABERBACH, J. D., 127, 190, 199, 472 ARONOFF, J., 7, 54, 232, 473
ABERLE, D. F., 308, 472 ARONSON, E., 60, 63, 69, 357, 362, 363, 381,
ABRAHAM, K., 298, 472 382, 467, 471, 473, 502
ABRAMSON, P., 123, 124, 472 ASCH,S. E., 11, 68, 71, 336, 474
ADELSON, J., 120, 125, 155, 163, 167, 431- ATKINSON, J., 336, 417, 418, 428, 435
432, 472 AUSTIN, G. A., 106, 477
ADKINS, M., 151,516 AXELROD, J., 162, 474
ADLER, A., 4, 6, 8, 472 AXELSON, L. J., 197, 486
ADLER, N., 74, 120,472,516 AZRAEL, J., 118, 123, 474
ADORNO, T., 17, 52, 61-62, 73, 78, 84, 92-93,
14111, 151, 152,378,417,425,451,472 BAIN,R., 297, 474
AGGER, R. E., 45, 344, 472 BAKER, J. M., 156,478
ALGER, C. F., 261, 264, 267, 383, 385 BALDWIN, A. L., 314, 474
ALKER, H. A., 29, 35, 472 BALES,R. F., 209, 474
ALLINSMITH, B., 348, 472 BALINTFY, J:L., 386, 410, 507
ALLINSMITH, w., 348, 472 BANFIELD, E., 84, 88, 329, 474
ALLPORT, F., 343, 350, 472 BARAKAT, H., 179, 474
ALLPORT, G. w., 30, 31, 33, 57, 60, 83, 84, BARBER, J. D., 25, 41-42, 45, 48, 13411, 228,
304, 313, 315, 323, 447, 459, 472-473 230-232, 298, 303, 307, 310, 360, 391,
ALMOND, G. A., 22-23, 24, 39, 45, 81-82, 84, 403, 407, 410, 45-58, 460, 474
129, 135, 183, 195, 226227, 324, 334, BARCLAY, J., 155, 497
337, 341-342, 345-347, 354, 435, 450, BARDACH, E., 392, 474
473 BARKER, C. H., 73, 516
ANDERSON, B., 54, 473 BARKER, E., 52, 84, 474
ANDERSON, L., 131,512 BARNARD, C. I., 210, 474
ANDERSON, R., 349, 486 BAUMOL, W. J ., 406,474
ANDREWS, F. M., 17,475 BAUMRIND, D., 158, 382, 474
ANGELL, R. c., 182, 315, 473, 489 BAY,c., 26, 15311, 474
528 Index Index

CHOMSKY, N., 443, 479 CUMMINGS, E., 135, 481 DURKHEIM, E., 16, 172, 173, 174, 176, 180,
CHRISTIE,R., 61, 73, 153n, 154, 155, 335, CUTLER, N., 134, 481 181, 192,444, 453, 484
360, 425, 451, 453, 468, 479 CUTLER, S. J., 190, 481 DUTTON, J. M., 385, 386, 403, 410, 411, 484
CHU,K., 386, 410, 507 CUTRIGHT, P., 367, 481 DYE, T. R., 24, 484
CITRIN, J., 454, 520 CYERT, R. M., 386, 405, 479, 481 DYK,R. B., 79, 524
CLARDY, M., 409, 521 DYSON, J. w., 373, 484 .
CLARK, L. P., 50, 298, 319, 479 DAHL,R. A., 83, 440, 481
CLARK, R., 417, 421, 479, 503 DAHLGREN, H., 126,503
CL4RKE, J., 129, 479 DAMANKOS, F. J., 180,502
CLARKSON, G. P. E., 384, 400, 479 DARLEY, J. M., 63, 500
CLAUSEN, A. R., 370-371,479 DAVIDS, A., 420, 481
CLINARD, M. B., 172, 176, 184, 479 DAVIES, J. c., 1-27, 200, 223, 234-260, 446,
CLOWARD, R. A., i76, 479 447448, 456, 464, 482
CNUDDE, c., 47, 54, 363, 479 DAVOL, S. H., 180, 482
COBB,E., 151, 516 DAWES, R. M., 368, 509
COBB, W. J., 381, 494 DAWSON, P. A., 373, 482
COE,R. M., 400, 479 DAWSON, R. E., 24, 39, 120, 131, 133, 482
COHEN, A. K., 176, 479 DAY, R. a.,410, 482
COHEN, K. J., 386, 479 DE CHARMS, R., 46, 82, 421, 482
COHEN, R., 357, 471 DE FLEUR, M. L., 184, 336, 432,523
COHN,T. s., 154, 155, 158, 479 DE GRAZIA, s., 173, 298, 482
COKER, F. w., 84, 479 DE LAMATER, J., 264, 282, 482
COLBY, K. M., 390, 393, 479480, 522 DEAN, D. G., 180, 190,482
COLEMAN, J. s., 131, 133, 182, 324, 331, DE RIVERA, J., 266, 483
386, 403, 407, 480, 502 DE VINNEY, L. c., 61, 520
COLES, R., 304-305, 480 DE VOTO, B., 298, 31411,483
COLLINS, B. E., 59-60, 63, 72, 357, 365, 367, DELGADO, J. M . R., 257, 482
381, 382, 478, 497 DELPRATO, D., 381,498
COLLINS, N. J., 368, 509 DENNIS, J., 118, 120, 121, 123, 129, 135,
COMSTOCK, c., 157, 161, 16&169, 480, 516 482-483, 484
CONNELL, R., 119, 480 DEUTSCH, M., 156, 210, 264, 285, 483, 524
CONVERSE, P. E., 21, 44, 61, 67, 69, 77, 83, DEUTSCHER, I., 50, 483
84, 91, 92, 98-101, 103n, 105, 114, DEVERUUX, G., 30511, 483
129, 134, 135, 158, 227, 324, 335, DEXTER, L. A., 315, 317, 455, 483
340n, 349, 354, 361, 431, 468, 477, DI PALMA, G., 54, 194, 454, 483
480, 514 DI RENZO, G. J., 48-49, 52, 456, 483
CONWAY, M. M., 468, 480 DICKS,H. v., 153n, 313, 466, 483
COOK, P., 153n, 451, 479 DICKSON, W. J., 367,513
COOLEY, R., 131, 520 nres, R., 427, 483
COOLIDGE, c., 231, 456 DILLEHAY, R. c., 62, 153x1, 154, 161, 451, 463,
COOMBS, s. L., 398, 480 498, 505
COOPERSMITH, s., 39,481 DODGE, R. w., 360, 368, 375, 485
CORNBLETH, c., 120,481
DOLLARD, J., 12, 243, 303, 308, 314, 483
CORNFIELD, J., 370, 481
DOOB, L. w., 58, 243, 268, 289-290, 463
COSER, L., 86, 481 DOOLEY, L., 298,483 GABENNESCH, 3., 451, 487
COTTRELL, L. s., JR.,61, 520 GAGE,N., 154, 487
DOUVAN, E., 45, 82, 483
COUCH, A., 154, 156, 481, 491-492 GAGNON, J. H., 357, 4-75
cox, D.R., 359, 375, 481 DOWNS, A., 397-398, 460, 483
GAHAGAN, J. P., 284, 286, 287, 288, 511
CRANE, w., 331,494 DOZIER, G., 191, 489
GALLUP, G., 335, 522
CRAWFORD, T. J., 198, 481
DRIVER, M. J., 408409, 483484, 517, 521 GALTUNG, J., 264, 277, 280, 487-488
DRUCKMAN, D., 264, 265, 408, 484
CRECINE, J. P., 405, 481 GAMSON, W. A., 227, 268, 488
CRITTENDEN, J., 134, 481 DUBERMAN, M., 313, 484
GARCIA,c., 120, 488
CRONBACH, L., 31, 37, 481 DUCKLES, R., 168-169, 480
GARDNER, R., 417, 488
GROTTY, w. J., 367, 368, 481 DULLES, A. w., 300, 484 GARRATY, ~.~.,297,299,301,314,315,316,
CROW, w. J., 268, 285, 287, 481, 512 DUPEUX, G., 129, 480 318, 488
CRUTCHFIELD, R., 156, 481 DURBIN, E. F. M., 262,484 GAUDET, H., 17, 48, 61, 324, 327, 344, 500
 Index I Index 531

GEBHARD, P. H., 357, 498 HERZON, F. D., 451, 492 JANOWITZ, M., 45, 46, 155, 156, 495
GEIS,F. L., 73, 360, 468, 479 HESS, R. D., 23, 120, 124, 126, 130-131, 324, JAROS, D., 120, 123, 328, 495
GEORGE, A. L., 25, 46-47, 48, 50, 204, 229, 328, 333, 336, 484, 492 JENNINGS, H. H., 210, 495
298, 299, 307, 391, 410, 458, 462, 488 HILL, R., 344, 475 JENNINGS, M., 120, 122, 123, 126, 127, 128,
GEORGE, J. L., 25, 48, 50, 204, 229, 298, 299, HIMELHOCH, J., 155, 492 130, 135, 136, 328, 332, 450, 495, 500
307, 391, 410, 488 HAAN, N., 39, 54, 435, 475. HIRSCH, H., 120, 123, 128, 133, 328, 492 JESSOR,R., 156, 174, 495,512
GERTZOG, I. N., 360, 368, 373, 488 HAAS,E. B., 293, 491 HITSCHMANN, E., 303, 315, 492 JESSOR,S. L., 174, 495
GERWIN, D.,405, 488 HAEFNER, D., 156, 491492 HOFFMAN, M., 419, 518 JOHNSON, A., 297, 495
GETZELS, J., 422, 430, 434, 491 HAIRE,M., 336, 491 HOFSTADTER, R., 84, 142, 200, 316, 492 JOHNSON, c., 424, 495
GIERKE, o., 84, 488 HALL, C. s., 29, 33, 448, 449, 491 HOLLANDER, E. P., 156, 230, 311, 492-493 JOHNSON, G. w., 298, 495
GILBERT, D., 155, 157, 488 HALL, E., 10, 491 HOLLINGSHEAD, A., 119, 133, 493 JONES,E., 298, 303, 495
GILBERT, G. M., 302, 312, 313, 318, 488 HALPERIN, M. H., 285,516 HOLST,E. VON, 256, 493 JONES, H. M., 297, 298, 318, 495
GILSON, c., 395, 488 HAMBLIN, R. L., 182, 491 HOLSTI,O. R., 267, 286, 312, 314, 493, 508 JONES, M., 39, 495
GLAD,B., 50, 267, 296-321, 443, 445, 459, HAMILTON, R., 350, 491 HOLT,R. R., 167, 168, 301, 302, 308, 314, JOSEPHSON, M., 298, 496
466, 488 HANFMANN, E., 422, 430, 434, 491 319, 493 JULIAN, J. P., 230, 493
GLAZER, N., 326, 512 HANNA, J. F., 404, 491 HOLTZMAN, W., 35, 493 JUNG,c., 6, 296, 414
GLENN, N., 134, 135, 488 HANSFORD, E. A., 286,518 HOLZBERG, J. D., 304, 493
GLOCK, c., 352,488 HANSON, R. c., 174, 495 HOLZMAN, P., 417, 488 KAGAN, J., 7, 39, 43, 418, 419, 496
GOLDBERG, s., 156, 425, 489 HARDING, W. G., 231, 455 HOMANS, G., 210, 493 KAHN,R. L., 17, 214, 475, 496
GOLDHAMER, H., 455, 489 HARGROVE, E. c., 228,460,491 HONZIK, M., 39, 495 KAMIN, L. J., 361, 371,496
GOLDSTEIN, M., 45, 472 HARLOW, H. F., 7, 12, 13, 15, 24, 25, 255 HOOK,s., 209, 226, 493 KANOUSE, D. E., 395, 496
GOODENOUGH, D. R., 79, 524 HARLOW, J. K., 12,491 HORKHEIMER, M., 139-140, 141n, 154, 164 KAPLAN, A., 2, 18-19, 82, 248, 251n, 500
GOODNOW, J. J., 106, 477 HARNED, L., 155, 491 HORNEY, K., 46, 296, 307, 318-319, 493 KAPLAN, M. A., 300, 301, 496
GORDEN, R. L., 315, 318,489 HARRINGTON, c., 120, 472 HOROWITZ, E., 415, 493 KARDINER, A., 84, 93, 496
GORDON, L., 199,489 HARRIS, D., 159, 491 HOROWITZ, I. L., 194, 493 KARIEL, H. s., 300, 496
GOSNELL, H. F., 368,489 HARRIS, S. E., 85, 521 HOROWITZ, R., 415, 493 KARLINS, M., 64, 409, 449, 496, 521
GOTTERSFELD, H., 191, 489 HARRISON, R., 414, 419, 423, 434-435, 491 HORST,P., 308, 493 KARP,S. A., 79, 524
GOTTFRIED, A., 50, 298, 299, 301, 305, 307, HART,C. w., 381, 494 HORTON, J. E., 47, 180, 191, 493, 522 KATZ, D., 33, 34, 59, 74, 156, 169, 203-233,
314, 489 HARTMAN, D. A., 297, 298, 491 HOVLAND, C. I., 11, 12, 62, 64, 66, 68, 71, 264, 282, 323-324, 329, 344, 350, 351,
GOTTSCHALK, L., 315,489 HARTMANN, G. w., 360, 368, 370, 375, 491 72, 360, 449, 466, 493, 518 448449, 454-455, 456n, 457, 458,
GOUGH, H. G., 155, 158, 409, 489, 491 HARVEY, O. J., 84, 92, 491 HOWARD, J. w., JR., 157, 29711, 307, 494 459, 461, 482, 496, 516, 523
GRANBERG, D., 264, 489 HATCH, w., 162, 474 HOWREY, P., 410, 494 KATZ,J., 162, 474, 496
GRAVES, T., 174, 495 HATHAWAY, s., 164, 491 HUFFMAN, J., 54, 155, 507 KAUFMAN, w., 172, 177
GREEN, R., 428-429, 436, 489 HAVEL, J., 154, 155, 479 HUGHES,H. s., 316n, 320, 494 KAYSEN, c., 85, 521
GREENACRE, P., 303, 316, 489 HAWKINS, B. w., 361, 371, 502 HUNT,D. E., 84, 92, 491 KELEJIAN, H. H., 410,494
GREENBERG, E., 118, 120, 123, 489 HAWLEY, W. D., 174, 379, 487, 491 HUNT,J., 38, 494 KELLEY, H. H., 11, 62, 64, 68, 360, 365
GREENBLATT, M., 157, 489 HAY,J., 297, 316, 491 HUNT,w., 331, 494 KELLY, E., 31, 79, 496
GREENSTEIN, F. I., 23-24, 39, 44, 45, 49, 75, HAYS,W. L., 364,491 HUNTINGTON, S. P., 101, 494 KELMAN, H. c., 78-79, 155, 261-295, 382,
88, 118, 120, 124, 153n, 204, 228, 300, HAYTHORN, w., 156, 491-492 HYMAN, H. H., 23, 24, 60, 119-120, 126, 130, 4604164, 482, 497
305, 312, 324, 328, 336, 342, 413n, HEAD, K., 335, 341n, 513 155, 158, 159, 305-306, 322-355, 361, KENISTON, K., 154, 171, 172, 178, 179, 180,
422-423, 431, 434, 435, 436, 438469 HEAD, K. B., 17, 475 381, 443, 450, 453, 466, 467, 494 363, 481, 497
GREGG, L. w., 403, 412, 490 HEATON, E., 357, 471 KENNAN, G. F., 316n, 497
GRIMES, M., 135, 488 HEFNER, R., 264, 335, 512-513 KENYON, s., 123, 497
GRONSETH, E., 129, 490 HEIDER, F., 58, 69, 71, 492 KERR, H. H., 264, 265, 497
GRUBER, J . E., 379, 491 HEIST, P., 162, 164, 492, 523 KEY,V. o., JR., 80-81, 84, 232, 324, 343, 455,
GUEDALLA, P., 297, 31411, 317, 320, 490 HELMREICH, R. L., 365, 478 463, 497
GUETZKOW, H., 264, 267, 268, 271, 294, 383, HEMPEL, C. G., 309, 492 KIELL,N., 298, 313, 497
385, 386, 403, 408, 490, 516 HENNESSY, B., 456,492 KIESLER, C. A., 59-60, 63, 72, 357, 365, 367,
GUGGENHEIM, F., 423, 424, 490 HENRY, w., 135, 413n, 481, 492 JACOB, H., 41, 46, 84, 229, 428, 457, 477 381,382,497
GUILFORD, J., 31, 490 HERMAN, S. N., 264, 492 JACOBS, R., 52, 156, 501 KILLIAN, L. M., 223, 498,.522
GULLAHORN, J., 400,490 H E R M A N NF., , ' ~ 267,
. 271, 408410, 463 JAHODA, G., 125, 495 KINSEY, A. c., 357, 358, 498
GULLAHORN, J. E., 400, 490 HERMANN, M. G., 267, 271, 408410, 463 JAHODA, M., 15311, 425, 479 KINTZ,B., 381, 498
GURIN, G., 45, 182, 336, 428, 477, 490, 523 HERO,A., JR., 349, 492 JAKOBOVITS, L. A., 357, 4-95 KIRBY, D., 184, 196, 498
GURIN, P., 182, 490 HERRING, E. P., 105, 492 JAMES, w., 5-6, 461, 495 KIRSCHT, J. P., 62, 153n, 154, 161, 451
GURR, T. R., 224, 246, 248, 249, 486, 490 IIERSEY,J., 157, 492 JANIS,I. L., 11, 48, 61, 62, 64, 66, 68, 72, KISH,L., 363, 369, 372, 498
GUSFIELD, J. R., 193, 490 HERTZMAN, M., 427,524 266, 271, 360, 450, 458, 493, 495 KLECKA, w., 134, 498
532 Index index 533

KLEIN,G., 417, 488 LAWRENCE, L. H., 281,497 LUPFER, M., 367, 51 1 154, 168, 232, 247, 248, 251, 302, 319,
KLEIN,W. w., 44, 155, 156, 168, 506 LAZARSFELD, P. F., 17, 21, 45, 48, 61, 324, LUTTERMAN, K. G., 182,502 432, 456, 505
ICLINEBERG, o., 125, 262, 264, 300, 498, 499 327, 328, 329, 342, 344, 475, 500 LYLE, W. H., 155, 158, 502 MATTHEWS, D., 324, 335, 505
KLOPFER, w., 434, 435,498 LAZARUS, R., 30, 417-418, 419, 500 LYND, H. M., 17, 21, 84, 503 MATTHEWS, R., 126, 503
KLUCKHOHN, c., 94, 315, 489, 498, 507 LE VINE, R., 118, 131, 501 LYND,R. s., 17, 21, 84, 304, 503 MATZA, D., 199, 505
KLUCKHORN, F. R., 84, 115, 498 LEGGETT, J. c., 200, 501 LYNN, D., 129, 503 MAY, H. F., 188, 505
KNUTSON, J., 7, 27, 28-56, 68, 94, 111, 112,LEITES, N., 84, 462, 501 LYONS,s., 120, 123, 503 MAY,M. A., 262, 505
120, 121, 15311, 166, 168, 229, 232, LENSKI, C. E., 174, 487 MAY, R., 239, 505
247, 320, 413-437, 446, 447, 448, 449, LENTZ, T. F., 264, 484 MC CALL, G. J., 469, 503 MAZLISH, B., 299, 310, 505
450, 454: 456, 467, 468, 498 LERNER, D., 18, 19, 39, 44, 54, 500, 501 MC CANDLESS, B., 155, 478 MEAD, M., 94, 442n, 505
KOCH, s., 68,498 LERNER, M., 204, 249, 439, 445n, 490 MC CLELLAND, D., 6 7 , 417, 421, 428, 482 MEEHL,P., 37, 481
KOENIG, K., 136, 498, 508 LEVENS, H., 191, 501 MC CLINTOCK, c., 156, 169, 496 MEEKER, R. J., 286, 518
KOESTLER, A., 239, 241-242, 498 LEVENTHAL, H., 52, 156, 501 MC CLOSKY, H., 46, 53, 54, 78, 126, 159, 180, MEIER, D. L., 182, 505
KOGAN, N., 35, 453, 498 LEVIN, M., 126, 130, 133, 501, 508 181, 194, 201, 264, 277, 322, 335, MEISSNER, P., 427, 524
ICOHLBERG, L., 13, 14, 100, 113, 168, 4 2 6 LEVINE, C., 333, 494 454,483,503 MELIKIAN, L., 155, 157, 158, 511
427, 499, 521 LEVINSON, D. J., 17, 44, 52, 61-62, 73, 78, MACCOBY, E., 126, 503 MERELMAN, R., 120, 131, 450, 505
KOHN,P. M., 451,499 84, 92-93, 119, 141, 143, 150, 152, MACCOBY, M., 168-169, 430-431, 487, 503 MERRIAM, C. E., 18, 84, 119, 300, 440, 505
KOMORITA, S. s., 287, 499 155, 157, 167, 378, 417, 425, 427, MC CONAHAY, J. B., 356382, 466-477, 491 MERTON, R. K., 79, 84, 172, 173, 175, 176,
KOPIN,I. J., 255, 499 446447, 451, 468, 472, 488, 489, 494, MC CONAHAY, S. F., 379, 491 184, 505
KORNBERG, A., 399,516 501, 509 MC CONAUGHY, J. B., 298, 503 METTE,D., 381, 498
KORNER, A., 419-420, 435, 499 LEVI'TT, E. E., 155, 158, 502 MC CORMACK, T., 344, 502 MICHEL, J. B., 451, 463, 505, 522
KORNHAUSER, w., 192-193, 226, 499 LEVY,D. M., 312, 501 MC CRONE, D., 129, 483 MICHELS, R., 220, 505
KOTLER, P., 386,490 LEVY, L. H., 296, 501 MC DILL, E. L., 191, 503 MIDDLETON, x., 126, 179, 180, 182, 502, 505
KRAMER, G. H., 360, 367, 368, 499 LEWIN,K., 30, 58, 61, 75, 355, 501 MC DOUGALL, w., 5-6, 19, 59, 60, 503 MIDLARSKY, M., 249, 505, 521
KRAUSS, R. M., 285, 483 LEWIS, H., 427, 524 MC FARLAND, A. s., 204, 217, 503 MILBRATH, L. W., 21-22, 44, 45, 155, 156,
KRAUT, R. E., 363, 370-372, 499 LEWIS,o., 93, 501 MAC FARLANE, J., 39, 495 168, 207, 230, 506
KRECH, D., 13, 24, 25, 499, 511 LIEBOWITZ, M., 194, 493 MC GEE,H., 156,503 MILGRAM, s., 156, 256, 360, 485, 506
KROEBER, A. L., 98, 160, 499 LIEBSCHUTZ, s., 42, 120, 122, 123, 501 MC GUIGAN, F. J., 381, 503 MILLER, C. R., 182, 506
KROEBER, T., 44, 51, 499 LIFTON, R. J., 73, 84, 306, 313, 501 MC GUIRE, W. J., 59, 60, 63, 64, 66, 69, 10311, MILLER, D. R., 79, 506
KTSANES, T., 350, 476 LIKERT, R., 60, 61, 389, 501, 507 471, 503 MILLER, H. M., 182, 518
KUBIE,L. s., 302, 319, 499 LINDLEY, D..v., 372, 501 MAC HOVER, K., 427, 524 MILLER, J. C. P., 372, 501
KUDIRKA, N., 52, 156, 501 LINDQUIST, E. F., 375, 501 MC ILLWAIN, C. H., 84, 504 MILLER, J. G., 91, 94, 506
LINDZEY, G., 29, 33, 414, 415, 448, 449, 467, MAC NER,R., 35, 85, 180, 504 MILLER, N. E., 59-69, 63, 72, 243, 357, 365,
LAMARE, J. W., 373, 499 491, 501-502 MC KINLEY, J. c., 164, 491 367,381,382,483,497
LAMBERT, w., 125, 264, 499 LIHMN, H., 417, 488 MAC KINNON, W. J., 155,504 MILLER, R. B., 151, 516
LAMPRECHT, F., 255, 499 LINTON, R., 31, 444, 502 MC PHAIL,c., 197, 504 MILLER, S. M., 158, 159, 418, 436, 506, 512
LANDAU, D.,462, 499 LIPPMANN, w., 84, 450, 502 MCPHEE, W. N., 45, 328-329, 342, 344, 396- MILLER, W. E., 21, 44,45, 61, 77, 126, 134,
LANDECKER, W. s., 174, 487 LIPSET,S. M., 54, 159, 188, 221, 249, 323, 397, 475,504 135, 158, 227, 324, 330, 331-332, 335,
LANE,R. E., 22, 23, 26, 27, 36, 44, 46, 49, 324, 350, 502 MAC RAE,J., 410, 504 340n, 349, 354, 397, 477, 506, 520
50, 52, 83-116, 120, 126, 129, 155, LIPSITZ, L., 159, 178, 502 MC WHINNEY, W. H., 400, 504 MILLETT, K., 357, 506
156, 189, 298, 307, 312, 326, 450, 461, LIPSKY, M., 197, 502 MALONE, D., 315, 316n, 317, 504 MINER, H. M., 174, 487
464, 499, 500 LITT, E., 120, 130, 502 MANN,F. c., 214, 504 MISCHEL, w., 31, 35, 37, 506
LANGER, W. c., 7, 48, 50, 51, 299, 307, 459, LITTMAN, R. A., 441-442, 450, 457, 459, 502 MANN, M., 193, 504 MISIAK,H., 29, 506
500 LOEHLIN, J. c., 400, 502 MANNHEIM, K., 84, 450, 504 MITRANY, D., 293, 506
LANDER, W. L., 297, 313, 316, 500 LOEVINGER, J., 38, 167-168, 502 MARCH, J. G., 91, 386, 405, 481, 504 MIZRUCHI, E. H., 180, 182, 506
LANGHAM, P., 156, 491-492 LORENZ, K., 239-241, 502 MAKCUSE, H., 468, 504 MODIGLIANI, A., 264, 265, 268, 488, 506
LANGTON, K., 24, 129, 130, 133, 500 LORINSKAS, R. A., 361, 371, 502 MARK, V. H., 257, 504 MONYPENNY, W. F., 297,506
LAO,R. c., 182, 490 LOTH, D., 195, 485 MARTIN, C. E., 357, 498 MOORE, C. G., 412, 506
LAPONCE, J. A., 356, 500 LOVEJOY, A., 309,502 MARTIN, J. G., 155, 168, 504 MORGAN, J., 335, 424-425, 435-436, 506
LASKI,H., 84, 500 LOWE, C. M., 180,502 MARTIN, W. c., 183, 504 MORLEY, J., 297, 506
LASSWELL, H. D., 2, 18-21, 22, 26, 38-39, LOWE, F., 344, 502 MARTIN, W. E., 158, 491 MORRISON, H., 421, 482
46, 49-50, 62, 82, 84, 87, 203, 228, LOWELL, E., 417, 503 MARVICK, D., 45, 46, 155, 156, 330, 495 MORROW, c., 380,506
248, 251n, 298, 300, 301, 309, 311, LOWENSTEIN, K., 85, 502 MARX, , 15, 16, 20, 25,55, 112, i 7 6 MORTON,
~ . , 5 9, A., 126, 503
313, 325, 326, 332, 346, 355, 435, LOWENTHAL, L., 152, 502 177, 223, 245-246, 248, 251, 453 MORTON, I., 424, 506
440, 457, 460, 500, 513 LUMSDAINE, A. A., 62, 360, 493 MASLOW, A. H., 7-8, 14, 17, i9, 32, 34, 37, MOSS,H., 33, 496
LATANE, B., 63, 500 LUMSDAINE, M. H., 61,520 39, 42, 43, 47, 50, 54, 111, 140, 142, ILIOWRER, O. H., 243, 314, 483
 Index 535
Index
ROPER, E., 190, 339-340, 342, 343, 513, 524
RORER, L. c., 154, 513
RORSCHACH, H., 414, 416, 430
ROSENAU, J. N., 266, 268, 292, 513
ROSENBERG, M. J., 45, 60, 62, 63, 68, 69, 71,
168, 270, 274, 346, 352, 382, 471
ROSENTHAL, H., 396397, 410, 514
ROSENTHAL, R., 63, 381, 514
ROSENZWEIG, M. R., 13, 51 1
ROSNOW, R. L., 63, 514
ROSS, H. L., 364, 366, 380, 478
ROSSI, P. H., 367, 481
ROSSITER, c., 84, 514
PADIOLEAU, J. G., 410, 509 ROSTOW, w. w., 18, 514
PAIGE, G. D., 266, 31611, 509 ROSZAK, T., 196, 514
PAMMETT, J., 123,509
PARKER, E. B., 264, 266, 274, 523
PARROTT, G., 52, 509
PARSONS, T., 85, 91, 94, 160, 172, 173, 178,
206,499,509
PASSINI, F., 31, 510
PATCHEN, M., 275, 510
PATRICK, J., 130, 510
PAVLOV, I. P., 8-9, 12, 25 1, 440, 5 10
PAYNE, J., 450, 510
PAYNE, R., 255, 510
PEAK, H., 36, 335,510
PEAR, T. H., 262, 510
PEARL, s., 45, 472
PEARLIN, L. I., 179, 510
PELZ, D. c., 213, 510
PERRY, S. E., 313, 510, 520
SANFORD, R. N., 434,487
PERSONS, c., 381, 498
SAPIN, B., 266, 520
PETTEE, G. s., 246, 510
SARBIN, T. R., 74, 516
PETTIGREW, T. F., 42, 46, 159, 169, 510
SARNOFF, I., 33, 73, 156, 169, 217, 421, 425,
PIAGET, J., 13, 100, 119, 125, 510
496, 516
PIERCE, E. L., 297, 510
SAWREY, w., 129, 503
PILDNER, H., 420, 481
SAWYER, J., 271, 516
PILIAVIN, I., 378, 510
SCHAAR, J. H., 46, 159, 180, 181, 454, 503
PILIAVIN, J. A., 378, 510
SCHACHT, R., 172, 177, 516
PILISUK, ~ . , ' 2 6 7 ,287, 407, 510
SCHANCK, R., 350, 496
PIN, E., 188, 510
SCHANK, R., 390, 480
PINE, F., 157, 51D-511
SCHAPPE, R., 381, 498
PINNER, F., 129, 51 1
SCHAULAND, H., 399, 516
POLSBY, N. w., 264, 266, 274, 523
SCHEIN, E. H., 73, 516
POMEROY, W. B., 357, 498
SCHELLING, T. c., 285, 516
POOL, I. DE s., 396397,511
SCHILD, E. o., 385, 476
POPKIN, s., 396397,511
SCHILDER, P., 319, 516
POPPER, K. R., 412, 51 1
SCHLESINGER, A. M., JR., 215, 516
POWELL, C. A., 408, 410, 51 1
SCHLESINGER, J. A., 390, 517
POWELL, G., JR., 47, 508
SCHLIEWEN, 'R., 410, 525
PREWITT, K., 24, 39,47, 120, 131, 133, 135,
SCHMIDL, F., 298, 320, 517
482, 508, 511 SCHMIDT, G., 357, 518
PRICE, D. E., 367, 51 1
SCHNEIDER, J., 297, 517
PROCTOR, C. H., 182, 51 1
SCHNEIDER, P., 45, 350, 485
PROPPER, M., 420, 511
SCHNEIER, I., 73, 516
PROSHANSKY, H., 420, 5 11
 Index Index

SCHOENBERGER, R. A., 53-54, 201, 517 SKINNER, B. F., 9-10, 20, 58, 67, 81, 251,
STRAUSS, H. J., 25, 521
SCHRODER, H. M., 84, 92, 408-409, 491 443, 519 STRAUSS, L., 84, 521
SCHULTZ, R. I., 386, 490 SKOLNICK, P., 267, 287, 510
STREUFERT, s., 408-409, 517, 521
SCHUMAN, H., 331, 345, 477, 517 SLIDER, c., 357, 471 STREUNING, E. L., 179, 521
SCHUMANN, H., 69, 480 SMELSER, N. J., 306, 312, 519
STRODTBECK, F. L., 84, 115, 498
SCHWARTZ, D. c., 247-248, 456, 517 SMELSER, W. T., 306, 312, 519
STUART, E., 119,520
SCHWARTZ, M., 264, 517 SMITH,C. u., 223, 498 SUCHMAN, E. A., 61, 520
SCHWARTZ, R. D., 366, 467, 523 SMITH,G . H., 348,5 19 sucr, G. J., 59, 336, 409, 423, 509
SCIOLI,F. P., 373, 484 SMITH, L. B., 459, 519
SUEDFELD, P., 409,521
SCOTT, W. A., 59, 60, 5 17 SMITH, M. B., 11, 33, 34, 39, 4243, 45, 54,
SULLIVAN, H. s., 8, 155, 296, 521
SCOTT, W. R., 91, 403, 475, 5i7 57-82, 95, 96, 119, 168, 169, 298, 307, SUOMI, S. J., 12, 255, 491
SEARS, D. o., 43, 61, 64, 67, 69, 81,97, 120, 312, 313, 417, 443, 446, 448449, 450, SUTTON, F. x., 84, 85,521
130, 323, 324, 347, 358, 360, 361, 453-456, 461, 462, 464, 468 SWANSON, G. E., 79, 506
362, 468, 487, 500, 517, 525 SMITH,R. B., 396, 504 SWINGLE, P., 267, 285, 521
SEARS, R. R., 243, 483 SMITHBURG, D. w., 386,518
SYKES, G., 171, 521
SECHREST, L. B., 366, 467, 523 SMOKER, P., 356, 385, 408, 410, 500, 504
SEELEMAN, v., 415, 424, 517 SNIDERMAN, P. M., 454, 520
TAJFEL, H., 264, 521
SEEMAN, M., 46, 168, 172, 178, 179, 180, SNOW,E., 24,520 TANENHAUS, J., 131, 520'
182, 183, 190, 191, 363, 507, 517, 518 SNYDER, R. c., 131, 264, 266, 267, 285, 383,
TANNENBAUM, P. H., 59, 60, 63, 69, 336,
SEIDENBERG, B., 154, 155, 479 385, 490, 51 1, 512, 520 409, 423, 471, 509
SODEUR, w., 451,513
SELZNICK, G. J., 155, 159, 160, 162, 453 TANSEY, P. J., 385, 521
SOKOL, R., 200, 520 WAHLKE, J. c., 229, 331, 485, 494
SERENO, R., 312, 517 TANTER, R., 249, 505, 521 WALKER, A. M., 45, 82, 483
SEXTON, v., 29, 506 SOMERS, R., 157, 494 TAPP,J., 426427,521 WALKER, J. L., 199, 472
SOMIT, A., 131, 520
SHAFFER, W. R., 397-398, 517 TAPPER, E., 129, 521 WALLACH, M., 35, 498, 523
SHAND, A. F., 59, 517
SOREL, G., 84, 245n, 520 TARROW, s., 120, 124, 336,422-423, 431, 434,
SOROKIN, P., 112, 243, 520 WALTER, J., 129, 523
SHAPIRO, M. J., 383, 398, 404, 478, 517 435, 436, 490 WALTZ, K. N., 461, 523
SPENCE, D., 417, 488
SHAVER, P., 264, 347. 476 TAVISS, I., 180, 521 WAPNER, s., 427, 524
SPENCE, K., 62
SHAW, J. c., 386, 392, 395, 399, 402, 411, TAWNEY, R. H., 6, 84, 521 WARNER, L. G., 184, 523
SPIRO,M., 446, 520 TAYLOR, G., 120, 123, 513
447, 508 WARNER, W. L., 84, 523
SPITZ,R. A., 7, 12-15, 24, 255, 520 TAYLOR, J., 168, 521
SHAW,R. E., 410, 518 SROLE, L., 46, 168, 180-181, 182, 191, 520 WAXWICK, D., 136,508
SHEATSLEY, P. B., 155, 158, 159, 347, 350, TEMPLETON, F., 191,521 WATSON, J. B., 9, 10, 16, 74, 440, 523
STAGEY, B., 428429, 436, 489 TERHUNE, K. w., 270,408,522
361, 453, 494 STAGNER, R., 140, 520 WEBB, E. J., 366, 467, 523
SHEFFIELD, F. D., 62, 360, 493 TESLER, L., 390, 522
STANAGE, s., 200, 471 WEBER, M., 6, 206, 216: 523
SHERIF, C. w., 71, 518 TEST,M., 378, 477
STANLEY, J. c., 308, 357, 359, 364, 366, 367, WEBSTER, H., 162, 164, 167, 476, 516, 523
SHERIF, M., 11, 62, 71, 72, 73, 518 THIBAUT, J. w., 112, 522
370, 375, 380, 478, 520 THOA, N. B., 255, 499
WEDGE, B., 310, 523
SHILS,E. A., 91, 94, 509 STANTON, A. H., 313, 520 WEICK, K. E., 403, 523
SHNEIDMAN, E., 416, 418-419, 518
THOMAS, L., 128, 522
STAPEL, J., 335, 520 THOMAS, W. I., 1617, 19, 60, 246, 522
WEIDERMANN-SUTOR, I., 357, 5 18
SHURE, G . H., 286, 518 STAR,S. A., 61, 520 WEIL,A., 125, 510
SIEGEL, A. E., 378, 518
THOMPSON, R. c., 451, 522
STARBUCK, W. H., 385, 386, 403, 410, 411
THOMPSON, V. A., 386, 518
WEINSTEIN, E. A., 313, 524
SIEGEL,s., 378, 518 STARK, s., 312, 520 WELDON, T. D., 89, 524
THOMPSON, W. E., 47, 180, 191, 493, 522
SIGEL,I., 419, 518 STARK, w., 187-188, 520 WESSLER, R., 168, 502
THORNDIKE, R., 335, 522
SIGEL,R., 44, 49, 118, 120, 518 STEINBERG, s., 155, 159, 160, 162, 453 WESTBY, D. L., 197, 524
THORSON, s., 120, 482
SIGUSCH, v., 357, 518 STEMBER, C. H., 381, 494 WESTIE,F. R., 155, 168, 504
THURSTONE, L. L., 60, 522
SILBERSTEIN, F. B., 168, 518 STERN, G., 156,425,489
TILLER, P., 129, 522 WHEELER, s., 134n, 476
SIMMONS, J. L., 54, 179, 469, 503, 518 STEVENSON, H., 119, 520
TINNEY, E. H., 410, 482 WHITE,K., 156, 523
SIMON,H. A., 384-390, 392-393, 395, 399, STINNETT, N., 129,523 TOBIN, J., 85, 521 WHITE, R. K., 75, 78, 95, 96, 266, 284, 298,
402, 403, 411412, 440, 447, 479, 490 STOKES, D. E., 21, 44, 61, 77, 126, 129, 134,
TOC~JEVILLE, A. DE, 88, 223, 246, 522 307, 312, 313, 417, 519, 524
SIMON, J., 368, 518 135, 158, 227, 324, 531-332, 335, MLLEY,H., JR., 122, 522 WHITING, J., 129, 477
SIMON, W., 357, 475 340n, 349, 354, 397-98, 477, 520 TOLMAN, E. c., 9, 262, 522 WICKER, T., 460, 524
SIMPSON, E. L., 32, 112, 518 STOLL, c., 385, 494 TORNEY,J., 120, 126, 130-131, 324, 333, WILKE, w., 131,520
SIMPSON, G. E., 200, 202, 518 STORM, T., 232, 478 336, 492 WILKINSON, R., 452, 524
SIMPSON, M., 451, 518 STOTLAND, E., 59, 74, 496 TOZZER, A. M., 297, 522 WILLIAMS, R., 348, 524
SIMPSON, R. L., 182, 518 TROTSKY, L., 222, 229, 299, 301, 307, 522 WILLIAMS, R. B., 255, 499
SINGER, E., 353, 494 TROW,M., 324,502 WILLIAMS, R. H., 157, 489
SINGER, J. D., 261, 410, 51&519 TRUEBLOOD, C. K., 297, 303, 309, 318, 320 WILLIAMS, R. M., JR., 61, 520
 Index

WYATT, F., 320, 525

vey methodology, 61; and theoretical Cognitive dissonance, and political attitudes,
WILLNER, A. R., 312, 524
relevance, 67-75
WYLIE, L., 124, 525 69-7 1
WILSON, J. g., 197, 301, 357, 524
WYSZYNSKI, A., 427428, 507
Authoritarianism : in contemporary perspec- Cognitive phenomena, in man-machine sim-
WILSON, R. A., 180, 199, 524 tive, 139-1 70; in environmental de- ulations, 408-409
WILSON, w., 1, 25, 41, 48, 50, 204, 218, 228,
terminism, 17; and ideology, 52-53; Cognitive process, simulation of, 389-392
229, 231, 298, 299-300, 307, 310, 311, major personality dispositions of, 143- Computer, in simulation, 384-385, 392-394
316, 456,458,524 149; as personality trait, 4546; and Conflict, theories of, 4-5. See also Interna-
WINER,B. J., 357, 359, 364, 370, 372, 524 political attitudes, 73; problems in tional conflict
WIRTH,L., 192, 524 study of, 451-453, 454; researcl Conflict resolution : experiments in, 289-291 ;
WISH,M., 264, 524 needed on, 163-1 70 problems in study of, 463-464
WITKIN,H. A., 79, 427, 524 Authority, in core belief system, 111 Conservatism: patterns of, 142; personality
WOLFENSTEIN, E. v., 228, 229, 298, 299, 301,
Autokinetic effect, in situationalism, 11 variables in, 53-54; scales of, 201;
307, 311, 524
utility of, 98, 100, 104-105, 112-113,
WOLFERS, A., 261, 271, 524 Behavior: democratic, and personality, 36; 114
WOLFINGER, R. E., 367, 524
environment as direct influence on, Constraint in analysis of political beliefs, 93,
WOODWARD, j., 190, 339-340, 343, 524
8-14; environment as mediated influ- 98-105
WOODWARD, W. E., 298,525 ence on, 14-18; equation for, 2-3, 26, Control of independent variables, 367-369
WOSHINSKY, O. H., 450, 510
446; equation for, related to aggres- Conventionalism as personality disposition,
WRIGHT, c., 333, 344, 348, 475, 494, 524 sion and violence, 251-255; individ- 144, 146
WRIGHT, J . D., 451, 525
ual, simulation of, 386-396; lawful Core belief system in analysis of political be-
WRIGHT, Q., 242-243, 246, 249, 525
and routinized, 97-98; organism as liefs, 109-1 16
determinant of, 3-8; pathological, re- Culture, political, and psychobiography, 312
Subject lated to psychobiography, 299-300,
302-303; personality's influence on, Decision making, simulation of, 386, 405-
Achievement, need for, 7; projective mea- - - political, terms and modes of, 83-
Analysis,
38-44 406
sures of, 421, 428 85 Behavior, electoral : computer simulations of, Deindividuation, study of, 360
Acquiescence, effect of on F scale, 154-155 Anomie: alienation and political behavior 396-399; in environmental determin- Demand characteristics, 381-382
Activity, political: distribution of, 339-343; related to, 171-202; concept of, 172- ism, 17; experiments in, 360-361; pro- Deprivation: effects of related to aggression,
level of related to personality, 45, 52; 176; defined, 173-174; field-theoreti- jective measures of, 428429; research 255-256; relative, and violence, 248
measures of, 343-345; projective mea- cal view of, 184-186; as individual in, 21-22; simulation of representa- Depth psychology, related to psychobiogra-
sures of, 427428 quality, 186182; measurement of, tives', 404; studies of, 370-372 phy, 296, 302, 307, 321
Affect, political, and political character, 345- 174-176; as personality trait, 46; po- Behavior, political: anomie and alienation Destructiveness as personality disposition,
347 litical effects of, 186-202; problems in related to, 171-202; cross-national 145, 146
Affiliation, motivation for, projective mea- study of, 453454; responses to, 176 studies of, 22-23; explanations of, Developmentalism in environmental deter-
sure of, 428 Anti-intraception, as personality disposition, 351-355; research trends in, 21-25; in minism, 12-14, 15
Aggression : authoritarian, as personality dis- 144, 146 stable countries, 20-22 Deviation, political: and alienation, 193-
position, 144, 146; defined, 235~236; Anti-Semitism, related to authoritarianism, Belief systems and political character, 348- 195; of blacks, 197-199; to the left,
and frustration related, 243-245; gen- 141-143, 144, 146, 148, 159 349 195-200; to the right, 200-202; of
eral explanations of, 237-241; inte- Associations, secondary, role in mass socie'y, Beliefs, political: functions of in systems youth, 196-197
grating principle of, 251-255; physio- 226 theory, 95-97; misunderstandings pos- Discontent, political, studies of, 249-250
logical research on, 255-256; problems Attitudes: change of, 378-379; model of sible in analysis of; 88-89; patterns Dispositions in political attitudes, 57-59
in study of, 464-466; projective tech- structure of, 394-395; problems in of, 83-116; problems in study of, 450; Dogmatism: and ideology, 52-53; of politi-
niques in study of, 419; theories on, study of, 44-49; projective mea- stability of, 86 cal leaders, 4849; as political per-
234-260 sures of, 429430 Blacks, political deviation of, 197-199 sonality trait, 4 7 4 9
Alienation: anomie and political behavior Attitudes, political: change theories for, 62-
related to, 171-202; concept of, 176- 63; and cognitive theories, 68-72; con- Character, political: and beliefs and atti- Economizing: in analysis of political beliefs,
184; defined, 178; dimensions of, 179- cept of, 57-60; dispositions in, 57-59; tudes, 347-350; and individual differ- 105-109; bookkeeping-consistency
180; field-theoretical view of, 184- empirical generalizations on, 64-67; ences, 338-343; and political action, function of, 106-107; cost and gain
186; and mass society, 191-193; mea- 343-345; and political affect, 345- formula for, 107-108
history of concept of, 60-63; and judg-
surement of, 180-184; as personality 347; and political knowledge, 347 Education: and authoritarianism, 159-160,
ment theories, 71-72; and learning
trait, 46; political, concept of, 189; Charisma : as leadership pattern, 215-2 18; 162-163; as political socialization
theories, 67-68; and personality theo-
and political deviation, 193-202; po- in social movements, 221 agent, 129-132
litical effects of, 186-202; and po- ries, 72-75; and political character,
347-350; and political man, 80-82; Children, development of political authority Efficacy: as personality trait, 45; projective
litical participation, 189-191; prob- views of, 121-125 measures of, 427428
lems in study of, 453454; and reli- and psychobiography, 3 12; research
Cognitive consistency, and political attitudes, Ego development and authoritarianism, 168
gion, 187-189; scales of, 181-183; on, 63-75; and self-reconstitution the-
69, 71 Electoral behavior. See Behavior, electoral
sources of, 178-179 ories, 73-74; studies of, 22; and sur-
540 Index Index

Elite: analysis of, 19-20; political surveys of, Information-processing system, simulaticn of, Location and limits in systems theory, 97 Political beliefs. See Beliefs, political
331-332 387-389 Political character. See Character, political
Environment : as direct behavioral determi- Instincts: to McDougall, 5-6; to Pavlov, 8 Mass society and alienation, 191-193 Political deviation. See Deviation, political
nant, 8-14; influences from, 3; as me- Institutionalization in political socialization, Media, as political socialization agent, 133- Political-economic conservatism related to
diated behavioral determinant, 14-18 122, 123 134 authoritarianism, 142
Instruments, verbal, 335-336
Epistemology in core belief system, 115-1 16
Ethnocentrism: authoritarianism related to, Integration, political, 279-282 i Mothers: and developmental interaction, 12;
influence of in political socialization,
Political integration, 279-282
Political interpretation in core belief system,
142, 146, 148-149; projective mea-
sures of, 423427
Events: explanation of, in core belief system,
Interaction in developn~entalisrn,12-14
Internal-external control related to aliena-
tion, 182, 191, 198
Internal-external scale and political atti-
I
I
128, 129, 133
Motivation: of leaders, 227-232; of political
leaders, 456-457; for power, 41, 46-
115
Political participation. See Participation, po-
litical
114-1 15; as political socialization 47, 428 Political psychology: history of, 1-27; as a
agent, 134 tudes, 82 Motives related to projective techniques, pluralistic universe, 438-469; prob-
Experimentation: advantages and disadvan- Inter-Nation Simulation (INS) as man-ma- 417418 lems in study of, 447-467
tages of, 362-364; benefits of, 358- chine simulation, 407-410 Multiple measures in experimentation, 380- Political science: psychology related to, 439-
359; complicating factors in, 380-382; International conflict: aggressor-defender 381 447; sociology related to, 441443
ethics of, 382; methodology for, 364- model of, 284-286; conflict-spiral Political socialization. See Socialization, po-
380; problems of, 466-467; studies model of, 286-287; nature of, 284- Needs : hierarchy of, 15-16, 27, 32, 111-1 14; litical
using, 359-362 291; structural-change model of, 287- hierarchy of, related to aggression and Politicization in socialization, 121, 123
Experimenter effects, 381 29 1 violence, 251-252; instrumental, 7-8; Politics: personal involvement model, 278-
International politics: assumptive frame- Maslow's list of, 7; Murray's list of, 284, 295; personality in study of, 28-
F scale: correlates of, 152-153; development works in, 261-295; individual contri- 6, 8; satisfaction of related to revolu- 56. See also International politics
of, 149-152; in political attitudes con- butions to, 268-273; problems in study tion, 246247 Power as personality disposition, 145, 146
cept, 61-62; research based on, 153- of, 461464; public opinion and, 273- Power, motivation for: as personality trait,
163 284 Obedience, study of, 360 41, 46-47; projective measures of, 428
Family: influence of on authoritarianism, International relations: attitudes toward, Operant conditioning, 9 Predispositions in behavior determination,
146-147; as political socialization 264-265; social-psycholugical study of, Opinions, change in, 11, 78-79 18-19
agent, 125-1 29 262-268; theory of, and simulations, Oreanism
" as determinant of behavior, 3-8 Presidents: motivations and character of,
Foreign policy, determinants of, 277-278 408-409 Organizations, leadership related to hierarchy 230-232; personality of, 41-42; as
Frustration related to aggression, 243-245 International system: functionalist model, of, 210-218 problem solvers, 391; problems in
Functionalism as political attitude theory, 293-295; national model, 291-293; study of, 455, 456, 457458, 460
74-50 structure of, 291-295 Participation, political: and alienation, 189- Prisoner's Dilemma, 399402, 407
Interpretation, political, in core belief sys- 191; measures of, 190; studies of, 21- Projective techniques: advantages of, 433-
Games as small group simulations, 399-402 tem, 115 23 437; direct methods related to, 420-
Goals in core belief system, 111-1 14 Partisanship, family influence on, 126 423; history of, 414-416; inferences
GRIT strategy, 286-287, 288, 290 Peers as political socialization agent, 133 of, 419-420; politically relevant uses
Group processes, simulation of, 399-406 Perception, group distortion of, 11 of, 423433; problems of, 467; quan-
Leaders: availability of, 49-52; dogmatism Personality: authoritarian, 5, 139-1 70; be- tification in, 418-419; responses in,
Hierarchy: establishment of in political be- of, 48-49; motivation and personality havior influenced by, 38-44; concepts 416-420; validity of, 422
liefs, 93; leadership related to, 210- of, 227-232 of, 29-35; defined, 30; dispositions, Projectivity as personality disposition, 145,
2 18; motivations along, 229 Leadership : and charisma, 215-218; and related to international politics, 270- 146
Human sets in core belief system, 110-1 11 cognitive skills, 214-215; in demo- 271; holistic model of, 32-34, 37-38; Protestant ethic, to McClelland, 6 7
cratic structures, 224-227; great man in man-machine simulations, 409- Psychobiography: assumptions about, 299-
Idealization in political socialization, 121- theory of, 208-209; organizational hi- 410; measurement of, 35-38; political 302; as competitor of political and so-
122, 123-124 erarchy related to, 210-218; patterns relevance of, 44-55; and politics, ciological study, 301, 304-306; con-
Identity in core belief system, 110 of, 203-233; political leadership re- model of, 39-43; problems in study tributions of, 310-313; critique of,
Ideology: machine, as simulation, 394-396; lated to, 204-208; problems in study of, 449-450; in study of politics, 2% 302-307; data collection and analysis
political, 85; political, direction of re- of, 454-460; responsive and self-pro- 56; traits model of, 31-32, 33-34, in, 313-318; defined, 296; early works
lated to personality, 52-55; produc- pelled, 344-345, 346; of revolutions, 45-49 of, 24-25; generalizing from, 309-
tion of, 85-86; self-consciousness of, 221-224; of social movements, 220- Personalization in political socialization, 121, 310; history of, 297-299; interpreta-
104-105 221; social settings related to, 218- 123 tions in, 318-320; and leadership
Idiosyncrasy related to psychobiography, 220; targets of influence of, 205-206; Phenomenology and political beliefs analysis, availability, 50, 53; problems in, 24-
300-301, 303-304 task- or support-oriented, 209-218; in 87-90 25, 459; utility of, 307-310
Immigration, effects of, 1 6 17 totalitarian structures, 224-227. See Political activity. See Activity, political Psychology related to political science, 439-
Independent variables, control of, 367-369 also Psychobiography Political analysis, terms and modes of, 83-85 447
Industrialization, effects of, 16 Liberalism in political belief patterns, 98, Political attitudes. See Attitudes, political Psychopolitical assumptions, in international
Informants in surveys, 332-333 100, 104-105, 112-113, 114 Political behavior. See Behavior, political politics, 272-273
 Index
Public opinion: international politics related problems in study of, 450-451; pro-
to, 273-284; polls of, 274-275; static jective measures of, 431-433; and so-
model of, 274-277 cial change, 136-138; studies of, 23-
24, 117-138; surveys of, 328-329, 332
Quasi-experiments, 380 Socioeconomic class and authoritarianism,
Questionnaires in surveys, 333-334 158-159
Sociology-of-knowledge in political belief
Randomization: errors in, 373-377; in ex- analysis, 86-87
perimentation, 369-377; in field set- Stability: political, measures of, 249; politi-
tings, 377-380 cal socialization related to, 124-125,
Rational voter, theory of, 397-398 136138
Reasoning contrasted with constraint, 98--105 Stereotypy as personality disposition, 145,
Recruitment and psychobiography, 311-312 146
Regnant centers in systems theory, 91, 93-94 Stimulus-response: in environmental deter-
Religion and alienation, 187-189 minism, 8-10; relevance of to politi-
Research: experimental, 356382; program- cal attitudes, 67-68
matic, 377 Submission, authoritarian, as personality dis-
Response, equation for, 18 position, 144, 146
Revolution: defined, 236-237; leadership of, Superstition as personality disposition, 145,
221-224; need satisfaction related to,
146
246-247; theories on, 234-260
Surveys: advantages and disadvantages of,
Role performance and psychobiography,
310-311 357-358; as experiments, 337-338; in-
struments of, 333-337; national, 327;
Selflessness related to aggression, 241-242 nature of, 324-326; in political psy-
Sex, concern with, as personality disposition, chology, 322-355; problems of, 466;
145-146 strategic universes in, 327-333; varie-
Simulation: of cognitive processes, 389-392; ties of, 326-328
by computer, 384-385; by computer, Systems: simulations of, 406410; theory in
of electoral behavior, 396399; in con- political beliefs analysis, 90-98
flict, study of, 267-268, 271, 285, 287;
defined, 385; history of, 385-386; ide- Therapy related to psychobiography, 300,
ology machine as, 394-396; of indi- 303
vidual behavior, 386-396; of informa- Thermidor reaction to revolution, 222-223
tion-processing system, 387-389; of
large group processes, 403-406; prob- Validity: internal and external, 364-367; in
lems in, 410-412, 467; reasons for, simulation, 410-41 1
389, 392-394; of small group pro- Values: in core belief system, 111-114; in-
cesses, 399-403; of systems, 406-410 tergenerational changes of, 54
Situation related to authoritarianism, 156- Violence: defined, 235-236; endocrines re-
158 lated to, 257-259; integrating princi-
Situationalism in environmental determin- ple of, 251-255; nerves related to,
ism, 10-12 256-257; problems in study of, 464-
Situational pressures in international politics, 466; roots of, 259-260; social theory
271-272 of, 245-246; theories on, 234-260
Social movements: leadership of, 220-221; Voting behavior. See Behavior, electoral
motivations in, 227
Social setting, related to leadership, 218-220 War: defined, 237; general explanation of,
Socialization, political: and adult learning, 242-243; problems in study of, 464-
134-136; agents of, 125-134; of black 466; theories on, 234-260
children, 42; history of research on,
119-1 21; manifest and latent, 39-40; Youth, political deviation of, 196197