“This is an outstanding volume with a star-studded lineup of editors

and authors, with strong credibility in behavioral science and its

progression. The editors mount a compelling case for the merits of
behavioral science as an academic and applied field, and one in
which these integrate coherently. For authors or students wishing to
update on what behavior analysis has done since Skinner, this
volume is a must. While it is dense and accurate, it is readable and
digestible. An excellent addition to a complex area of psychology,
but one which the editors and authors convince us is worth grappling
with.”
—Yvonne Barnes-Holmes, PhD, associate professor in
behavior analysis, and senior research fellow at Ghent
University

“A truly remarkably and needed book that provides a comprehensive

analysis of fundamental behavioral processes and contemporary
theory, research, and practice published in the areas of language and
cognition. It synthesizes Skinner’s analysis of verbal behavior with
contemporary relational frame theory (RFT), and progressively
introduces important core concepts such as stimulus equivalence
classes, response generalization, referencing, perspective taking,
problem-solving, and rule-governed behavior. Acceptance and
commitment therapy (ACT) is presented as a contemporary behavior
analytic therapy that addresses the implications of the science of
language and cognition for practice and instruction of applied
behavior analysis. This book should be required reading in every
graduate program in psychology and behavior analysis.”
—Martha Pelaez, PhD, professor in the department of
educational psychology at Florida International University

“I highly recommend this book to academicians, researchers, and

scientific practitioners who are interested in the advancement and
application of behavior science. Central threads in behavioral
accounts of sociocultural phenomena highlight communication, and
ways language plays a fundamental role in human behavior. The
selection of chapters coedited by Fryling, Rehfeldt, Tarbox, and
Hayes illustrate the power of behavior science and collective
abilities of our scientific group to respond to the emerging
opportunities for scientific impact. In that regard, the authors lay
substantial groundwork for the advanced analyses of language and
cognition with workable implications for development and
dissemination of associated technologies.”
—Ramona A. Houmanfar, PhD, professor in the department
of psychology, and director of the behavior analysis program
at the University of Nevada, Reno
 Publisher’s Note
This publication is designed to provide accurate and authoritative information in regard to the subject
matter covered. It is sold with the understanding that the publisher is not engaged in rendering
psychological, financial, legal, or other professional services. If expert assistance or counseling is
needed, the services of a competent professional should be sought.
Distributed in Canada by Raincoast Books
Copyright © 2020 by Mitch Fryling, Ruth Anne Rehfeldt, Jonathan Tarbox, and Linda J. Hayes
Context Press
An imprint of New Harbinger Publications, Inc.
5674 Shattuck Avenue
Oakland, CA 94609
www.newharbinger.com
Cover design by Amy Shoup
Acquired by Ryan Buresh
Edited by Rona Bernstein
Indexed by James Minkin
Figure 16.1. is used by permission from Springer Nature. Adaptation of “Derived Stimulus Relations
and Their Role in a Behavior Analytic Account of Human Language and Cognition”; Barnes-
Holmes, Finn, McEnteggart et al; Perspectives on Behavior Science. Copyright January 1, 2017.
All Rights Reserved

Library of Congress Cataloging-in-Publication Data on file

 Contents

Chapter 1: An Introduction to Applied Behavior Analysis of Language and

Cognition
Chapter 2: Basic Concepts
Chapter 3: Basic Verbal Operants
Chapter 4: Complex Verbal Behavior
Chapter 5: Identification and Establishment of Bidirectional Verbal Operants
Chapter 6: The Scope and Significance of Referential Behavior
Chapter 7: Human Behavior Is Referential Behavior
Chapter 8: Observational Learning
Chapter 9: Generative Responding Through Contingency Adduction
Chapter 10: Equivalence-Based Instruction: Designing Instruction Using
Stimulus Equivalence
Chapter 11: Relational Frame Theory: Basic Relational Operants
Chapter 12: Assessing and Teaching Complex Relational Operants: Analogy
and Hierarchy
Chapter 13: Rule-Governed Behavior and Verbal Regulation
Chapter 14: Problem Solving
Chapter 15: Second- and Third-Wave Behavior Therapy
Chapter 16: Implicit Cognition and Social Behavior
Chapter 17: Perspective Taking, Empathy, and Compassion
Glossary of Acronyms
References
Index
 Chapter 1:

An Introduction to Applied Behavior

Analysis of Language and Cognition

Linda J. Hayes
University of Nevada, Reno

Progress toward a natural science of behavior has been delayed by a

number of difficulties. The first of these—the attribution of causal status to
nonmaterial entities—has been overcome by the exclusion of hypothetical,
nonnatural elements and their powers from the events making up the subject
matter of a science of behavior (Kantor, 1924). While not all behavior
scientists have forfeited the right to postulate unobserved entities for
purposes of explanation, the need for such entities and their forces is
proportional to the absence of information derived from observational
sources. Hence, advances in observational methods and technologies may
be expected to make such distractions increasingly unnecessary.
A second difficulty pertains to the definition of a unit behavioral event.
The units of analysis problem has been resolved, for the most part, by
sacrificing the uniqueness of individual responses for the affordances of
response class constructions. More than one such unit has been proposed,
though Skinner’s (1938) concept of the operant has proven particularly
valuable. The criteria proposed by Skinner for the membership of responses
in operant classes included their conditions of occurrence, a provision
enabling not merely prediction but also control over class member
occurrences. Further, because prediction and control of class member
occurrences are more readily demonstrated when such occurrences are taking
place at high frequencies, Skinner (1953) proposed that preliminary
investigations focus on simple responses of arbitrary form that could occur at
high frequencies without significant fatigue for extended periods of time. The
notion was that this preparation would permit for the development of laws
and principles of behavior that, subsequently, would be applicable to more
complex, nonarbitrary forms of responding. With this aim and upon this
premise, the science of behavior proceeded.
The absence of sophisticated observational technologies at the time of
these developments placed severe limitations on the automated detection of
even simple responses, however, representing a third factor delaying progress
of the natural science of behavior. Automated detection of responses
demanded that they be localized on particular mechanical operanda (e.g.,
levers, keys), the operations of which were substituted for measures of
responding. As a result, the nuances of individual responses fell below the
threshold of instrumental precision. Still, Skinner’s experimental preparations
(e.g., Skinner, 1938; Ferster & Skinner, 1957) resulted in a corpus of
principles (e.g., stimulus control, reinforcement, extinction) that proved
applicable to the behavior of a wide variety of species and led to the
development of a powerful technology of behavior change presently being
employed in many domains of human activity. All the research presented in
this volume, with the exception of that in chapter 7 by Emilio Ribes, rests on
a foundation of the principles of behavior as construed by Skinner.

Challenges Related to the Subject of Behavior

The promise of an experimental analysis of unrestricted, nonarbitrary forms
of responding, and their elaboration into more complex forms, has not been
fully realized. For the most part, this circumstance is owing to the subject
matter of behavior science. Behavior has a number of features that
complicate its scientific investigation. In the first place, behavior is not a
thing localizable in an organism, but is rather a function resulting from
responding and stimulating, taking place in a field of many other factors. As
such it is not possible to identify or classify behavior in the absence of
information about the setting in which it is occurring. Measurement
practices thereby must be capable of capturing functional relations between
behavior and environmental variables, not merely the occurrence of
responses with particular formal properties.
Secondly, behavior does not come in discrete units, but rather occurs as
a continuous stream of activity. Hence, where one event ends and another
begins must be determined on the basis of arbitrary criteria, presenting
significant difficulties for the replication of findings. While substituting
switch closures for responses solves this problem of measurement, it also
distorts the character of the events measured. More worrisome is that the
prediction and control of behavior—operations achieved by way of a
frequency-based interpretation of probability—require that the character of
behavior be distorted in this way. Practices capable of depicting behavior as a
continuous stream of activity are needed to solve this problem.
Added to this, behavior is the action of the whole organism, and in this
regard, it is always a complex phenomenon involving the simultaneous
participation of multiple response systems (e.g., muscular, neurological,
sensory, glandular). Complete descriptions of behavior thereby entail
consideration of its relevant organismic components, implying the need for
measurement in real time.
Finally, behavior is corrigible. This is to say it is a continuously
changing repertoire, becoming more elaborate and more varied with respect
to contextual cnnircumstances over the course of its development. As such,
response events initially selected for measurement are likely to have evolved
into different forms and to have developed different relations with environing
events over the course of their investigation. This circumstance is of
particular concern when the behavior at issue is verbal in kind. Verbal
behavior is especially difficult to investigate due to its substitutive character
(Kantor, 1936, 1977), conventionality—and thereby arbitrariness—of
response forms (Kantor, 1982), and enormous topographical variation.
Perhaps it is for this reason that more than one conceptualization of the
verbal repertoire, as revealed in the chapters of this book, have been
developed in the science of behavior over the last half century.
Most behavior analysts are familiar with at least some aspects of B. F.
Skinner’s (1957) analysis of this repertoire. More specifically, most are
familiar with the elementary verbal operants, particularly those in which the
stimulus products of verbal responding show formal similarity with prior
verbal stimuli (i.e., echoic, copying text), those in which the stimulus
products of responding share point-to-point correspondence with prior
stimuli (i.e., taking dictation, textual) and those occurring under the control
of conditions other than prior verbal stimuli (i.e., tact and mand). Less
familiar, and less often investigated and validated in application, are the
elementary operants characterized by more complex contingencies (i.e.,
intraverbal and autoclitic). The elementary verbal operants were
differentiated by the unique contingencies into which response instances
could be organized as members. These classes have enormous numbers of
members, and their identification was not Skinner’s primary aim in writing
the book Verbal Behavior. Rather, his primary aim was to consider the
contributions of historical conditions and current stimuli such as to foretell
the topographical and dynamic features of a “single forthcoming instance of
verbal behavior” (1957, p. 28). How he proposed to pursue this aim is not
obvious until after the eighth chapter of Verbal Behavior where he discusses
multiple causation and supplementary stimulation. With respect to this aim,
and in the present author’s perspective, Skinner comes closer to its
achievement than any other theorist.

Criticisms of B. F. Skinner’s Position

As is evident in some of the criticisms of Skinner’s position, not everyone
working on this topic has shared Skinner’s aim. Critiques of Skinner’s
position (from within the behavioral community at least) have focused on
two aspects of his work. First, it has been argued (e.g., Hayes, 1996) that
Skinner focused exclusively on the occurrence and characteristics of the
speaker’s behavior, neglecting the same of the listener save the latter’s role
in the mediation of consequences for the speaker’s behavior—much of
which had to be assumed as opposed to directly observed. While Skinner
did not in fact ignore the verbal aspects of the listener’s behavior entirely
(e.g., see Skinner, 1957, pp. 357–367), he did argue that the behavior of the
speaker could be analyzed without considering the behavior of the listener
(1957, p. 2).
Interestingly, it might be argued that relational fame theory (RFT) has
focused almost exclusively on the behavior of the listener, to the neglect of
the speaker. In much of the early RFT research, in which a match-to-sample
procedure was employed, the responses measured were acts of selecting a
particular stimulus from an array of stimuli. Acts of this sort demonstrate a
“knowing” of relations among stimuli—which is an act of listening, perhaps,
or understanding, but not an act of speaking. For example, when a child
points to stimulus X upon being asked to “show me X” in the presence of an
array of stimuli in which X is included, the response of pointing to X is not
an instance of speaking. By contrast, upon being presented with stimulus X
and being asked to say what it is, the response of naming the object is one of
listening and understanding as well as speaking.
The same is true of the early work on stimulus equivalence. In this
case, though, some attempts to evoke topographically distinct response forms
were made or were at least imagined to be ongoing (Horne & Lowe, 1996).
Added to these is Kantor’s (1936, 1977) interpretation of psychological
linguistics, which is not a critique of Skinner’s position per se but rather an
independently developed alternative to it. From Kantor’s perspective, as
exemplified in chapter 6 by Patrick Ghezzi, linguistic acts of the referential
type were held to involve simultaneous stimulation from two sources, one of
which was always the listener. Accordingly, it was not possible to focus on
either the speaker’s or the listener’s behavior in isolation.
The other major criticism of Skinner’s analysis of verbal behavior has
had to do with his conceptualization of stimulus events. Stimulus events have
come to be understood as having both functional and object properties (e.g.,
see Hayes, Barnes-Holmes, & Roche, 2001). While Skinner (1957) did
distinguish the function of a stimulus from its status as a physical object, and
he did make use of this distinction in addressing particularly complex
episodes of verbal behavior involving autoclitics, he did not do so
consistently. For the most part, stimuli were conceptualized as objects, which
were or were not present in a given circumstance, giving rise to such claims
as “Apart from an occasional relevant audience, verbal behavior requires no
environmental support (Skinner, 1974a, p. 100), and such analyses as “seeing
in the absence of the thing seen” (Skinner, 1957, p. 363). It was this issue that
prompted Skinner to propose a category of operant behavior that, unlike all
other categories of operant behavior, was not contingency shaped, namely
rule-governed behavior. Rule governance was a less-than-convincing solution
to the problem, however (see Parrott, 1987). The problem was accounting for
the control exerted by stimuli in their absence, and the solution was to
promote the distinction between the functional and object properties of
stimuli. It was this issue that gave rise to an important body of research,
generally referred to as stimulus equivalence. And it was this distinction and
its implications, in turn, that gave rise to RFT. Interestingly, the distinction
between stimuli as objects and stimuli as functions is one of the hallmarks of
Kantor’s analysis of behavior (1924, 1958), including behavior of the
linguistic type (1936, 1977).
 The implications of making this distinction are profound. The
distinction makes it possible to assume that the functional properties of a
stimulus are possible in the absence of its physical object properties. This is
to say the functional properties of stimuli may operate from nonoriginal
sources which, according to Kantor, is an outcome of historical relations
between original and nonoriginal sources. Kantor (1936, 1977) refers to this
phenomenon as substitute stimulation. Other theorists, as is apparent in
most of the remaining chapters of this volume, account for this phenomenon
in different ways and refer to it with different terms. Nonetheless, all are
pointing to the same phenomenon and all are recognizing its implications for
the understanding of exceedingly complex forms of human behavior.
The development of a natural science of behavior has been impeded by
several factors, including the attribution of causal properties to hypothetical
entities, the definition of the unit of a behavioral event, and the lack of
sophisticated observational technologies. B. F. Skinner’s radical behaviorism,
including its analysis of verbal behavior, resolved many of these challenges.
Conceptual difficulties remained with Skinner’s analysis, however, including
an understanding of the behavior of the listener and referential behavior. The
conceptualization of rule-governed behavior and the study of derived
stimulus relations (and later RFT) permitted for an analysis of the behavior
of both the speaker and listener. J. R. Kantor’s analysis of substitute
stimulation made it possible to assume that the functional properties of a
stimulus are possible in the absence of its physical object properties. These
developments, including the notion of substitute stimulation, permitted for
the understanding of complex human behavior.

About This Book

The aims of the present book flow from this general understanding of the
corrigibility of behavior science. Essential to this evolution is the
incorporation of new materials into our educational curricula. Applied
Behavior Analysis of Language and Cognition provides an analysis of the
most innovative and contemporary research published in the area of
language and cognition to date. Authors share important perspectives that
will guide the reader in their own research or clinical work in the areas
subsumed under each chapter. The book is designed to maximize readers’
grasp of the complex issues involved in this subject. At the beginning of
each chapter is a “Link and Preview” box, in which the editors provide a
brief overview of the topic at hand and its link to previous chapters. Each
chapter—written by experts in their respective areas—contains an in-depth
discussion of the topic, a consideration of implications and future
directions, a summary of the chapter’s key points, and a list of study
questions to facilitate mastery of the concepts presented in the chapter.
Finally, as our field is teeming with terminology and acronyms, a glossary
of acronyms is included at the back of the book for readers’ reference.
It is the editors’ belief that the most important new materials for
practitioners of behavior science pertain to matters of language and
cognition. This book reflects our best efforts to make these materials
available and accessible.
 Chapter 2:

Basic Concepts

Linda J. Hayes
University of Nevada, Reno
Kenneth W. Jacobs
University of Nevada, Reno
Matthew Lewon
University of Nevada, Reno

Link and Preview

While the majority of this text is devoted to specific topics related
to language and cognition in applied behavior analysis (ABA), as
described in the introduction, the present chapter describes the
basic concepts and principles involved in ABA. We believe it is
important to provide an overview of these concepts and principles
at the beginning of the text given their foundational role in
understanding behavior.

Basic concepts seem to be involved in everything in behavior analysis, from

the simple to the complex. We begin our overview with brief comments on
the historical context of behavior analysis, followed by a description of
respondent and operant processes. The chapter also provides an overview of
select advanced topics related to stimulus control and motivation.
Historical Context
In the one hundred plus years since John B. Watson (1913) regarded
psychology as a science of behavior, and the over fifty years since Skinner’s
(1963) restatement in terms of radical behaviorism, there have been two
enduring conditioning paradigms within the science of behavior:
respondent conditioning and operant conditioning. Ivan Pavlov (1927)
established the former, and B. F. Skinner (1938) the latter. The operant
conditioning paradigm directly spawned the development of ABA (Baer,
Wolf, & Risley, 1968), whereas both respondent and operant conditioning
informed behavior therapy (Wolpe, 1976a, 1976b). Barring some overlap,
these learning processes and their associated applications remained
relatively independent of one another (Williams, 1987). The supposed
distinction between respondent and operant conditioning, for instance, was
“shrouded with ambiguity” (Kazdin, 1979, p. 632). More recently, there has
been a convergence of thought, where respondent and operant conditioning
have been conceptualized as complementary rather than distinct (e.g.,
Delgado & Hayes, 2013, 2014; Domjan, 2016; Rehfeldt & Hayes, 1998).
In addition to the classical and operant paradigms, J. R. Kantor (1924)
developed the field-theoretical position known as interbehavioral
psychology. A similarity fundamental to both operant and interbehavioral
psychology is their adherence to behavior as the proper subject matter of a
natural science of psychology (Morris, 1984). Kantor’s lifelong pursuit was
to construct a natural science of psychology impervious to the mentalistic
vernacular of nonscientific institutions (Clayton, Hayes, & Swain, 2005).
Although it had few followers to begin, the explicit and thoroughgoing
explication of interbehavioral psychology’s philosophical assumptions has
guided theoretical developments in behavior science to this day (Fryling &
Hayes, 2015; Hayes & Fryling, 2009, 2014, 2015; Smith, 2016).
An even more recent addition to the ABA, behavior therapy, and
interbehavioral movements is contextual behavioral science. The
philosophy of contextual behavioral science, functional contextualism, is
aligned with and also extends Skinner’s (1945) radical behaviorism (Hayes,
Barnes-Holmes, & Wilson, 2012; Hayes, Hayes, & Reese, 1988). The
contextual behavioral science approach to clinical application (Hayes, Levin,
Plumb-Vilardaga, Villatte, & Pistorello, 2013) and language and cognition
(Hayes, Barnes-Holmes, & Roche, 2001) is unique, but most importantly, is
consistent with and committed to a natural science of behavior. The
assumption that a natural science of behavior is possible is a thread that
unifies each of behavior science’s respective movements. We mention these
theoretical foundations because they are a part of the larger context from
which the basic concepts reviewed in this chapter are derived.

Respondent Conditioning
Respondent behavior describes responses (unconditioned responses;
URs) that occur in the presence of particular environmental events
(unconditioned stimuli; USs) without a history of learning. While the
relationship between the US and UR is relatively constant (i.e., given the
US, the UR will always occur), the UR may in some cases be modified by
consequences (Domjan, 2005). In other cases, the repeated and consistent
presentation of the US may gradually reduce the magnitude of the UR to the
US, a process called habituation. For example, you may notice the loud
sounds of airplanes coming and going every time you visit your friend who
lives near the airport. Your friend, however, may hardly notice those
sounds, if they notice them at all. You might say that habituation has
occurred in this case (your friend no longer responds to the loud sound of
planes, the US, or at least not as strongly as they used to). The inverse,
dishabituation, is when the magnitude of the habituated UR recovers due
to the subsequent presentation of the US following exposure to a different
stimulus (Groves & Thompson, 1970). In other words, if your friend was
not exposed to the sounds of the airplanes for a while, and was then
exposed to them again, your friend might again respond to the sound and
find it disruptive.
Respondent conditioning, also known as Pavlovian or classical
conditioning, is when one stimulus acquires the functional properties of
another. This occurs when two stimuli consistently occur together, either in
space (i.e., near each other) or time (i.e., during or right before or after each
other). The most reliable conditioning is obtained when two stimuli
frequently occur together and infrequently occur independent of one another
(Rescorla, 1967, 1988). A prototypical example is the acquisition of the
eliciting function of a US by a stimulus that previously did not elicit that
response (we often call this stimulus, prior to conditioning, a neutral
stimulus). When the previously neutral stimulus elicits a response similar to
the UR elicited by the US, we call it a conditioned stimulus (CS), which
elicits the response now called a conditioned response (CR). Fears and
phobias are examples of the outcomes of respondent conditioning, where an
initial experience with a US (e.g., a prick from a needle) causes a previously
neutral stimulus (e.g., the doctor’s office or the sight of a needle) to become a
CS. It is important to note that stimuli may also acquire other functions via
respondent conditioning, including reinforcing (conditioned reinforcers;
Williams, 1994), aversive (conditioned aversive stimuli; Azrin & Holz,
1966), discriminative (Hayes, 1992), and motivational functions (conditioned
motivating operations; Michael, 1993).

Operant Conditioning
Antecedent, behavior, and consequence constitute the three-term
contingency. Antecedent refers to an event or condition that precedes a
behavior, and consequence refers to an event or condition that happens
after a behavior. Each component of the contingency is defined in terms of
the other (Timberlake, 2004). This is to say that where there is behavior
there are consequences, and those consequences coordinate behavior with
stimuli (Skinner, 1974a). While consequences increase or decrease the
probability of a behavior occurring, antecedents regulate the probability of
that behavior occurring in a given setting or situation.
Antecedent, behavior, and consequence are open-ended terms in the
sense that no stimulus is guaranteed to function as an antecedent or
consequence and no behavior is guaranteed to function strictly as a
respondent or operant (Domjan, 2016; Timberlake, 1988). Antecedent,
behavior, and consequence represent candidate elements: stimuli and
responses that may or may not function as conditioned stimuli, discriminative
stimuli, reinforcers, or punishers (more on these terms below). Determining
what does or will function as an antecedent or consequence is a matter of
observation and experimentation.
In the sections that follow we describe the processes and outcomes of
antecedents and consequences as they relate to operant behavior. We will
elucidate the conditions under which candidate antecedents and
consequences have discriminative or reinforcing functions. Furthermore, we
will describe the conditions under which motivation and emotion influence
the three-term contingency as a unit of analysis.
 Consequential Control
Consequential control refers to changes in behavior due to
reinforcement or punishment. Below is a thumbnail take on how behavior
analysts speak of reinforcement and punishment, then and now (cf. Killeen &
Jacobs, 2017a, 2017b). Afterward, we describe the particulars of
reinforcement and punishment procedures and the process of extinction.

STRENGTHENING
Early in his career, Skinner (1938) described reinforcement as the
process by which behavior is strengthened. The converse, punishment, was
described as the process by which behavior is weakened (Skinner, 1953).
“Reinforcement builds up these tendencies; punishment is designed to tear
them down” (Skinner, 1953, p. 182). The “tendencies” Skinner referred to are
behavior: the acts of whole organisms in context. Those acts are
“strengthened” or “weakened” in the sense that there is an observed change
in their probability of occurrence. In lay terms, behavior that occurs a lot, or
intensely, might be said to be strong, and that which occurs infrequently, or at
a very low magnitude, might be said to be weak.
The notions of response strengthening and weakening are metaphors.
They are figures of speech that represent an assumption fundamental to any
experimental or applied behavior analysis (i.e., behavior is probabilistic). The
occurrence or nonoccurrence of a response is indicative of its probability, its
likelihood of occurrence on any given occasion. Rate of responding is a
measure of response probability that has been deemed the basic datum of the
science of behavior (Skinner, 1956, 1966b). Other measures such as count,
percent, duration, and latency can be indicative of response probability as
well, so our applied analyses of behavior are not limited to rate of responding
alone. This means behavior change can be measured in a variety of ways.
The probability that a response will or will not occur depends on its
consequences within a given setting. We can predict the likelihood that a
response will occur, and can even control its occurrence, but we do so in
terms of response classes. This is to say that behavior analysts do not predict
the occurrence of particular response forms on any given occasion. Instead,
behavior analysts predict whether or not a particular class of behavior will
occur.
 Within any given class of behavior are many variations of response
form. For instance, one might request water when thirsty in various ways:
“Water, please?” “Can I have water?” “I want water,” and so on. Regardless
of how they request that water, the consequence is the same: access to water.
When the consequence for a variety of response forms is the same, behavior
analysts functionally class those responses; they group those responses into
classes on the basis of their consequences. In addition to the water-related
response class there are food-related response classes, attention-related
response classes, escape-related response classes, and so on.
In the case of a thirsty client, the applied behavior analyst can make a
class-based prediction: When in the presence of listeners who may have
water, a water-deprived client will engage in one or more various forms of
verbal behavior that have produced access to water in the past. We cannot
predict with certainty that the individual will ask, “Water, please?” versus
“Can I have water?” but we can predict that the individual will request water
in some form.

SELECTION
Later in his career Skinner (1981) dropped the notions of strengthening
and weakening in favor of selection. In a masterstroke, Skinner (1981)
described genetic, behavioral, and cultural processes in terms of selection by
consequences. Consequences are responsible for the traits of a species, the
behavioral repertoires of a species’ members, and the cultural practices of a
group. Variation at all three levels is key, as it makes the selection of an
adaptive trait, behavior, or practice possible (Hayes, Sanford, & Chin, 2017).
At the behavioral level, consequences are reinforcers and punishers that
differentially select responses, which constitute the functionally defined
classes of behavior already described. When a response class is stereotypic,
and therefore restricted in terms of variation, applied behavior analysts are
tasked with expanding that response class or shaping and maintaining new
ones (Hayes & Sanford, 2015).
In keeping with Skinner (1981), we will describe the processes and
outcomes of consequential control in terms of selection. Evolutionary science
is a natural ally to the behavioral sciences (Wilson, Hayes, Biglan, & Embry,
2014), as consequential control is a product of our “evolved susceptibility to
reinforcement” (Skinner, 1981, p. 501).
 Reinforcement
The outcome of reinforcement is an increase in the probability of a
response on future occasions. There are two sorts of reinforcement: positive
reinforcement and negative reinforcement. Positive reinforcement entails
the presentation or addition of a stimulus, whereas negative reinforcement
entails the removal or subtraction of a stimulus. Both positive and negative
reinforcement increase behavior based on the contingent presentation or
removal of a stimulus, respectively. Reinforcers, however, are not exclusively
stimuli. Events or activities may also function as reinforcers (Jacobs,
Morford, King, & Hayes, 2017; Premack, 1959; Timberlake & Allison,
1974). For the sake of brevity we will speak in terms of stimuli, but we will
indicate which candidate reinforcers are stimuli, events, or activities when
those designations are applicable.

POSITIVE REINFORCEMENT
If contingent access to a stimulus follows a target response and increases
that target response, then positive reinforcement is observed. An example of
positive reinforcement is the case in which a parent provides a child with
access to a preferred toy (stimulus) as long as the child says “please” (target
response). We say the toy functioned as a reinforcer if and only if its
contingent presentation was observed to increase the rate of the target
response (i.e., saying please).

NEGATIVE REINFORCEMENT
If the contingent removal of a stimulus follows a target response and
increases that target response, then negative reinforcement is observed. An
example of negative reinforcement is if a teacher removes an academic
demand (activity) on the condition that the child appropriately requests a
break (target response). The academic demand is said to have functioned as a
negative reinforcer if and only if its contingent removal was observed to
increase the rate of the target response (i.e., requesting a break). Another
example is the case in which the removal of a headache (event) is contingent
upon your taking medicine (response). If the probability of taking medicine
increases on future occasions, then the removal of that headache is said to
have functioned as a negative reinforcer.
 Punishment
The outcome of punishment is a decrease in the probability of a
response on future occasions. There are two sorts of punishment: positive
punishment and negative punishment. As is the case with reinforcement,
positive and negative refer to the presentation or removal of a stimulus,
event, or activity contingent upon a particular behavior.

POSITIVE PUNISHMENT
If the contingent presentation of a stimulus follows a target response and
decreases that target response, then positive punishment is observed. An
example of positive punishment is if a parent requires a child to excessively
clean up a room (activity) on every occasion the child engages in property
destruction (target response). We say that excessive cleaning functioned as a
positive punisher if and only if its contingent presentation was observed to
decrease the rate of the target response (i.e., property destruction). This
procedure is also known as overcorrection (see Timberlake & Farmer-
Dougan, 1991).

NEGATIVE PUNISHMENT
If the contingent removal of a stimulus follows a target response and
decreases that target response, then negative punishment is observed.
Negative punishment is the case in which a teacher, for example, removes
access to playtime (activity) on every occasion the child hits another
individual (target response). Playtime is said to have functioned as a negative
punisher if and only if its contingent removal was observed to decrease the
rate of the target response (i.e., hitting). This procedure is also known as
time-out. Additionally, and critically, negative punishment requires the
removal of a positive reinforcer for its procedures to be effective in
decreasing behavior.

Extinction
Up until this juncture we have described what happens to behavior when
natural or scheduled events follow behavior on particular occasions. We have
described the processes and outcomes of reinforcement and punishment. The
question before us now is: What happens to behavior when previously
obtained consequences are lacking or withheld?
Extinction is a reduction in responding that no longer produces
reinforcing consequences. In procedural terms, a reduction in responding is
observed when reinforcement for a particular response is lacking or withheld.
For the effects of extinction to be observed, reinforcement must be lacking or
withheld for responses that have produced reinforcing consequences in the
past. Like punishment, a reduction in behavior occurs, but not because of the
contingent removal of a reinforcing stimulus (negative punishment). Whereas
negative punishment is a contingency, extinction is the decoupling of
contingencies. In other words, extinction involves the discontinuation of the
relationship between responses and their reinforcers.
Importantly, we can only say that extinction has occurred when there is
an observed decrease in a previously reinforced behavior. Here the word
“extinction” may be used to refer to a procedure (i.e., no longer providing the
reinforcer) and an outcome (i.e., the reduction in behavior). Besides its
primary effect of reducing behavior, the process of extinction has at least
three additional effects: an initial burst in responding, variability in
responding (operant variability), and variants of response recovery.

EXTINCTION BURST
When reinforcement for a previously reinforced response is withheld, a
rapid increase in responding may be observed initially. An extinction burst
refers to that initial rapid increase in responding. If reinforcement for socially
appropriate requests is withheld, then a burst in socially appropriate requests
is likely to ensue. For example, if a child has learned to say “Mom” to get his
mother’s attention, he may say “Mom” repeatedly, and perhaps more loudly,
when attention is no longer forthcoming. As extinction proceeds, those
requests for attention may also vary in form, which brings us to the next
effect of extinction.

OPERANT VARIABILITY
A client may not emit the same socially appropriate request when
reinforcement for that request is withheld. For example, the child may say,
“Look at me,” “Pay attention,” or other responses after the response “Mom”
is no longer followed by attention. Extinction is said to induce operant
variability in such a way that new response forms, and variations of old
response forms, can be selected by their consequences.

RESPONSE RECOVERY
Subsequent to an extinction burst or operant variability, the response
form for which reinforcement is withheld will eventually subside. A
particular request, after extinction, may now have a zero probability of
occurrence within a particular setting. With the passage of time or change in
context, however, the previously extinguished response may recur at or near
its original baseline level of occurrence. After a period of rest between
extinction sessions, for example, extinguished responding might recover due
to extraneous sources of control within the setting where extinction took
place (Pierce & Cheney, 2013). Continuing with our previous example, it is
possible that the child will resume yelling “Mom” in a few days. This effect,
known as spontaneous recovery, is only one of the many different forms of
response recovery after extinction. We briefly describe those other forms of
recovery, but encourage readers to see Bouton, Winterbauer, and Todd (2012)
for references and review.
Renewal, resurgence, and reinstatement are terms that describe the
recurrence of responses previously extinguished. Bouton et al. (2012)
describe these types of response recovery as relapse effects. In the case of
renewal, extinguished responding recurs due to changes in context. If a
response was selected by its reinforcing consequences in context A, but
extinguished in context B, the extinguished response will renew (i.e., recur) if
the individual returns to context A. This is known as the ABA renewal
effect. Resurgence occurs when an initially reinforced response is put on
extinction while an alternative response is now reinforced. If you then
extinguish the alternative response, the previously extinguished initial
response will resurge (i.e., recur; see Lattal et al., 2017, for a review of the
definition of resurgence). Lastly, reinstatement is the case in which a
response recurs because the previously extinguished response is reinforced
on a later occasion, which reinstates its occurrence.

Antecedent Control
Antecedent control is a general term used to describe a change in the
probability of a particular response when in the presence of certain stimuli.
Whereas consequential control refers to changes in the probability of
behavior due to events that follow it, antecedent control refers to changes in
the probability of behavior due to several classes of events that precede it.
Antecedent control includes both stimulus control and motivating operations,
on which we elaborate below.

STIMULUS CONTROL
Stimulus control is a continuum on which stimulus discrimination and
stimulus generalization are opposite endpoints (Dinsmoor, 1995a; Rilling,
1977). On one end of this continuum, stimulus discrimination occurs when
measures of behavior vary systematically across two or more different
settings. On the other end, stimulus generalization occurs when consequences
for behavior in one setting affect measures of behavior in another setting. In
applied situations, either can be desirable or problematic depending on the
nature of the performance. The processes and outcomes of each, as well as
the factors that affect them, are described below.

Stimulus discrimination. Stimulus discrimination refers to a change in

the probability of an operant response in one setting versus another. That
change in the probability of a response is due to discrimination training,
which entails a history of differential reinforcement, punishment, or
extinction across settings. In procedural terms, discrimination training is
when a given response is consistently followed by one outcome, on a
particular schedule in one setting, while that same response is followed by
either a different outcome or the same outcome, on a different schedule in
another setting (Dinsmoor, 1995a; Terrace, 1966). Such training may be
arranged intentionally or occur adventitiously.

SD & S-delta. The most common discrimination training procedure

involves reinforcing a response in one setting and extinguishing it in
another. Other combinations of reinforcement, punishment, extinction, or
different schedules of reinforcement/punishment for a single response—
across different settings—may also bring about stimulus discrimination. In
the prototypical procedure, the stimulus in the presence of which a response
is reinforced is referred to as the discriminative stimulus (abbreviated SD
or S+). Due to a history of reinforcement in its presence, the presence of the
SD increases the probability of the response. The stimulus in the presence
of which the response is extinguished is typically referred to as S-delta (SΔ
or S-), and its presence serves to decrease the probability of the response.
While the term “discriminative stimulus” implies that only the SD exerts
discriminative control, both SD and SΔ may be considered discriminative in
the sense that both affect the probability of a response: the former increases
it while the latter decreases it. Stated generally, the probability of a given
response in a particular setting is related to that response’s history of
reinforcement/punishment/extinction in that setting.

Stimulus classes. In addition to the stimulus control exerted by a particular

setting, specific properties of stimulus objects or events may come to
acquire discriminative functions under the relevant contingencies. A
stimulus class, or concept, is defined as a group of stimuli that share a
specific property and that are functionally related to responding in the same
way (Chase & Danforth, 1991; Herrnstein, Loveland, & Cable, 1976;
Zentall, Galizio, & Critchfield, 2002). The property shared by all members
of a given stimulus class may be established as an SD via concept training:
a given response is followed by a particular consequence in the presence of
members of one stimulus class, and is followed by a different consequence
in the presence of members of another stimulus class. The outcome is that
the response will become more or less probable in the presence of the
property that is shared among all members of the given stimulus class, even
if various members of the class differ in other regards. An example of this is
reinforcing the verbal response “red” in the presence of many different red
objects, while extinguishing the same response in the presence of objects of
other colors. Subsequent to this training, the response “red” will be more
likely in the presence of all red objects and less likely in the presence of
non-red objects.

Stimulus control factors. Practical applications are often concerned with

establishing specific stimuli or stimulus properties as discriminative (i.e.,
we want behavior to occur in the presence of certain stimuli but not in the
presence of other stimuli). A number of factors affect the extent to which
this is achieved. The first is stimulus salience, which pertains to the
magnitude of the difference between an intended SD and other stimuli
present during training (Dinsmoor, 1995b). Stimuli that are more
distinguishable from background stimulation are more likely to acquire
discriminative functions under the appropriate contingencies. Salience is
related to many factors, including but not limited to the intensity of a
stimulus relative to the setting in which it occurs (Mackintosh, 1976),
sensory capacity (i.e., what an individual is capable of detecting; Heffner &
Heffner, 2007; Horowitz, 1987), and the biological relevance of stimuli
(Domjan, 1983).
Another factor affecting discrimination training is stimulus disparity,
which is the extent to which an intended SD for a given response differs
along some physical dimension from SΔ(s) during training (Dinsmoor,
1995b). The greater the difference between an SD for a given response and
various SΔs present during training (i.e., SDs for other responses, often
called “distractors”), the more readily the former will acquire the appropriate
discriminative function.

Stimulus generalization. Stimuli that are similar in certain ways tend to

evoke similar patterns of responding, or stimulus generalization. This term
refers to circumstances in which the consequences of a given response in
one setting affect its probability in another similar setting. The conditions
under which stimulus generalization occurs have been studied extensively
in the laboratory (Honig & Urcuioli, 1981; Rilling, 1977). In applied
settings it is often important to explicitly program for generalization such
that performances established in treatment occur in the appropriate
circumstances outside of treatment (Stokes & Baer, 1977). In some
situations, though, generalization may be undesirable (e.g., it may be
important that a given response occur in only one of two similar settings),
and appropriate discrimination may require further training. Terrace (1966)
described how this may be expedited by beginning training with SDs and
SΔs that are dissimilar and gradually making them more similar across
training, that is, until the response is only occurring in the appropriate
setting. This process is known as fading or errorless learning and is the
basis for various prompting and transfer of stimulus control technologies
(Green, 2001).

Overshadowing. When a response is reinforced, punished, and/or

extinguished, a plethora of environmental stimuli are present, but not all
stimuli will necessarily acquire discriminative functions. Only those stimuli
that are consistently correlated with a particular consequence, for a
particular response, will acquire such functions (Mackintosh, 1977).
Furthermore, if two equally relevant stimuli are present when a response
produces a particular consequence, one of them may acquire discriminative
functions while the other does not (e.g., Reynolds, 1961). This is known as
overshadowing. Overshadowing may also occur when an already-
established SD (S1) and another stimulus (S2) are both present when a
response is reinforced or punished. Under these conditions, S1 may prevent
S2 from acquiring discriminative functions unless the consequences or
schedule of reinforcement for the response change with the introduction of
S2 (Kamin, 1969; Rescorla & Wagner, 1972).
In light of the above, it is important to recognize that a variety of stimuli
are present during discrimination training and that any of these may
adventitiously acquire some discriminative functions in the course of
training. The delivery of reinforcers (Bouton & Trask, 2016; Franks & Lattal,
1976) or aversive stimuli (Ayllon & Azrin, 1966; Holz & Azrin, 1961), states
produced by drugs or other motivating operations (Lubinski & Thompson,
1987; Overton, 1984), and the stimulation provided by an individual’s own
behavior (Killeen & Fetterman, 1988) are all capable of acquiring and
exerting discriminative control over behavior. In some cases, it is even
possible for these events to interfere with the discriminative control
established by a practitioner.

MOTIVATING OPERATIONS
The study of motivation in behavior analysis pertains to changes in the
probability of behavior that occur due to the influence of environmental
events functioning as motivating operations. Motivating operations (MOs)
are antecedent events that have two functions. First, they alter the value of
reinforcers and aversive stimuli, increasing or decreasing their
reinforcing/punishing efficacy. Second, they alter the probability of behavior
relevant to those events/activities as consequences. The former has been
referred to as the value-altering effect and the latter as the behavior-
altering effect (Laraway, Snycerski, Michael, & Poling, 2003).
There are two subclasses of MOs: establishing operations (EOs) and
abolishing operations (AOs; Laraway et al., 2003). We take each in turn.

Establishing operations. EOs are events that increase the efficacy of

reinforcers or aversive stimuli and increase the probability of behavior. If an
individual is exposed to an EO for a reinforcer, the probability of all
responses that have produced that reinforcer in the past increases. If an
individual is exposed to an EO that increases the extent to which an event
functions as aversive, the probability of behavior that has produced the
reduction or removal of that aversive stimulus in the past increases.

Abolishing operations. AOs have the opposite effect: they reduce the
efficacy of reinforcers or aversive stimuli and decrease the probability of
behavior. An AO for a particular reinforcer will decrease the probability of
responses that have produced that reinforcer in the past, and an AO for an
aversive stimulus will decrease the probability of responses that have
resulted in the removal/reduction of that stimulus in the past.
Note that EO and AO are relative terms. Whether a particular event
functions as an EO or AO depends on the motivational condition(s) to which
it is compared. For example, six hours of food deprivation is an EO relative
to no food deprivation, but is an AO relative to twelve hours of food
deprivation.

Types of motivating operations. A wide variety of events function as

MOs, and various MOs for different reinforcers/aversive stimuli may
operate concurrently (Sundberg, 2013). Unconditioned MOs function as
MOs without a prior history with those events, and conditioned MOs are
events that acquire motivational functions through correlation with other
events (Michael, 1993). The types of events that function as MOs include
various deprivation operations (e.g., food, water, sex, or sleep deprivation),
aversive stimulation (a given aversive stimulus can serve to punish the
response that precedes it and also function as an MO for its
reduction/removal; Michael, 1993), drug or alcohol intake (Valdovinos &
Kennedy, 2004), pain or illness (O’Reilly, 1997; O’Reilly, Lacey, &
Lancioni, 2000), verbal stimuli (sometimes referred to as augmentals;
Hayes, Barnes-Holmes, & Roche, 2001), events associated with emotions
(Lewon & Hayes, 2014), and the repeated presentation of reinforcers or
aversive stimuli (known as habituation, described earlier in this chapter;
Azrin & Holz, 1966; McSweeney & Murphy, 2009).

TERMINOLOGICAL DISTINCTIONS
Since the terms MO and SD both fall under the umbrella that is
antecedent control, it is worth emphasizing their distinctive functions for
technical as well as practical purposes. Michael (1982, 1993) distinguished
between the effects of SD and MO on the basis that the former relates to the
differential availability of reinforcement for a given response while the latter
relates to the differential effectiveness of a particular reinforcer. This is often
a useful distinction in practical applications, as behavioral excesses or
deficits related to the presence or absence of SDs will require different
intervention strategies from those related to MOs.
Despite their technical differences, a variety of SDs and MOs are always
present and interact in the evocation of behavior. Research has shown that
EOs increase the probability of a given response in the presence of an SD,
but to a lesser extent they also increase the probability of the response in the
presence of other stimuli not explicitly established as SDs (Edrisinha,
O’Reilly, Sigafoos, Lancioni, & Choi, 2011; Lotfizadeh, Edwards, Redner, &
Poling, 2012). This is to say that a particular response is most likely when
both the SD and an EO for the relevant reinforcer are present, and least likely
when both are absent.

Implications and Future Directions

The principles of classical (i.e., respondent) and operant conditioning are
the theoretical foundations of behavior science movements both new and
old. No matter how foundational, though, those principles are verbal
constructions. Even though they are the products of meticulous
observations, they remain susceptible to misinterpretation in the form of
mentalisms and dualisms. This is to say that the principles of behavior are
not static, as they evolve in accordance with the behavior of scientists,
practitioners, and the cultural milieu within which they operate. Testaments
to such evolution are Skinner’s (1963) take on the law of effect, Rescorla’s
(1988) take on classical conditioning, Premack’s (1959) probability-
differential hypothesis, Timberlake and Allison’s (1974) response-
deprivation hypothesis, and Baum’s (2012) rethinking of reinforcement.
While the principles of behavior are fundamental, they are subject to
refinement and evolution over time.

Conclusion
While behavior science is diverse in its research and application of the
principles of behavior, there is a common thread that unites such diversity:
the pursuit of a natural science of psychology, or more specifically, a natural
science of behavior. The aim of this chapter was to provide an overview of
fundamental behavioral processes, including some relatively advanced
topics. These processes serve both as a foundation for the remaining
chapters and as a starting point for the behavior analysis of language and
cognition pursued in this text.

Study Questions
1. Provide an example of habituation.
2. What is dishabituation? Elaborate on your example from question 1
and explain how dishabituation may occur.
3. Give an example of respondent conditioning, being sure to describe
a previously neutral stimulus, a US and UR, and how the neutral
stimulus became a CS eliciting a CR.
4. What are the differences between positive and negative
reinforcement and punishment?
5. How is extinction different from punishment? Explain with an
example.
6. What are the effects of extinction as reviewed in the chapter?
7. Distinguish between SD and S∆. Provide an example of each.
8. How might overshadowing interfere with the establishment of
stimulus control during discrimination training?
9. What are the two effects of MOs?
10. How are MOs distinguished from SDs?
11. Provide an example of both an EO and an AO.
 Chapter 3:

Basic Verbal Operants

Rocío Rosales
University of Massachusetts Lowell
Yors A. Garcia
The Chicago School of Professional Psychology
Sebastian Garcia
Southern Illinois University, Carbondale
Ruth Anne Rehfeldt
Southern Illinois University, Carbondale

Link and Preview

While chapter 2 focused on general behavioral processes, this
chapter begins to consider language from the perspective of
behavior analysis. Specifically, the authors focus on the basics of
B. F. Skinner’s (1957) approach to verbal behavior, including an
overview of fundamental concepts, research, and implications for
practice. These principles are important in their own right, and also
set the stage for more complex topics to be addressed later in the
book.

Behavior analysis is a science dedicated to identifying functional relations

between behavior and environmental variables (Delprato & Midgley, 1992).
A behavioral account of language (Skinner, 1957) places special emphasis
on the conditions that affect the occurrence of verbal behavior: motivation,
discriminative stimuli, and reinforcement. The focus on the pragmatic
function of language distinguishes Skinner’s approach from more
traditional accounts of language that focus on the topography or structure of
language (i.e., Chomsky, 1966).

Verbal Operants as Defined by Skinner

Skinner (1957) defined verbal behavior as “behavior reinforced through the
mediation of other persons [who] must be responding in ways which have
been conditioned precisely in order to reinforce the behavior of the speaker”
(p. 2, p. 225). He accordingly described a taxonomy of several elementary
verbal operants:
1. The mand is “a verbal operant in which the response is reinforced
by a characteristic consequence and is therefore under the functional
control of relevant conditions of deprivation or aversive
stimulation” (pp. 35–36). That is, the mand (a sign or gesture,
selection of a picture or activation of a vocal-output device, or a
vocalization) specifies its reinforcer and is influenced by relevant
establishing or motivating operations (Laraway, Snycerski, Michael,
& Poling, 2003; Michael, 1988).
2. The tact is a response (i.e., vocalization or sign) evoked by a
particular nonverbal discriminative stimulus (object, event, or
property of an object or event) and is maintained by generalized
conditioned reinforcement such as praise or attention from a listener
(pp. 81–82).
3. The intraverbal is a verbal response (i.e., written or vocal) under
the control of a verbal stimulus with no point-to-point
correspondence or formal similarity to the response, and is
maintained by generalized conditioned reinforcement (p. 71).
4. The echoic is a response (i.e., imitation or vocalization) under the
functional control of a verbal stimulus with point-to-point
correspondence and formal similarity to the response, and is
maintained by generalized conditioned reinforcement.
5. Textual behavior is a response (i.e., overt or covert vocalization)
evoked by a verbal stimulus (i.e., written text) with point-to-point
 correspondence but no formal similarity (p. 65).

Michael (1982) further expanded upon the last two operants proposed
by Skinner, echoic and textual responding, by describing the codic and
duplic. He suggested that we adopt the term “codic” for relations preceded
by a verbal stimulus with point-to-point correspondence but no formal
similarity (i.e., textual and taking dictation) and the term “duplic” for
relations preceded by a verbal stimulus with both point-to-point
correspondence and formal similarity (i.e., echoics, imitation, and copying a
text). For example, if a learner is presented with the auditory stimulus “D-I-
N-O-S-A-U-R” and accurately writes each corresponding letter on a
whiteboard, this is an example of a codic. If the same learner is presented
with the visual stimulus DINOSAUR in a book and proceeds to write each
corresponding letter on a whiteboard, this is an example of a duplic.
According to Michael, this arrangement resulted in useful categories and
prevented confusing extensions such as referring to Braille as textual
behavior and imitation of signs as echoic behavior.
Importantly, the topography, or form, of verbal behavior is not limited
to vocalizations. Rather, all of the verbal operants may be emitted as a
gestural sign, a picture exchange, or a touch or point to activate a speech-
generating device. The topography is not relevant because it does not provide
us with information about the function of the response. According to Michael
(1985), verbal behavior can be either selection based or topography based.
Selection-based verbal behavior (and selection-based communication
systems) requires the same response topography across many responses (i.e.,
a point, a touch, a press of a button, or a picture exchange). Topography-
based verbal behavior (and topography-based communication systems)
requires a unique response form that has point-to-point correspondence with
the relevant response product (i.e., a vocalization, a written response, or a
manual sign). Skinner’s analysis was primarily focused on the function of a
verbal response because the topography of a response does not help to inform
us about the functional relationships that exist between antecedent and
consequent stimuli that evoke a specific response. In other words, Skinner’s
analysis of verbal behavior was a functional analysis, which has important
implications for intervention development.
In the last several decades, a large and growing body of empirical
support has emerged for Skinner’s analysis of verbal behavior as a
framework for teaching language to individuals with developmental delays
(Aguirre, Valentino, & Leblanc, 2016; Carr & Miguel, 2012; Dixon, Small, &
Rosales, 2007; Sautter & LeBlanc, 2006). Continued study of verbal behavior
is important for advancing the analysis of language and cognition from a
behavioral perspective. Skinner’s interpretation of language also has copious
practical value for the clinical services provided to individuals with autism
spectrum disorder (ASD) and other related disorders (Sundberg & Michael,
2001). For example, traditional approaches to language assessment do not
fully capture the function of the learner’s repertoire. This may be due in part
to the lack of consideration of motivating operations and automatic
reinforcement, and the overemphasis on words and their meanings (Sundberg
& Michael, 2001). Collectively, these factors can lead practitioners to
underestimate the complexity of verbal relations and make inaccurate
assumptions about a learner’s existing repertoire. Inaccurate assumptions can
include attributing the learner’s failure to acquire skills to a diagnosis rather
than to an incomplete analysis of the environmental variables that may
contribute to such failure (Sundberg & Michael, 2001).

Verbal Behavior Research

In this section we turn our attention to the empirical support for and clinical
implications of three of the elementary verbal operants: echoics, mands, and
tacts.

Echoics Defined
The echoic is one of the earliest verbal operants observed in young
children and is said to be dependent on the exposure to speech sounds
produced by the verbal community (Horne & Lowe, 1996). When a child
readily imitates novel verbal behavior in the absence of contingent social
reinforcers, they are emitting generalized echoic responding. Generalized
echoic responding is fundamental to the development of more complex
behaviors such as self-instructions, listener behavior, and naming (Horne &
Lowe, 1996), and it is considered a behavioral cusp because it facilitates
many new interactions (Rosales-Ruiz & Baer, 1997). In behavioral
interventions for children with language delays, an echoic repertoire is useful
for teaching other verbal operants (Barbera & Kubina, 2005; Kodak,
Clements, & Ninness, 2009). Thus, a generalized echoic repertoire is
fundamental for the development of more complex verbal behavior.
Skinner (1957) asserted that echoic behavior should not be confused
with self-reinforcing responses that are not socially mediated. For example,
an infant’s babbling is automatically reinforcing if it persists in the absence
of feedback from parents (p. 58). Automatic reinforcement occurs when
reinforcement is the product of a response and is not socially mediated
(Vaughan & Michael, 1982). From the time of an infant’s birth, parents’
sounds and words that are paired with reinforcing activities (i.e., feeding)
become conditioned reinforcers for the infant. These sounds, when produced
by the child, will help strengthen the muscle movements that are necessary to
produce babbling. As a result, infants will babble the sounds that are paired
with socially mediated reinforcement more frequently.
The process of automatic reinforcement helps strengthen a child’s vocal
verbal repertoire and prepare them for speaking in words and eventually full
sentences (Sundberg, Michael, Partington, & Sundberg, 1996). The process
of automatic reinforcement involves two steps: (1) sounds and words heard
by a young child become conditioned reinforcers when paired with the
parent’s positive feedback and interactions (e.g., as mother talks to her infant
while she feeds her); and (2) production of the sounds by the child is
strengthened by the product of her verbal behavior in the form of auditory
stimuli (e.g., the infant makes vocalizations that begin to sound like her
mother’s voice, and these auditory stimuli serve as reinforcers for more
vocalizations). The closer the sound production (e.g., the infant’s
vocalization) is to the sound conditioned as a reinforcer (e.g., the mother’s
voice), the higher the value of that reinforcer (Sundberg et al., 1996).

ECHOIC INSTRUCTION
When children with language delays fail to develop an echoic repertoire,
vocal models help to prompt simple vocalizations (Barbera & Kubina, 2005;
Bourret, Vollmer, & Rapp, 2004). If the child does not emit any
vocalizations, echoic instruction may involve a shaping procedure whereby
the instructor provides a vocal prompt and differential reinforcement for
successive approximations to each target sound (Cook & Adams, 1966;
Harris, 1975). For example, when teaching a child to say “mama,” the mother
may initially place emphasis on the beginning sound of the word (“mmmm”)
and initially reinforce any approximation to this sound with enthusiasm. The
shaping procedure would require the mother to fade out this enthusiastic
reinforcement over time and as additional opportunities to echo the sound
were made available. After a few instances of the child reliably saying
“mmmm,” the mother may change her reaction to this vocalization and
provide enthusiastic praise for a closer approximation (e.g., child says
“mmmma”). This process may be difficult and slow if vocalizations of any
topography occur at very low rates. An alternative procedure with moderate
empirical support is the stimulus-stimulus pairing (SSP) procedure
(Petursdottir, Carp, Mathies, & Esch, 2011; Shillingsburg, Hollander, Yosick,
Bowen, & Muskat, 2015).

STIMULUS-STIMULUS PAIRING
Stimulus-stimulus pairing is designed to increase vocalizations using the
same process that is thought to result in increased infant babbling (Yoon &
Bennett, 2000). Many variations of SSP have been reported to date. The
essential feature of SSP is an adult presenting a vocal sound paired with
delivery of highly preferred tangible or social stimuli. No overt response is
required by the child in this procedure. Through the repetition of sounds
paired with preferred stimuli, as emitted by the therapist, the goal of SSP is to
establish unique and specific vocal sounds as conditioned reinforcers (as
described above). Following repeated presentations, the child may produce
sounds that are the same or very similar to what they hear (Sundberg et al.,
1996). The increase in any vocalization has important clinical implications
because these behaviors provide an opportunity for reinforcement, shaping,
and mand instruction. In other words, once the child is producing initial
vocalizations through the use of SSP, those vocalizations may then be
expanded upon using additional interventions.
To date, the empirical support for the efficacy of SSP is mixed. A recent
literature review by Shillingsburg and colleagues (2015) suggests that
conclusions are difficult to draw from the evidence base because of varied
participant characteristics, information reported on participants’ verbal
repertoires, and reported procedural differences (e.g., number of pairing
trials, number of experimenter-emitted sounds, and type of preferred items
used during instruction). As such, clinical recommendations for the specific
use of SSP are difficult to establish. In general, the results of the literature
review show higher success with SSP when participants were younger (i.e.,
under age five), with use of a delayed pairing procedure (i.e., delivery of the
preferred item overlapped with presentation of the target sound), and when
there was control for adventitious reinforcement (i.e., the experimenter
withheld access to preferred items for a specified time if the participant
emitted a sound immediately following presentation of the vocal model). As
noted by the authors of the review, although this last finding is surprising, it
could be partially accounted for by the fact that the majority of studies that
controlled for adventitious reinforcement also happened to be conducted with
younger participants. Additional research is needed to control for these
variables. Applied researchers are encouraged to continue working on this
important line of work, especially in identifying specific participant
characteristics that can predict success with this procedure (Shillingsburg et
al., 2015). One important avenue for future research may be to establish
initial vocalizations with SSP and then implement more traditional echoic
and manding teaching procedures.

Mands Defined
In nontechnical terms, a mand is a request that occurs under specific
motivation and specifies the reinforcer to be delivered. As discussed in
chapter 2, a motivating operation (MO; Laraway et al., 2003) is defined as a
change in the environment that momentarily increases (establishing
operation [EO]) or decreases (abolishing operation [AO]) the value of a
reinforcer and evokes behavior that has previously resulted in access to the
reinforcer (Michael, 1982). For example, if a child has been denied access to
their favorite toy dinosaur, the reinforcing value of the dinosaur is
temporarily increased, and behaviors that have produced access to the
dinosaur are evoked (i.e., the child says, “Give me the dinosaur” or simply
“dinosaur”).
Selection of mand modality is an important component of clinical
assessment and intervention for learners with language delays. Several
studies have systematically evaluated outcomes for topography- versus
selection-based mand forms in learners with autism spectrum disorder
(ASD) and related disorders. In aggregate, the results of these studies provide
support for the use of selection-based communication systems (Adkins &
Axelrod, 2001; Barlow, Tiger, Slocum, & Miller, 2013; Chambers &
Rehfeldt, 2003; Gregory, DeLeon, & Richman, 2009; Lorah, Parnell, Whitby,
& Hantula, 2015; Tincani, 2004; Ziomek & Rehfeldt, 2008), but there are
idiosyncrasies across findings and the number of participants included is
generally small. Therefore, further research is needed to draw firm
conclusions. Shafer (1993) outlined considerations for selection- and
topography-based systems (described above in the chapter introduction). We
offer some clinical recommendations for the selection of a mand modality
based on the information discussed by Shafer and the empirical evidence to
date.

SELECTION-BASED MAND SYSTEMS

First, a widely used selection-based system with ample empirical
support is the Picture Exchange Communication System (PECS; Frost &
Bondy, 2002). A meta-analysis conducted by Tincani and Devis (2011)
revealed that PECS has been moderately effective in establishing mands up
to the fourth phase (of six total phases) and that PECS has been shown to
facilitate vocalizations in some participants. A second form of selection-
based verbal behavior is the use of speech-generating devices, which now
encompass the use of more accessible high-tech devices such as the iPad, but
results from studies investigating the efficacy of this modality relative to
others are mixed (see Lorah et al., 2015; Still, Rehfeldt, Whelan, May, &
Dymond, 2014).
Potential disadvantages for the use of selection-based responding
systems include prerequisite skills that may be required for successful
implementation (i.e., scanning and making conditional discriminations).
Gregory et al. (2009) evaluated the relationship between matching and motor
imitation skills with regard to the acquisition of both manual sign and
exchange-based systems for six children with developmental disabilities.
Following an assessment to determine if motor imitation and matching were
already in each participant’s repertoire, participants were taught to mand for
the same preferred items using both systems. The results showed strong
correspondence between motor imitation and manual sign acquisition, but
also between matching and manual sign. In addition, the correspondence
between both assessments and exchange-based communication was almost as
strong. However, three of the six participants did not learn either
communication system. These results support the claim that these skills may
expedite the acquisition of these communication forms.
A second noted disadvantage of selection-based responding systems is
that the learner must have the pictures or device available at all times, which
may limit its use in natural settings (Sundberg & Partington, 2013). This
second disadvantage is at least partially addressed by the use of high-tech
portable devices that are more readily available and commonly used by
neurotypical peers for other purposes (i.e., iPhone or mini iPad). In this
regard, the use of high-tech devices may be more socially acceptable because
virtually everyone in today’s society makes frequent use of such devices and
therefore their use may “stand out” less than manual sign (Lorah et al., 2015).

TOPOGRAPHY-BASED MAND SYSTEMS

Topography-based responses such as manual sign have been described
as conceptually similar to speech (Sundberg & Partington, 2013) because
there is point-to-point correspondence for each response with its relevant
response product. The use of manual sign has demonstrated efficacy as a
viable communication system for learners with language delays, with some
studies showing increased vocalizations following instruction that is paired
with tacts of the requested items (i.e., “total communication”; Barrera,
Lobato-Barrera, & Sulzer-Azaroff, 1980; Barrera & Sulzer-Azaroff, 1983;
Brady & Smouse, 1978; Carbone et al., 2006; Carbone, Sweeney-Kerwin,
Attanasio, & Kasper, 2010). A commonly noted disadvantage for the use of
manual sign is that the verbal community that can interact with the individual
leaner will be limited. In addition, manual sign may not be a good fit for
learners with ASD who also often present with motor impairments (Green et
al., 2009). Manual sign has the advantage of requiring only simple
discrimination repertoires, and the response topography is easily identified
across different individuals, so long as the individual is capable of executing
the standard sign (as opposed to a unique modified version of the sign).
Studies that have evaluated preference for a mand topography
demonstrate mixed results, with some showing a clear preference for speech-
generated devices (Lorah et al., 2015; McClay et al., 2016) and others
indicating that some learners prefer a picture exchange system (van der Meer,
Sutherland, O’Reilly, Lancioni, & Sigafoos, 2012). Recent studies have
demonstrated that although learners may show a preference for specific mand
topographies, this preference does not always impact the success of
interventions designed to decrease challenging behavior (Winborn-
Kemmerer, Ringdahl, Wacker, & Kitsukawa, 2009). Other considerations that
should be made in mand modality selection include the response effort
required by the learner (Horner & Day, 1991; Torelli et al., 2016), proficiency
in use of the mand topography (Ringdahl et al., 2009), and the history of
reinforcement for use of a specific modality (Matter & Zarcone, 2017).
The collective results from studies that have evaluated preference for
mand modality indicate that a mand topography assessment should be
conducted with all learners using augmentative communication systems. A
mand topography assessment is conducted with procedures similar to typical
preference assessments with actual objects or pictures. Guidelines for
implementation of this type of assessment are beyond the scope of this
chapter; interested readers should reference studies that have outlined the
steps to perform this type of assessment (see LaRue et al., 2016; van der
Meer et al., 2012). Future research on mand modality selection should
consider the skills that learners must present with in order to benefit from
multimodal communication systems (Shafer, 1993) and how teaching
multimodal systems may help to facilitate vocalizations and emergent verbal
relations, a topic we will discuss later in this chapter.

MAND INSTRUCTION
Mand instruction should be an essential component of any early
intervention curriculum because it increases the child’s control over their
environment (Sundberg & Michael, 2001). In addition, if therapists begin
treatment planning with mand instruction, it provides them with an
opportunity to pair themselves with unconditioned and previously established
conditioned reinforcers. Through mand instruction and the pairing that
occurs, therapists are likely to become conditioned reinforcers themselves.
The goal of mand instruction is to teach independent requests that are under
the control of relevant MOs. Therefore, ensuring the relevant MOs are
present during mand instruction is of paramount importance.
Two common strategies designed to ensure relevant MOs are present
during instruction include capturing learning opportunities as they occur in
the natural environment (Fenske, Krantz, & McClannahan, 2001) and
contriving or creating an opportunity by manipulating other relevant stimuli
in the environment (Hall & Sundberg, 1987). For example, incidental
teaching requires caregivers to identify MOs that occur naturally in the
child’s environment, such as when a child shows interest in a toy, and then
use a series of prompts to encourage manding (Fenske et al., 2001). A
common strategy used to contrive MOs during incidental teaching includes
withholding reinforcers for specified periods before a mand instruction
session begins (to create sufficient levels of deprivation, an EO). O’Reilly,
Aguilar, et al. (2012) evaluated levels of deprivation for preferred items prior
to mand instruction sessions. Results confirmed that limiting access to the
preferred item prior to the teaching session produced evocative and
reinforcing-establishing effects (an EO), whereas having access to the
reinforcer immediately before the start of the teaching session produced an
abating and reinforcer-abolishing effect (an AO; Michael, 1988; O’Reilly,
Aguilar, et al., 2012).
A second evidence-based approach for contriving MOs is the
interrupted behavior chains procedure (Carter & Grunsell, 2001; Duker,
Kraaykamp, & Visser, 1994; Rosales & Rehfeldt, 2007; Sigafoos, Kerr,
Roberts, & Couzens, 1994). This procedure involves contriving transitive
conditioned establishing operations (CMO-T). Michael (1993) defined
CMO-Ts as previously neutral conditions whose occurrences alter the
effectiveness of another stimulus and evoke responses that produce or
suppress that stimulus. The procedure begins with the therapist teaching a
sequence of steps (e.g., getting dressed to go outside). Once the child learns
the sequence of steps, the therapist interrupts the sequence by removing a
needed item to complete the behavior chain (e.g., moving the shoes out of
sight). Following a brief pause to allow the child an opportunity to respond
independently (e.g., “Shoes please”), the therapist provides a response
prompt (e.g., “Say…”). Following the missed step, the child completes the
remaining steps and the therapist delivers the reinforcer (i.e., going outside).
A major strength of the interrupted chains procedure is that it can
momentarily transform a previously neutral stimulus (e.g., shoes) into a
reinforcer, because it is necessary to continue the chain and access the larger
reinforcer at the end, thereby providing the opportunity to teach a learner to
mand for the item that they normally would not be motivated to mand for.
Transfer of stimulus control (ToSC) procedures are also effective for
establishing a mand repertoire. These procedures can be used in combination
with those outlined above (Thomas, Lafasakis, & Sturmey, 2010). ToSC
requires the teacher or therapist to introduce and then gradually fade prompts
in one of several ways (i.e., time delay, fading the stimulus prompts gradually
over time, use of a less intrusive prompt, or changing the topography of the
prompt). In this manner, a stimulus that already exerts control over the target
behavior is systematically faded until the child produces the target response
independently.
 Teaching simple mands (e.g., single words) is a prerequisite for teaching
more advanced language skills such as mands for information (e.g., “Where
is the milk?” or “When is my birthday?”; Lechago & Low, 2015). Previous
studies on this topic have combined CMO-T with an interrupted behavior
chain to produce mands for information. For example, Landa, Hansen, and
Shillingsburg (2017) taught the mand for information “When” (e.g., “When
can I have…”) to three children with ASD. They alternated two conditions:
presence of EO (e.g., therapist denied access to items or activities) and
absence of EO (e.g., therapist denied access to reinforcers and requested
completing an additional activity). When the therapist withheld relevant
information from the learner, the learner produced mands for information
exclusively.
Other researchers have examined the effects of procedures to establish
variability in manding. This is important because individuals with ASD often
engage in repetitive and restrictive language. For example, a child with ASD
may emit the same mand form (e.g., “I want…”) without engaging in an
alternative mand form (e.g., “I would like…”) to gain access to preferred
items. One strategy with empirical support for teaching mand variability is
the use of a lag schedule of reinforcement, in which the wording of a mand
cannot be identical twice in a row, in order for the learner to receive
reinforcement (Lee & Sturmey, 2006). A second strategy with empirical
support is script-fading procedures. Scripts are visual or auditory supports
that serve as discriminative stimuli for the student to emit a response.
Brodhead, Higbee, Gerencser, and Akers (2016) investigated the effects of a
script-fading and discrimination-teaching procedure on mand variability in
individuals with ASD. Scripts were faded progressively (e.g., “I would like
”, “I would ”, “I ”) in the
presence of a discriminative stimulus using lag schedules of reinforcement.
Sellers, Kelley, Higbee, and Wolfe (2016) also evaluated script training and
fading for the acquisition and maintenance of varied mand frames (i.e., “May
I have ”, “Please give me ”) for preschool
children with ASD. If participants did not demonstrate increases in their
mand variability following a continuous schedule of reinforcement, the
researchers implemented post script-fading extinction to induce variability.
This procedure was effective for all but one of the participants in the study.
These results demonstrate that this is an effective procedure that should
continue to be evaluated.
 In summary, mand instruction is an essential component of early
intensive intervention for young children with language delays, but it should
also be prominent when intervening on problem behavior for learners of any
age. Several studies support the use of assessment-based intervention for skill
acquisition, but there is a paucity of research to identify the most effective
intervention to establish verbal operants for a given individual. Future
research should evaluate how such an assessment may be conducted.

Tacts Defined
In nontechnical terms, tacts are labels or descriptions of the world
around us, and “tacting” refers to this behavior of labeling and describing.

BENEFITS OF A TACT REPERTOIRE

Tacts are of primary benefit to the listener because they convey
information on environmental stimuli. Stimuli that evoke tacting can include
all of the senses (visual, auditory, olfactory, sensory, and gustatory). Tacts
may serve as the basic building blocks for conversation. For this reason, they
are often targeted in early curricula for learners with ASD and related
disorders (Partington, 2008; Sundberg, 2008). Additionally, the development
of a tact repertoire may help reduce the frequency of nonfunctional language
emitted by children with ASD (Karmali, Greer, Nuzzolo-Gomez, Ross, &
Rivera-Valdes, 2005). Initial targets for tact instruction may consist of
learning tacts for objects and persons, followed by representative pictures of
these items. Mastery of these initial targets is typically followed by tacts of
feature, function, and class of common items; prepositions and actions; and
relational descriptors (e.g., soft/rough, more/less, hot/cold, wet/dry;
Partington, 2008; Sundberg, 2008).

TACT INSTRUCTION
 When echoic and mand repertoires are well established, these verbal
operants can be used as prompts in ToSC procedures (Barbera & Kubina,
2005). For example, in the echoic-to-tact transfer, a stimulus is presented and
the instructor presents an echoic prompt (i.e., tacts the stimulus and waits for
the learner to respond). If the learner does not tact the stimulus independently
following the initial echoic prompt, a second prompt may be added (i.e., “Say
”) followed immediately by the discriminative stimulus once
more (i.e., “What is it?”). Barbera and Kubina (2005) successfully
demonstrated this prompting sequence as an effective method to establish
stimulus control for their selected targets.
It is important to note that the supplemental question “What is it?” is not
part of Skinner’s (1957) definition of a tact, and this additional discriminative
stimulus may sometimes interfere with the teaching procedure (Marchese,
Carr, LeBlanc, Rosati, & Conroy, 2012). The introduction of a supplemental
question in tact instruction can be problematic because the learner may not
respond appropriately to stimuli in their environment (considered a “pure”
tact form). That is, the learner may not emit a response until a question is
posed (Marchese et al., 2012). A related problem with presenting
supplemental questions during tact instruction is that the question may
acquire intraverbal control over early responses and lead to interference with
the acquisition of new responses (Partington, Sundberg, Newhouse, &
Spengler, 1994). One other potential problem with inclusion of a
supplemental question during tact instruction is that learners may begin to
imitate part of the question. This is especially likely if the learner engages in
echolalia under other conditions. While additional intervention can remediate
this problem (McMorrow, Foxx, Faw, & Bittle, 1987), it could lead to
unnecessary delays in targeting other important treatment goals. Sundberg,
Endicott, and Eigenheer (2000) targeted signed tacts for two young children
with ASD who presented with previous difficulty acquiring tacts. They
compared two conditions: object with supplemental question versus object
with a prompt to “sign [object name].” Results of this study demonstrated
superior performance in the sign-prompt tact acquisition condition.
The instructional procedures outlined above for mands and tacts may be
best suited for learners with an established or emerging speaker repertoire.
An alternative procedure that has received attention in recent years is listener
training (Fiorile & Greer, 2007). Listener training involves teaching
conditional discriminations to a learner via match-to-sample procedures and
then testing for the emergence of speaker responses (i.e., mands and tacts). A
related procedure that has received empirical support is multiple exemplar
instruction (MEI). We will discuss both of these topics below.

Functional Independence of Verbal Operants

The notion of functional independence of verbal operants has been
presented as a major tenet of Skinner’s analysis of verbal behavior
(Lamarre & Holland, 1985). The concept of functional independence
suggests that establishing one verbal operant within the repertoire of a
learner will not necessarily lead to emission of any other verbal operant for
that learner. For instance, a child who reliably emits a mand “chocolate”
under the pure control of MOs would not be expected to say “chocolate” in
the sole presence of a nonverbal stimulus (i.e., they would not necessarily
be expected to tact just because they mand). The same is true with the
opposite relation, acquiring the mand response “chocolate” after learning to
tact for the same item. Whether verbal operants are functionally
independent has been a matter of debate in recent years. Some empirical
support exists for functional independence across verbal operants (Hall &
Sundberg, 1987; Lamarre & Holland, 1985; Partington et al., 1994; Simic &
Bucher, 1980; Twyman, 1996), while other studies demonstrate
interdependence of mands and tacts (Finn, Miguel, & Ahearn, 2012;
Gilliam, Weil, & Miltenberger, 2013; Kooistra, Buchmeier, & Klatt, 2012;
Wallace, Iwata, & Hanley, 2006) and other elementary verbal operants
(Grow & Kodak, 2010).
Fryling (2017) recently suggested that Skinner neither called for
research to confirm the functional independence of verbal operants nor
openly declared that the verbal operants were functionally independent. In
fact, Skinner included a full chapter on multiple causation (Michael, Palmer,
& Sundberg, 2011), which can be interpreted as an implication that verbal
behavior is controlled by different antecedent variables. Regardless of
whether Skinner urged for clarifying the role of functional independence
across verbal operants, the data so far are equivocal (Gamba, Goyos, &
Petursdottir, 2015). In their review of the literature on functional
independence of mands and tacts, Gamba and colleagues (2015) reported
procedural differences in training and testing probes that were employed in
the published studies on this topic. For this reason, they conclude that the
reported lack of transfer between mands and tacts in the published literature
(and vice versa) could be attributed to alternative explanations, such as
problems in capturing the relevant controlling variables to evoke the mand or
tact response (i.e., contriving or capturing MOs) or establishing trained
responses that do not function as mand or tacts, but are instead controlled by
other variables present in the training situation (i.e., presence of an item
during mand probes). Thus, Gamba and colleagues suggested that the
existing support for functional independence presents problems related to
issues of construct validity, and as such, no single published study has
reported exceptionally strong evidence for mand-tact independence.
Gamba and colleagues (2015) suggest that unless future research on this
topic controls for critical variables (e.g., appropriate participants, use of
standardized language assessments, separate antecedent and consequential
control for mands and tacts), more research seems unnecessary. These
authors suggest designing alternative strategies that do not involve teaching
one operant and then testing for the other, as it may be extremely difficult to
design such a study that does not leave room for alternative interpretations,
regardless of the outcome. Instead, future research may evaluate existing
verbal repertoires in a functional analysis to determine how each operant is
used (Lerman et al., 2005). Alternatives are to evaluate the emergence of
both mands and tacts following listener training (Ribeiro, Elias, Goyos, &
Miguel, 2010) or to test for the functional independence of verbal operants by
manipulating schedules of reinforcement for each verbal operant that is
assessed.
The goal of identifying functional independence or interdependence in
applied research and in practice is to establish effective and efficient
communication for learners with language delays. Research in this area has
been primarily focused on establishing emergent verbal relations. Emergent
relations are defined as responses that are acquired under the controlling
conditions for one verbal operant (i.e., a tact) that transfer to conditions of
another verbal operant (i.e., a mand) with no further teaching or direct
instruction (Finn et al., 2012; Gilliam et al., 2013; Kooistra et al., 2012;
Wallace et al., 2006). Practitioners who are well informed on the definition
and importance of functional independence and interdependence (i.e., the
conditions under which functional interdependence has been observed) will
be better equipped to develop and implement treatment plans that will
establish emergent relations (Egan & Barnes-Holmes, 2010). This translates
into better, more efficient intervention, potentially enabling teaching more
skills in the same amount of time.
To this end, we offer some clinical recommendations: First, practitioners
should conduct frequent assessment and probes to evaluate the functional
independence or interdependence of verbal operants. Clinicians are cautioned
against assuming functional interdependence as this may lead to failures in
curriculum development and disproportionately increase one verbal
repertoire over the other (Carr & Miguel, 2012; Sautter & LeBlanc, 2006).
Second, practitioners should consider teaching mands first under pure MO
control, and if tacts do not emerge, then incorporate ToSC procedures. It
would be wise to use visual, vocal, or textual stimuli, along with the
presentation of the manded item, to bring the response under nonverbal
discriminative stimulus, and then deliver generalized conditioned
reinforcement instead of the preferred item (Lechago & Low, 2015). Third,
practitioners should use mixed operant training, that is, combine mands and
tact trials in one single teaching block (e.g., say “dinosaur” under SD and
MO control), then test for the emergence of either verbal operant (Arntzen &
Almås, 2002).
Fourth, clinicians are advised to use MEI. For example, Nuzzolo-Gomez
and Greer (2004) evaluated the effects of MEI on the emergence of untaught
mands or tacts of novel adjective-object pairs in four children with ASD or
other developmental disabilities. The experimenter provided initial probes for
mands under appropriate levels of deprivation (i.e., the participants had
access to the stimuli used in the study only during the experimental
conditions, and if a child did not make a selection during an initial probe, the
experimenter terminated the session until a state of deprivation could be
established). Once the child made an initial selection, the experimenter
placed the item out of view in a cup, bowl, or box as the child observed. They
then placed the item in an array in front of the child with two other containers
of the same type but a different size (i.e., small, medium, and large cup).
Tacts consisted of the experimenter’s placing the three items in front of the
child and pointing to the object to be tacted with no further verbal
antecedents. The experimenter explicitly taught the participants to mand and
tact using up to three different adjective-object response forms that were
different from the original set used during baseline probes. MEI included
learning opportunities for mands and tacts. Participants received instruction
for the mand with one adjective-object pair followed by instruction for the
tact response with the same adjective-object pair. This instruction was
followed by probes for the original untaught responses. If the participant did
not meet the mastery criterion, they were exposed to a second and then a
third adjective-object pair for both mand and tact response forms. Results of
the study showed that MEI produced high levels of correct responding on the
untrained operant.
Most of the published research on functional independence has been
conducted with mands and tacts. More work is needed to assess functional
independence across other verbal operants (Aguirre et al., 2016; Carr &
Miguel, 2012). There may be prerequisite skills that facilitate the emergence
of verbal operants, but those data are not yet available. Future research
should focus on targeting prerequisite skills before assessing for functional
interdependence of verbal operants (Aguirre et al., 2016).

Synthesizing Skinner’s Analysis of Verbal Behavior

with Relational Frame Theory
As we have discussed thus far, Skinner’s (1957) account of verbal behavior
advanced the field of behavior analysis by incorporating a topic that had not
previously been discussed from a behavioral perspective. Since the
publication of Skinner’s book, other behavior analysts have expanded upon
his analysis (Greer & Speckman, 2009; Hayes, Barnes-Holmes, & Roche,
2001; Horne & Lowe, 1996; Sidman, 1994). Barnes-Holmes, Barnes-
Holmes and Cullinan (2000) proposed a synthesis of Skinner’s work with
relational fame theory (RFT) that has led to applied research with
significant clinical implications for learners with autism and related
disabilities (Rehfeldt & Barnes-Holmes, 2009; see chapters 11 and 12 in
this volume for more elaborate overviews of RFT). The main benefit of
combining these two approaches is the development of teaching
applications that are focused not only on contingency-based learning, but on
derived or emergent (not directly taught) skills (Murphy, Barnes-Holmes, &
Barnes-Holmes, 2005). In addition, the nature of derived or emergent
relations can help account for the generativity of human language (Murphy
& Barnes-Holmes, 2009a).
Rehfeldt and Root (2005) evaluated a procedure to establish derived
mands, defined as mands not directly taught but that emerged from a history
of conditional discrimination instruction. Specifically, Rehfeldt and Root first
taught three adults with developmental disabilities to mand for preferred
items using pictures. This was followed by conditional discrimination
instruction to establish two relations: dictated name to picture and dictated
name to corresponding text. Next, the experimenters conducted probes for
derived mands (i.e., mands for preferred items using text instead of pictures).
All participants demonstrated derived manding following this procedure.
Rosales and Rehfeldt (2007) also evaluated derived manding for adults with
moderate developmental disabilities. This study extended the previous
findings by evaluating “pure” mands, where the reinforcer was not in view of
the participant during the derived mand probes. This was accomplished
through the utilization of the interrupted chained task procedure (Hall &
Sundberg, 1987). Results showed that all participants were successful in the
derived mands probes although performance declined at follow-up.
A series of studies by Murphy et al. (2005) and Murphy and Barnes-
Holmes (2009a, 2009b, 2010) also demonstrated derived manding skills in
typically developing children and children with ASD. The focus of these
studies was on mands for “more” or “less” tokens of various amounts. For
example, Murphy et al. (2005) taught children with ASD derived mands for
single tokens in the context of a game. They first taught participants to mand
for a specific number of tokens (X1 and X2) by presenting a corresponding
card to the experimenter. Importantly, the stimuli were arbitrary symbols and
nonsense syllables printed on cards. They then implemented a conditional
discrimination procedure to establish stimulus classes with three nonsense
syllables (i.e., A1-B1/A2-B2 and B1-C1/B2-C2) and directly taught
participants to mand for tokens using only the original stimuli (X1 and X2).
On subsequent test probes, participants demonstrated derived mands using
the stimuli that were conditionally related during training (i.e., C1 and C2).
A follow-up study by Murphy and Barnes-Holmes (2009a) included
derived mands for the addition or removal of a single token, and a third study
showed derived manding skills for more or fewer tokens of specific amounts
(Murphy & Barnes-Holmes, 2009b). Collectively, these results support the
synthesis of Skinner’s analysis with RFT to establish derived manding skills
in children with and without ASD and in adult learners with moderate
developmental disabilities.
The teaching approach employed in this series of studies should be
further explored in verbal behavior programming for children with ASD in
applied settings. The evidence from applied studies indicates these teaching
procedures are likely to prove effective and efficient for teaching multiple
mand responses. However, absent the data to support these procedures in
applied settings, this remains an empirical question. Future studies should
also evaluate the effectiveness of the procedures with individuals with less
sophisticated verbal and cognitive repertoires. Another suggestion for
practitioners working with this population is to establish flexible responding
from the beginning of instruction by teaching multiple mand targets that lead
to the same outcome simultaneously (i.e., water, drink, and juice) instead of
teaching one at a time to criterion (Murphy & Barnes-Holmes, 2009b).

Implications and Future Directions

Understanding Skinner’s analysis of verbal behavior is important for
practitioners tasked with the development of appropriate treatment planning
for individuals with language delays. For this reason, practitioners should
be familiar with the variety of topographies and teaching techniques
available for alternative communication. Selecting appropriate targets and
learning to manipulate variables to evoke vocalizations and other
topographies of verbal behavior are skills that practitioners should develop
in order to provide the most enriching learning environment for their
clients.
Skill-acquisition programming should include a goal for functional
communication that begins with teaching a learner how to mand for preferred
and needed items. Starting with an assessment of the existing verbal behavior
repertoire, capitalizing on the use of MEI, and testing for emergent relations
are all of utmost importance in order to make the most impact with limited
instructional time. We also recommend that practitioners familiarize
themselves with the different strategies for contriving and capturing MOs and
automatic reinforcement, two variables that will have an impact on the
development of early verbal repertoires (Sundberg & Michael, 2001).
The issue of functional independence and interdependence should be
explored with procedures different from those of past studies (Gamba et al.,
2015). Another area with a dearth of empirical support is language
intervention for individuals exposed to and/or learning multiple languages
(Lang et al., 2011). The U.S. Department of Education, National Center for
Education Statistics (2017) estimates that 13.8% of the total English language
learner (ELL) population enrolled in US public elementary and secondary
schools are students with disabilities. Given these statistics, it is evident that
teachers and practitioners will be required to develop intervention plans for
learners who are bilingual or multilingual. Practitioners and researchers alike
should aim for cultural competency when working with families from diverse
backgrounds and develop a better understanding of when the learner will
benefit from one language versus the other. For example, some researchers
have recently started to evaluate whether ELLs have a preference for the
home language (Aguilar, Chan, White, & Fragale, 2017), the impact of the
language of instruction on rates of challenging behavior (Padilla Dalmau et
al., 2011), and rates of acquisition for new skills when instruction is
presented in a bilingual format (Leon & Rosales, 2017).
Leon and Rosales (2017) evaluated the effects of tact training when
instruction was presented only in English and compared this training to
instruction presented in a bilingual format with both English and the
participant’s home language. This study was the first to directly compare
rates of acquisition in the context of bilingual training. The results showed
faster acquisition when instruction was in English, but better generalization
and maintenance for stimuli trained in the bilingual instruction condition. The
results should be evaluated with caution given the participant’s age (six
years, eight months) and history of reinforcement for listening to and
speaking in English only in his school environment. Applied researchers
should continue researching bilingual modifications to verbal behavior
instruction as it is an area ripe for investigation. For example, it would be
interesting to continue to evaluate rates of acquisition for mands and tacts
when instruction is presented in one or multiple languages, and also to
evaluate the learner’s preference for monolingual or bilingual instruction.
Skinner (1957) described his analysis of verbal behavior as an “exercise
in interpretation rather than a quantitative extrapolation of rigorous
experimental results” (p. 11). While only a handful of researchers applied his
interpretation to working with individuals with ASD and related disabilities
early on, there is now a robust body of evidence to support Skinner’s analysis
of verbal behavior (Aguirre et al., 2016; Dixon et al., 2007). Basic and
applied research in behavior analysis has been focused on designing
procedures that incorporate Skinner’s analysis to teach relevant language
skills to learners with and without disabilities, although arguably the focus
has been on young children with disabilities. Stewart, McElwee, and Ming
(2013b) pointed out that despite the upsurge of research on the basic verbal
operants, the generativity of language has been addressed far less. They
described language generativity as the “ability to produce sentences never
before said, and to understand sentences never before heard” (p. 137). While
the research that has been described in this chapter has been largely focused
on elementary verbal operants with some discussion on how procedures can
be designed to promote the emergence of untaught verbal operants, the
research agenda established by Stewart et al. (2013b) will only further
enhance the utility of a behavior analytic interpretation of language.

Conclusion
This chapter provided a synopsis of Skinner’s conceptualization of verbal
behavior, including an overview of his taxonomy of verbal operants, which
is based upon a functional analysis. A distinction was made between
selection-based and topography-based verbal behavior. Three elementary
verbal operants were defined according to Skinner’s taxonomy—echoics,
mands, and tacts—along with research in support of Skinner’s
conceptualization of each operant. Research by applied behavior analysts
has focused on identifying best practices in teaching manding, tacting, and
echoic behavior, and recent research has bridged Skinner’s analysis of
verbal behavior with relational frame theory.

Study Questions
1. Provide an example of an echoic.
2. What is the difference between a codic and a duplic?
3. Briefly outline two procedures to teach echoic responding.
4. What is a mand? Distinguish between a selection-based and
topography-based mand.
5. Describe two strategies to teach mands.
6. What is a tact? How is a tact distinguished from a mand?
7. How is an echoic-to-tact transfer procedure used to teach tacts?
 8. Provide an example of the functional independence of mands and
tacts.
9. Use the same example to describe the functional interdependence of
mands and tacts.
10. What is a derived mand? Give an example.
11. What does it mean to establish flexible responding?
 Chapter 4:

Complex Verbal Behavior

Yors Garcia
The Chicago School of Professional Psychology
Rocío Rosales
University of Massachusetts Lowell
Sebastian Garcia-Zambrano
Southern Illinois University
Ruth Anne Rehfeldt
Southern Illinois University

Link and Preview

The previous chapter provided an overview of the conceptual
importance and empirical support for three of the elementary
verbal operants defined by Skinner: echoics, mands, and tacts.
Beyond those elementary verbal operants, Skinner defined more
complex forms of verbal behavior, including, for example, the
intraverbal, various extensions of the tact, the speaker serving as
his or her own listener, and autoclitics, to which he devoted a full-
length chapter. These complex forms of verbal behavior will be the
focus of the current chapter.

Skinner’s Verbal Behavior (1957) was a theoretically rich and

comprehensive account of language from a behavioral perspective. It is
evident that Skinner’s analysis was not limited to the elementary verbal
operants, yet much of the empirical work on verbal behavior has been
largely focused on these operants (de Souza, Akers, & Fisher, 2017; Dixon,
Small, & Rosales, 2007; Dymond, O’Hara, Whelan, & O’Donovan, 2006;
Petursdottir & Devine, 2017). Two recent literature reviews specifically
focused on the simple intraverbal (Aguirre, Valentino, & LeBlanc, 2016)
and multiply controlled intraverbals (Stauch, LaLonde, Plavnick, Bak, &
Gatewood, 2017), showing a steady increase in research on this particular
verbal operant.
One possible explanation for the dearth of applied research on the more
complex verbal operants is that these operants typically involve multiple
sources of control and therefore cannot be addressed until the relevant
prerequisite skills are established (Eikeseth & Smith, 2013; Michael, Palmer,
& Sundberg, 2011; Sundberg, 2016). In addition, language assessments that
help guide treatment planning have not always been helpful in guiding
language intervention on more complex repertoires (Gould, Dixon,
Nadjowski, Smith, & Tarbox, 2011), although a more recently developed
behavioral language assessment has begun to address this gap (Dixon, 2014a,
2014b, 2015). Applied researchers and practitioners alike must not lose sight
of the importance of programming for more complex forms of verbal
behavior. Learners with autism spectrum disorder (ASD) and related
disabilities require individually tailored and comprehensive treatment
programming to meet their unique needs. Complex forms of verbal behavior
are related to important skills such as academic performance (Greer, Yaun, &
Gautreaux, 2005), problem solving, (Sautter, LeBlanc, Jay, Goldsmith, &
Carr, 2011), conversation, (Beaulieu, Hanley, & Santiago, 2014), and
obtaining employment (O’Neill & Rehfeldt, 2017).
In this chapter, we will define and review the empirical literature on
more complex forms of verbal behavior, beginning with the intraverbal and
ending with dictation taking, copying text, and speaker-as-own-listener
behavior (chapter 5 focuses on the speaker-as-own-listener). There is ample
empirical support for a variety of transfer of stimulus control (ToSC)
procedures that result in the acquisition of intraverbals, and research is
emerging on techniques that result in derived or untaught intraverbal
behavior for learners with and without developmental disabilities (Aguirre et
al., 2016; de Souza et al., 2017; Grannan & Rehfeldt, 2012; May, Hawkins,
& Dymond, 2013). Fewer studies, however, have evaluated methods to
improve intraverbal behavior related to the conversational speech of typically
developing adults (de Souza et al., 2017; Petursdottir & Devine, 2017).
Likewise, research on intraverbal behavior has expanded into areas such as
simple and conditional discrimination processes, compound stimulus control,
divergent and convergent stimulus control, and reverse and derived
intraverbals. In the next section we will review some of this research and
provide recommendations for clinicians.

Research on Complex Verbal Operants

A number of conceptually interesting and clinically important areas of
research may fall within the purview of “complex verbal operants,”
including the intraverbal, dictation taking and copying text, and speaker-as-
own-listener. We begin with the intraverbal.

Intraverbal
Skinner (1957) defined the intraverbal as a verbal operant that is under
the stimulus control of a preceding verbal stimulus, without point-to-point
correspondence, and is maintained by generalized conditioned reinforcement.
For instance, a child says “Alex” when asked, “What is your name?” and this
response is followed by praise or acknowledgment by an adult or peer. In this
example, there is no topographical point-to-point correspondence between
the SD “What is your name?” and the response product “Alex.” Intraverbals
range from simple chains of verbal stimuli (e.g., 1, 2…, A, B…) to fill-in-
the-blank responses (e.g., “You buy things with ?”),
answering questions (e.g., “How old are you?”), and categorization (e.g.,
“Tell me some mammals?”).
Palmer (2016) called attention to two different types of intraverbals,
intraverbal control and intraverbal operant. In the case of the intraverbal
operant, a single verbal stimulus (written, spoken, gesture) evokes an
intraverbal response “as the result of a history of reinforcement for emitting
that response in the presence of that stimulus” (p. 97). We observe multiple
examples of intraverbal operants when we respond to specific questions
presented to us (e.g., “What’s your phone number?” “Where do you live?”).
Nevertheless, as Palmer (2016) remarks, intraverbal responses are sometimes
under multiple sources of stimulus control (i.e., intraverbal control), for
example, when verbal stimuli interact with supplemental stimuli to facilitate
problem solving (e.g., self-prompts, auditory imaging; Aguirre & Rehfeldt,
2015; Kisamore, Carr, & LeBlanc, 2011; Mellor, Barnes, & Rehfeldt, 2015).
A number of instructional methods for teaching intraverbal responses
have been empirically validated. One of the most widely used procedures is
ToSC. In this procedure, the instructor concurrently presents a stimulus
(prompt) that already exerts some control over the target behavior along with
the target verbal stimulus and then systematically removes (fades) the
stimulus until the learner produces the target response independently.
Multiple studies have used echoic (e.g., saying the correct response vocally
and providing an opportunity for the participant to repeat the correct
response; Watkins, Pack-Teixeira, & Howard, 1989), textual (e.g., showing
the correct response as a written word; Krantz & McClannahan, 1993), tact
(e.g., showing the correct response as a visual or auditory stimulus;
Goldsmith, LeBlanc, & Sautter, 2007), and echoic prompts combined with
error correction (Kodak, Fuchtman, & Paden, 2012) to increase precise
stimulus control over intraverbal responses.
Other teaching methods have also demonstrated efficacy for establishing
intraverbal repertoires. These include instructive feedback (IF; Carroll &
Kodak, 2015), multiple exemplar instruction (MEI; Lechago, Carr,
Kisamore, & Grow, 2015), peer-mediated interventions (PMIs; Beaulieu et
al., 2014 ), compound tact instruction (Devine, Carp, Hiett, & Petursdottir,
2016), stimulus pairing (Byrne, Rehfeldt, & Aguirre, 2014; Vallinger-
Brown & Rosales, 2014), lag schedules of reinforcement (Contreras &
Betz, 2016), chaining procedures (Valentino, Conine, Delfs, & Furlow,
2015), differential observing responses (DORs; Kisamore, Karsten, &
Mann, 2016), distributed trials (Haq & Kodak, 2015), and precision
teaching (Cihon et al., 2017). Furthermore, Aguirre and colleagues (2016)
reported that research has expanded into derived intraverbals that are the
product of a different operant or conditional discrimination instruction (e.g.,
Grannan & Rehfeldt, 2012; May et al., 2013) and from the instruction of
other intraverbals (e.g., Dickes & Kodak, 2015). Research has shown that
intraverbal responses may be under simple stimulus control (“jump,” “1,
2,…”), conditional discrimination (“If your name is Charlie, say your
ABCs!”), and compound stimulus control (“get up” and “jump”; Axe,
2008; Eikeseth & Smith, 2013; Sundberg, 2016). In the coming sections we
will review some of the conceptual and empirical work that has been recently
conducted on simple and multiple controlled verbal operants. In addition, we
will present some of the procedures that have been more effective in teaching
direct and derived intraverbals in individuals with ASD and related
disabilities, and offer some clinical recommendations.

SIMPLE VERBAL DISCRIMINATION

Skinner (1957) defined a verbal stimulus as a product of prior verbal
behavior. For example, when an individual sends a text message to someone
else saying, “How are you?” this stimulus product may evoke in the listener
either a written verbal response product (e.g., typing “good”) or a vocal
response product (leaving a voicemail message; Sundberg, 2016). Verbal
stimuli may also control private verbal responses, as when the speaker-as-
own-listener engages in thinking (e.g., I have to pay my bills). Verbal stimuli
may control simple discriminations; for example, the written word “STOP”
on a traffic sign evokes an observing response (looking to both sides of the
street before crossing it). In the case of learners with ASD, lacking simple
verbal discrimination skills will impede the acquisition of more advanced
intraverbals. For instance, intraverbal questions such filling in the blank (e.g.,
“A, B…,” “1, 2…”) or animal sounds (e.g., “A dog says ?”)
require learners to discriminate between “C” and “D” or “ruff-ruff” and
“meow,” respectively. In particular, “A, B” is the SD and “C” the intraverbal
response. If a participant fails to respond “C,” one potential solution would
be to teach both “C” and “D” under simple discrimination format until the
learner is responding reliably (Eikeseth & Smith, 2013; Sundberg, 2016).
ToSC is one of the most effective procedures for teaching simple verbal
discriminations in children with ASD (e.g., Finkel & Williams, 2001;
Ingvarsson & Hollobaugh, 2011). For instance, Watkins et al. (1989) taught
simple intraverbal responses to three individuals with intellectual disabilities
using echoic prompts and a simple verbal stimulus (e.g., “name a color,”
“name a size,” or “name a texture”). Similarly, Finkel and Williams (2001)
used textual and echoic prompts to facilitate intraverbal responses in a child
with ASD. In this study, the child’s stimulus control was enhanced through
echoic prompts (e.g., “I like to eat muffins”). Results regarding the efficiency
of each prompting strategy are mixed. Some studies report that a history of
reinforcement (Coon & Miguel, 2012), visual imaging occasioned by the
presence of pictures (Ingvarsson & Hollobaugh, 2011), and preexisting
repertoires and individual preferences (Vedora & Conant, 2015) may impact
the effectiveness of each form of instruction.
 Sundberg and Sundberg (2011) noted that some of the problems related
to the acquisition of intraverbals stem from the complexity of the verbal
discriminative stimuli and the lack of generalized tact, mand, and listener
skills established in the learner’s repertoire. Although the research on
intraverbal prerequisite skills is lacking (Axe, 2008), some studies suggest
that simple discrimination skills may facilitate the acquisition of simple and
complex intraverbals (Guerrero, Alós, & Moriana, 2015; Stauch et al., 2017).
In this case, starting with simple nonverbal discrimination and instruction
following, for example, might involve the clinician’s presenting two blocks
together, where one stimulus, “green,” is correlated with reinforcement
(SR+), and the other stimulus, “red,” is placed under extinction (i.e., S-delta,
or SΔ). The clinician repeats the same procedure with spoken instructions
(e.g., to clap hands upon hearing the instruction “clap your hands”). Once
these skills are in the participant’s repertoire, simple intraverbals may be
taught (e.g., “A kitty says ,” “One, two…”, “How old are
you?”; Sundberg, 2016). We recommend that clinicians assess prerequisite
skills such as simple visual discriminations, echoic responses, and learning
history prior to conducting intraverbal instruction.

VERBAL CONDITIONAL DISCRIMINATION AND COMPOUND

STIMULUS CONTROL
Advanced intraverbal responses are under conditional discrimination
and compound stimulus control (Eikeseth & Smith, 2013; Sundberg, 2016).
This type of antecedent control may further be subsumed into what has been
named divergent and convergent control (see Michael et al., 2011). In
divergent control, a single discriminative stimulus (SD) controls different
response topographies (e.g., hearing “car” may evoke different responses,
such as “drive,” “wheels,” “vehicle,” or “auto”; see Feng, Chou, & Lee,
2017; Lee, Chou, & Feng, 2017). The opposite occurs in convergent
control; that is, multiple discriminative stimuli contribute to the strength of a
single response topography (e.g., “the person who does your nails is a
”). In this last example, the response “manicurist” is under
the control of two stimuli: “person” and “nails” (see Devine et al., 2016;
Haggar, Ingvarsson, & Braun, 2018). Two specific verbal stimuli are under
convergent control, verbal conditional discrimination (VCD) and a
compound stimulus.
 These stimulus control procedures have one significant difference. In
VCD, one verbal stimulus affects the evocative function of another verbal
stimulus (e.g., “What do you wear that is brown?”). In this example, the
verbal stimulus “wear” alters the function of “brown”; that is, saying “shoes”
is under the control of two different verbal stimuli. On the contrary, under
compound stimulus control, two independent verbal stimuli are combined to
evoke a particular response, for instance, “talk slow” and “talk fast,” “eat
slow” and “eat fast” (see Eikeseth & Smith, 2013). Although a compound
stimulus can occur during conditional discrimination instruction, in
compound stimuli one stimulus does not determine the function of the other
stimulus.
Prior to teaching complex intraverbals, practitioners should assess
appropriate conditional stimulus control. This means that a correct vocal
response (e.g., “shoes”) must be evoked by the presence of the verbal
stimulus (e.g., “brown”). If other responses are evoked (e.g., “sweater,”
“watch”), further simple verbal discrimination teaching may be needed to
bring both verbal stimuli under conditional control (Eikeseth & Smith, 2013;
Kisamore et al., 2016; Sundberg & Sundberg, 2011).

PREREQUISITES AND INTRAVERBAL ASSESSMENT

Little research has been conducted on the identification of the specific
prerequisite skills learners must acquire prior to intraverbal instruction.
Nonetheless, Axe (2008) and Sundberg and Sundberg (2011) provide some
general guidelines. First, one should ensure that the learner has reliable
mand, echoic, tact, and listener repertoires prior to beginning intraverbal
instruction (e.g., Vedora & Conant, 2015). Second, the instructor should be
sure that simple discrimination skills are in the child’s repertoire before
teaching basic intraverbal questions such as fill in the blank and animal
sounds. Third, teaching advanced intraverbal questions (e.g., “What do you
wear that is green?”) requires conditional discrimination and compound
stimulus control; therefore, one should teach these skills first in case the
repertoire is absent. Prerequisite skills that facilitate conditional
discriminations are simple and compound tacting and sorting pictures (Axe,
2008; Devine et al., 2016; Ribeiro, Miguel, & Goyos, 2015). Fourth, a
practitioner should ensure that the learner has a basic listener repertoire of
responding by feature (i.e., parts of items, descriptions), function (i.e., use or
purpose), or class (i.e., category). For instance, the learner will point to a
picture of a cookie when asked to identify it by its features (“round,”
“crunchy”), function (“you eat it”), and class (“food” or “snack”). Fifth, one
should use available verbal behavior curricula to assess current verbal skills
(Dixon, 2014a, 2014b, 2015; Partington, 2008; Sundberg, 2008). With these
guidelines in mind, we now turn to intraverbal instruction.

INSTRUCTIONAL PROCEDURES
In this section, we describe the different instructional procedures to
teach intraverbals under divergent and convergent control. We also present
current research on establishing derived intraverbals along with some
applications for teaching more advanced social skills.

Divergent control and intraverbals. In divergent control, as noted earlier,

a single verbal SD occasions diverse intraverbal responses (e.g., intraverbal
categorization). Some of the most effective procedures to promote
intraverbals under simple verbal control are IF, lag schedules of
reinforcement, and ToSC. Instructive feedback typically involves adding
additional nontarget stimuli to the antecedent or consequence events of
direct instructional trials (Carroll & Kodak, 2015). In each trial, the
instructor presents a target instruction, provides an opportunity to respond,
delivers the appropriate consequence, and presents an additional
nontargeted IF stimulus. The learner is not required to respond to the IF
stimulus, and if they do, the instructor does not deliver consequences. For
example, the instructor may present an intraverbal question “What are some
fruits?” When the learner says, “apple,” the instructor delivers praise and an
IF stimulus such as “Right! Banana is a fruit too.” In this case, banana is the
IF stimulus. Multiple responses may be added to the IF so that divergent
control may be established.
Several studies have demonstrated the effect of IF in the establishment
of intraverbal responses. For example, Carroll and Kodak (2015) found that
using prompt delay with IF increased response variability when intraverbal
categories were taught. Tullis, Frampton, Delfs, and Shillingsburg (2017)
found similar effects when IF was implemented to teach problem-solving
skills. In this study, the instructor presented a verbal question to the
participant (“What is the problem?”), provided reinforcement for correct
problem selection (e.g., “shoe with no laces”), and added IF in the form of
problem explanation to the consequence (i.e., “That’s right. This is a problem
because the laces are missing so you cannot tie the shoe”). One participant
acquired the secondary target without explicit teaching, whereas the other
two participants required additional instruction for the problem explanation
section. Haq, Zemantic, Kodak, LeBlanc, and Ruppert (2017) reported that
the efficacy of IF may depend on the learner’s repertoire before instruction
begins, such as attending and echoic behavior. Thus, during instruction,
prompting attention to the target stimulus and using vocal prompts may
facilitate the acquisition of novel intraverbal responses (Carroll & Kodak,
2015; Haq et al., 2017).
Implementing lag schedules of reinforcement is a second alternative for
fostering divergent intraverbals. Lag schedules differentially reinforce
intraverbal variability by making reinforcement available for responses that
differ from preceding responses based on the value of the lag. For example,
on a Lag 2 schedule, a response is reinforced if it differs from the
immediately preceding two responses (e.g., saying, “I like apple” and “I like
ice cream” when asked, “What do you like to eat?”). In a recent study by
Contreras and Betz (2016), individuals with ASD were taught to provide
variable responses to questions using a lag schedule of reinforcement. For
example, participants were exposed to Lag 1 and Lag 3 (e.g., “Tell me things
that fly” with potential responses under Lag 3 being “flies,” “birds,” and
“eagles”). The results showed an increase in varied responding in two
participants, while the other participants required additional variability
instruction. The results are promising in terms of increasing varied vocal
responding using different lag schedules of reinforcement (Lee & Sturmey,
2014; O’Neill & Rehfeldt, 2017; Wiskow, Matter, & Donaldson, 2018).
Lastly, ToSC procedures have also facilitated the acquisition of
divergent intraverbals. Diverse procedures such as tact prompts, errorless
learning, schedule thinning, prompt delay, and intraverbal prompting
have yielded positive results (Feng et al., 2017; Goldsmith et al., 2007; Lee et
al., 2017).
In sum, to increase intraverbal variability (i.e., divergent control), start
with simple verbal discrimination instruction using ToSC procedures. If
prompt dependency is a problem during instruction, use an alternative prompt
method (e.g., Ingvarsson & Hollobaugh, 2011; Mueller, Palkovic, &
Maynard, 2007) while also implementing more efficient error-correction
procedures (Kodak et al., 2016). We also recommend inserting IF trials early
during the instruction and probing for novel intraverbals. Finally, including
lag schedules of reinforcement during instruction to maximize intraverbal
categorization and beginning with a simple Lag 1 prior to requiring more
response variability (e.g., Lag 3) may also be effective.

Convergent control and intraverbals. In convergent control, multiple

antecedent variables influence the strength of a single response (Michael et
al., 2011). For instance, if a learner says “parrot” in response to “Tell me a
green animal,” in this case, the response “parrot” is under the control of two
stimuli, “green” and “animal” (i.e., conditional discrimination). Several
studies have used methods such as conditional discrimination instruction
(e.g., Guerrero et al., 2015; Kisamore et al., 2016; Pérez-González &
Alonso-Álvarez, 2008), multiple tact instruction (Belloso-Díaz & Pérez-
González, 2015; Devine et al., 2016; Mellor et al., 2015), and intraverbal
chains (Valentino et al., 2015) to teach complex intraverbal responses to
individuals with developmental disabilities; we discuss these below.

Conditional discrimination. The purpose of this strategy is to establish

discriminative control under additional stimuli. Take for example two
complex questions that, though relatively similar, are difficult to
discriminate for some learners with ASD: “What do you eat?” and “What
do you eat with?” In this case, each question leads to a different accurate
response, “chicken” and “fork,” respectively. However, for learners with
ASD who may lack complex conditional discrimination skills, these
answers will take numerous trials to master, unless they are taught under
specific stimulus control. In this case, the correct response would have to
come under differential stimulus control of the words “eat” and “with.”
Similarly, asking participants to repeat the words before providing the
intraverbal response may facilitate auditory discrimination in multiple
controlled intraverbals.
Kisamore et al. (2016) used conditional discrimination teaching
procedures and the requirement of a DOR to teach multiply controlled
intraverbals to individuals with ASD. For instance, when the experimenter
asked “What’s a fruit that’s green?” the participant was required to repeat
part of the question “fruit green.” If the participant did not repeat the verbal
stimulus, the experimenter presented a vocal prompt (e.g., “Say, fruit green”).
Results showed that conditional discrimination instruction with the DOR was
more effective than the prompt-delay procedure with error correction in
establishing convergent intraverbals in four of the seven participants.
 Blocked-trial procedures have also been effective in establishing
intraverbal relations to promote complex intraverbal relations under
conditional discrimination control (Haggar et al., 2018; Ingvarsson, Kramer,
Carp, Pétursdóttir, & Macias, 2016). In this procedure, the number of trials is
gradually reduced contingent upon specific criteria, and the procedure ends
with sample stimuli randomly interspersed. Ingvarsson et al. (2016) used
constant prompt delay and error correction to teach responses to intraverbal
questions through a blocked-trial procedure. Question pairs were selected
that took the form “What do you ?” versus “What do you
with?” A total of eight question pairs were included for
each participant. Presentation of each question was randomly alternated until
ten consecutive answers occurred. Contingent on meeting mastery criterion
(i.e., four consecutive trial blocks with no more than two errors), participants
proceeded to the next phase. In the subsequent phases, the number of correct
consecutive answers and accuracy criterion were lowered. In the last two
phases of the teaching procedure, questions were counterbalanced until
fifteen consecutive correct answers occurred. In the last step, two questions
were presented in a quasi-random fashion. This procedure was effective in
establishing novel intraverbal responses under conditional discrimination
control, although two participants required additional error correction to
acquire the first discrimination. Additional studies found that adding
criterion-level probes after each step of the blocked-trial procedure increased
accurate responding (Haggar et al., 2018).
Practically speaking, these studies show that teaching learners to
discriminate components of the intraverbal questions using DOR and
blocked-trial procedures increases the occurrence of intraverbal responses.
Further research is warranted to study some of the behavioral mechanisms
(e.g., blocking effect, salience of SD, density of reinforcement, auditory
stimulus control, visual imaging) that could explain the effectiveness of these
procedures (Haggar et al., 2018; Ingvarsson et al., 2016; Kisamore et al.,
2016). Future studies should also compare the blocked-trial procedure to
other potentially more efficient teaching procedures (e.g., varied-trial
instruction; Cariveau, Kodak, & Campbell, 2016).

Multiple tact instruction. In this procedure participants are required to tact

the name of the stimulus while providing the category of the item (e.g.,
Belloso-Díaz & Pérez-González, 2015; Devine et al., 2016; Grannan &
Rehfeldt, 2012). For instance, the instructor presents a target stimulus (e.g.,
picture of dog) to the participant and asks, “What is this?” Then, after the
participant emits a correct response, the instructor asks, “What else is it?”
The instructor then conducts intraverbal probes (e.g., “Tell me some
animals”). Most of the empirical literature shows that intraverbals do not
emerge after multiple tact instruction, though performance improves upon
implementing ToSC procedures (Aguirre et al., 2016). On the contrary,
recent studies report that using compound tact instruction—that is,
presenting two stimuli simultaneously (e.g., “dog” and “cat”)—enhances
the emission of intraverbal responses (Belloso-Díaz & Pérez-González,
2015; Devine et al., 2016). Although results are positive, further research
needs to clarify the different variables that affect the emergence of complex
intraverbals under the control of nonverbal stimuli.
An additional strategy that has shown effective results in acquiring
multiple controlled intraverbals is using mediating responses (e.g., visual
imaging instruction; Aguirre & Rehfeldt, 2015; Kisamore et al., 2011; Mellor
et al., 2015; Sautter et al., 2011). Such responses also allow the individual to
manipulate variables covertly and to “prompt and probe his own behavior”
(Skinner, 1957, p. 442) to increase the probability of a solution. For example,
when someone asks, “What does your school look like?” the response is
partially prompted by precurrent responses that involve private visual stimuli
(e.g., imagining classrooms and buildings) and the question asked. In
particular, Mellor et al. (2015) reported that intraverbals (e.g., saying
“camera” when asked, “What makes the sound of a click?”) and
categorization intraverbals (e.g., “What are some sounds that you know? Tell
me as many as you can”) emerged after multiple tact instruction (e.g., saying
“camera” when presented with the recording of a camera clicking) and
auditory imagining (e.g., “I know that a camera goes [camera clicking]; oh
right, a camera goes click”). Results showed that multiple tacts and
imagining were sufficient to yield criterion-level intraverbals in three and
two of four participants, respectively. However, two of four participants
needed additional ToSC instruction. Although these studies have
demonstrated effective results in neurotypical individuals, further research
needs to be conducted in individuals with ASD and related disabilities.

Reverse intraverbals. Intraverbals can be derived from other intraverbals,

as is the case with reverse intraverbals. For example, teaching an
intraverbal response “What is the opposite of tall?” (“short”) may also
promote the reverse intraverbal (“What is the opposite of short?” “tall”). A
number of studies have failed to produce the emergence of reverse
intraverbals in neurotypical individuals and individuals with developmental
disabilities (Pérez-González, Garcia-Asenjo, Williams, & Carnerero, 2007;
Petursdottir, Carr, Lechago, & Almason, 2008). Positive outcomes were
obtained only after participants were exposed to MEI (Pérez-González et
al., 2007).
One recent study found that teaching original intraverbals produced the
emergence of reverse intraverbals in some participants (Dickes & Kodak,
2015). In this study, the researchers taught compound stimuli across
categories (e.g., object/function relations, opposites, and animal sounds)
using a multiple baseline design across categories (e.g., saying “ooo-aaahh”
when asked, “A monkey says?”). In the probe sessions, the researchers
assessed directly taught and reverse intraverbals (e.g., “Ooo-aaahh says the
?”). Teaching direct intraverbals produced emergence of
reverse intraverbals in two out of three participants; however, results were
variable when reverse intraverbals were taught directly. In other words, direct
intraverbal instruction was sufficient to produce emergence of reverse
intraverbals. As the authors noted, results should be interpreted with caution
due to problems with experimental control; the researchers did not introduce
the independent variable in all categories. Future research should explore the
role of prior histories with verbal conditional discrimination in establishing
reverse intraverbals.

Derived intraverbals. Recent studies have demonstrated that intraverbal

responding may also emerge from other basic verbal operants in the
absence of direct instruction (Allan, Vladescu, Kisamore, Reeve, & Sidener,
2015; Daar, Negrelli, & Dixon, 2015; Dixon, Belisle, Munoz, Stanley, &
Rowsey, 2017; Dixon, Belisle, Stanley, Speelman, et al., 2017; Grannan &
Rehfeldt, 2012; May et al., 2013; Shillingsburg, Frampton, Cleveland, &
Cariveau, 2018). For instance, a child may be taught a tact relation (e.g.,
“What is the name of this animal?” [dog]), followed by derived intraverbal
probes (e.g., “What is a dog?”). Unlike the other instructional protocols
discussed previously, generalized contextually controlled patterns of
responding mediate the emergence of derived intraverbals (see Stewart,
McElwee, & Ming, 2013a, for further details). This is worth mentioning
because the behavioral mechanisms responsible for derived intraverbals
may increase the likelihood of novel verbal operants emerging in the
absence of direct instruction. Due to limited space, we describe only some
of the most relevant findings.
In one study, Grannan and Rehfeldt (2012) demonstrated the emergence
of intraverbal categorization through direct tact instruction. Specifically,
participants were taught to tact categories when presented with a picture
(e.g., the instructor presented a picture of vehicles and asked, “What are
some things that take you places?” [cars, planes]), followed by visual-visual
conditional discrimination (e.g., matching pictures of vehicles). Results
showed that derived intraverbal responses emerged in all participants (e.g.,
saying “car” when asked, “What are four things that take you places?”). In a
similar study, May et al. (2013) also found that derived intraverbals emerge
after tacts are directly taught. When derived intraverbals do not emerge,
adding listener instruction (Smith et al., 2016), MEI (e.g., Greer, Yaun, &
Gautreaux, 2005; Pérez-González et al., 2007) or stimulus-pairing
procedures (Byrne et al., 2014; Vallinger-Brown & Rosales, 2014), or
overlapping SDs between directly taught and emergent relations (e.g.,
Shillingsburg et al., 2018), may facilitate further emergence of untaught
intraverbals.
Additional studies have shown that conditional discrimination
procedures may be used to establish derived intraverbals such as novel wh-
questions (e.g., who, where, what; Daar et al., 2015), metaphorical
intraverbal responses about feelings (Dixon, Belisle, Munoz, et al., 2017),
and derived intraverbal categorization (Dixon, Belisle, Stanley, Steelman, et
al., 2017). For instance, Dixon, Belisle, Munoz, and colleagues (2017)
showed that teaching metaphorical tacts (e.g., “If you felt like this [picture of
a tornado], how would you feel?” [sad, mad, or upset]) promoted the
emergence of metaphorical intraverbals (e.g., the instructor asked, “[Name]
was told to clean up his/her toys and feels like a tornado. How might [name]
feel?”). In this example, the correct response would be “mad.” Results of the
study showed that all three children answered metaphorical questions during
the first instructional phase; however, during the second phase when novel
stimulus sets were presented, one participant required a modification (i.e.,
reduced stimulus array). Taken together, these studies show that advanced
intraverbals may emerge from tact instruction under conditional control (i.e.,
picture plus auditory stimulus). Although most of the studies on this topic
have reported positive results, some have shown that it is important to
include alternative procedures to enhance derived responding in children with
ASD (see Shillingsburg et al., 2018). In the next section, we will explore one
additional procedure that may facilitate more advanced intraverbal
repertoires such as that involved in social communication.

Peer-mediated interventions. PMIs consist of teaching typically

developing peers how to model and prompt targeted social behaviors for
their peers with a diagnosis of ASD or related disabilities. Research on the
evaluation of social skills intervention has highlighted PMIs as one of the
most promising methods to improve this target behavior in this population
(Chang & Lock, 2016). Krantz, Land, and McClannahan (1989)
demonstrated the efficacy of PMIs with three adolescents diagnosed with
ASD serving as peer prompters for three other peers with ASD. The
procedure consisted of a peer’s delivering a prompt to talk about a topic of
interest (e.g., sports). Generalization probes showed similar levels of
performance with peers not involved in the study and in settings outside of
the teaching environment, and also showed maintenance of the behavior
once the peer prompter was withdrawn.
More recently, Beaulieu et al. (2014) evaluated a PMI behavioral skills
training (BST) procedure to improve the conversational skills of a college
student with a learning disability. The focus of the study was not isolated to
intraverbal responding, but rather included several additional indicators of
advanced conversation skills, including time spent as listener versus speaker
and number of interruptions. Peers were taught to provide instructions,
modeling, role-play, and feedback to the participant during all teaching
sessions. For interruptions, they delivered feedback visually at first, and then
vocally, if necessary. Following the introduction of BST, the participant
emitted fewer interruptions, increased the time spent as a listener during the
sessions, and showed a decrease in the duration of high-specificity content
(defined as an excess of details provided on a topic). The effects of the
intervention were maintained with naïve peers and compared favorably with
normative data on the conversational skills of three undergraduates without
disabilities.

Dictation Taking, Textual Behavior, and

Copying Text
 Three additional complex verbal operants were described in Verbal
Behavior (Skinner, 1957), including dictation taking, textual behavior, and
copying text. Dictation taking corresponds to what is known as writing;
Skinner defined dictation taking as a written response controlled by a vocal
verbal stimulus (e.g., writing a sentence as someone says it). Textual
behavior is vocal responding controlled by a visual verbal stimulus (e.g.,
reading the newspaper). Both dictation taking and textual behavior involve
point-to-point correspondence and no formal similarity between the verbal
SD and the response product. In point-to-point correspondence, all
components of the response match the components of the SD, for instance,
writing “apple” upon hearing the word “apple,” or saying “apple” when
seeing the written word “apple.” In both cases each letter of the SD matches
each letter of the verbal response “apple,” and there is no formal similarity
between the SD and response product; both are in a different modality (e.g.,
SD is a written stimulus and the response is vocal). These operants are
among some of the most advanced verbal skills and are underresearched
topics in verbal behavior (Dymond et al., 2006). In this section, we will
briefly review some of the procedures to promote the development of these
complex verbal operants and provide some clinical recommendations for
practitioners.
 Some studies have used simple stimulus-control procedures (Asaro-
Saddler, 2016; Pennington, 2016; Pennington & Delano, 2012) and derived
relational responding technology (de Souza, de Rose, & Domeniconi, 2009;
de Souza & Rehfeldt, 2013; Greer, Yaun, & Gautreaux, 2005) to teach
written (i.e., dictation-taking) and spoken spelling skills in individuals with
ASD (Aguirre & Rehfeldt, 2015). For example, de Souza et al. (2009)
developed a research program that starts with teaching simple words with
consonant-vowel sequences (e.g., “bad,” “fan,” “nap”), followed by teaching
complex sequences of words (e.g., “bread, chair, snake”) and then integrating
reading (textual behavior) and writing skills in a network of derived relations
to assess reading comprehension. Additional studies have found that teaching
dictation (de Souza & Rehfeldt, 2013), teaching written and spoken spelling
skills (Greer, Yaun, & Gautreaux, 2005), and promoting collateral responses
(Aguirre & Rehfeldt, 2015) enhance the occurrence of untaught spoken and
written skills in individuals with intellectual disabilities. In addition, de
Souza and Rehfeldt (2013) evaluated the effect of conditional discrimination
instruction (i.e., matching printed words to their printed synonym) on the
emergence of untaught vocal spelling. Aguirre and Rehfeldt (2015) examined
the effects of visual imaging and instruction on correct written spelling
responses in three adolescents with learning disabilities. In the instructional
phase, the examiners presented participants with a flash card (i.e., “The word
is ”), which they then removed. They instructed participants
to imagine the word for three seconds and later asked them to write the word.
An error correction plus reinforcement procedure was added in case
participants did not meet mastery criterion of written spelling. Only one
participant met the mastery criterion during the visual imaging only
condition, and the other two participants required error correction plus
reinforcement.
Miguel, Yang, Finn, and Ahearn (2009) taught children with ASD to
select pictures and printed words when given their dictated name via match-
to-sample. Then, without further instruction, all participants demonstrated
emergent textual behavior (i.e., reading aloud written words). Sprinkle and
Miguel (2013) obtained similar results. These studies show that textual
behavior may be established indirectly through instructional procedures that
promote derived stimulus relations between printed words, pictures, and
dictated names. These findings are particularly relevant for practitioners who
wish to minimize instructional time. Likewise, constructed-response match-
to-sample tasks have been implemented to teach vocal and copying skills to
students with intellectual disabilities (Hanna, de Souza, Rose, & Fonseca,
2004). In this procedure, an instructor presents a pool of letters (e.g., “C,”
“B,” “A,” D,” “R”) simultaneously with the sample stimulus (e.g., written
word “CAR”) to the participant, and then asks the participant to construct a
word identical to the sample. If the sample stimulus is a printed word, the
task is essentially one of copying. In contrast, if the sample is a spoken word,
then the participant is required to construct and spell the correct word.
Additional procedures such as BST (Pennington, Foreman, & Gurney, 2018),
video modeling (Marcus & Wilder, 2009), praising qualifying autoclitics
(Hübner, Austin, & Miguel, 2008), and ToSC (Fabrizio & Pahl, 2007; Haq &
Kodak, 2015) have also been implemented to teach textual responses.
Taken together, these studies show that arranging for derived stimulus
relations, using MEI, and establishing mediating behavior such as visual
imaging may increase the emergence of untaught textual (Miguel et al.,
2009), dictation-taking (de Souza & Rehfeldt, 2013), and written and spoken
spelling skills (Aguirre & Rehfeldt, 2015; Greer, Yaun, & Gautreaux, 2005).

The Speaker-as-Own-Listener
Thus far, we have reviewed how separate speaker (e.g., mand, tact,
intraverbal) and listener (pointing, attending, following instructions)
repertoires may be established. Skinner (1957) described several repertoires,
including thinking, problem solving, self-editing, and self-awareness, that
emerge when the speaker mediates their own responding as a listener.
Speaker-as-own-listener is an advanced repertoire that allows an individual to
achieve significant independence and engage in complex social skills (e.g.,
writing, reading, perspective taking; Greer & Speckman, 2009; Spradlin &
Brady, 2008). For example, when someone is reading a text, they not only
read the written words but also listen to what is read. Reading consists of a
speaker-listener relationship under the control of print stimuli or pictures.
Similarly, when an individual is writing, the speaker and listener within the
same skin are under the control of the auditory and textual stimuli. Taking
perspective also requires that the individual discriminate their own point of
view where events take place and respond to them from the same perspective
(e.g., when asked, “How do you feel?”). We only mention this topic briefly
here as it is the focus of the next chapter.
Implications and Future Directions
The material presented here on complex verbal behavior has a number of
important implications for practitioners and researchers. First, it is not clear
what preexisting repertoires must be intact prior to initiating intraverbal
instruction, though some efforts have been made in this regard (Partington,
2008; Sundberg, 2008; Sundberg & Sundberg, 2011). More research is
warranted to determine the minimal number of tacts, echoics, and listener
skills necessary before introducing intraverbal instruction. One interesting
research line explored the use of psychometrically sound assessments to
evaluate verbal deficits in individuals with ASD (Dixon, 2014a, 2014b,
2015). In the meantime, we recommend that practitioners follow best
practice—that is, teach tacts, mands, and listener skills before proceeding to
intraverbal teaching (see Sundberg & Sundberg, 2011).
Second, results are mixed regarding the best prompting strategy to teach
intraverbals (e.g., Coon & Miguel, 2012; Ingvarsson & Hollobaugh, 2011;
Vedora & Conant, 2015). Therefore, we encourage practitioners to assess the
type of prompting method that is most suitable for the learner prior to
intraverbal instruction (see Seaver & Bourret, 2014, for an example of a
prompt assessment). Future studies might also include error analyses (see
Devine et al., 2016; Kisamore et al., 2016) and compare echoic and tact
prompts with and without error correction during intraverbal training (see de
Souza et al., 2017).
Third, research on multiply controlled intraverbals has been conducted
in recent years (e.g., Stauch et al., 2017). This area offers important
opportunities and challenges for practitioners and researchers alike. Palmer
(2016) made an important conceptual distinction between the traditional
definition of intraverbal (i.e., intraverbal operant) and a multiply controlled
intraverbal (i.e., intraverbal control). In the former case, studies have focused
on teaching simple intraverbals such as question answering, fill-in-the-blank
responding, and categorization skills using ToSC procedures (see de Souza et
al., 2017; Petursdottir & Devine, 2017). In these studies, a single verbal
stimulus was established as discriminative for the intraverbal response (e.g.,
“What is your name?”). Other studies have explored the role of collateral
stimuli and verbal conditional discriminations to teach intraverbal responses,
primarily in neurotypical individuals (e.g., Aguirre et al., 2016; Kisamore et
al., 2011; Mellor et al., 2015; Valentino et al., 2015). For instance, some
studies have demonstrated that using visual imaging may enhance problem-
solving (Kisamore et al., 2011) and written spelling skills (Aguirre &
Rehfeldt, 2015).
Similarly, other studies have used conditional discrimination procedures
to teach intraverbal responses, for instance, block-trial procedures (i.e., the
same sample stimulus is presented repeatedly across trials instead of
alternating each trial; Ingvarsson et al., 2016) and DOR (e.g., naming or
touching the sample stimulus before selecting the comparison stimulus;
Kisamore et al., 2011). Practically speaking, this means that practitioners
must teach simple discrimination repertoires prior to teaching basic (e.g., A,
B…) and complex (e.g., “What are some red fruits?”) intraverbal responses.
Then, they can teach advanced repertoires such as problem solving or covert
intraverbal behavior using supplementary stimuli such as visual imaging.
Although more research is needed in this area, currently available results are
promising.
Fourth, once simple and multiply controlled intraverbals are in the
participant’s repertoire, it is important to target intraverbal variability (de
Souza et al., 2017; Haq et al., 2017; Lee & Sturmey, 2014) and derived
intraverbals (Grannan & Rehfeldt, 2012; May et al., 2013). Although the
results have been very consistent showing that lag schedules and IF increase
intraverbal variability, there is still room for additional research in assessing
the controlling variables in these procedures. Similarly, research in derived
relational responding has demonstrated that more efficient protocols can be
implemented to teach derived intraverbals (Daar et al., 2015; Shillingsburg et
al., 2018).
Lastly, data on more elaborate repertoires such as dictation taking,
textual behavior, copying text, and speaker-as-own-listener are emerging in
the field. We recommend that clinicians assess some prerequisites (e.g.,
listener skills, fine-motor coordination, and self-regulatory skills such as self-
control) prior to teaching dictation-taking skills. Similarly, evidence-based
practices such as using speech-generating devices, self-management, visual
support (e.g., scripts, pictures), and peer-mediated instruction have been
found to be effective in facilitating the acquisition of writing skills in learners
with ASD (Asaro-Saddler, 2016; Pennington et al., 2018). Results of other
studies promote the use of derived relational responding protocols in teaching
written and spoken spelling repertoires (de Souza & Rehfeldt, 2013; Greer,
Yaun, & Gautreaux, 2005). Finally, teaching naming and say-do
correspondence repertoires facilitates the achievement of important
independent skills. Research in this area is well developed (see Greer &
Ross, 2008, and the following chapter for more details).
We conclude this chapter with some recommendations for future
research in complex verbal behavior. Research is lacking on intraverbal
responses focused upon social questions (e.g., “Who is your best friend?”);
most research has concentrated on teaching categorization, feature, function,
and class responses (de Souza et al., 2017; Stauch et al., 2017). Similarly,
there is limited research exploring intraverbals focused upon emotions
(Dixon, Beslise, Munoz, et al., 2017) and using peer-mediated interventions
with middle and high school students with ASD. Research should also
incorporate generalization and treatment fidelity measures (Chang & Locke,
2016). Furthermore, future research should investigate the role of different
problem-solving strategies (e.g., organizing and grouping stimuli, visual
imagery, observing one’s own environment, and covert intraverbal behavior)
on the acquisition of complex intraverbal categorization in individuals with
developmental disabilities (e.g., Aguirre et al., 2016; Mellor et al., 2015;
Petursdottir & Devine, 2017).
Likewise, future research should compare whether instructive feedback
or lag schedules of reinforcement is more effective in establishing response
variability for children with ASD (Carroll & Kodak, 2015; de Souza et al.,
2017). Also, evaluating the role of different types of lag schedules of
reinforcement (e.g., Lag 3) in intraverbal instruction (Lee & Sturmey, 2014)
and investigating the effects of different reinforcers on intraverbal variability
is in order.
Divergent and convergent control is another area that requires further
exploration. For instance, research on divergent control is needed to evaluate
the efficiency of different instructional methods (e.g., formal prompts,
instructional feedback; Feng et al., 2017; Lee et al., 2017). In convergent
control (e.g., conditional discrimination and compound stimulus control), a
need exists for further evaluation of the effectiveness of blocked-trial
procedures and DORs in establishing intraverbal responses. This might
involve evaluating generalization across people, environment, and different
responses using blocked-trial procedures and whether DOR alone or
combined with a blocked trial-procedure is more effective in establishing
convergent intraverbals (Haggar et al., 2018; Kisamore et al., 2016).
Likewise, future research should extend the use of derived relational
responding technology in establishing more complex intraverbal responses,
the role of covert verbal responses on overt responses such as academic tasks,
and the establishment of vocal spelling with different verbal repertoires (de
Souza & Rehfeldt, 2013; Grannan & Rehfeldt, 2012).
What started as a friendly challenge introduced to Skinner by the
philosopher Alfred North Whitehead, who said, “Let me see you account for
my behavior as I sit here saying ‘No black scorpion is falling upon this
table’” (Skinner, 1957, pp. 456–457), has turned into a vibrant research
program with effective applications, mostly in the area of developmental
disabilities (de Souza et al., 2017; Petursdottir & Devine, 2017). Sixty years
have passed since Skinner published his most important work. However,
buried in his book still lie many areas awaiting empirical testing, for instance,
the role of echoic behavior and supplementary stimulation in remembering,
metaphorical extensions of the tact, tacting private behavior, and multiple
causation (Schlinger, 2017). More research is still needed, particularly with
typically developing individuals. Skinner (1969) once mentioned,
“Behaviorism, as we know it, will eventually die—not because it is a failure,
but because it is a success. As a critical philosophy of science, it will
necessarily change as a science of behavior changes” (p. 267). The field of
verbal behavior will only prosper and survive by standing on the shoulder of
the giants, as was the quest of this chapter.

Conclusion
Skinner’s taxonomy of verbal behavior includes several operants that are
the foundation of a number of complex skills. Unlike the operants described
in the preceding chapter, many such operants are multiply controlled, and
for this reason a number of prerequisite skills may be established before
such operants are acquired. This chapter explored the processes involved in
intraverbal, dictation taking, textual, and copying text repertoires and
discussed research on strategies for teaching such repertoires in individuals
who may lack such skills.

Study Questions
 1. What is an intraverbal? Explain by distinguishing it from one of the
other verbal operants reviewed in this text so far.
2. Describe how the same word could be an echoic, mand, tact, or
intraverbal.
3. How have transfer of stimulus control procedures been used in
intraverbal instruction?
4. Give an example of both divergent and convergent control.
5. What are three prerequisites described in the section titled
“Prerequisites and Intraverbal Assessment”?
6. Describe the differential observing response procedure and how it is
used in intraverbal instruction.
7. What is multiple tact instruction?
8. Distinguish between dictation taking and copying text.
9. Describe one procedure that has been explored to promote dictation
taking or copying text.
10. Comment on two of the implications for practice as described at the
end of the chapter.
 Chapter 5:

Identification and Establishment of

Bidirectional Verbal Operants

R. Douglas Greer
Columbia University Graduate School of Arts and Sciences and Teachers
College, New York, NY
Peter Pohl
Child Psychology Practice, Garmisch-Partenkirchen, Germany
Lin Du
Columbia University Graduate School of Arts and Sciences and Teachers
College, New York, NY
Jennifer Lee Moschella
Columbia University Graduate School of Arts and Sciences and Teachers
College, New York, NY

Link and Preview

Up to this point the text has provided an overview of fundamental
behavioral processes (chapter 2), basic verbal operants (chapter
3), and complex verbal operants (chapter 4). The present chapter
goes a step further in the area of complex verbal operants, and in
particular focuses on bidirectional verbal operants, including the
speaker’s functioning as their own listener. Importantly, the chapter
is written from the perspective of behavioral development and
pertains to the building blocks for some of the most complex forms
of verbal behavior.
We now understand a great deal more about the developmental sequence
needed for children to become fully verbal. Research from our laboratory
and schools (http://www.cabasschools.org) has resulted in the verbal
behavior developmental theory (VBDT; Greer & Ross, 2008; Greer &
Speckman, 2009). VBDT builds on Skinner’s (1957) theory of verbal
behavior, as well as relational fame theory (Hayes, Barnes-Holmes, &
Roche, 2001), research on derived relations, and other research in verbal
behavior. Central to all of this research is Skinner’s idea that in the
development of verbal behavior in children, the listener (observing or
“perceptual” responses) and speaker (speaking or producing responses) are
initially independent. There is growing agreement on this point across
disciplines. Developmental research in neuroscience has identified
correlates in brain physiology that shows that responding as a listener
precedes speaking (Imada et al., 2006). Both behavior science and
neuroscience now concur that experience is key. In fact, the Imada et al.
study provides snapshots of brain changes as a function of experience.
What verbal behavior development contributes to these disciplines is the
identification of the kinds of experiences that make the development possible,
the relation of these experiences to the fundamental principles of behavior,
and how children come to have the wherewithal to contact or benefit from
those experiences. Verbal behavior development findings have led to
increased numbers and sophistication of tools available to professionals to
establish components of development children might be missing. These
components of development are called behavioral developmental cusps
(Rosales-Ruiz & Baer, 1997). The onset of cusps allows children to (a) learn
what they could not learn before, (b) learn faster, or (c) learn in new ways
(i.e., cusps that are new learning capabilities such as the stimuli controlling
learning from observation or exposure).

Bidirectional Operants
The focus of this chapter encompasses some of the cusps that are the
bidirectional operants that mark the particular intercept of the listener and
speaker. These can be separated into three different categories: (a) verbal
episodes between persons, (b) the speaker-as-own-listener (Donley &
Greer, 1993; Greer & Speckman, 2009; Skinner, 1957), and (c) the learning
of word-object relations as speaker and listener incidentally. All of these are
examples of the joining of the speaker and listener. Some even argue that
we are only fully verbal at the point where the speaker and listener are
joined (Barnes-Holmes, Barnes-Holmes, & Cullinan, 2000; Greer & Du,
2015a). Bidirectional verbal operants are involved in talking with others,
talking to oneself, and listening to oneself. Social reinforcers are involved
in all three types, but they are only conspicuous in the verbal episodes
between persons.
Skinner (1957), in his seminal work, stated that verbal behavior is social
behavior. He and others (e.g., Tomasello, 2008) have theorized that verbal
behavior evolved because of the need for the human species to collaborate in
order to hunt and gather and later cultivate and harvest. In a review of a body
of research, Greer and Du (2015b) argued that behavioral developmental
cusps emerge in many cases with the onset of newly learned social
reinforcers (i.e., conditioned social reinforcers). Verbal behavior differs from
language in that the former pertains to the function of communicating, while
the latter pertains to the structure, or the lexicon, of communication. This
does not mean that the two disciplines are necessarily adversarial; rather, they
should be complementary. Verbal behavior also differs from the structural
aspect of language in that it includes social interactions such as frowns,
smiles, grimaces, and gestures. In essence, verbal behavior encompasses the
range of interactions in which individuals seek to affect—or be affected by or
benefit from—the behavior of others.
What may have initially begun with a need to collaborate for mutual
primary reinforcement (i.e., to emit and respond to mands) may have
eventually evolved to direct reinforcement of communicative exchanges by
the behaviors of the parties involved (i.e., conversational units or tacts as
contact reinforcement). One research endeavor focuses on verbal episodes
involving communicative exchanges between children or children and adults.
This consists of analyses of bidirectional verbal operants involving
exchanges between two or more individuals, where an individual is
reinforced both as speaker and listener in one exchange. This is referred to in
the literature as a conversational unit (Baker, 2014; Donley & Greer, 1993;
Greer & Ross, 2008; Lodhi & Greer, 1989; Schmelzkopf, Greer, Singer-
Dudek, & Du, 2017; Sterkin, 2012). More frequent conversational units
speak to the magnitude of the social repertoire of the individual. Still another
social reinforcement dimension of the conversational unit is the initiation of
episodes. Initiating an interaction and responding to an episode initiated by
others are each measures of social reinforcement. The complementary
phenomenon in linguistics is “turn taking.” What verbal behavior adds to the
concept of turn taking is that the conversational unit doesn’t necessarily
require that exchanges “make sense” to an observer; semantics and syntax are
irrelevant to these bidirectional operants. Moreover, a verbal listener or a
verbal speaker response need not involve a lexicon. Evidence of
reinforcement is the continuation of exchanges and few or no lulls in
conversation. The conversational unit includes within it other verbal
operants.
In our laboratory and practice we use the presence or absence of
conversational units as the essential measure of social behavior in children.
This measure is used in interventions that have proved successful in
establishing, or enhancing, conversational exchanges by children with adults
and with children. These appear to result in establishing the natural but
learned reinforcement that is characteristic of social verbal behavior. The use
of traditional behavior analytic interventions to increase socially appropriate
speech in children with autism—scripts or videos, for example—have a long
track record (e.g., Charlop & Milstein, 1989; Taylor, Levin, & Jasper, 1999).
In these cases, the focus has been on behavior as the target. While these
interventions are successful in many cases, there have been problems with
maintenance. Our research program has shown that targeting specific
behavioral cusps may lead to more expansive verbal behavioral development.

Methods to Identify and Establish Bidirectional

Verbal Operants
For each bidirectional operant—verbal episodes between persons, verbal
behavior within the skin (also referred to as speaker-as-own-listener or
talking to oneself), and incidental learning (i.e., bidirectional naming)—we
will describe the cusp and give a general description of how to identify and
establish bidirectional verbal operants.

Verbal Episodes Between Persons

Lawson and Walsh (2007) developed a protocol to establish social
reinforcement for talking to others, seeking them out, and being sought; we
term this social listener reinforcement (SLR; Baker, 2014; Greer & Ross,
2008; Lawson & Walsh, 2007; Sterkin, 2012). This protocol is used
specifically when children, who have characteristics like those described in
the studies, emit tacts and mands but do not initiate conversational units or
tacts with adults or other children. However, when a child demonstrates that
social reinforcement is present but lacks the behaviors (i.e., does not initiate
conversational units), then the use of scripts, videos, and other similar
interventions to establish language repertoires is needed. But, if the behaviors
are not directly reinforced by communicating with others and being
communicated with by others, then establishing the cusp that results in new
social reinforcement for bidirectional operants, the SLR protocol, is
suggested. When maintenance does not result because reinforcers other than
social reinforcers have been used, the establishment of the relevant social
reinforcers also may be needed. In some cases, after the cusp is established,
scripts and videos can be used to expand the variety of verbal behavior.
Below we describe three procedures that have been shown to be
effective in increasing verbal episodes with other individuals: (a) social
listener reinforcement involving collaboration contingencies, (b) the
intensive tact protocol, and (c) observational conditioning. Each appears to
act to condition reinforcers for bidirectional operants between children and
adults or between children, but different presenting repertoires of children
determine which are appropriate for any given child.

SOCIAL LISTENER REINFORCEMENT

The SLR protocol involves a sequence of increasingly complex games,
ending with an empathy phase, that use a yoked contingency game board to
establish social listener reinforcement. A game board is designed with two
paths (Greer & Ross, 2008; Stolfi, 2005), one for the students and one for the
teacher/experimenter. Children must follow the rules of the game and
collaborate with a partner in order to “move up” the path toward an agreed-
upon prize and avoid the teacher’s beating them to the prize. The pair must
ask each other questions and listen to each other in order to “move up.” For
example, the peer confederate holds an unknown item while the target child
is blindfolded, and the target child must ask the peer questions (speaker
behavior). If the target child emits a correct response by telling the
experimenter or teacher exactly what the peer confederate said, the team
“moves up” the game board (first listener behavior and then speaker
behavior). The procedure is designed to build reinforcement for speaking and
listening to peers and to increase the child’s conversational units with others.
The ages and repertoires of the children determine which games are selected.
The types of interactions involving the game board include peer tutoring,
video monitoring, and board games to increase verbal operants emitted by
participants with a wide range of prerequisite verbal skills (Baker, 2014;
Sterkin, 2012). Baker found that the SLR procedure led to more increases in
conversational units when compared with traditional video modeling,
supporting the notion that the SLR procedure established reinforcement for
verbal exchanges, a critical component of truly social behavior and audience
control. As the sequence and sophistication of the games progress, the
reinforcement should come to incorporate “winning” together as a new
reinforcement effect in itself.

INTENSIVE TACT PROTOCOL

The emission of tacts appears to be preliminary to, or a building block
for, conversational units. There is growing evidence that the tact is under the
control of social reinforcement and not just any type of generalized
reinforcers (Eby & Greer, 2017; Schmelzkopf et al., 2017). An intervention
that we call the intensive tact protocol (Greer & Du, 2010) has proved
successful in increasing the initiation of tacts and conversational units. It
appears that the procedure has different effects for different children
depending on their initial repertoire. For children who initiate few mands and
tacts, the procedure increases mands and to a lesser degree tacts, but does not
establish conversational units (Baker, 2014; Schmelzkopf et al., 2017). For
children with more tacts in their repertoire, the initiated mands decrease and
the tacts and conversational units increase, demonstrating the shift toward
more social reinforcement (Eby & Greer, 2017). The decrease in mands and a
corresponding increase in tacts indicates a shift in reinforcement control to
social contact and is a sought after and desirable outcome (Costa & Pelaez,
2014).
The intensive tact intervention builds social listener reinforcement
(Delgado & Oblak, 2007; Greer & Du, 2010; Pistoljevic & Greer, 2006;
Schauffler & Greer, 2006; Schmelzkopf et al., 2017) and consists of daily
presentation of 100 learn units of pure tact opportunities, where the
participant receives social praise as reinforcement for emitting tacts without a
vocal-verbal antecedent. That is, a stimulus is presented without the use of
vocal instructions such as “What is it?” Target stimuli can consist of social
studies content (e.g., national monuments) and can be determined by
participants’ age, grade-level curricular requirements, and grade-level
achievement. The procedure can also incorporate autoclitic functions.

OBSERVATIONAL CONDITIONING
An observational conditioning intervention (i.e., Singer-Dudek, Oblak,
& Greer, 2011) is another effective protocol to condition social attention and
praise as a reinforcer. Greer, Singer-Dudek, Longano, and Zrinzo (2008)
conditioned adult praise as a reinforcer by presenting a simple task to target
participants and peer confederates, while delivering praise only to the peer
confederate (the target participant was denied praise for completing the task).
Schmelzkopf and colleagues (2017) also found increases in conversational
units with peers as a function of this protocol. (See chapter 8 on
observational learning for more on interventions that capitalize on the
repertoire of learning by observation.)
The three behaviors discussed above—social listener reinforcement, the
intensive tact protocol, and observational conditioning—establish or enhance
adult attention as a reinforcer which, in turn, results in more tacts and, more
significantly, more conversational units. Once the speaker and listener are
joined, the repertoires for listening to others (being “communicated to” as a
function of reinforcement stimuli) and speaking to others (communicating to
as a function of reinforcement stimuli) are embedded in the stimuli in the
interaction. Of course, the onset of the cusp is the very beginning of this
repertoire. But, once it is present, the experiences in communicating with
others lead to a more sophisticated repertoire because the reinforcers are
present.
Children also take on the roles of speaker and listener when they talk
aloud as they are alone and engaged in fantasy play. Skinner (1957) argued
that this overt behavior, speaker and listener roles, becomes covert as
children are made “self-conscious” as a result of certain audiences. We also
argue that it is a critical speaker-as-own-listener cusp leading to more
complex behavior, which we discuss below.
 Verbal Behavior Within the Skin (Talking to
Oneself)
While it is empirically difficult to test whether one talks to oneself
covertly, it is apparent and observable in young children when they engage in
fantasy play, where they talk aloud to themselves. The phenomenon is
described as a critical stage by many developmental psychologists who study
play and language development (Vygotsky, 1967). Behavior analysts have
investigated this phenomenon in children demonstrating the alternation of
speaking and listening, where the roles of listener and speaker can be directly
observed. Lodhi and Greer (1989) demonstrated that typically developing
five-year-old girls engaged in self-talk conversational units (i.e., alternated
roles as speaker and listener), providing evidence for Skinner’s (1957)
“speaker and listener within the skin.” Many of the self-talk exchanges in
Lodhi and Greer’s study also included the say-do correspondence repertoire
that Paniagua and Baer (1982) identified early on. A child may act as a
speaker in saying that they, or an object of play, will engage in an action, and
then the child may act as a listener by doing what they said they would do.
This demonstrates that the listener and speaker repertoires are joined in the
child’s verbal behavior development. Say-do correspondence and self-talk
signal that bidirectional verbal operants are present in the child’s verbal
repertoire. These functions appear to lay the foundation for more complex
listener/speaker interactions within the skin, such as thinking verbally. Then,
by extension, once print control becomes part of the relations, one functions
as writer and own reader, or self-editor.

Incidental Learning (Bidirectional Naming)

The third bidirectional operant and developmental cusp is what Horne
and Lowe (1996) called naming. We concur with Miguel (2016) that the term
bidirectional naming, and the abbreviation BiN, is less confusing. By
extension, we suggest that the term unidirectional naming (UiN) refer to the
demonstration of the listener component. While it is possible, and even
probable, that one can be taught to emit speech sounds (parroting or
canonical babbling) before responding as a listener, we argue that this is not
really verbal from a developmental and a verbal behavior perspective. For
example, one can shape speech such that a child says the name of an item and
receives the item; however, if the child did not first acquire the listener
relation, it is not truly verbal. When the speech sound is taught or learned
first, it may function as a duplicative response, as in singing or parroting
under the control of the correspondence rather than the verbal function (see
Gladstone and Cooley, 1975, for the seminal study). There are numerous
examples of individuals with autism who are not verbal but can sing the
lyrics of an entire song. They are presumably under the control of the
correspondence between their responses and what they heard before, yet none
of the responses are verbal in the sense that they affect the behavior of others
in verbal functions. However, if the speech sounds are in repertoire,
interventions may join the listener to the speaker.
The specific aspect of BiN that we focus on is how this developmental
cusp (i.e., a cusp that is also a new learning capability that changes how
children can learn) results in children’s learning language and verbal
behavior functions without direct instruction (i.e., incidental learning). We
believe that the body of existing research suggests that BiN, together with the
two other cusps associated with the joining of the listener and speaker within
the skin, represents the point at which the child becomes fully verbal. The
subsequent complex behaviors associated with reading and writing, using the
methods of science, authority, and logic, are built on the emergence of these
cusps.
We now know that some children who did not demonstrate BiN can do
so after they undergo environmental interventions, including multiple
exemplar instruction (MEI) across listener and speaker responses with
training sets of stimuli. After receiving MEI, children from eighteen months
to six years of age (neurotypical children and children with language delays)
demonstrated BiN with novel stimuli (Corwin & Greer, 2017; Fiorile &
Greer, 2007; Gilic & Greer, 2009; Greer, Corwin, & Buttigieg, 2011; Greer,
Stolfi, Chavez-Brown, & Rivera-Valdez, 2005). Children who demonstrate or
have BiN (i.e., learn words for things incidentally) for novel but familiar
types of stimuli (i.e., novel animals, cartoon characters, vegetation) may not
demonstrate BiN with unfamiliar stimuli (e.g., constructed symbols such as
chemical symbols). However, they can develop BiN for these unfamiliar
stimuli after contact with exemplar experiences involving those types of
stimuli (Lo, 2016). Children with basic BiN require additional experiences to
simultaneously learn the names of the stimuli and at the same time the
actions paired with the stimuli (Cahill & Greer, 2014). Similarly,
simultaneously learning additional auditory, tactile, and olfactory relations
requires additional experiences consisting of pairings with these senses.
However, once basic BiN is present, the additional relations appear to
accrue simply by an automatic reinforcement conditioning process.
It has been more than fifteen years since we first identified how to
establish BiN. Our understanding of the cumulative research, in the years
since, is that the initial accrual of reinforcement for observing the visual
stimuli along with the words (Longano & Greer, 2014) allows those stimuli,
in turn, to condition the new actions, scents, tactile stimuli, and sounds as
reinforcing stimuli for observing responses, and the new stimuli join with
those already in repertoire with repeated experience (Frias, 2017; Kleinert,
2018; Lo, 2016). No instruction or additional reinforcement is needed; simple
exposure alone appears to result in the accrual of the new relations. From the
perspective of the principles of behavior, this appears to explain the rapid
growth of language observed in young children. The stimuli are now clear for
“spontaneous” language development (Chomsky & Place, 2000). It is not
spontaneous; rather it is under the control of learned reinforcers.
Understanding the developmental phenomenon of BiN has significant
applied utility, in addition to the obvious contribution to basic science. Once
BiN is present, children can learn in typical education settings, while without
it they struggle or cannot (Corwin & Greer, 2017; Greer et al., 2011). In our
laboratory and inclusion classrooms, we identified that BiN is an essential
cusp needed for children to be successful in general education. Once BiN is
present, the expansion of verbal repertoires is no longer dependent on direct
instruction. That is, children’s responding no longer needs to be directly
reinforced or corrected for each skill. Instead, children can observe and learn
a response to stimuli as a listener and speaker simultaneously.
Below, we discuss the identification and establishment of bidirectional
operants involved in appropriate self-talk, and then we examine different
types of BiN.

BIDIRECTIONAL OPERANTS INVOLVED IN APPROPRIATE

SELF-TALK
Skinner (1957) described the phenomenon of individuals’ talking to
themselves as a rotation of speaker and listener responses. He proposed that
children speak to themselves aloud until the behavior is punished by the
audience (parents or peers), at which point the behavior goes beneath the
skin, not unlike going from reading aloud to reading silently. The rotation of
listener and speaker roles within the skin is, from a radical behaviorist
perspective, “thinking.” The indications of a child’s being truly verbal and
rotating the listener and speaker roles within the skin are (a) say-do
correspondence (Farrell, 2017; Paniagua & Baer, 1982) and (b) self-talk in
fantasy play (Lodhi & Greer, 1989). These verbal cusps, or components of
the speaker-as-own-listener cusp, may be prerequisites or co-requisites to
BiN and SLR (Greer & Ross, 2008). The research on these relations remains
to be done. Self-talk refers to the emission of conversational units by one
individual, where each of the roles of speaker and listener intersect in a
conversational unit. Each behavior function, listener and speaker, is
reinforced in an exchange. These are bidirectional operants between listener
and speaker roles. These responses are distinguished from vocal stereotypy
and palilalia because the function of self-talk is the conditioned
reinforcement for exchanging of behaving as a speaker and listener and
usually involves young children’s fantasy play with toys that have
anthropomorphic characteristics (e.g., animals, dolls, puppets; see Lodhi &
Greer, 1989).
In a recent dissertation, Farrell (2017) told two- and three-year-olds,
“Say what I say” (echoic) and “Do what I do” (imitation). The experimenter
modeled one action figure saying, “Let’s go save the world!” and the other
figure saying, “Okay, let’s go!” followed by the experimenter’s moving the
figure. The experimenter then gave the participant an opportunity to do the
same and reinforced their echoic and imitating responses. Participants
demonstrated first instances of say-do and self-talk conversational units
during free-play settings. Intraverbal mutual entailment emerged as a
function of the intervention (i.e., untaught responses of example to category
or category to example). Say-do correspondence is usually implicit in self-
talk and conversational units.
When an individual has say-do correspondence, they can state what they
will do (Paniagua, 1990; Paniagua & Baer, 1982), as when a child tells a
caregiver, “I’m going to go so high on the swings!” and runs over to do so.
Similarly, Farrell (2017) reported that do-say correspondence is the same
cusp, when children say what they did (i.e., “I went on the slide!”). In
essence, the individual behaves as a listener in response to their own speech.
Say-do correspondence and self-talk are the prerequisites for more complex
repertoires such as problem solving and self-awareness (Greer & Du, 2015b).
Say-do correspondence, self-talk conversational units, conversational units
with others, and BiN are all examples of bidirectional operants. Say-do, self-
talk in fantasy play, and BiN are developmental cusps that are components of
the joining of the speaker and listener within the skin.

TYPES OF BIN
BiN is of utmost importance for one’s educational and social
development. It is one of the cusps demonstrating the joining of the speaker
and listener that is necessary to become fully verbal. BiN allows the learning
of word-object relations incidentally and allows children to learn from
demonstration and observation. Typically developing children generally
acquire BiN without direct instruction as a result of experiential contact with
thousands of verbal interactions. However, if children cannot contact those
experiences, even though the experiences are present, they are missing the
BiN cusp.
Research suggests there are several effective interventions that can lead
to the establishment of full naming or the listener half of naming. One
protocol, but not the only, is MEI across speaker and listener responses (i.e.,
Corwin & Greer, 2017; Feliciano, 2006; Fiorile & Greer, 2007; Greer, Stolfi,
et al., 2005; Rosales, Rehfeldt, & Lovett, 2011). MEI involves the rotation of
presentations of stimuli across the four response topographies: (1) visual
match-to-sample while hearing the tact of object/picture (observing), (2)
point to (observation response), (3) tact (production response), and (4)
intraverbal tact responses (producing) to the same set of stimuli. This
exposure can result in the child’s responding as both speaker and listener to
untaught responses to novel stimuli. The rotation component between listener
and speaker for the same stimuli in MEI was shown to be both sufficient and
necessary for the acquisition of BiN for children who lacked this capability.
The QR code below is a link to a video that provides a sample of two MEI
interventions to induce BiN.
 However, those who have acquired basic BiN under observational
conditions cannot necessarily learn the names of things by exclusion (Greer
& Du, 2015b). Therefore, the participant is asked to discriminate the
unknown stimulus (i.e., Chinese logograph) against four other known ones
(i.e., common home or school items), as well as point to, tact, and intraverbal
tact the novel stimuli. In Greer and Du (2015a), instruction was structured in
the same rotated fashion as in traditional MEI; the participant was reinforced
sequentially first by observing the stimulus and simultaneously attending as a
listener to the word for the object, and emitting listener and speaker
responses to the object. This exemplar experience was shown to be effective
in the establishment of BiN by exclusion.
Recent research shows that this kind of transformation of stimulus
function across listener and speaker responding could also occur without
direct instruction (Lo, 2016; Longano & Greer, 2014). Longano and Greer
(2014) tested the effects of stimulated natural conditions on the emergence of
BiN with preschoolers with disabilities. The experimenters presented the
novel stimulus to the participant and tacted the stimulus while holding or
pointing to it. The participant was not required to emit any responses, and
thus no consequence was delivered. They found that exposure to this type of
tact word-object relation experience led to increases in the children’s echoics
and in turn resulted in the emergence of BiN. In a recent dissertation, Lo
(2016) presented repeated unconsequated probes to young preschoolers with
disabilities until they demonstrated mastery (90% accuracy or above) for
each of the four responses, including listener responses to visual stimuli,
speaker responses to visual stimuli, listener responses to auditory stimuli
associated with the visual stimuli, and speaker responses to the additional
auditory stimuli. Lo found that the exemplar experiences with the repeated
pairings of visual and auditory stimuli functioned to establish conditioned
reinforcement for spoken and nonspoken auditory stimuli. Lo’s study
suggested that the increases in stimulus control may be responsible for the
onset of traditional BiN as well as BiN to auditory nonspeech stimuli.
 Additional stimulus control accrues from further experience once an
individual demonstrates basic BiN as a result of pairings. Recent verbal
behavior development research suggested that such multisensory experience
may be responsible for the establishment of other forms of BiN (Cahill &
Greer, 2014; Frias, 2017). Cahill and Greer (2014) found that simultaneously
presenting actions and names hindered participants’ learning of names, but
pairing reinforcement with multiple responses across imitation actions,
responses to “Find ,” intraverbals, and tacts induced this
type of BiN. Another study investigated the relation between naming and
additional auditory stimuli, olfactory stimuli, and tactile stimuli entering the
naming “frame” (Frias, 2017). The experimenter provided the participant
with naming experience of stimulus-stimulus pairing by presenting a stimulus
for the participant to observe by seeing (visual), hearing (auditory), touching
(tactile), and smelling (olfactory) together with the name of the stimulus
(Pérez-González & Carnerero, 2014). Stimulus relations accrue for the names
of stimulus modalities after such experience (Frias, 2017).
These studies provide support for the differential development of BiN
across varied stimulus modalities. What is especially interesting and
important about the expansion of the stimuli accrued from BiN experiences is
that once the basic BiN repertoire is established, the additional stimuli are
learned without direct instruction. This finding suggests the source of BiN
resides in the pairing of existing reinforcers for observing responses with the
new stimuli, consistent with research by Longano and Greer (2014) and
others showing the strong association between the onset of new conditioned
reinforcers and demonstrations of several cusps (Greer & Du, 2015a).
The identification of verbal developmental cusps was driven by
seemingly unsolvable learning problems presented by children with language
and social deficits and guided by the basic science of behavior. The findings
are inherently translational in that applied investigations promote basic
findings. The theory and findings of verbal behavior development also
suggest relations with biology, a type of relation that has a long history in
behavior analysis. We conclude the chapter with a consideration of the
potential relevance of operant bidirectionality as a behavioral phenotype in
the evolution of verbal behavior.
The Bidirectional Operant as Behavioral
Metamorphosis
Together with the foundational cusps of observing and producing as part of
the child’s preverbal cusps, we have regularly established incidental
learning of verbal bidirectional operants in typically developing children
earlier than expected (Gilic & Greer, 2009). At the same time, in many
other children with language delays, we have induced bidirectional verbal
operants experimentally. In both cases, the resulting emission of
bidirectional operants was striking and seemed to indicate a fundamental
transformation of the child’s subsequent development. We felt that this
remarkable emergence of operant bidirectionality in the child’s behavioral
development merited special consideration (Greer, Pohl, Du, & Moschella,
2017; Pohl, Greer, Du, & Moschella, 2018): Accordingly, we report on
comparative evidence that suggests that the bidirectional operant may be
functionally homologous to the biological phenomenon of metamorphosis
and a behavioral phenotype in the evolution of verbal behavior.
In pursuing this hypothesis, we follow Skinner (1966a) and Donahoe
(1996), both of whom have provided strong arguments for the functional
similarities between natural selection in phylogeny and selection by
reinforcement during ontogeny. Donahoe’s (2012) review of a collection of
papers concerned with developments in evolutionary biology during the 150
years since publication of Darwin’s (1859/2013) On the Origin of Species
concludes that “our understanding of the details of selection by reinforcement
is arguably at least as complete as was Darwin’s understanding of natural
selection” (Donahoe, 2012, p. 256).
Metamorphosis (meta, meaning “change” + morphe, meaning “form”)
as a biological process is extreme life-stage modularity and means that a
single genome produces at least two highly distinctive morphological and
behavioral phenotypes, which occupy very different habitats. It is ubiquitous
throughout nature, repeatedly evolving in the history of multicellular life, and
many species—butterflies, bees, beetles, ants, moths, flies, and wasps, as
well as many marine invertebrates (amphibians and fish)—undergo dramatic
phenotypic remodeling during development. In a symposium of fourteen
comparative biologists primarily concerned with metamorphosis across a
number of taxa in the animal kingdom, Bishop et al. (2006) loosely
conceived metamorphosis as a component of organismic ontogeny. However,
despite considerable differences in the definition of the term, no distinctive
life-cycle transitions comparable to metamorphosis in insects (e.g., caterpillar
to butterfly) have been documented for mammals.
In her groundbreaking treatise, Developmental Plasticity and Evolution,
developmental evolutionary biologist and insect social behavior expert Mary
Jane West-Eberhard (2003) devotes a chapter to “phenotypic recombination
due to learning” (pp. 337–352), wherein she focuses on the importance of
learning for the origin of novel adaptive phenotypes. West-Eberhard (2003)
notes, “Most people, including most biologists, probably underestimate the
importance of learning in the biology of nonhuman animals… So learning
itself has not always received the attention it deserves as a phenomenon of
general evolutionary interest” (p. 337). Recent evidence supports her insight
at the cellular level (Robinson & Barron, 2017). And at the level of behavior,
a study on behavioral development in dogs explicitly referred to “behavioral
metamorphosis” (Scott & Nagy, 1980), and an editorial by Oppenheim
(1980) considered metamorphosis more generally from the perspective of the
phenomenon’s adaptive advantage.
In his Biographical Sketch of an Infant, Darwin (1877) observed that his
son, at exactly twelve months of age, began to vocalize for food in a prosodic
manner employing the consonant-vowel “mum,” with “a most strongly
marked interrogatory sound at the end” (p. 293). Darwin related that
observation with farsightedness to a view he had entertained six years earlier
in The Decent of Man (Darwin, 1871), namely that before any prehuman
hominid species began to use articulate language, it uttered notes in a true
musical scale similar to the vocalizations of a number of nonhuman primates
including gibbons. That the nonhuman primate vocal tract can produce and
use sounds to communicate involving both speaker and listener functions,
such as in male-female duets, has been well documented by a gibbon
vocalization field study carried out by Geissmann (2002).
A recent observational study with baboons (Papio papio) suggests that
this species of nonhuman primates produces vocalizations comparable to
human vowels (Boë et al., 2017). The researchers employed acoustic
analyses of vocalizations coupled with an anatomical study of the tongue
muscles and dynamic modeling of the acoustic potential of the vocal tract.
The data suggest that baboons are capable of producing at least five
vocalizations with the properties of vowels, in spite of their high larynx, and
that they can combine them when they communicate with members of the
same troop. The finding suggests that the last common ancestor of humans
and baboons may have possessed the anatomical prerequisites for speech—
hinting at a much earlier origin of language than heretofore believed. Could it
be that correspondence is also an embedded reinforcement effect at play here
as it is with children (Gladstone & Cooley, 1975)?
Regarding speech perception from a comparative perspective, valuable
support would be given to the hypothesis developed here if, for example,
functional ear asymmetries like those typically found in humans were
established experimentally in nonhuman primates. In the first study to
document ear advantages for acoustic stimuli in a nonhuman primate species,
employing an operant conditioning, same-different auditory discrimination
task, Pohl (1983) found in four baboons (Papio cynocephalus) highly
significant and reproducible ear advantages in the monaural discrimination of
pure tones, three-tone musical chords, synthetically constructed consonant-
vowels (CVs), and vowels in each of the animals with marked individual
differences in the direction of ear asymmetry. The ear advantages found in
these baboons under monaural conditions resembled those obtained with
dichotic presentation in human subjects (Pohl, Grubmüller, & Grubmüller,
1984) and thus suggest, especially when considered together with the recent
Boë et al. (2017) findings on the vowel-like quality of baboon vocalizations
mentioned above, that the baboon may provide a valuable nonhuman primate
model of functional asymmetry in the central auditory system.
In a subsequent study on ear advantages for temporal resolution in the
same four animals, Pohl (1984) hypothesized that the temporal resolution,
which underlies an individual’s ear advantage for speech perception, would
predict an ear advantage for a temporal resolution task to correlate precisely
with an ear advantage for the discrimination of CVs that differ in their
temporal features. Using the same methodology as in the previous study to
test the hypothesis in all four baboons, Pohl employed a gap-detection task
that required the resolution of brief intervals of no noise (between seven and
twenty milliseconds, depending on the animal) in bursts of random noise.
The finding clearly confirmed the hypothesis in that the probability of
obtaining by chance the same direction of ear advantage in four animals
under two quite different task conditions is <.01. The results suggest that a
single perceptual mechanism based on fine temporal analysis underlies the
CV discrimination and gap detection tasks in both humans and baboons and
that this mechanism may be asymmetrically organized within the central
auditory system of primates.
Any attempt to understand behavioral metamorphosis during speech
development from a biological perspective must incorporate a set of
established facts from studies of birdsong development. The late ethologist
Peter Marler seems to have left the science of biology with an unusual
legacy. In an early comparison of the ontogeny of birdsong and speech,
Marler (1970a) stated, “There may be basic rules governing vocal learning to
which many species conform, including man” (p. 669). Soha and Peters
(2015) list six basic rules, elaborated on the basis of Marler’s extensive
research in songbirds, which underlie the ontogeny of vocal behavior in both
humans and white-crowned sparrows (Zonotrichia leucophrys). We cite these
rules here as they are consistent with what has been discovered about verbal
behavior development (see Soha & Peters, 2015):
1. Exposure to certain sounds (normally, the vocalizations of adult
conspecifics) during development has a crucial impact on the adult
repertoire.
2. This exposure has the greatest effect during a certain developmental
time range, that is, learning proceeds most readily during an early
critical period.
3. Vocal learning is guided by predispositions to proceed in a certain
way. In particular, human babies preferentially attend to speech
sounds and young songbirds are predisposed to pay particular
attention to the vocalizations of their own species.
4. Individuals must hear their own output during vocal development,
as auditory feedback guides the matching of this output to what was
heard and memorized previously (italics ours). Auditory feedback is
then less important; in Marler’s words, it “becomes redundant”
(Marler, 1970b, p. 23) after vocal development is complete.
5. At least some vocal learning apparently proceeds without external
reinforcement, suggesting that internal reward reinforces the process
of matching vocal output to memorized models. [It should be
mentioned here that this is consistent with the Horne & Lowe (1996,
p. 207) theory and research findings from Longano and Greer (2014;
insertion in brackets ours)].
 6. Vocal motor development proceeds in stages. Young individuals
begin by making vocal sounds that do not resemble those of adults
(italics ours), and even deafened individuals produce the earliest
versions of these sounds (babbling or early subsong). In normal
individuals, the effects of auditory experience then become apparent
at the next stage. (Soha and Peters, 2015, p. 934)

In the interim, Patricia Kuhl (2004) and her associates have amply filled
this empirical gap between birds and infants with their multidisciplinary
research on social and speech development in early childhood. Pertinent to
the argument developed here, Kuhl (2003) notes that “Both babies and birds
must…rehearse and refine their communicative repertoires, actively
comparing and gradually matching (through auditory feedback) their
productions to the sound patterns stored in auditory memory” (p. 9645,
italics ours).
What, then, is Marler’s legacy? It hinges on the empirically grounded
fact that songbirds need to hear their own auditory output during vocal
development as a necessary precondition to the production of typical species-
specific singing as juveniles and adults. As we have seen, this fundamental
fact of vocal learning during development also holds for the human infant.
For babies, the task involves learning the phonemic units and prosodic
characteristics (rhythm, stress patterns, and intonation) of their mother
tongue. For songbirds, the developmental requirement is to learn the specific
notes, syllables, and prosodic characteristics typical of their species’ song
repertoire. Taken together and combined with the primate data cited
previously, these findings suggest that a class of operant behavior underwent
selection by reinforcement, enabling the modular design of a bidirectional
behavioral phenotype as a necessary precondition for the evolution of human
speech development.
We hypothesize that in joining the listener and speaker together in the
same skin, the bidirectional operant would seem to represent complete
behavioral metamorphosis. To use the classic example from insect ontogeny,
if the caterpillar’s habitat is the ground, branches, and leaves, and the
preverbal child’s niche is the immediate physical and social environment,
then the butterfly’s new habitat is the air and plants, and the verbal child’s
novel environment is the symbolic universe of a verbal community. Whereas
the pre-BiN child behaves more like a nonhuman primate in a proximate
physical and social environment, the child who has acquired the bidirectional
operants becomes a member of a verbal community, which is the equivalent
of a novel environment of symbolic relationships.
The evolutionary prerequisite of that profound behavioral
transformation is the ontogenetic selection of separate observing and
producing phenotypes and their modular integration into the bidirectional
operant through social learning. In this context and with a view toward “a
more comprehensive understanding of language,” Greer (2008) mentions
music and dance in addition to verbal behavior and the visual arts as
evolutionary outcomes of the ontogenetic selection of separate observing and
producing responses and the cultural joining of them (p. 370). This sets the
stage for the hypothesis proposed here concerning the evolutionary origins of
the bidirectional operant as an instance of behavioral metamorphosis in
Homo sapiens. Based on the recent discovery of the bidirectional operant
through behavioral research, Marler’s legacy is that through his comparative
studies of birdsong development, he has enabled our understanding of the
bidirectional operant as an exemplar of complete behavioral metamorphosis
during human verbal development. It is conceivable that in the years to come
this insight could enable the approximation of two subdisciplines of biology,
developmental evolutionary biology and developmental verbal behavior
analysis, and spawn future interdisciplinary research on behavioral
metamorphosis.

Implications and Future Directions

The present chapter has important implications for research and practice. In
taking a verbal behavior developmental theory approach, the chapter draws
attention to the concept of behavioral developmental cusps, which expand
an individual’s behavioral repertoire in meaningful ways. This chapter
provides an overview of several instructional protocols that have been
developed and researched in efforts to understand how to best promote the
emergence of behavioral developmental cusps. More research is needed on
the procedures described, including replications as well as studies with
individuals with different prerequisite skills, in different settings, with
respect to different topographies of behavior, and more. This research
program has substantial implications for clinicians, especially considering
assessment and intervention efforts in the area of language development.
Effective practice is guided by knowledge of not just how to teach various
skills, but also when to teach them with particular learners, what to do next,
and so on. Given this, the behavioral sequence needed to establish fully
verbal behavior, as described in this chapter, has remarkable clinical value.

Conclusion
Verbal behavior development builds on a variety of theoretical frameworks,
all of which can be used to identify the kinds of experiences, or cusps, that
make verbal development possible. This chapter discussed behavioral cusps
known as bidirectional operants, which include verbal episodes between
persons, speakers-as-own-listeners, and the learning of word-object
relations. The joining of speaker and listener repertoires is critical for verbal
behavior development. This chapter disseminated research on assessment
and instructional methods for identifying and establishing such behavioral
cusps.

Study Questions
1. What three things do behavioral developmental cusps allow children
to do?
2. Provide two examples of bidirectional verbal operants as mentioned
in the text.
3. What is the premise of the social listener reinforcement protocol?
When is it most likely to be helpful?
4. Describe the intensive tact intervention and the type of generalized
reinforcer that is to be used.
5. What does it mean for an individual to function as both a speaker
and a listener?
6. What is say-do correspondence? How does say-do correspondence
relate to the speaker’s functioning as a listener?
7. What is self-talk? How does self-talk relate to the integration of the
speaker and listener repertoires?
 8. How do the chapter authors explain “spontaneous” language
development?
9. What does it mean to become fully verbal from the perspective of
the chapter authors?
10. What is the difference between unidirectional and bidirectional
naming?
11. How is multiple exemplar instruction used to establish BiN?
12. Is there an animal model of an individual’s functioning both as a
“speaker” and a “listener”? In which species?
13. Describe bidirectional operants between individuals and
bidirectional operants within individuals.
14. Define the conversational unit. How does it serve as a measure of
social behavior?
 Chapter 6:

The Scope and Significance of Referential

Behavior

Patrick M. Ghezzi1
University of Nevada, Reno

Link and Preview

Thus far we have covered basic processes, Skinner’s analysis of
verbal behavior, including both fundamental and more complex
aspects of it, and the analysis of bidirectional operants. This
chapter and the next pertain to a relatively less well-known
approach to language within behavior analysis, specifically
referential behavior. The analysis is derived from interbehavioral
foundations and is ripe with theoretical, investigative, and practical
implications.

“What precisely happens when one speaks?” J. R. Kantor (1936) answered

his question in this way: “In effect, the speaker refers, and the hearer is
referred to some thing, person, or event” (p. 75). Kantor offered the
example of one person, A, saying to a second person, B, “The car is
wrecked.” Person A is the speaker in this case who refers person B, the
listener, to the wrecked car. “It’s ruined,” replies B, thereby completing the
interaction initiated by A, the speaker.
What is plain to see in Kantor’s example is that the two people are
talking to each other about something. A layperson calls it the “topic of
conversation,” and Kantor calls it the referent, that is, “some thing, person, or
event” to which a speaker refers a listener. The speaker who refers the
listener to the wrecked car is stimulated in the moment to say something
about it, and the listener, in turn, is stimulated to respond to what the speaker
said about it. A layperson would say the two people are “having a
conversation.” Technically, the episode is termed a referential interaction.
A distinguishing feature of referential behavior is how pervasive it is.
The same can be said about other complex human behaviors—thinking and
imagining come to mind—yet what further distinguishes referential behavior
is how conspicuous it is. It is almost always public in the overt, psychological
sense of the term and in the sense that it so often takes place openly for all to
see and hear. Individual instances are sufficiently distinct to observe and
record, and measures of count and time, long favored in the natural sciences,
apply seamlessly to it.
Referential behavior is further distinguished by how early it develops
and expands throughout the life span. By two to three years of age, noted
Bijou (1989), a young child is fully capable of engaging in it. The expansion
of referential behavior is explosive, actually, eventually becoming the
cynosure of a person’s personal and professional life. It is no overstatement
to say that referential behavior is a cornerstone of society, and further, that it
constitutes a remarkable adaptation by the human species.
An additional feature of referential behavior is how understudied it is in
behavior science. How could something so common, so conspicuous, and so
fundamentally human fly so far under the proverbial radar? What is it about
an ordinary, everyday, directly observable and objectively measurable
behavior that seems so easy to ignore?
The antilogy is even harder to fathom when it comes to young children
who experience difficulties engaging in referential behavior and suffer a
lifetime because of it. The behavior develops quickly for most youngsters
and flourishes naturally throughout adolescence, adulthood, and well into old
age. This trajectory is impeded early in life for some children, notably young
children with intellectual and developmental disabilities. A common
treatment objective for these youngsters, indeed, is to learn how, when,
where, and with whom to engage in referential behavior.
The aim of the present chapter is to elaborate upon these features of
referential behavior. We begin with a brief survey of J. R. Kantor’s writings
on language and linguistics to reveal the origins of the analysis of referential
behavior, and move from there to discuss referential behavior in detail. Three
advancements in the analysis of referential behavior since Kantor’s death in
1984 are highlighted next, followed finally by suggestions for future research
and development in the field.

Origins of the Analysis

Judging from the chronology of his publications, Kantor took an early
interest in language. In a series of articles published in the 1920s (1921,
1922, 1928, 1929), he argued for an objective analysis of meaning and
language, and in 1926 devoted a lengthy chapter to language in the second
volume of his Principles of Psychology. Here, he (1) differentiated
psychological data from other language-related things and activates, (2)
categorized and analyzed language behavior situations, and (3) speculated
on the development of language in relation to cultural and social conditions.
Several years later, in another textbook on general psychology (1933), he
presented a brief account of his approach to language. Three years after
that, in 1936, the breadth and depth of Kantor’s views were developed fully
in his first volume on language, An Objective Psychology of Grammar. In it
he emphasized the crucial difference between speaking and listening and
the products of these behaviors, namely, words and sentences. The former
was taken to be the proper domain for a functional psychology of language,
and the latter, the focus of structural linguistics.
After An Objective Psychology of Grammar, Kantor published another
article on language, this on the role of language in logic and science (1938).
He returned to the study of language some thirty years later, when in the
1970s he published four papers under the authorship “Observer” (1970,
1971a, 1971b, 1976). In those articles he reiterated that authentic
psychological language, consisting of referential and symbolizing behavior,
should be distinguished from the products of these behaviors. His position
was subsequently outlined in a chapter of his second textbook of general
psychology (Kantor & Smith, 1975). In the same year in an article published
in the first issue of the Mexican Journal of Behavior Analysis, Kantor (1975)
labeled his position “psychological linguistics.” The fact that “linguistics”
was then, as now, synonymous with the study of words and sentences did not
deter Kantor, nor did the fact that the hybrid term “psycholinguistics” was
and still is infused with doctrine antithetical to his approach. Regardless,
Kantor (1977) published his second and final book on language,
Psychological Linguistics.
Shortly before his death, Kantor published another series of articles on
language. In them he argued that (1) the confusion between speaking and
listening and the products of these behaviors was due to the disunity in
psychology regarding the basic premises of a natural science of psychology
(Kantor, 1981), (2) the use of metaphors in psychology and language (e.g.,
the brain conceptualized as a computer) has had a regressive effect on both
disciplines (Observer, 1983a), and (3) the concept of meaning in psychology
and linguistics is misunderstood due to the longstanding confusion between
events and constructs (Observer, 1983b).
It is clear that Kantor had much to say about language. Over some sixty
years he wrote on every conceivable topic in psychology, maintaining
throughout that all psychological activity can be construed as the observable
activities of organisms interacting with the things, objects, people, and events
collectively called the environment. As a specialized area of study within the
field of scientific psychology, language was for Kantor distinguished only by
its complexity and preponderance among human beings.
One of the great ironies in the history of behavior science is that to this
day, Kantor’s writings on language are neither widely acknowledged nor
appreciated. There is a notable exception to this, a review by the esteemed
behavior scientist Nate Shoenfeld in 1969 of Kantor’s (1936) An Objective
Psychology of Grammar. Shoenfeld proclaimed the volume a “stunning” and
“historic” book and urged his colleagues and students to study it. Shoenfeld’s
advice went unheeded, yet neither Kantor nor Skinner seemed to mind.
Kantor did not refer to Skinner’s (1957) Verbal Behavior in any of his
writings, and Skinner, who died in 1990, never referred to Kantor’s (1977)
Psychological Linguistics in any of his writings. It is common knowledge
that Skinner’s approach ascended during his lifetime, while Kantor’s
approach languished during his life in relative obscurity (Ghezzi & Lyons,
1997).
The two approaches to language have since been compared and
contrasted in detail by several accomplished behavior scientists, notably
Parrott (1984a, 1987) and Morris (1982, 1984). This is not the place,
however, to repeat or expand on the issues. Suffice it to say that the
similarities between the two approaches are greater than the differences, and
that a synthesis of the two, as the present chapter will show, is both possible
and desirable. We turn now to referential behavior, our main focus.

Referential Behavior in Detail

What laypeople call the topic of conversation, or what is known technically
as the referent, is not the only stimulus to which a speaker (and listener)
responds. “At the basis of referential behavior,” wrote Kantor (1977), “is
the psychological process of bistimulation” (p. 62). The concept of
bistimulation recognizes that, in addition to the referent, individual
speakers and listeners themselves are a source of stimulation for referential
behavior. In Kantor’s (1936) example of the wrecked car, imagine a
scenario in which a teenage driver is sobbing and apologizing profusely to
his father about what happened to the car, and the father, in turn, is loudly
chastising his son for careless driving. When the boy speaks to his mother
about the accident, in sharp contrast, he is defiant and his mother, demure.
It is clear to see in ourselves and other people that whom we talk to
influences what we talk about and how, when, where, and why we talk about
it. Determining the direction and magnitude of the change in referential
behavior as a function of the characteristics of individual speakers and
listeners is a fertile area for research, one that we will have more to say about
later in the chapter.
In addition to bistimulation, Kantor recognized a third factor, setting
conditions, which Bijou (1996) defined as “the general surrounding
circumstances that operate as inhibiting or facilitating conditions in a
behavioral unit” (p. 149). Bijou and Baer (1978) and Bijou (1996) have
compiled a fairly comprehensive list of biological, physical, and sociocultural
setting conditions that can influence operant and respondent behavior.
Referential behavior, likewise, can be affected by the biological status of a
speaker and listener (e.g., ill, tired, hungry), the physical environment in
which a speaker and listener interact (e.g., a stadium, church, bedroom), and
the sociocultural practices of the group to which a speaker and listener
adhere. As to the latter type of setting event, the imaginary interaction
between the teenager and his father involving the car wreck would be more
civilized, presumably, if the two were at a busy restaurant. This type of
sociocultural setting condition would have an inhibiting effect on the
interactions between father and son; after all, it is unacceptable in polite
society for a parent to chastise their child in public.
The diagram in figure 6.1 depicts a single episode or unit of referential
behavior and interaction with all four factors: speaker, listener, referent, and
setting conditions. Note that the speaker, in Phase 1, initiates an interaction
by responding simultaneously to a listener and a referent. In Phase 2, the
listener responds, in turn, to the speaker and to the speaker’s referent, thereby
completing the interaction initiated by the speaker. Note, too, that the
interaction between the speaker and listener occurs under the influence of
setting conditions.

Figure 6.1. A single episode of referential behavior and interaction.

In this figure, the speaker is one person, the listener another. Speakers
and listeners can differ along countless dimensions such as age, gender, race,
education, occupation, avocation, social status, and level of development.
The referent may involve all manner of things, objects, people, pets,
relationships, places, events, and activities. It may be real or imaginary, fact
or fiction, present or absent, concrete or abstract, timeless, timeful, or
ephemeral.
Referential interactions may be narrative, mediative, or both. In
narrative referential interactions, one person’s referential behavior is
coordinated with another person’s referential behavior, as in casual
conversation, the prototype referential interaction. The content and manner of
narrative interactions range widely, to say the least, from the simple,
everyday exchanges between two coworkers over the weather to two heads of
state talking about the growing threat of nuclear annihilation. The sheer
volume and breadth of narrative interactions over a person’s lifetime, in a
word, is astounding. It is hard to imagine a life devoid of narrative
interactions. It would be hard to differentiate Homo sapiens from other
species without this fundamentally human proclivity, and harder still to
envision a society in which people never talk and listen to one another about
the things and events that matter to them.
A mediative referential interaction, in contrast to a narrative
interaction, occurs when a speaker’s referential behavior is coordinated with
a listener’s nonreferential behavior. There are two major types of mediative
interactions, each distinguished by whether a speaker’s referential behavior
precedes or accompanies a listener’s nonreferential behavior. When a speaker
asks or tells a listener to do something such as open a door or pass the salt
and the listener complies, a preceding mediative interaction is identified.2
An accompanying mediative interaction, by comparison, commonly
involves a speaker and a listener who are each engaged in a nonreferential
activity that stimulates one or the other person to comment on it. Consider
the mother who says to her daughter as they garden together, “You’re doing
great!” and the child continues to garden without speaking or gesturing. In
addition to identifying this as an accompanying mediative interaction, it may
be important, for example, to determine whether or not the mother’s
referential behavior functioned to reinforce more frequent or longer bouts of
gardening with her daughter in the future.
A speaker’s (and listener’s) referential behavior, whether mediative,
narrative, or both, varies in content and style according to each person’s
history of learning and present circumstances. Encompassed in this history
and circumstance is the role that society plays in developing a speaker and a
manner of speaking that is at once both conventional and idiolectical. As a
matter of living in a group, a person learns, along with other members, the
customs, norms, traditions, values, beliefs, and manners of the group and,
along with other members, learns to speak a common language and to talk to
certain people in certain situations according to convention. It is cultural
behavior, in Kantor’s view (1936, 1982), a mode of action that prominently
includes referential behavior between the members of the group on a regular
basis and with respect to a common set of customs, values, and the like. At
the same time and also as a matter of living under conventional
circumstances, a person learns to speak in a style or manner that is instantly
recognized by family, friends, associates, and admirers as that person.
Speaking is much like a person’s fingerprints in this regard, absolutely
unique, utterly distinctive, and seldom mistaken for someone else’s.
 Before turning to what qualifies as language behavior but not as
referential behavior—namely, symbolizing behavior—it will be helpful to
take a closer look at the distinction Kantor drew between the actual behaviors
of speaking and listening, on the one hand, and the products of these
behaviors, on the other hand. Understanding Kantor on this point is vital to
understanding his approach to language in general and referential behavior in
particular.

The Trouble with Words

It is customary for psychologists and linguists, in the ancient tradition of
philology, to study language behavior and its development in terms of
written words and sentences and their syntactic and semantic properties.
The approach is unproductive, in Kantor’s view, because it confuses the
products of behavior with the behavior itself. Consider the following chain
of events: (1) a language sample of two people talking to each other is
taken, (2) the sample is either transcribed into ordinary words and sentences
or transformed into a notational system (e.g., International Phonetic
Alphabet; Bloom, 1993), (3) various counts and calculations are performed
on the reconstituted sample, and then, once the results are in, (4) an account
is given as to how and why the two people in the sample were able to speak
and understand what was on each other’s mind. Subjecting the behavior of
two people talking to each other to this type of analysis and then calling it a
psychological analysis is the problem, according to Kantor, and the only
way around it is to replace words and sentences as the unit of analysis with
referential behavior and interaction.
We turn next to a distinction Kantor drew between referential behavior
and symbolizing behavior. The symbolic type of language behavior, as with
the referential type, is inherently interactive and fundamentally
communicative. However, while the referential type normally involves
verbal-vocal behavior and/or gestures, as in casual conversation, the
symbolic type centers on creating and using symbols. A distinction is
therefore made between two types of language interactions, referential and
symbolic, each of which requires a separate analysis.

Symbolizing Behavior
The symbolizing type of behavior involves a person responding to a
stimulus that has been designed or created to substitute, stand for, or point
to something else, either by that person or by some other person. The
behavior of a person with respect to a thing that denotes some other thing or
event, in short, is called symbolizing. Symbols take many forms—pictures,
diagrams, figures, words—and serve many functions—identify, inform,
remind, direct, warn, and so on. The property common to all symbolizing is
that there is a stimulus that, ideally, corresponds unambiguously to a given
circumstance and regularly produces the correct or appropriate response(s)
to it. Said another way, there is ordinarily a firm relationship, first, between
the symbol and the thing symbolized, and second, between the symbol and
the behavior for which the symbol was created. Consider the road sign that
reads, DANGEROUS CURVE AHEAD. The sign has a fixed relationship
to the condition of the road ahead and, assuming a driver attends to the sign
and is oriented to what it signifies, the behavior of slowing down is
virtually guaranteed.
A second example of symbolizing behavior involves the exit sign above
a door at a movie theater. The sign is placed above the door to show people
where to leave from once the show is over and where to go in case of an
emergency. A person’s response to this type of stimulus may be simply
orientational, such as noticing the sign upon entering the theater and taking a
seat, or it may entail a more extended response such as noticing the sign and
then leaving the theater through the door under the sign after the movie is
over. Speculating on the origins of this type of symbol, it is conceivable that
someone actually stood guard at the door to a theater in an earlier era, say, in
Shakespeare’s time, greeting patrons when they arrive to the performance,
showing them where to leave from when they depart and where to go in case
of an emergency. The exit sign in a modern movie theater serves essentially
the same function, now unaccompanied by the referential behavior on which
these types of sign are based.
The symbolizing type of language behavior is at least as prevalent,
public, conspicuous, and central to life in society as the referential type. The
similarity between the two types ends there, however. With symbolizing
behavior, the stimulus-response relation is linear in operation and rigid in its
function; only a narrow range of responses and stimuli are tolerated by virtue
of the function for which they are designed. This differs greatly from
referential behavior, where multiple sources of stimulation arising from
speakers, listeners, referents, and setting conditions do not symbolize,
substitute, or stand for anything, and where considerable variation in
conventional and idiolectical behavior especially is allowed and even
nurtured.
We turn now to the last section of the chapter on selected advancements
in research on referential behavior. The advancements include (1) the
codevelopment of a method for analyzing referential behavior and a
procedure for conducting descriptive and experimental research, (2) research
on interventions for remediating deficits in the referential behavior of young
children with intellectual and developmental disabilities, and (3) the results
of a longitudinal study on the development of referential behavior in a young
child with autism over the course twenty-five months of intensive behavioral
treatment.

Selected Advancements in the Analysis of

Referential Behavior
The development of a method for analyzing referential behavior and
interaction, together with the development of a procedure for conducting
research that complimented the method of analysis, is significant for this
one reason: Kantor’s approach to referential behavior, until then, was long
on conceptual analysis and short on research. With a method of analysis and
a procedure for conducting research available for the first time, the path was
clear, as we shall see, to investigate a variety of topics.

Analyzing Referential Behavior and Interaction

Shortly after Kantor’s death, Bijou and his colleagues published two
articles describing a method for analyzing referential behavior (Bijou, Chao,
& Ghezzi, 1988; Bijou, Umbreit, Ghezzi, & Chao, 1986). A third article
served as a manual for teaching people how to code videotapes of research
sessions based on the method (Ghezzi, Bijou, & Chao, 1991).
The method was designed as a direct observation system, suitable for
descriptive and experimental research in any setting involving two people, a
speaker and a listener, who talk to each other while being filmed for later
analysis. The analysis entails two phases. In the first, trained raters identify,
in successive one-minute intervals, (a) the interactions initiated by the
speaker and responded to by the listener, termed “complete interactions,” and
(b) the interactions initiated by the speaker to which the listener did not
respond or responded irrelevantly, termed “incomplete interactions.” This
first phase distinguishes between referential and nonreferential behaviors and
interactions, provides details on the number of complete and incomplete
referential interactions initiated by each person in the dyad, and tracks the
continuation or change in the referent(s) over the entire course of the
conversation.
In the second phase of analysis, the experimenter records the duration
and modality (verbal-vocal and/or gestural) of a speaker’s initiation,
identifying any accompanying concurrent and repetitive behaviors, such as
“fidgeting,” and determining the referential content of the initiation. This
includes time frame (past, present, future, or no time frame), actuality (real,
imaginary, or a combination of both), and references to persons (self, listener,
others), animals, objects, places, activities, and the like.
The experimenter also identifies the prevailing setting conditions during
the second phase. Three categories—physical, social, and interactional—are
used. Research conducted in a small room at an elementary school, for
example, would constitute one set of physical (room) and social (school)
setting conditions, while research conducted in a room of a child’s home
would constitute a different set of conditions. The interactional setting, which
refers to situations wherein two people agree to communicate according to
some set format, as in taking turns telling knock-knock jokes or reciting
movie lines, may vary over the course of a conversation, in which case the
initiation is classified accordingly.
For each interaction initiated by a speaker, there is a corresponding
response by the listener; thus the listener’s response is also analyzed. The
analysis is nearly identical to the analysis performed on the speaker: the
experimenter records the duration of the listener’s response and classifies it
according to modality; notes any concurrent behavior that may have
accompanied the listener’s response to the speaker’s initiation; documents the
referential content of the listener’s response to each initiation; and records
whether the interaction was narrative, mediative, or a combination of the two.
The experimenter also determines whether or not the interaction that
immediately follows the interaction under study was initiated by the same
person—the speaker—and if so, whether it was a continuation of the referent
in the preceding interaction (the speaker stayed on a topic), or whether the
referent was changed (the speaker changed the topic). Alternatively, the
listener may change the referent, thereby becoming the speaker, provided the
other person in the dyad responds relevantly to the new referent, in which
case he or she now becomes the listener. The value of these designations is
that they show a person’s tendency to sustain interactions on the same
referent, as opposed to either changing referents frequently or letting the
other person in the dyad initiate interactions by introducing their own
referents. These designations, incidentally, are aspects of referential behavior
that figure prominently in the section below on teaching referential behavior.
The first experimental test of the method (Ghezzi, Bijou, Umbreit, &
Chao, 1987) centered on the well-established finding that a child’s language
is altered in response to the age of the person they are addressing (e.g., Shatz
& Gelman, 1973). Ghezzi, Bijou, Umbreit, and Chao (1987) developed a
procedure whereby a preadolescent spoke to either a six-year-old child, an
age-peer, or an adult for ten minutes a day over several successive days. The
dyad met at the same time each day in a small room with decorated walls and
some furnishings at the elementary school the children attended. A session
began with a brief familiarization period conducted separately for each dyad
in which the individuals were allowed “to look at the camera and television”
before taking their seat on a bean bag chair in front of a video recorder and
TV. Once seated, the dyad was told that they would be watching a short
segment of a popular television show and that after it was over, they could
“talk about the segment or anything else.” The experimenter left the room
and returned ten minutes later to end the session.
Both the procedure and the method of analysis exceeded our
expectations. The results confirmed and elaborated on the finding that a
child’s language—what is technically referential behavior—is altered in the
presence of people who differ in age. The preadolescents in the Ghezzi,
Bijou, Umbreit, and Chao (1987) study talked for a short time to a six-year-
old child about a large number of referents, yet spoke at length about a
comparatively smaller number of referents to a thirty-year-old adult. As to
the referents, the preadolescents talked mostly about themselves or their
family and friends when speaking to the adult, yet talked most often about
the child or the child’s family and friends with the six-year-old.
 The combination of (1) creating a method for analyzing referential
behavior and interaction, (2) developing a procedure that fit the method, and
(3) using the method and procedure to confirm a well-established finding in
traditional child language research was significant. Notable, too, was that it
stimulated additional research on the effects of varying the composition of
speakers and listeners in a dyad.
Lyons and Williamson (1988), in a study involving dyads of adults with
schizophrenia and typical adults talking to one another in a lounge-like room,
found that the average duration of interactions initiated by both typical
speakers and those with schizophrenia was longer when the listener had
schizophrenia. Speakers with schizophrenia, compared to typical speakers,
initiated interactions of the longest duration regardless of the listener. Lyons
and Williamson also found that typical speakers talked much less about
themselves than the speakers with schizophrenia did, who talked about
themselves almost exclusively and in equal measure with both typical
listeners and those with schizophrenia.
Chiasson and Hayes (1993) formed dyads of undergraduate first-year
students, undergraduate seniors, and graduate students to investigate how the
“social status” of a member affects one another’s referential behavior and
interaction. The first-year students initiated roughly twice as many
interactions when the listener was a fellow first-year student compared to
when the listener was a senior or a graduate student. Chiasson and Hayes also
reported that the average duration of each interaction was longest when the
first-year students spoke to one another, intermediate when they spoke to a
senior, and shortest when they spoke to a graduate student.
This collection of findings by Ghezzi, Bijou, Umbreit, and Chao (1987),
Lyons and Williamson (1988), and Chiasson and Hayes (1993) shows clearly
that a speaker’s referential behavior is altered by the characteristics of the
listener with whom the speaker is interacting. That this conclusion is drawn
from results based on a method and procedure founded on Kantor’s analysis
of referential behavior is sufficient reason to take the method and procedure
in a different direction. The direction Bijou and his colleagues took was
teaching referential behavior to young children with intellectual and
developmental disabilities.

Teaching Referential Behavior

 In a paper on the development of referential behavior, Bijou (1989)
described an experiment on social skills training by Ghezzi, Bijou, and Chao
(1987; Ghezzi & Bijou, 1994). The study involved three elementary school-
aged children (ages six to nine) with mild intellectual delays whom school
officials described as unable to carry on a conversation with their typically
developing classmates and who were suffering socially because of it. The
experimenters tested and developed two interventions. One intervention (A)
involved showing a child selected excerpts of videotapes that depicted
exemplary and nonexemplary instances of the child initiating a referential
interaction or sustaining a series of referential interactions while talking to a
typical classmate. An adult trainer taught the target child to eventually
identify, without assistance, the exemplary and nonexemplary instances and
to practice socially desirable alternatives.
The second intervention (B) involved teaching a child to initiate and
sustain a series of referential interactions on referents either supplied by an
adult trainer (e.g., plan a party, build a clubhouse) or offered by the child
(e.g., favorite sports personalities, toys). The adult, through role-playing
exercises, taught the child to talk about, and elaborate on, the adult-supplied
or child-offered referents. Toward the end of a session, the adult gave the
target child two assignments: (1) to have a conversation with the typically
developing member of the dyad (an age-peer) about the same referent(s)
introduced and discussed during role-playing, and then, right after the
upcoming conversation with the typically developing child was over, (2) to
report to the adult on the details of what they had just talked about.
The effects of both interventions were evaluated during five-minute
probe sessions in which the dyad (the target child and the typically
developing peer) was asked to “have a conversation about anything you
like.” For the A intervention, both children in the dyad were escorted back to
their classroom at the end of the session. For the B intervention, the target
child stayed with the adult for about fifteen minutes to report the details of
the conversation the child just had with the typically developing member of
the dyad.
Intervention A produced three main outcomes: (1) an increase in the
number of referential interactions initiated by a target child (i.e., the child
started more conversations with the typically developing peer in the dyad),
(2) an increase in the proportion of the interactions initiated by a target child
that either related to or differed from the referent(s) the child first introduced
(i.e., the child either continued talking about a topic or changed the topic
altogether), and (3) an increase in the number of relevant responses by a
target child to the referential interactions initiated by a typically developing
peer (i.e., the child responded relevantly to more of the typically developing
child’s initiations, thereby completing more interactions). In contrast to the
first two increases, the third increase occurred without specific instruction.
Intervention B produced two main outcomes: (1) an increase in the
average duration of a referential interaction initiated by a target child (i.e., the
child spoke longer while starting a conversation, and (2) an increase in the
average duration of a target child’s responses to interactions initiated by the
typically developing child in the dyad (i.e., the child spoke longer in response
to the typically developing child’s initiation). This second outcome occurred
without teaching.
The analysis of the details of the referential behavior and interactions
within and across dyads revealed no large or consistent differences between
the two interventions. The children spent roughly half the time talking about
each other’s interests, relatives, friends, and pets, and spent the remaining
time talking about themselves, other people, places, and activities.
The results of this study were significant for several reasons. The two
interventions improved a target child’s referential behavior as a speaker in
three ways: the child introduced more referents, talked more about them, and
talked about them for longer periods of time. The interventions also produced
two uninstructed improvements in a target child’s behavior as a listener: the
child responded more often, and for longer periods of time, to the interactions
initiated by the typically developing member of a dyad.
That the method for analyzing referential behavior and interaction
yielded measures that were highly sensitive to the effects of the interventions
paled in comparison to the revelation that referential behavior is
exceptionally meaningful to the social development of a child. It is widely
acknowledged that a great many children with (and without) disabilities have
a problem interacting with their age-peers and are socially disadvantaged
because of it (Guralnick, 2001). It so happens that many of these children
differ from their age-peers in their preference for solitude over
companionship (Wang, 2015). Under prolonged circumstances, a minimal
repertoire of desirable social behavior would develop, as friendships and
personal relationships would be difficult to acquire and maintain, leading
eventually to isolation, loneliness, depression, and, according to some
experts, self-harm and suicidal thoughts (Endo et al., 2017).
More often than not, according to Ghezzi (1999), modifying the social
behavior of a child is a matter of (1) assessing how the child interacts
referentially with other children and adults, (2) intervening on those aspects
of the child’s referential behavior that are deemed deficient, and (3)
evaluating the effects of the intervention on the child’s referential behavior
and interactions with adults and age-peers. It is a familiar process, one that
applied researchers and practitioners alike follow assiduously. What is
unfamiliar about the process as it pertains to referential behavior is the most
important part: how to measure referential behavior and interaction for
purposes of its ongoing assessment, intervention, and evaluation.
A study by Willmoth, Lewon, Taylor, and Ghezzi (2017) is a recent
development in the analysis of referential behavior that illustrates the
process. A five-year-old boy with autism was nearing the end of nearly two
years of early intensive behavioral intervention (EIBI; Klintwall & Eikeseth,
2014). Informal observations of his referential behavior with his home-based
adult tutor revealed that while he initiated now and then, he seldom continued
to keep initiating on the same referent(s). He would say that he was looking
forward to going to the park this afternoon, for instance, and then, after the
adult completed the interaction (e.g., “That will be fun!”), he would either
say nothing more about the park or would initiate the very next interaction by
introducing an entirely new referent (e.g., “Hot dogs are my favorite food”).
Willmoth et al.’s (2017) study began with the assumption that a “good
speaker” continues to initiate interactions on a referent at a frequency
appropriate to the listener, setting, and circumstance. Following a series of
baselines sessions, described below, the researchers decided that the
frequency at which the child continued to initiate interactions on a referent(s)
he himself introduced while talking to his tutor was low. The researchers
developed a referential teaching procedure to improve his behavior as a
speaker by increasing the number of times he continued to initiate an
interaction with his tutor on a referent he himself introduced. Recall that the
same aim was set for the young children with intellectual disabilities in the
study reported by Ghezzi, Bijou, and Chao (1987; Ghezzi & Bijou, 1994) and
discussed by Bijou (1989), the details of which we discuss and interpret
below as meaningful to the social development of a child. The route taken
with the young child with autism differed from the route taken by the
children with intellectual disabilities, however.
Willmoth et al. (2017) conducted baseline sessions over eleven days for
five minutes each day in a small room in the child’s home. The tutor told the
child that they would be “having a conversation” and that they could “talk
about whatever they liked.” The tutor refrained from taking on the role of
speaker by initiating interactions of her own during baseline, and instead
replied to the child’s initiations without elaboration and without regard to
whether or not he continued to talk about the referent(s) he introduced. If the
child remained silent, so, too, did the tutor, until either the child initiated a
referential interaction or the five-minute session timed out.
The procedure followed a changing criterion protocol whereby social
praise was contingent on the child’s initiating a series of interactions on a
referent he introduced. If the child started talking about a video game, for
example, the tutor would (1) reply to his initiation (e.g., “Video games are
amazing!”) and (2) provide praise (e.g., “What you told me about playing
those games is awesome!”), at first contingent on a relatively smaller number
of continuations and later, on a comparatively larger number, depending on
the number of continuations recorded the last time he had a conversation with
his tutor.

Figure 6.2 Mean number of continued referential interactions by the

child as speaker.
 The mean numbers of referential interactions continued by the young
child per five-minute session throughout baseline and over twenty-three days
of the referential behavior training are shown in figure 6.2. The average
remained low during baseline at roughly four per session, meaning that, on
average, the boy continued to initiate about four consecutive interactions on a
referent he introduced. He exceeded that average across the first criterion
and, for the most part, across each and every one of the five subsequent
increases in the criterion for contingent praise. Near the end of intervention,
the child was initiating roughly eight interactions per session with his tutor
on referents he introduced, twice as many as in baseline.
It became apparent that while the child was improving as a speaker, his
behavior as a listener was “poor” in the sense that he seldom continued to
talk about a referent introduced by his tutor. The tutor would start talking, for
example, about her dog (“My dog loves to play”), and the child’s typical
response was to make a simple statement (e.g., “Dogs bark”) or to ask a
simple question (“Does your dog bark?”) and leave it at that without
additional follow-up or continued referential behavior relevant to his tutor’s
dog and what she said about it. It was also apparent that the procedure
developed to improve his behavior as a speaker—sustaining a series of
interactions on the referent(s) that he introduced—had no discernable effects
on his behavior as a “good listener,” someone who, presumably, continues to
talk about a referent introduced by a speaker at a rate appropriate to the
speaker, setting, and circumstance. The baseline rate of this behavior was
considered low for this child, which stimulated a procedure designed to
increase the number of times he continued to talk about a referent(s) his tutor
introduced.
Baseline sessions were conducted over eight days for five minutes each
day in the same small room in the child’s home. The adult tutor told the child
that they would be having a conversation and that he could talk to her about
whatever he liked. The tutor assumed the role of speaker by initiating
interactions with the child on referents that she selected from a list generated
from an informal assessment of his most preferred things and activities (e.g.,
games, toys, movie characters, bike riding, baseball). The tutor refrained
from encouraging the boy to continue interacting about her referent(s), and
instead simply introduced a referent (e.g., “Your bike must be fun to ride!”),
waited for a response (e.g., “It is!”) to complete the interaction, and then
proceeded either (1) to initiate another interaction on the same referent (in
which case the boy continued to talk about the tutor’s referent), or (2) to
initiate a new interaction on a different referent (in which case the boy did
not continue to talk about the tutor’s referent). If the child initiated an
interaction by introducing his own referent (e.g., “Do you eat spaghetti?”),
the tutor, now listener, would complete the interaction (e.g., “Yes”) but
without encouragement to the child to continue interacting with her on his
referent.
The procedure followed a changing criterion protocol, as before, but this
time with praise contingent on the child’s continuing to respond to a referent
introduced by his tutor in her role as speaker. For example, if the tutor started
talking about Spiderman (e.g., “Spiderman is awesome!”), she would
continue to initiate interactions about the character so long as the child
continued to give a response that was relevant to the referent (e.g., “He’s the
best!”). When the child met or exceeded the criterion number of
continuations set for the day, the tutor would praise the child and then
continue to initiate another interaction on the same referent (e.g., “Thanks,
buddy, for talking more to me about Spiderman! I wonder if his mom and dad
know he’s Spiderman.”). This cycle of events continued throughout each
session, provided the child kept giving a response that was relevant to his
tutor’s referent. If his response was irrelevant, if he made no response at all,
or if he responded to his tutor’s referent by introducing a referent of his own
(e.g., “It’s cold outside”), the tutor would look away from him and wait for a
period of roughly ten seconds of silence to elapse before initiating an
interaction on a referent from her list of his favorite things to talk about.
The mean number of continuations by the young child as a listener
throughout baseline and intervention are shown in figure 6.3. The average
decreased from nearly eight continuations per five-minute session at the start
of baseline to about four per session thereafter. He exceeded that average
once intervention began at the first criterion (four continuations) and, with
one exception (Session 25), kept meeting or exceeding the criterion that was
set each day for continuing to talk about his tutor’s referent. Near the end of
the intervention, the boy continued to interact with his tutor on her referent
about fourteen times on average, roughly a three-fold increase above
baseline.
Figure 6.3 Mean number of continued referential interactions by the
child as listener.

How the combination of initiating interactions on the child’s favorite

referents, not responding to initiations on referents that he himself
introduced, and changing the criterion for contingent praise contributed to the
results of the Willmoth et al. (2017) study are questions for further study. The
main point is that as far as referential behavior is concerned, it is entirely
possible, and desirable, to assess, teach, and evaluate the effects of teaching
referential behavior as a speaker, a listener, or both.
We turn now to a third and final area of advancement in referential
behavior research involving the development of referential behavior in a
young child with autism over the course of twenty-five months of EIBI.

The Development of Referential Behavior

Bijou (1983, 1989) published two papers on the development of
referential behavior from infancy through the preschool years. He described
the typically developing child as progressing through four stages from
infancy to roughly four years of age: (1) preverbal-vocal communication, (2)
primitive referential verbal-vocal interactions, (3) first-approximation
referential interactions, and (4) second-approximation referential interactions
and the beginning of symbolizing behavior. The mayor types of behaviors
and interactions taking place during these periods would populate each stage,
according to Bijou, and each stage, in turn, would be further marked by its
own transition from the previous stage to the next stage. Later stages of
referential behavior would extend through the early elementary school years
and beyond into adulthood and eventually old age.
The process of developing a repertoire of referential behavior unfolds
naturally and without delay for most young children. For a significant
minority of children, however, the process can be slow or even stopped in its
tracks. When it is, parents worry about delays and deficiencies in the
development of their child’s language, communication, and social behavior.
Chronic impairments in these three domains in particular can lead to a
diagnosis of autism and, in turn, to treatments designed to address a child’s
deficits in these same three domains.
Because referential behavior entails all three domains, a child’s social
behavior and functional speech, and perhaps his or her grammatical behavior
as well, should develop markedly as the deficits in the child’s referential
behavior are replaced with more appropriate or desirable referential
interactions with age-peers and adults. It is a working hypothesis, one that
gains support by the results of the studies on teaching referential behavior
reviewed above. It is also a hypothesis that acknowledges the simple yet
powerful truth that at the end of the day, what a parent of a young child with
autism wants most is to have a conversation with their child (Pituch et al.,
2011; Richards, Mossey, & Robins, 2016).
A. Lewon (2019), in an ongoing longitudinal dissertation study, asked
these three questions: How does a young boy with autism speak to his
mother? How does the mother speak to her son? How do these referential
interactions develop over the course of an EIBI program? Lewon posed these
questions at the start of a young boy’s program, which began shortly after his
third birthday and ended just after his fifth birthday, two years later. The
child’s mother, also a wife and homemaker, reported at intake that her and
her husband’s first priority was to communicate and be more socially
engaged with their son.
Once a month for twenty-five consecutive months, an adult assistant
arrived at the family home to videotape a five-minute conversation between
the mother and her son while they were seated on a couch in the family living
room. The assistant instructed the child and his mother “to stay seated for the
next five minutes and have a conversation about anything you like.” The
assistant left the room and returned five minutes later to end the session.
 There were no formal contingencies in effect at any time during the
videotaped sessions that related either to the mother or her son, nor were
there ever any procedures in place over the course of the child’s EIBI
program that isolated referential behavior as a treatment target. The videotape
sessions, then, were probes designed to track the development of the child’s
referential behavior for the duration of his intensive intervention program for
autism.
The top panel in figure 6.4 displays two data streams, one representing
the frequency of complete referential interactions initiated by the child’s
mother (open squares) and the second, the frequency of complete referential
interactions initiated by the child (closed squares). What constitutes a
complete interaction, to reiterate, is a relevant response by a listener to a
referent introduced by a speaker, and what constitutes a speaker is the one
who initiates, or continues to initiate, an interaction on a referent that they
introduced.
During the early months of intervention, as the top panel in figure 6.4
shows, the mother was far more likely to initiate a referential interaction with
her son than her son was to initiate an interaction with her. Said another way,
the mother had plenty to say to her son, but he had nothing to say to her as a
speaker. This imbalance changed gradually but steadily over the course of
treatment, culminating at eighteen months in a fairly even mix of interactions
initiated by the child and his mother. Proportionately, the child initiated about
52% of the interactions with his mother during the last three months of
intervention, compared to roughly 26% during the first three months of
intervention.
 Figure 6.4 Frequency of complete interactions (top panel) and
incomplete interactions (bottom panel) initiated by the mother and
child as speakers in each monthly sample.

The bottom panel in figure 6.4 contains two data streams spanning the
same twenty-five consecutive months of the mother and child’s interactions.
One data stream shows the frequency of incomplete referential interactions
initiated by the mother (open squares), and the second stream shows the
frequency of incomplete interactions initiated by her son (closed squares).
The mother’s data show that over the first seven months of treatment, she
initiated between fourteen and twenty-four referential interactions each
monthly session to which her son either did not respond or responded
irrelevantly. The mother, in other words, had plenty to say to her son, yet he
had nothing to say back to her as a listener. The number of these incomplete
interactions fell steadily over the course of intervention to near-zero levels by
the end of intervention.
The child’s data in the bottom panel of figure 6.4 show that the number
of referential interactions he initiated that his mother failed to complete,
either by not responding or by responding irrelevantly to her son’s
referent(s), remained at near-zero levels throughout the study. The mother
completed all but a few interactions initiated by her son over the entire course
of treatment, as might be expected of any parent who wants their child to talk
to them. The contrast between the child and his mother on this measure is
striking at first, given how many more interactions she initiated that he failed
to complete compared to the number he initiated that she failed to complete.
The child improved greatly in this regard such that by eighteen months into
intervention and for the final seven months of intervention, he completed
most if not all of the referential interactions initiated by his mother.
These positive developments in the child’s referential behavior occurred
without any direct or specific instruction. They were emergent effects, in
other words, clearly the result of the cumulative effects of over two years of
EIBI, and yet just as clearly a result that is impossible to attribute to any one
aspect of the intervention. The child with autism in the Willmoth et al. (2017)
study, on the other hand, also received over two years of EIBI and yet his
referential behavior was poor by comparison and did not “emerge” for the
better until a program was implemented near the end of intervention to
improve his performance as a speaker and a listener. Accounting for
differences of this sort in the context of the aims and values of EIBI
constitutes a potentially lively area for research.

Implications and Future Directions

At the start of this chapter I made the claim that referential behavior is
understudied. It is, in fact, woefully understudied in comparison with the
scope and significance attached to it. One speculation is that the analysis of
referential behavior in Kantor’s time was entirely conceptual and that a
clearly identified unit of behavior and the numerical measures pertaining to
it was unavailable (Ghezzi, 1999). That has changed, and now there is a
functional unit of analysis for referential behavior with standard measures
of count and time attached to it. The frequency with which a speaker
initiates a complete referential interaction, and the time the listener spends
responding to the speaker’s initiation, is one example of two metrics
available to researchers within a single referential interaction. In a nutshell,
the methods and measures described here open a door to countless
experimental and descriptive studies on the development of referential
behavior from infancy to old age.
In keeping with the theme of this book and the need, in my view, to
introduce new and alternative procedures for enhancing the social
competence of young children with developmental disabilities, we shall keep
mostly to applied matters of this sort.
How the analysis of referential behavior might inform the techniques of
contingency management, and vice versa, is an area that deserves close
attention. The use of contingent praise enabled the young child with autism in
the Willmoth et al. (2017) study both to initiate a greater number of complete
interactions as a speaker and to continue interacting with his tutor as a
listener. It is a fine example of what is possible when episodes of referential
behavior and interaction are treated as units to which contingent
consequences can be applied as a means to develop or refine a repertoire as a
speaker, a listener, or both.
The tradition in applied behavior analysis is not just to demonstrate a
reliable change in behavior but to show a socially relevant and personally
meaningful change as well. Social validity, in other words, is a value to
which behavior analysts have long adhered (Wolf, 1978). As this pertains to
modifying referential behavior, it is presently unknown whether or not a
change in a child’s referential behavior is related to a change in that child’s
status as a friend or playmate, for instance, or whether or not the child or the
child’s parent would regard the change in referential behavior as socially
significant or personally desirable. Improvements in a child’s social
competence may be assessed by means of peer nomination and rating scales
(e.g., Ghezzi, Robles, & Bijou, 1992), by asking a child’s parent or the child
themself to rate their level of satisfaction with the change in behavior (e.g.,
Hood, Luczynski, & Mitteer, 2017), or by asking total strangers to judge the
social appropriateness of a child’s behavior (e.g., Peters & Thompson, 2015).
One possible variable in this regard centers on the two cases of teaching
referential behavior described above. The instructors decided to increase the
number of interactions the child initiated to his tutor as a speaker in the
Willmoth et al. (2017) study, and once that was achieved, the instructors
increased the frequency of his continuations as a listener to his tutor’s
initiations. It seems self-evident that a person who strikes an even balance
between initiating referential interactions and responding to the initiations of
the other person in a dyad would be judged as a “good speaker” and a “good
listener,” respectively. This may near the mark, or not, but in any case it
seems to be a fruitful area for research, one that could provide practitioners
with guidelines on the characteristics of referential behavior and interaction
that facilitate making and maintaining friendships, for instance, or that foster
a child’s popularity among age-peers.
How referential behavior develops over the lifespan is another area ripe
for research and full of applied significance. Bijou (1983; 1989) provides the
framework for exploring the conditions under which an infant moves from
gestures and babbling to holophrastic speech and the beginnings of
referential behavior. The stages Bijou outlines may be useful guideposts for a
young child’s social development especially, and may be useful as well to
people in the field of early intervention who work closely with parents to
identify and ameliorate deficits in the initial development of a child’s
language behavior before problems arise at later stages.
One last direction for the future pertains to the aims and values of EIBI.
Most young children develop an adequate repertoire of referential behavior
and enjoy interacting with their peers, siblings, parents, and other adults
throughout childhood, adolescence, and adulthood. Other young children fall
below this trajectory, but the child with autism falls farther below most. For
this child, as well as for the child’s parents, referential behavior can be
considered the applied endpoint (Pituch et al., 2011). If this is anywhere near
the mark, then establishing a treatment environment that effectively teaches
the referential behaviors necessary for a child to become a competent speaker
and listener at home, in school, and out in the community could be the key to
a successful outcome.

Conclusion
The study of referential behavior was inspired by the many writings of J. R.
Kantor, the founder of interbehaviorism. In lay terms, referential behavior
refers to two people talking to each other about something. Research has
shown that referential behavior is established at a very young age in
children, and then expands greatly over the course of an individual’s
interaction with their verbal community. In fact, the repertoire is at the heart
of many aspects of personal and professional life. This chapter identified
the different types of referential interaction, including mediative and
narrative, and provided examples of each. A number of studies have
analyzed children’s referential interactions, results of which demonstrate the
important role of referential behavior in social competence.

Study Questions
1. Describe referential behavior.
2. Distinguish between the referor and the referee with an example.
3. What is “bistimulation”? Explain with an example.
4. Describe two setting conditions that could influence the occurrence
of referential behavior.
5. What is a narrative referential interaction?
6. What is a mediative referential interaction?
7. What is the problem with studying sentences and words as a means
of understanding language in behavioral perspective?
8. What is symbolizing behavior?
9. What is the main difference between referential and symbolic
behavior?
10. What is the general finding of the series of studies by Ghezzi, Bijou,
Umbreit, and Chao (1987), Lyons and Williamson (1988), and
Chiasson and Hayes (1993)?
11. Why is referential behavior a socially significant target behavior?
12. What are the results of the longitudinal study by Lewon described in
the chapter?
 Chapter 7:

Human Behavior Is Referential Behavior

Emilio Ribes-Iñesta
University of Veracruz

Link and Preview

The previous chapter provided an overview of both referential
behavior and symbolic behavior, and in particular described the
clinical implications of work in this area. The current chapter
continues with the very general topic of referential behavior but
addresses the topic of language in science more generally. This
includes a discussion of terms within psychology and applied
behavior analysis specifically. This chapter will be of particular
interest to those who are interested in delving further into
theoretical issues pertaining to the topic of language in science.

“Language” and “behavior” are not technical terms, although they may be
used technically for given purposes. This is one of the reasons why
confusion often arises in psychology. Indeed, when people talk about
“psychological” terms, they often think they are talking about the same
things because they are using the same words, but this is often not the case.
Words in ordinary language are multisensical; that is, they are used in
different senses in different situations and contexts. Ryle (1949; 1962)
pointed out that oftentimes people suppose that they are sustaining a
dialogue about a common issue when, actually, two simultaneous
monologues about different issues are taking place. In ordinary language
words do not have a specific and constant referent or correspondence with
things. However, in technical languages some words have specific uses
regarding only some things. Clearly that is an important distinction between
ordinary language and the technical language of the sciences.
“Language” and “behavior” are also not univocal terms, even in the
scientific domains of linguistics and psychology. I will try to delimit the ways
in which these terms will be used from now on. “Language,” as a term,
always has to do, directly or indirectly, with the social practice of human
beings. In fact, “language” is the only term inherently reflexive; we can talk
about language in language or through language. Language originated with
human society, that is, with the organized division of labor and appropriation
of goods and mutual services between individuals (Ribes, 2001). This
explains why communication in some animal species does not constitute true
language, and why it is absurd to postulate “private” languages of different
sorts. Language emerged as conventional practice and cannot be conceived
as an individual product or process. On the contrary, as I shall argue later,
human individuals become differentiated from one another, not by language,
but in language. Language is not external to human life and practice; it is not
a means for relations among individuals, but the medium itself in which
human practice and social relations take place and, therefore, are sensical or
meaningful.
Language and social human practice are one and the same thing and are
intrinsically interlocked. No human society without language is known. In
this sense, language is not only talking and so forth, but as Ludwig
Wittgenstein (1953, p. 19) pointed out, “to imagine a language means to
imagine a form of life.” Any social form of life as language embraces a
diversity of activities, relations, products, and dispositions. It would be
impossible to set a single frame to classify them. Nevertheless, following J.
R. Kantor (1936, 1977), a useful first step may consist of distinguishing
“alive” from “dead” language, that is, distinguishing between language
taking place as acts from the vestiges of those acts.

On Human Behavior as “Language”

“Dead” language is related to the vestiges and products from actual
practice, as are recordings of all sorts: marks, drawings and pictures,
inscriptions, writings, books, codex, tapes, and so forth. Dead language
products are not equivalent to the practices from which they resulted, nor
are they an outcome of these practices, although sometimes dead language
products may provide clues from which to infer part of the conditions
prevailing when actual practices took place. Nevertheless, in the same
manner by which words and expressions do not correspond to things or
events, linguistic vestiges should never be confounded with the actual
practices that gave rise to them. By contrast, “alive” language always
consists in part of a social practice involving more than one individual—a
practice articulated with the multiplicity of interindividual relations
occurring in society. The relations so involved do not necessarily have to be
simultaneous in time and place, especially when some type of graphing or
writing is involved.
In behavioral psychology, language has been considered a special type
of operant behavior (Skinner, 1938, 1957) and has been equated with so-
called verbal behavior (also see chapters 3 and 4). However, this is a
conceptual mistake and leads to confusion. Verbal behavior deals with speech
—that is, with orally emitted language—and, although it is in some sense the
dominant modality of language practices, it is not the only one nor the most
important in all occasions. Because of this, the term “linguistic behavior” will
be used to refer to any kind of individual language activity, and “verbal” will
be restricted to speech. Alive language is the underlying practice of social
relations between individuals and involves different modes or modalities of
occurrence as behavior. These modes can be identified as active modes or
reactive modes, although they usually take place simultaneously, not only
between individuals but also with a single person. Some active and reactive
modes involve inseparable pairs, whereas some active modes usually occur
together as functional patterns.
The active modes are gesturing, speaking, and writing, while the
reactive modes are observing (not looking), listening (not hearing), and
reading (not texting). It is obvious that gesturing-observing, speaking-
listening, and writing-reading are complementary pairs, and that normally
they occur at the same time with a single individual and in other individuals
in relation: gestures are observed (sometimes as proprioceptive observation),
speech is listened to, and writing is read. These relations are synchronic and
most of the time we are not aware of them, but as developmental studies
show (King & Quigley, 1985; Marschark, 1993; Marschark, Mourandian, &
Halas, 1994), children with impaired vision and/or hearing are seriously
affected in learning to speak and in conventional social gesturing (and
reading). Individuals who do not read are unable to write, even if they copy
letters or words. On the other hand, when individuals interact with others,
active modes of linguistic behavior do not occur in isolation of each other.
While speaking and reading we gesture, and while writing we speak
“silently” or loudly in the form of reading. Not only do linguistic modes not
occur in isolation, but they are always part of an act or activity by the
individual in a given circumstance, in such a way that they are patterned as
components of situational interactions.
Linguistic behavior, in this perspective, includes aggregate behavior
patterns of the individual as part of a social practice, even though some
components of the patterns do not look linguistic given their morphology.
Behavioral response systems include biological, natural reactions and actions
that are common to all individuals as members of the human species. In fact,
psychological behavior consists of the development of new functions and
organization of biological response systems (Kuo, 1967) as a result of the
interaction of the individual with different kinds of environmental
contingencies related to objects, events, and other individuals. But human
beings, in addition to their biological responses systems, develop novel kinds
of responding derived from them, arbitrary in their morphology. These novel
response systems are, precisely, linguistic response systems. Arbitrariness in
morphology means that, from a biological point of view, there is no
necessary relation between the form of the responses and the physical and
chemical circumstances in which they occur. However, the arbitrariness in
form does not mean arbitrariness in function. On the contrary, to the extent
that linguistic response systems are conventional and shared in nature, their
functions are delimited and established by social practice. Social practices
constitute the rules “governing” language functions. Grammar came after
writing as a formal description of language uses and functions. These
functions always involve and entail necessarily social relations with other
individuals articulated with objects and events in the environment.
Given their conventional nature and arbitrariness in form, linguistic
response systems, in contrast to purely biological ones, are detachable from
the original conditions in which they are (were) first emitted or “learned.”
This feature of linguistic response systems allows individuals not only to act
in similar ways in different situations, but also to respond “in the absence” of
the circumstances (individuals, objects, and events included) in which
linguistic behavior once took place. Given this detachability of linguistic
behavior systems, human beings may relate to past and future times, different
places, absent persons or objects, and “constructed” conventional objects
(concepts, myths, nonexistent beings, and so forth). It is also possible to
relate to the linguistic behavior of absent individuals, both as alive and as
dead language.
In the history of humankind, and in individual human development, the
first modes of language were (are) gesturing and speaking, with writing being
the last one to appear (6,500 years ago and when children are taught to read
and write). With the advent of writing, words were created as transcriptions
of the articulated sounds when speaking. Words do not represent objects or
events. Words represent articulated speech sounds, and words, as
grammatical units, did not exist before writing. Like speech itself, words, as
graphic representation of speech, are a human construction. Because of this,
the advent of writing also represented the existence of linguistic
(conventional) objects in the form of words, pictograms, hieroglyphics, and
so forth. These linguistic objects coexist with “natural” objects and with
artifacts constructed by means of crafts and technology. Writing, as
patterning of stimulus objects and conditions, fills spatial and temporal gaps
in the relations between individuals under common circumstances (personal
interactions). Writing also consists of complex linguistic stimulus objects in
different modes of knowing (ordinary, scientific, artistic, formal, religious,
technological, and others), with which individuals may interact in different
places and time (impersonal interactions).

Understanding
Although what we have identified as reactive modes of language
(observing, listening, and reading) have been usually related to
understanding, this is not an accurate interpretation. Understanding is not a
different “process” or moment from speaking and writing. The separation of
understanding from the active modes of language is a mistake that leads,
most of the time, to the assumption that understanding is a special kind of
covert (not implicit) behavior, usually consisting of some sort of self-
speaking or self-observing, as in introspection (e.g., Schlinger, 2008).
Actually, understanding always involves “overt” behavior, both linguistic and
nonlinguistic in form. Understanding entails completing an episode, in
concordance with a social convention. Understanding is nothing other than a
shared social practice by at least two individuals. Linguistic behavior always
occurs in episodic situations, involving at least two individuals, although on
occasion an individual may perform two different functional roles, as in a
soliloquy. In this last case, it would be absurd to predicate some sort of “self-
understanding.”
Language, given its inseparability of social practice, wraps all human
relations and interactions. It is the medium in which all human behavior takes
place. Extending Kantor’s (1924, 1926) concept of contact medium, we could
say that language, articulating social institutional practices as interindividual
relations, is the contact medium exclusive of human behavior, a conventional
contact medium. Wittgenstein (1980, p. 678) emphasized this binding
between language and human life when saying “We’re used to a particular
classification of things. With language, or languages, it has become second
nature to us.”
Linguistic behavior (as equivalent to human behavior) is conventional in
nature (Ribes, 1993, 2006a) and, to that extent, always takes place in
concordance with those participating in any interaction or relation.
Concordance, inherent to convention, presupposes understanding. Nobody
speaks, gesticulates, or writes attempting to be understood, but actually does
so because they are always understood. When a person speaks, gestures, or
writes to another person, they are sure, beforehand, that it is being
understood. Otherwise, people would probe all the time before contacting
others. In fact, “understanding” is used in ordinary language practices only
when, by special reasons, individuals show inappropriate behavior in relation
to what is being indicated, spoken, or written to or for them. Philosophers
and psychologists, who have created or followed the myth of the “cognitive”
individual as the origin of knowledge and language, are responsible for
raising the false problem of understanding.

Behavior and Conduct

In addition to understanding, a second important concept to be examined
is that related to behavior. “Behavior” and “conduct” are neither technical
terms nor exclusive of psychological descriptions or attributions. These terms
are used in everyday language, and in Romance languages, these words (e.g.,
comportamiento and conducta in Spanish) derive from the Latin words
comportare and conducere, meaning “what is brought with,” and “to lead
together,” respectively). These terms are ordinary terms, employed not only
in reference to the doings of people, but also to changes related to direction in
inanimate objects or events. Disciplines other than psychology also employ
the term “behavior” to describe changes in conditions of particles, physical
bodies, electricity, molecules, cells, groups, institutions, markets, polls,
movement, temperature, pressure, direction, and many others depending on
the subject matter under scrutiny. “Behavior” and “conduct” are generic
terms that apply not only to movements of organisms, but to changes in the
activity or conditions of a diversity of “entities” or things. In the case of
psychology, behaviorism (Watson, 1913) naturalized the term to define its
subject matter (or at least an index of it). Nevertheless, “behavior” has no
single meaning within behaviorism.
Kitchener (1977) identified a diversity of definitions of behavior among
the various behavioristic formulations, pointing to different conceptual
subject matters despite using the same term. This is not a surprising situation,
since words and concepts are not the same. Words are the morphological
units composing the lexicon of a natural language, while concepts are the
functions that exert words. The same word may have different conceptual
meanings, that is, pertain to different conceptual frames, and at the same time
different words may be conceptually similar. Ribes (2004a) stressed that the
term “behavior” is not a neutral, descriptive, or ostensive term corresponding
to or denoting “psychological” events or phenomena. Rather, as used in
psychology, the term is an abstraction related to different ordinary uses of the
word and other akin terms that show “family resemblance” (Wittgenstein,
1953) according to some explicit or implicit criterion. Although, as noted
above, other disciplines use the term “behavior” in their descriptions, only in
psychology is “behavior” considered a term indigenous to its subject matter,
without the necessary conceptual awareness that this is not the case.
Kantor (1963) called attention to the fact that “behavior” is a term used
by different disciplines, and that psychology adopted it from biology through
the conceptual tradition of reflex physiology. Kantor was especially critical
about the organocentric conception of behavior, as an activity flowing from
the organism, either as reaction (provoked or “elicited”) or as action
(emitted). To get away from this reductionist conception, he proposed to
distinguish between biological and psychological behaviors, and afterward,
these two from cultural or institutional behavior. Psychological behavior
consists of functional contacts involving an acting/reacting individual and the
stimulus properties of objects and events in the environment, including other
individuals and, in some cases, the same individual interacting with their own
linguistic behavior. Psychological behavior is not movement, or actions, but
functional relations between individuals and particular objects (sometimes
other individuals).
Nevertheless, psychological behavior is always to be viewed from the
standpoint of the individual and not from the stimulus object or event,
especially when this is another individual also behaving. The analysis of
relations between individuals, as a functional unit, pertains to the domain of
social science. Interindividual relations constitute institutional behavior and
may be approached as a multidisciplinary enterprise between social science
and psychology (Ribes, Pulido, Rangel, & Sánchez-Gatell, 2016).
Psychological behavior takes place only as individually framed or oriented
episodic components forming part of ecological (survival) or social (living
together, coexistence) relations. There are no autonomous psychological
contacts, foreign to ecological and conventional media. Psychological
contacts constitute a special set of relations between individuals showing
differentiated reactional systems and particular stimulus objects and/or other
individuals. However, only some of these relations between individuals and
stimulus objects and other individuals qualify as psychological, always being
components of ecological or social relations. Psychological behavior is
concerned only with those relations that can be identified from the
perspective of the individual, and not from the stimulus objects or as a two-
component indivisible relation.
Psychological relations, not surprisingly, usually correspond to those
ordinary language practices including or consisting of “mental” terms and
expressions. “Mental” terms and expressions do not occur in isolation. They
are always part of social episodes, in which terms characterize the
circumstances of the individual acts in reference to other individuals, objects,
and events. “Mental” terms do not describe, denote, or refer to special
activities taking place “inside” the individual. The so-called mental terms and
expressions are constitutive of the social practice and are part of the
circumstances that constitute the episode in which they take place.
Psychological behavior always consists of an episodic relation in ordinary
language (Ryle, 1949); it involves episodes in which an individual may be
characterized, by himself and others, as remembering, learning, feeling,
perceiving, and so on. These references are relative not to something
happening in the individual, but to something taking place between the
individual and other individuals, objects, or events as the circumstances of
the relation. This is why psychological behavior may be considered as a
collection of episodic segments of ordinary language practices (Wittgenstein,
1969). These segments are part of the referential nature of ordinary language.
Reference means that when talking we always talk about something (i.e., the
talk is referring to a thing but is distinct from that thing itself) to somebody
(and in special circumstances to ourselves). Speaking (and obviously
gesturing and writing) is not a mere denotative or descriptive accompaniment
to the reactions and actions taking place when individuals behave. As an
accompaniment, language would be dispensable or redundant.
Human behavior is functional because it is integrated with language,
even in its reactive modes. Because of this, psychological “language” should
be understood as language about the individual circumstances involved in
social episodes, and nothing else, irrespective of the fact that this talking
about the individual-in-relation is in singular first, second, or third person (I,
you, or they). For example, when “remembering” talk occurs, remembering
consists of the relations taking place in the episode. Remembering talk is not
an index, description, or “reference” of something taking place in the
individual. Remembering talk is what remembering is about. The same can
be said of any “mental” term or expression corresponding to psychological
phenomena in ordinary language practices.
However, if psychological behavior is inherent in social relations
between individuals taking place in ordinary language practices, does this
mean that psychological behavior is exclusive of human behavior and cannot
be attributed to certain animal species? Usually, in daily life, we speak about
psychological reactions or acts in those animals that are domesticated or
deprived of their natural environment (e.g., zoos, circuses), animals with
which we have continuous or intermittent contact under socially established
circumstances. We attribute emotions and intentions to our pets, but we do
not do so with unknown, unfamiliar animals or with those pertaining to
invertebrate phyla, such as insects, arthropods, and echinoderms. Regardless
of the daily life extension of psychological attributes to animals, it is only
recently that the terms “mental” and “psychological” became part of ordinary
language practices, mostly as an effect of the social influence of the
psychiatric and psychological professions. In past centuries, “mind” and
“mental” were not ordinary terms. Even more, psychology, as a discipline
born at the beginning of the last century, became concerned with animal
“mind” processes, as an influence of the concepts advanced by Charles
Darwin (1974) and John Romanes (1888), in tracing back the conduct of
mankind to the evolution of species. Something similar occurred with the
incorporation of terms such as “physical,” “chemical,” and “biological” by
ordinary language practices. These terms label special technical languages
that are based on the referential functions of ordinary language but have
different purposes. The distinct domains of science do not deal with concrete
objects and situations but with general properties abstracted from them.
Scientific terms and concepts are concerned with properties and relations, not
with things and particular events.
Nevertheless, the case of psychology is distinct. Wittgenstein (1953)
commented that psychology involved experimental methods and conceptual
confusion, pointing to the fact that psychologists have incorrectly identified
ordinary “psychological” terms with alleged univocal references, in the form
of reports and descriptions of events or experiences supposed to occur within
individuals. In contrast to other sciences, psychology has assumed that terms
used in ordinary language, such as “perception,” “feeling,” “sensation,”
“memory,” “thinking,” “imagination,” and similar ones, constitute reliable
references to actual kinds of experiences, activities, or events taking place
“in” or “by” the individual. As a sequel to this historical condition,
psychology borrowed technical terms from other disciplines to become
“scientific,” while attempting, without fortune, to provide a technical status
to ordinary language terms by means of operational definitions and
constructed validity tactics. Conditioning theory, theories of cognition based
on information and computer models, and psychometric theories are
prominent examples of this historical tendency.
Although J. R. Kantor (1924, 1926) formulated a logical model specific
for psychological phenomena, he did not advance any theoretical concepts to
construct a technical language beyond the terms of ordinary language. In
order to construct a scientific theory, it is not sufficient to provide functional
interpretations of ordinary language terms. Functional analysis of language is
not equivalent to an abstract theory of the psychological phenomena inherent
in ordinary “mental” terms and expressions. A scientific theory requires a
technical language, denotative in nature, in which terms and expressions have
univocal meanings and correspond to abstracted properties, dimensions, and
relations of concrete objects and events. Nevertheless, such abstract terms do
not refer to particular concrete instances. Facts and events (Hanson, 1958) are
always “constructed” by abstraction of those events, objects, and conditions
dealt with in the daily practice of ordinary language. To the extent that
abstract technical concepts do not refer to particular events or objects, they
cannot be used as ostensive or descriptive terms in natural sceneries in daily
life. The functional grammar of abstract theoretical concepts is not the same
as the diverse grammars framing the sense of terms in ordinary language, and
it would be a serious logical mistake to establish or assume direct
correspondences between them. Scientific language is denotative, whereas
ordinary language is referential. It is hard to imagine a language in which
talking consists of naming and describing objects or events. Paradoxically, in
spite of its denotative function, nothing socially meaningful for living
together would be said to others. Naming and describing are not essential
dimensions in social communication.

Field Theory and Interdisciplinary Extensions

J. R. Kantor (1924, 1926) formulated a field model for the scientific
analysis of psychological behavior. His fundamental conception of the field
construct involved the identification of functional contacts between the
individual and particular stimulus objects or events, made possible by a
contact medium and promoted by setting factors in the situation and the
individual interactive history. Explanation of psychological behavior was
conceived as the description and systematic analysis of the interdependent
relations established between the components of the psychological field.
Nevertheless, the field system was not developed in order to abstract and
identify different classes or types of functional contacts (Ribes, 2018),
probably because Kantor’s efforts were devoted to a theory about psychology
and not to a theory in psychology.
Rather, the conceptual frame was used to examine ordinary
psychological terms as if they corresponded univocally to different classes of
psychological fields. In the best of cases, such analysis provided a functional
interpretation of ordinary language terms, but it fell short of satisfying the
requirements of a theoretical technical language. Ordinary terms may be
incorporated by the technical language of an abstract theory, but only when
they are provided with a univocal sense, under the logical frame of the
theory. Otherwise, the use of ordinary terms leads to misunderstandings and
confusion. As mentioned earlier, psychological terms are not “references” to
anything else. They are part of the individual episodes that constitute
psychological phenomena, and, therefore, it is absurd to consider them as a
special sort of self-denotative terms with “explanatory” functions. In the
same way, these terms, by themselves, isolated from the circumstances in
which they occur as part of ordinary language practices, do not represent
legitimate problems to be addressed by a scientific theory. Ordinary terms
and expressions are multisensical, depending on their meaning or sense from
the context in which they take place or are “used.” This is why they cannot
be directly translated to or from denotative abstract terms. Abstract
denotative terms may show a logical coverage including a diversity of
ordinary terms, and conversely, a single ordinary term may be analyzed from
the perspective of different denotative terms.

Logical Functions of a Psychological Theory

I have proposed that a balanced theory in psychology should include
four types of logical categories (Ribes, 2003): taxonomic, operational,
measurement, and representational. Taxonomic categories cover the
abstracted empirical domain to be considered as facts and their properties.
Operational categories are related to the observational and experimental
preparations, methods, procedures, and techniques through which facts and
their properties are set to occur, be manipulated, and be recorded, among
other operations. Measurement categories deal with the way in which
recorded facts, as events and properties meaningful for the theory, are
transformed into data, and therefore—in evidence of the functional relations
—implicitly assumed by taxonomic categories. Finally, representational
categories consist of the more or less formal schemas in which general
functional relations are “pictured” and structurally organized. Below I will
elaborate on taxonomic categories specifically.
Taxonomic categories, to the extent that they delimit the empirical
domain to be studied, are the first step in the development of a scientific
theory. Taxonomic categories frame the facts to be accounted for in terms of
processes and states. Kantor (1958) formulated a subject matter for
psychology and the logical model to address its study. This logical model
consists of a field analysis of functional contacts, made possible by a contact
medium and facilitated or interfered with by situational and historic setting
factors. However, Kantor did not take further steps to formulate a theory of
psychological behavior. To fill this initial conceptual gap, I have proposed a
taxonomy consisting of five different functional contacts in a field involving
interdependent contingencies between the individual’s behavior patterns and
the properties and dimensions of stimulus objects (Ribes, 1997, 2018). This
functional classification is based upon two central concepts: mediation, in
terms of the component articulating the field organization, and detachment,
referring to the progressive functional independence of properties present in
the situation in which the contact takes place. The taxonomy covers five
types of functional contact, two of which are exclusive to human beings but
at the same time build upon the rest of the contacts that are shared with other
animal species. The five types of contacts are coupling, alteration,
comparison, extension, and transformation of contingencies (see Ribes,
2018) and are analyzed in terms of molar measures describing directionality,
variation, preference, vigor, and persistence. Special operations and
experimental preparations have been gradually developed for their
experimental study.

Dealing with Natural and Social Settings

In the same way that ordinary psychological terms and concepts cannot
be directly related to abstract denotative terms, the latter cannot be directly
applied to describe psychological behavior in natural and social settings, or to
the complementary analysis required by multidisciplinary intersections,
especially with biology and social science. The “application” of scientific
findings demands an adaptation of the theoretical knowledge describing
abstracted conditions in the form of a technical language formulated to cope
with particular conditions characterizing specific situations in natural and
social settings. The application of scientific knowledge always entails the
empirical findings framed under the concepts and categories of a theory.
Theories in science do not deal with particular situations or concrete
events. In contrast, application of scientific knowledge has to do with
particular situations, actual individuals, and specific social criteria about the
goals, effects, or outcomes of such application. Applications of science
always deal with interdisciplinary fields, which consist of special
professional domains designed to solve socially important problems and
promote the achievement of institutional goals. Examples of interdisciplinary
fields are medicine and other health professions, administration and finance,
education-related professions, and a wide variety of engineering fields.
Interdisciplinary fields delimit the boundaries and characteristics of the
contributions to be made by different sciences, technologies, artistic
disciplines, and traditional practices. All those contributions must touch upon
the different dimensions of the problems to be solved or the goals to be
achieved. Contributing disciplines must adapt their knowledge to the
demands, requirements, and criteria of the interdisciplinary fields, or, in
exceptional cases, provide arguments and procedures to modify the ways in
which they are conceived. Consequently, scientific disciplines do not “apply”
straightforward knowledge and techniques when participating in an
interdisciplinary field or enterprise.
So-called applied behavior analysis (ABA; Baer, Wolf, & Risley, 1968)
arose as a direct application of the principles of operant conditioning
(equated with the experimental analysis of behavior) to social problems in
natural settings. The field of ABA included, in its beginnings, the behavior of
patients with mental illness (Ayllon & Azrin, 1968; Lindsley, 1960), children
with developmental disabilities (Azrin & Foxx, 1971), physical rehabilitation
(Parker et al., 1984), juvenile delinquency (Cohen, 1972; Hobbs & Holt,
1976; Wolf et al., 1976), elementary education settings (Barrish, Saunders, &
Wolf, 1969; Burchard & Barrera, 1972; Cossairt, Hall, & Hopkins, 1973;
Glynn, Thomas, & Shee, 1973, Holt, 1971; McDowell, 1968; Schutte &
Hopkins, 1970), the personalized instruction system, programmed
instruction, and precision teaching (Keller, 1968; Lindsley, 1992),
psychotherapy (Kohlenberg & Tsai, 1995), industrial and labor settings
(Hermann, Montes, Domínguez, Montes, & Hopkins, 1973) and some
extensions to community problems (Fawcett, Mathews, & Fletcher, 1980).
ABA was conceived as a direct application of principles identified in the
experimental laboratory, mostly with animal subjects, and afterward with the
replication of procedures from the animal laboratory with human subjects.
Principles consisted of operational categories relating procedures with effects
or outcomes, namely reinforcement, extinction, punishment, discrimination,
generalization, conditioned reinforcement, and so on (see chapter 2).
Experimental operations were transferred to natural settings under controlled
situations to assess effects on the frequency of different behaviors, either to
establish new behavior or to increase or decrease target behavior.
Nevertheless, the effectiveness of such procedures was limited to relatively
simple behaviors or interactions, in which the (high or low) frequency of
behavior was the relevant issue, not the behavior itself. Moreover, procedures
were used mainly under conditions in which it was possible to control the
interactions and outcomes of those behaviors. Thus, ABA procedures were
more akin to limited demonstrations than to a true technology for behavior
change, as it was assumed.
With the passage of time, it became obvious not only that the use of a
jargon based on operant conditioning operations had little to do with actual
procedures employed in human situations, but also that this jargon became
progressively mixed with nontechnical terms or terms borrowed from
different professions, disciplines, and even religious practices. The so-called
applied behavior analysis procedures required of simplified environments,
and their implementation, were not in correspondence with the original
definitions supporting them (Ribes, 1977, 2004b). This is to say that operant
principles, at least as they have been formulated for the experimental (and
theoretical) analysis of behavior, seem to apply to a rather limited set of
circumstances (e.g., those involving discrete repetitive responses and
stimulus changes related to deprivation or intensity parameters). Indeed, it
may be considered somewhat naïve to assume that individuals’ daily life is
“controlled” by “reinforcers,” which are to be identified as any condition or
event following behaviors varying in complexity, and which are difficult to
compare with each other (e.g., crying, reading letters or books, writing
novels, solving problems, consuming drugs). The world and daily life of
individuals is not arranged by discriminative stimuli and reinforcers. It is
difficult to attempt to conceive human behavior in natural settings in terms of
“response classes” being continuously affected, in discrete linear chunks, by
ad hoc “natural” discriminative and reinforcing stimuli. Skinner’s (1957)
analysis of language as verbal behavior clearly showed that interpreting
human behavior in terms of operant principles is, at best, a hermeneutic
exercise, full of contradictions and imprecisions (Ribes, 1999).
Scientific disciplines, in contrast to interdisciplinary fields, are able to
take distance from prevailing social ideologies and correlated dominant
practices. This does not mean that science is a neutral, pure, human enterprise
(see Kantor, 1953). Nevertheless, analytic thinking and empirical criteria for
confirmation of knowledge widens the horizons of sciences’ perspectives
about the determinants and solutions for problems arising in the context of
human social organization and activity. Therefore, the participation of
science within interdisciplinary fields is not a simple matter of adaptation and
application of available knowledge. It implies also a critical appraisal of the
social problems involved and challenging traditional views about their nature
and solution. In the case of education, unfortunately, the contributions of
psychology (and of the other participating disciplines), without any critical
perspective, have been assimilated by dominant social criteria to prolongate
the conservative nature of the “school.”

Competence: The Integration of Intelligence,

Learning, and Knowing
The final section of this chapter focuses on an example of an
interdisciplinary effort. Specifically, I will focus on the field of education to
examine the ways in which psychological theory may participate in an
interdisciplinary endeavor. Not only is education a relevant social field to
which psychology may contribute, but it also allows us to compare the ways
in which applied psychology in general, and ABA in particular, have
conceived their role. Education, as a general endeavor, mostly rests on the
“school”—the institution devoted to instructing individuals for the social
division of labor and for the reproduction of values inherent in a given
social formation. Education involves the documentation of the information,
abilities, and skills acquired by individuals. Since the French Revolution,
education has been designed to be offered in the school to individuals
grouped together according to different criteria: age, sex, type of
instruction, previous knowledge and ­abilities, and so forth. Individualized
instruction is an exception, even though this may sometimes take place in a
home or small work setting. The availability of electronic devices has
promoted impersonal educational environments, which, nevertheless,
should be conceived to be complementary but not equivalent to those that
require the explicit interaction of the learner with the teacher. In any case,
there are hierarchical social institutions prescribing, designing, and
regulating the educational programs offered in the school. These
institutional bodies make the decisions about what, how, where, and to
whom the varied contents of educational programs are to be taught. In spite
of the fact that teachers are only a part of the operation of the institutional
mechanism, they bear the responsibility of educational effectiveness,
conceived as the fulfillment of the outcomes demanded by the social
formation as a whole.
Two concepts are outstanding in reference to the psychological
dimensions involved in education, namely, intelligence and learning. These
two concepts emerged from the natural history of education and point to
aspects related to the individual being instructed, trained, or taught—the so-
called apprentice, pupil, or student. Whereas instruction depends on the
teacher, who is assumed to be knowledgeable about what is being taught and
the adequate procedures to teach it, the success of such effort depends in turn
on the capacities and abilities of the apprentice. Learning is the concept that
describes the final outcome of instruction. Psychology, from different
perspectives, has approached the teaching-learning process by focusing on
the construct of intelligence, its measurement, and its development, or, on the
other hand, by focusing on the design of the teaching procedures that
promote effective and fast learning. Psychologists construct intelligence and
learning tests and scales and devote themselves to the design of learning
environments and procedures, participating in the teaching of teachers.
Teachers are, thus, instructed to teach, to assess the intelligence and learning
of students, and to plan and administrate educational settings with these
purposes. Educational failure is attributed to teaching procedures and to the
inadequacy of students in showing basic skills or abilities. Nevertheless, it is
unusual to find explicit rationales of the educational process justifying what
is to be learned by apprentices and how this learning is to be incorporated in
the various functional domains of daily life. Learning is often conceived in
terms of “knowledge” about something or as basic skills related to reading,
writing, and calculating. The school curriculum, from the elementary school
to the university, is grounded on the assumption that learning consists of
knowledge acquisition and that the “mechanics” of reading, writing, and
calculating are mere instrumental devices for the acquisition or application of
knowledge. Psychology in general, and ABA in particular, have oriented
their contributions to improve the efficiency of the school system, without
questioning its adequacy, values, and implicit goals.
Intelligence, learning, and knowing (or cognition) are not technical
psychological concepts. These terms are part of ordinary language practices
and are used in a diversity of contexts. Nevertheless, in all cases, we can
identify a common functional sense that in no way makes them adequate for
theoretical purposes. “Intelligence” is a term with a long history (since
Aristotle) and is especially used in the context of evolutionary thinking as
equivalent to species-adaptive capacities. The measurement of intelligence,
as a measure of individual differences, emerged with two different purposes:
(1) with Galton (1899), to justify racial and class differences in humans, and
(2) with Binet and Simon (1916/1973), to measure differences in school
performance, initially to provide compensatory teaching, and later to select
individuals for different activities according to their capacities and abilities,
that is, to classify the affordances of social opportunities.
On the other hand, the term “learning” was seen as the process
responsible for behavioral changes and adaptation, in both animals and
humans, in such a way that learning almost became synonymous to behavior
(Bower & Hilgard, 1981; Thorndike, 1911), contrasting acquired
performance with so-called instinctive, inherited action patterns (Lorenz,
1965).
Finally, “knowing” is another ordinary term with a multiplicity of
explicit and implicit meanings or uses (Malcolm, 1963). Psychology (and
philosophy) incorporated knowing as the common function of other
nontechnical terms, namely, “perception,” “memory,” “thinking,” and
“consciousness,” all related to the subjective capture of the world by a
contemplative individual.

An Analysis of Terms
A conceptual analysis of the terms “intelligence,” “learning,” and
“knowing” helps to identify their functional uses, in order to formulate
interfacing concepts between them and those describing functional contacts
in interbehavioral theory. Ryle (1949) and Ribes (1981, 1989, 2007)
provided detailed analyses of the three aforementioned terms, so I will only
briefly review how they are used in ordinary language practices.
Intelligence is always predicated on doings and outcomes of
individuals. Strictly speaking, intelligence is always referred to as intelligent
behavior or performance. Intelligent behavior always consists of actions that
are effective in the context of their occurrence, usually solving a problem or
formulating its correct solution. Nevertheless, intelligence is never predicated
on a single occurrence of an effective performance by an individual, but
rather to a collection of occurrences by the same individual in a variety of
situations. A single effective behavior may be due to chance or fortune, and
the repetitive occurrence of the same performance in the same situation,
although showing effectiveness, does not show that with a change in the
situation the individual could be equally successful. Thus, an individual’s
behavior is described as intelligent only when the behavior is effective more
than once and in varied forms or situations.
 Ryle (1949) described intelligent behavior, emphasizing the variety of
possible ­effective performances, by saying that the crucial feature is knowing
the final destination and not a particular route to be taken. Contrary to
traditional views in psychology and education, intelligence should not be
conceived as something that is possessed in larger or smaller amounts, but
rather as a characteristic of performance when a problem must be formulated
and solved. Figure 7.1 shows a table with two axes: effectiveness of
performance and variety of behavior. Only the cell in which varied and
effective performance coincide corresponds to intelligent behavior. The other
cells show different types of behavior: creative behavior when performance
is varied but not effective, operative behavior (i.e., behavior that results from
training) when performance is effective but not varied, and “fool” behavior
when performance is neither effective nor varied. Not being intelligent does
not necessarily mean being a fool.

Figure 7.1 Different types of behavior according to the intersection of

variation and effectiveness continua.

Learning has been conceived by psychologists and neurobiologists as a

central process explaining the workings of behavior in animals and humans.
In the last century, learning theory became equivalent to behavior theory, and
the concept of learning supported the metaphor of the acquisition of
behavior. This metaphor allowed for the distinction between so-called innate
(or instinctive) behaviors and learned behaviors, a questionable and
misleading dichotomy. Writings about classical, instrumental, or operant
conditioning (among other procedural domains) commonly refer to the
salivary conditional response or to the key-pecking or bar-pressing responses
as acquired or learned responses. But it is well known that dogs normally
salivate, that pigeons peck on surfaces and objects, and that even rats, left by
themselves, press the bar in the operant chamber. In fact, those responses are
“available” before any conditioning procedure is developed and occur
without explicit elicitation. It is incorrect to speak about the emergence of
“new” responses, since what is taking place is the emergence of new
functions for response patterns already available. The dog salivates to the
signal of food, and the pigeon and rat extend the response patterns to food
seeking and consumption. The tone has become a component of a stimulus
segment determined by food delivery, while pecking and pressing have
become components of a behavior segment related to eating. Even in those
cases in which the specific response patterns do not seem to be available—as
in language learning or some motor skills, including walking—the potential
of such patterns is part of the reactional systems of the individual.
Kuo (1967) described development in terms of behavior gradients and
potentials. A good example of this is speaking, in that articulated language
involves a reduction or restriction of the sound frequencies originally emitted
by the newborn, and not new sound frequencies. Differences in phonetics
between the different natural languages is an outcome of this process. Based
on the acquisition metaphor, psychology has postulated learning to be an
underlying process taking place every time a new behavior function emerges.
Additionally, the consolidation of what is learned is attributed to a
complementary process: memory. However, this is an inaccurate description
of what actually happens when it is said that an individual is learning or has
learned. When learning is said to have happened, typically only one thing
occurs—the performance in a situation—as opposed to two things, the
performance and its learning and consolidation.
Learning is not identified by means of special activities, except for the
case in which the activities themselves are what is being learned. Learning is
said to occur when the performance of an individual fulfills a predetermined
criterion of mastery of an activity, or a criterion related to the results or
outcomes of a set of activities. When no criterion is prescribed or established,
it is nonsensical to talk about learning or nonlearning. The appeal to
“implicit” learning always consists of a post-hoc recognition of a prior
nonprescribed criterion. Therefore, it is safe to assert that the same behavioral
processes take place when individuals learn and when they do not learn. The
prescription of an activity/outcome criterion is the crucial difference between
both situations. Learning and not learning can only be identified in reference
to a criterion. The postulated stability or permanence of learned behavior is
the result of the stability and permanence of the contingencies required and
established by the corresponding ecological or social criteria.
Knowing is the third ordinary term dealing directly with the means and
goals of education as a social institution. It is assumed that the ultimate end
of schooling is to promote the “acquisition” and “use” of knowledge. The
terms “to know” and “knowledge” are used in a multiplicity of ways in
ordinary language, not only in reference to the mastery of a practice,
situation, or domain, but also in relation to doubt, estimation, and certainty,
among other uses. Knowledge usually applies to intellectual products as well
as to procedures and techniques. Knowing entails an episodic relation of an
individual with something or someone, in terms of exposure, information, or
mastery in practice, involving a great diversity of linguistic practices.
Knowing is not a special kind of activity but rather an outcome of direct and
indirect contact with entities (things, persons, living beings, institutions),
events, and doings, sometimes under specifically arranged learning
situations. Although knowing something is tantamount to being informed
about “that,” information implies only the availability of potential stimulus
conditions. Being informed is not the same as knowing what to do or
knowing how something or someone will behave, operate, or work.
Borrowing Austin’s (1955/1962) characterization of language
expressions, we may speak of constative and performative knowing (Ribes,
2007). Constative knowing consists of recognizing, pointing to, informing
about, describing, and confirming events, states of things, and previous acts.
Constative knowing may be an outcome of direct (e.g., talking, writing) or
indirect (e.g., observing, listening, and reading) contacts with what is known.
On the other hand, performative knowing consists of doing something, and
these doings may involve pure linguistic acts, as in theoretical practice
(reflexive language), or integrated patterns involving linguistic and effective
motor acts. Talking and writing, as Austin emphasized, are sometimes the
effective components of behavior. Performative knowing involves mastering
how to do things, including ways of talking and writing as acting.
Given that learning is learning of knowing how and knowing about, four
general types of learning may be identified (in this case, I take “saying” as
representative of the three active modes of language behavior: speech,
signaling/gesturing, and writing): (1) learning how to do and to say, (2)
learning how to say as a doing, (3) learning how to say about doing, and (4)
learning how to do as a saying.
These general types of learning may take place either as intelligent
behavior or as operative behavior, with the aim of schooling being to provide
the necessary conditions of these behaviors and learnings to occur in different
domains of social life. In the case of operative behavior, traditional wisdom
prescribes that practice, repetition, supervision, and appropriate feedback are
sufficient for its establishment and endurance. By contrast, these training
conditions, although necessary in certain occasions, are not sufficient for the
establishment and development of intelligent behavior. Indeed, even when
some intelligent behavior has developed, it is still important to diversify its
organization in accordance with different criteria of functional aptitude.
Functional aptitudes deal with different types of functional contacts of the
individual with the stimulus conditions involved in a problem. Different
aptitudes involve different skills, techniques, routines, and procedures, or a
different organization of the available skills, even when the “formal”
situation seems to be the same.
I have formulated the concept of functional competence (Ribes, 2006b,
2008) in order to deal with the learning of intelligent behavior in educational
settings. A competence is the functional organization of skills, techniques,
routines, and procedures in order to satisfy a functional aptitude criterion (see
figure 7.2). Aptitude criteria correspond to the five types of interbehavioral
contacts identified in the field theory: coupling, alteration, comparison,
extension, and transformation of contingencies (Ribes, 2018). Thus, a
competence is a relation between functional aptness or capacity, in reference
to a type of field contact, and intelligent behavior in the form of varied and
effective behavior. Competence is the interphase concept that “connects”
theory with practice. Through the analysis of competencies, it is possible to
look into intelligent behavior, learning, and knowledge a nonlinear way.
 Figure 7.2 Competence as the conjugation of operative behaviors
and functional aptitude criteria.

From this perspective, interdisciplinary intervention by psychologists

and educators is concerned with designing the conditions in which, for any
domain of knowing or life, individuals can deal with problems in any of the
five types of aptitude corresponding to interbehavioral functional contacts.
The “domain” in which individuals interact may be always the same (for
instance, learning about kinetics in physics), but the quality and functional
characteristics of those interactions may be completely different depending
on the type of interbehavioral contact involved. To diversify functional
aptitude in any domain, however, demands a profound change in the structure
of the schooling situation. It is not simply a matter of adapting teaching
procedures or introducing supplementary media. These are relevant only for
providing information or for training techniques and skills, as has been
traditional in media-concerned education or in ABA. The “contents” of
education must be replaced by competencies pertinent to specific domains of
life and knowledge. Competencies must be shown as actual behaviors that
are relevant for the mastery of a given domain—that is, doing something (and
this always includes linguistic interactions) that is a functional activity in the
situation. It is not behavior “external” to the situation involved, but behavior
“internal,” intrinsic to the usual practices involved in a given domain.
Competencies can only be learned from those who master them in a
given situation. One cannot simply be informed about competencies. Rather,
competencies have to be taught through example. To teach a competency
entails demonstrating it and guiding, prompting, and correcting the
apprentice with respect to the different criteria, prescribing the ways in which
to interact with the circumstances in a situation. Teaching a competency goes
beyond showing constative knowledge. Competencies always involve some
form of varied performative knowing, even in cases that are completely
linguistic in nature, as in theoretical practice. To the extent that learning a
competency always involves some kind of doing or activity, when teaching
competencies, no additional evaluation regarding its learning is needed after
it is demonstrated by the apprentice. Unlike traditional teaching, in which
knowledge is transmitted through textual or spoken language, the use of
media, or even demonstrations, competencies cannot be transmitted. The
teaching of competencies requires that both teacher and apprentice participate
in the actual activities involved in the exercise of performative knowing in a
given situation or domain. Competencies consist of episodic interactions and
not of specific, pinpointed behaviors or responses. In fact, competencies may
be exercised through very different activities, events, and stimulus
conditions. The domain and the aptitude criteria in a situation are the factors
that identify a competence as being the same, in spite of the diversity of
activities and elements involved in any episodic interaction.
I have previously discussed (Ribes, 2006b, 2011) why traditional
schooling does not allow or promote the establishment and development of
competencies and intelligent behavior by apprentices. Below, I will elaborate
on the reasons.

Flaws Inherent in Traditional Education Models

Formal education is most often based on the assumption that apprentices
learn “knowledge” as information, not as practical interactions within the
boundaries of disciplinary domains and life contingencies. The traditional
conception of teaching and learning, as transmission-contemplation, has to be
replaced by one in which to learn means to master the competencies involved
in the practice of any domain of life and knowledge. This entails replacing
the conception of having learned in terms of being informed about something
and being able to recognize, repeat, or reproduce it, by one in which learning
corresponds to effective interactions in practical circumstances involving
new knowing or new ways of showing such knowing. Learning has to be
approached in terms of what the apprentice must master as a practice, instead
of emphasizing what teachers have to inform students about. This is why
present schooling is, at least, deficient. Most learning is planned to take place
in a classroom, and sometimes in demonstration laboratories, where
apprentices observe and listen to the teacher. The teacher exposes students to
what is written in manuals and books and informs them about the actual
doings and learning of others. Leaning is evaluated through exams and tests,
in which apprentices have to recognize, repeat, or reproduce the information
received or the procedures involved in demonstrations.
Psychology and ABA have focused their efforts on improving the
“efficiency” of this model of schooling, emphasizing teaching and
instructional techniques, as well as assessment methods to evaluate “how
much” apprentices have “learned” to duplicate the information and
demonstrations provided to them. However, as previously mentioned, the
interdisciplinary intervention of scientific disciplines in education, at least,
should not be without criticism and provided on an “on demand” basis. Aside
from other arguments concerned with the social orientation of the present
universal model of schooling, an analysis—based on the nature of learning
from the perspective of interbehavioral field theory—leads to the formulation
of an alternative model centered on the apprentice and to design situations
promoting the development of functional competencies in the diverse
domains of knowledge and daily life. This alternative model should also
question the ultimate social goals of education and schooling. Following this
argument, I proposed an alternative model for preschool and elementary
education that can be extended to the complete schooling process (Ribes,
2008). This model involves the development of five kinds of competencies
for every segment of disciplinary knowledge in empirical sciences,
sequencing their organization in terms of increasing complexity of
phenomena (from physics and chemistry, to geology and biology, and finally
to social science).
In this new educational model, alive language (Kantor, 1936), in all its
modalities, is emphasized as an actual practice, and not as “rules” or
erudition. This approach to schooling includes the various arts as well as an
extended learning of mathematics as a transdisciplinary tool. Disciplinary
competencies are integrated with social domains of daily life, in order to
develop life competencies in health, social participation, communication,
environmental conservation, technological skills, and recreation. Sciences are
to be learned by doing and thinking in the conditions and situations in which
their practice take place (obviously not in the classroom). The same goes for
life competencies, which can also be integrated with language enrichment
and artistic activities.

Implications and Future Directions

The perspective described in this chapter has several implications for the
science of behavior, particularly as increased attention is given to topics
related to human language and cognition. As described in the chapter, the
language used throughout various areas of science has implications for the
conduct of subsequent work. That is, how we talk about what we do may
impact subsequent research, and ultimately how we approach various
applied topics. While this chapter is more theoretical in nature when
compared to other chapters in this volume, it is clear that several lessons
pertinent to both research and practice can be gleaned. More generally, as
the perspective described in this chapter is relatively less well known
among many applied behavior analysts, the potential implications of this
body of work remain to be seen.

Conclusion
Language is used in various ways throughout science. The present chapter
considered some of the ways in which language impacts science—in
particular psychology and behavior science—and provided some guidance
for the proper use of language in science. While largely a theoretical
chapter, it also considered applied issues and concluded with a
consideration of the topic of education and schooling. Behavior scientists of
all sorts are encouraged to analyze the language of science and practice and
consider the extent to which it impacts their work and various areas of
research and practice more generally.

Study Questions
1. How are technical and ordinary language different?
2. Explain why language is a conventional practice.
 3. How does the author distinguish between “alive” and “dead”
language?
4. Comment on the relationship between the products of language and
language itself.
5. What does the author mean when he suggests that linguistic
response systems become detached from the conditions in which
they were originally learned?
6. Why does the author have concerns about the term “understanding”?
7. How did the borrowing of language from biology result in a
reductionistic conception of behavior?
8. What is the author’s concern about Kantor’s logical model?
9. The author suggests that ABA did not amount to a true technology
of behavior change in applied situations. Explain.
10. What are some of the author’s concerns about operant principles in
applied circumstances?
11. Summarize some of the author’s concerns about the traditional
approach to education and schooling.
 Chapter 8:

Observational Learning

Mitch Fryling
California State University, Los Angeles
Rocío Rosales
University of Massachusetts Lowell
Natalia Baires
Southern Illinois University, Carbondale

Link and Preview

It probably goes without saying that we all learn a great deal from
observing things in our environment. While this is obvious to both
laypersons and scientists from various disciplines, the topic of
observational learning—the focus of this chapter—is often not
given extensive treatment in behavior analysis. The present
chapter focuses on observational learning, reviews conceptual
foundations and recent literature, and provides implications for
practice. We believe it is an important part of the story for
understanding a comprehensive account of language and
cognition in behavior analytic perspective.

Observational learning (OL), generally defined, involves behavior change

as a function of observing happenings in one’s environment. OL is a
conceptually challenging topic, one with significant theoretical and applied
implications. Indeed, most learning environments require learners to have
an OL repertoire. For example, teachers often model a particular task to a
classroom of students with the hope that students will engage in the
modeled behavior themselves at a later time. Social development also
seems to involve OL, as when a child learns how to play and interact with
their peers by observing others. Clearly, the absence of an OL repertoire
would be detrimental to one’s development; learning would have to occur
directly, with a trained therapist or teacher providing specific instructions,
prompting, and differential reinforcement. This is not only costly but also a
slow, cumbersome process that interferes with other learning opportunities.
For these reasons, developing the prerequisite skills needed to learn from
observation is an essential consideration for behavior analytic practitioners.
Moreover, learning from observation seems to be involved in day-to-day
living more generally. For example, humans live in social environments, and
their relationships with other people are greatly enhanced if they observe the
behavior of others in different situations, and if their behavior is subsequently
changed as a function of those observations. For instance, if we observe a
coworker to be overwhelmed at work, we may offer them help or wait before
asking them to take on an additional task. Alternatively, noticing that our
significant other is sad or disappointed about something that happened may
require the careful observation of their behavior with respect to the matter in
prior interactions. As these examples highlight, observing relations among
various factors in one’s environment fosters subsequent derived relational
responding. In this sense, OL may be considered a foundation of language
and cognition more broadly.
Our commitment to developing a comprehensive and coherent natural
science of behavior requires us to address OL more systematically in
behavior analysis. Moreover, as mentioned, OL is also a topic with
considerable clinical implications. Indeed, to the extent that individuals can
learn to learn from observation, they may develop skills more quickly,
including subtle repertoires that can be difficult to instruct directly (e.g.,
social skills), and adapt more efficiently to ever-changing environments. This
chapter provides an overview of behavior analytic research and scholarship
in the area of OL. We will begin by discussing conceptual work in the area,
followed by a review of various areas of research in OL. Special emphasis is
given to the emergence of derived stimulus relations as a function of OL, as
this is a particularly promising area for additional research. The chapter
concludes with implications for future research and practice.
Conceptual Analysis of Observational Learning
Our section on the conceptual foundations of OL begins with an overview
of the work by Bandura and colleagues, which was central to the social
cognitive perspective, followed by a review of behavior analytic conceptual
analyses of OL.

Bandura’s Theory
Generally, Bandura’s research program demonstrated something simple
yet very important—that behavior can develop as a function of observing a
model. Bandura’s work also demonstrated that it is not just modeling that is
important, but that the consequences of the model’s behavior also play a
critical role. In particular, participants who observe a model being punished
are found to demonstrate less behavior change relative to those who observe
a model being reinforced or receiving no programmed consequences at all
(e.g., Bandura, 1965). Another issue fundamental to Bandura’s work is the
delayed performance of the observer after observing a model; a great deal of
OL seems to be demonstrated long after the model has engaged in the
behavior. In this sense, behavior seems to occur in the absence of a current
environmental stimulus (we’ll address this more in the next section).
Following from this, Bandura’s research began to investigate cognitive3
factors to account for this delayed performance. For example, in one
experiment (Bandura, Grusec, & Menlove, 1966), one group of observers
was told to describe what they were observing, another was told to count
while they were observing, and a third group was not given any instructions.
The results showed that those who described what they were observing
engaged in more OL than those who didn’t, suggesting the importance of
“cognitive” factors in OL.
While the consequences of the model’s behavior have been found to
influence the extent to which observers engage in the observed behavior
themselves, observers can describe what they observe irrespective of this
(e.g., Bandura, 1965). To Bandura, this suggested that learning from
observation occurs regardless of observed consequences, whereas
performance, demonstrated by the engagement in the modeled behavior, is
dependent on the consequences of the model’s behavior. All of this led to the
development of a theory of OL (e.g., Bandura & Jeffrey, 1973). According to
the theory, the initial learning of behavior is accounted for by attention and
retention processes, whereas the performance of observed behavior is
determined by factors involved in motor reproduction and motivational
factors. Following an input-output cognitive-type model, the general theory is
that observed events are taken in by the individual and stored within the
individual, and that these stored experiences (e.g., images or symbols in
cognitive theory) later emerge to guide behavior.

Behavior Analytic Conceptualizations of

Observational Learning
Deguchi (1984) provided an alternative explanation of the outcomes
described in the OL literature based upon Skinner’s radical behaviorism
(Skinner, 1974a). Generally, Deguchi highlights that much of the literature on
OL has given little attention to the dynamic reinforcement history of the
observer. The majority of experiments conducted by Bandura involved
typically developing children, who are likely to have a rather extensive
history of reinforcement for imitating observed behavior. Moreover, that
history of reinforcement is likely to have involved an intermittent schedule of
reinforcement, rendering it particularly resistant to extinction and likely to
persist during brief postobservation probes during OL experiments. This sort
of reinforcement history may explain the popular notion of one-trial
learning found in the OL literature (i.e., behavior change that results from a
single observation).
Deguchi (1984) also accounts for the delayed performance found in the
OL literature by turning to reinforcement histories. Imitative behaviors may
not always be reinforced if they occur immediately after the modeling (as
when it is very obvious that one is simply imitating). In accounting for the
alleged cognitive events in Bandura’s theory, Deguchi describes the influence
of instructions on behavior. Bandura’s research showed that individuals who
were instructed to do various things during observations were more or less
likely to learn from observation, depending on the particular instructions. To
Deguchi (1984), this is simply seen as the impact of instructions on behavior
(more on this below). Finally, reinforcement histories are again used to
explain the role of observed consequences. Specifically, the consequences of
the model’s behavior likely have discriminative stimulus functions, with
observers having a probable history of reinforcement for engaging in
observed behavior only when the model is reinforced and especially not
when the model is punished. For example, observing someone receive social
reinforcement for engaging in a particular behavior may set the occasion for
an observer to engage in that behavior themself at a later time. Deguchi’s
work is especially noteworthy for emphasizing the reinforcement history of
the observer in accounting for OL.
Palmer (2012) discussed OL in the context of describing atomic
repertoires. Palmer suggests that much of our complex behavior can be
explained by appealing to smaller, fine-grained units of behavior, which may
occur both overtly (behavior others can see easily) and covertly (behavior
others cannot see easily). Atomic behavior may include imitative, echoic, and
tacting repertoires. Pertinent to OL, Palmer explains the one-trial learning
phenomenon by suggesting that atomic repertoires occur at the moment of
observation. Then, at a later time when part of the original observational
context is present, the atomic behavior occurs again and may set the occasion
for other behavior. For example, if someone is learning new job duties and
observes a model engaging in a particular behavior in a particular situation,
when that situation, or even part of that situation, arises again in the
observer’s experience, they may be more likely to engage in the observed
behavior themselves. To Palmer, the theoretical implication of this is that
behavior occurring as a function of a single observation is not occurring for
the first time with no history of reinforcement. Rather, the behavior has
occurred before, at least in some form, and was reinforced at that time, and
this history accounts for subsequent behavior change.
While the observer’s reinforcement history is central to the behavior
analytic account of OL, behavior analysts have also elaborated on the
stimulational aspects of OL. Masia and Chase (1997) provide an overview of
how conditional discrimination processes, generalization, and functional
classes (e.g., stimulus equivalence) may participate in OL. In considering
conditional discrimination, the authors provide an example of how
instructions can function as conditional stimuli that alter the extent to which
someone learns from observation. For example, instructions such as “You
will be tested on this later,” “Don’t do it this way,” or “Especially focus on
X” can alter the extent to which observers engage in observed behavior.
Masia and Chase also make the connection to rule-governed behavior, and
note that this literature may also inform our interpretation of OL. Considering
language and rules more specifically, Fryling, Johnston, and Hayes (2011)
note that observers likely have a long history of reinforcement for describing
relations in their environment (rule generation) and engaging in subsequent
rule following with respect to those descriptions. At the same time, Fryling et
al. discourage the assignment of causal or mediational status to this sort of
behavior. In other words, the authors do not suggest that this behavior causes
OL. Rather, this sort of rule generation is to be considered a product of
specific attention to what is observed, and such attention is surely required to
learn from observation.4
Generalization processes may also contribute to OL (Masia & Chase,
1997). Masia and Chase provide an example of an individual with a history
of getting burned on the stove who later observes someone get burned from a
campfire. In this case, the observer may avoid getting too close to the
campfire as a result of the observation and the similarity of the campfire to
their own experiences with the hot stove. Similarly, the functions of observed
consequences may also be impacted by generalization processes, as when we
observe someone become very afraid after some event, and as a function of
our own experiences of being similarly afraid in entirely different
circumstances, we learn to avoid what was observed (i.e., because the
consequences are morphologically similar). Generalization processes
highlight that it is not just observation, but observation and the similarities
between what is observed and our own experiences, that impacts OL.
Finally, Masia and Chase (1997) describe how the concept of functional
classes may contribute to OL. The stimulus equivalence literature has shown
how stimuli may develop the stimulus functions of other stimuli, even when
those stimuli share no physical similarities. Importantly, such stimulus
functions do not simply emerge, they are a product of an individual history of
responding with respect to relations among stimuli and responses. For
example, given a particular history, a number of very different stimuli may
develop the stimulus functions of being “bad” (e.g., smoking, failing school,
lying). At a later time, an individual with such a history may observe
someone engaging in a different behavior (e.g., swearing) while also being
told that the behavior is “bad.” As a consequence of this observation and the
derived stimulus functions of the word “bad,” the individual may avoid the
observed behavior and the context in which it occurred. As this simple
example highlights, derived stimulus relations are likely to participate in OL.
Elaborating on the stimulational aspects of OL, Fryling et al. (2011)
provided an analysis of OL derived from interbehaviorism (Kantor, 1953)
and interbehavioral psychology (Kantor, 1958). Fryling et al. (2011)
specifically emphasize the unique field orientation of the interbehavioral
position (Kantor, 1958), and especially highlight the dynamics involved in
the development of stimulus functions and the role of observing and
describing stimulus relations. As noted, the OL literature requires behavior
analysts to describe how past observations influence behavior at a later time,
often in the absence of an obvious stimulus. In the interbehavioral
perspective, this phenomenon can be explained with Kantor’s concept of
stimulus substitution. Generally, stimulus substitution occurs when one
stimulus develops the functions of a stimulus that is no longer present as a
result of an individual’s history of responding with respect to relations among
those two stimuli (such relations occur in space and/or time, and may occur
among stimuli, settings, responses, and variations thereof). For example, one
may think about an old friend when visiting a place that they frequented with
that friend (i.e., respond to the friend even when the friend is not physically
present). In this case we would say that aspects of the setting developed
substitute stimulus functions for the friend. Similarly, previously observed
behavior that occurred in a particular setting or situation may be occasioned
by stimuli that substitute for past observations. That is, previously observed
behavior may be present by virtue of stimulus substitution (see Fryling et al.,
2011, for a more detailed overview of stimulus substitution).
Conceptual work in behavior analysis has provided an alternative to
Bandura’s theory of OL. While the behavior analytic literature differs in
some ways regarding the conceptualization of OL, there is a similar focus on
environmental factors and how these influence OL. This focus on
understanding functional relationships in the natural environment is
conceptually coherent with our aims as a natural science and offers specific
practical advantages over cognitive perspectives. Most obviously, the
behavioral perspective points to environmental factors that may be altered
and studied in efforts to promote learning from observation in various
situations.

Observational Learning Research in Behavior

Analysis
While behavior analysts have begun to provide a comprehensive theoretical
account of OL, a number of behavior analysts have also conducted research
in the area of OL. The following sections provide an overview of behavior
analytic research on OL, including both general OL research and that which
specifically focuses on derived stimulus relations. Implications for research
and practice are provided throughout.

Assessment of Observational Learning

The majority of the behavior analytic research in the area of OL may be
considered assessments of OL, wherein researchers collect preobservation
baseline data, implement an observational intervention, and then conduct
postobservation assessments of behavior change. A number of topics have
been researched this way in the literature, with recent examples including
children’s preferences (Leaf et al., 2012; Leaf et al., 2016), the acquisition of
verbal operants (Storlie, Rehfeldt, & Aguirre, 2015), and parents learning
skills to work with their children (Eid et al., 2017). Researchers have also
examined factors that may impact OL. For example, Castro and Rehfeldt
(2016) assessed the extent to which learning from observation is influenced
by the observer’s relationship to the model (also see Singer-Dudek & Oblak,
2013, described below).

Observational Learning and Prerequisite Skills

Given the importance to day-to-day functioning of developing an OL
repertoire, it is important to consider prerequisite skills. As outlined by
Taylor and DeQuinzio (2012), examples of these prerequisite skills include
attending, delayed imitation, and identifying and discriminating among
contingencies. Related to attention, OL involves attending to multiple
stimuli (e.g., observing both a modeled response and the subsequent
consequence), dividing attention among various stimuli, and engaging in
sustained attention.
Imitation, specifically delayed imitation, is also pertinent to OL. Two
essential components are required in order for imitation to occur within the
context of OL: history of reinforcement and a generalized imitative repertoire
(Taylor & DeQuinzio, 2012). Imitation skills that are reinforced will be
maintained, while a generalized imitative response class will ensure that
novel responses occur without the need for direct instruction or a continuous
schedule of reinforcement. Delayed imitation is critical as responses often
need to be emitted long after the observation of a model. For example, a child
may observe a teacher solving a problem during instruction one day, and on a
later day be asked to solve a problem in a similar manner in the absence of
the model.
Finally, observers’ behavior must demonstrate discrimination between
observed contingencies of reinforcement and punishment. Such
discrimination facilitates the development of complex stimulus control such
that observers engage in modeled behavior that is reinforced, and not that
which is observed to be punished. For example, a child may learn to engage
in an appropriate social behavior (e.g., calmly asking for something) after
observing that behavior be reinforced, and not engage in a socially
inappropriate behavior (e.g., yelling) after observing such behavior fail to be
reinforced.

Developing Observational Learning Repertoires

Several researchers have pointed to the need for more research in the
area of developing OL with individuals who lack OL repertoires (e.g., Greer,
Singer-Dudek, & Gautreaux, 2006; Townley-Cochran, Leaf, Taubman, Leaf,
& McEachin, 2015). In one study that focused on developing OL repertoires,
Pereira Delgado and Greer (2009) investigated the effects of teaching
children without OL repertoires to monitor peers in order to learn from
observation. In Experiment 1, two children were taught to monitor peers
while those peers were being taught textual behavior (i.e., “reading” in
traditional terms). After the participants who were monitoring the peers
learned to select a green cube when the peer responded correctly and a red
cube when the peer responded incorrectly, both participants demonstrated
improvements in OL. Interestingly, while the participants only monitored a
peer learning textual responses, postmonitoring probes demonstrated that OL
improved while observing the same peer learning to tact fruits and while
observing a different peer learn textual behavior.
Experiment 2 of Pereira Delgado and Greer’s (2009) research replicated
the first experiment but with oral spelling as the target behavior. After
learning to monitor a peer’s behavior as in Experiment 1, all three children
improved oral spelling as a function of observing a peer being taught by a
teacher. Interestingly, two of the three observing participants in Experiment 2
learned at the same speed as or faster than their peer who was being taught
directly.
 Taylor, DeQuinzio, and Stine (2012) also investigated the effects of
teaching children to monitor their peers’ behavior while those peers were
instructed to engage in previously mastered textual behavior (i.e., reading
sight words). Monitoring conditions involved imitating peers’ responses after
being asked, “What did he say?” and engaging in a matching response (i.e.,
placing a chip under the correct word on a matching board). Specifically,
during training a teacher presented a sight word to the peer and said, “Read”
while the participant observed. Correct responses were reinforced with a
token and praise from the teacher. The teacher would then remove the word
card and ask the participant, “What did he say?” requiring an imitative
response. The teacher corrected the participant’s incorrect imitation responses
by re-presenting the word to the peer to read again and again asking the
participant, “What did he say?” If the participant still provided an incorrect
response, the teacher did not require a matching response and began a new
trial. For incorrect responses during the matching task, a least-to-most
prompting hierarchy was utilized, where tokens were not delivered and a new
trial followed. An exposure condition was also included with separate sight
words, with procedures identical to the monitoring condition with the
exception that monitoring responses were no longer required. Results
indicated that training improved all participants’ monitoring responses. Tests
for OL also demonstrated that all participants learned from observation when
they were required to monitor their peers’ behavior. Moreover, two of the
three participants demonstrated learning during the exposure condition,
suggesting that the effects of the training influenced behavior where
monitoring was no longer required.
More recently, MacDonald and Ahearn (2015) evaluated a broad
assessment and training package for OL. First a preassessment was
conducted, which involved assessing OL across five different tasks, each
with three variations. In tasks where OL was found to be deficient,
participants were taught specific prerequisite skills that have been suggested
in the literature (Taylor & DeQuinzio, 2012), including attending, imitation,
delayed imitation, and consequence discrimination. After participants
mastered a specific variation within a particular type of task, the
experimenters conducted an assessment to test for generalization across the
other variations of the task. If generalization across task variations did not
occur, the participant then went through the training with that particular
variation and, upon mastery, generalization across untrained variations was
again assessed. This pattern of testing and training continued until
participants demonstrated mastery across all task variations. Once this had
occurred, the researchers assessed generalization across task types. Finally, if
generalization across a particular task type did not occur, the researchers
initiated training for that task area, as described above. This continued until
the participants demonstrated mastery across all task types and variations.
(Yes, this study involved a lot of steps!)
Results demonstrated that the amount of training required for mastery of
each prerequisite skill varied across skills and participants. In addition, five
of the six participants demonstrated generalization across task types and
across variations of the task within each type. One participant did not
demonstrate generalization across task types, but did demonstrate
generalization across variations of tasks within each type. In general, the
protocol by MacDonald and Ahearn (2015) is noteworthy as it details how
OL might be assessed across skills, and also how generalization probes can
be systematically incorporated into an OL training program. Again, given the
importance of developing an OL repertoire, the evaluation of procedures to
develop OL after it has been found to be deficient is an important area for
further research in behavior analysis.5

Emergence of Conditioned Reinforcers Through

Observation
An additional type of OL involves the conditioning of reinforcers
through observation (Greer et al., 2006). Greer and Singer-Dudek (2008)
assessed whether neutral stimuli (discs and strings) would function as
conditioned reinforcers after children observed a peer be reinforced with
those items. First, the researchers demonstrated that the neutral stimuli did
not have an effect on the participants’ performance (skills already in the
participants’ repertories) or learning (acquisition of new skills). Then, the
researchers implemented an observational intervention that involved the
participants’ working side by side with a peer on the same task, with trials
being presented simultaneously. An opaque partition was between them,
however, such that the participants could see only their peers’ head and
shoulders and the cup where reinforcers were delivered, not their peers’
responses. The researchers placed discs or strings in the peers’ cup
contingent on correct responding, and did not correct errors. Participants’
cups remained empty after both correct and incorrect responses (i.e., their
access to the previously neutral stimuli was restricted). Findings
demonstrated that the observational intervention resulted in the conditioning
of the discs and strings as reinforcers.
Singer-Dudek and Oblak (2013) expanded upon the findings of Greer
and Singer-Dudek (2008) by specifically evaluating the role of peer presence
on the acquisition of conditioned reinforcer functions from observation. The
researchers compared two variations of the methodology used by Greer and
Singer-Dudek: a no-peer condition where participants worked side by side
with an empty chair, and a peer-present condition where participants worked
side by side with a peer. In both conditions the neutral stimuli were placed in
the translucent cup, in front of either the peer or the empty chair, while the
participants had no access to the stimuli (strings or toothpicks). Results
showed that conditioned reinforcement functions emerged for both
maintenance and acquisition tasks only after the peer-present observational
intervention, indicating peer presence is important. In addition, similar results
were obtained across participants who observed familiar (two participants)
and unfamiliar (one participant) peers, suggesting that peer familiarity was
not critical. While OL is generally regarded as learning new behavior as a
function of observing a model engage in that behavior, this line of research
demonstrates that reinforcers may also be conditioned through observation.
Importantly, derived stimulus relations have also been shown to develop as
a function of observation, and the final section reviews this literature
specifically.

Observational Learning Research in Derived

Stimulus Relations
A growing body of research in OL has examined the extent to which
untrained or derived stimulus relations emerge as a product of observation.
Sidman (1971) was the first to demonstrate the development of derived
stimulus relations when he taught an individual with severe intellectual
disability to relate dictated names to their corresponding pictures and also to
their corresponding printed texts. Following this direct teaching, the
individual labeled the pictures used in the training, read the corresponding
text when these stimuli were presented in isolation, and matched the
pictures to the text and vice versa. These additional relations were
performed without direct teaching. The emergence of these skills
demonstrated that a stimulus class consisting of dictated names, pictures,
and printed words was formed, and that the stimuli were functionally
equivalent (Sidman, 1971; Sidman & Tailby, 1982). This teaching paradigm
has demonstrated efficacy for teaching a wide variety of language and
cognitive skills to typically developing populations and those with
disabilities (Rehfeldt, 2011). The paradigm has wide implications because it
can lead to efficient use of instructional time (see chapter 10 for a specific
overview of stimulus equivalence). This efficiency can be further enhanced
when a participant observes another learner performing conditional
discriminations, and these observations lead to subsequent formation of
stimulus classes. To date, only a handful of studies have directly evaluated
the impact of OL on stimulus equivalence relations (MacDonald, Dixon, &
LeBlanc, 1986; Ramirez & Rehfeldt, 2009; Rehfeldt, Latimore, & Stromer,
2003).

Stimulus Equivalence and Observational

Learning
Training approaches to teach derived stimulus relations to learners with
developmental disabilities will often include a variety of match-to-sample
(MTS) tasks. During MTS tasks, a learner is presented with an array of
stimuli (termed comparison stimuli) and instructed to select the one that
corresponds to a stimulus that has been previously or simultaneously
presented (termed a sample stimulus). Responses are reinforced when the
learner makes a correct selection. Teaching may consist of two or more
directly trained relations (e.g., A-B and B-C) followed by tests for many
more untrained relations (e.g., A-A, B-B, C-C, B-A, C-B, A-C, and C-A).
MacDonald, Dixon, and LeBlanc (1986) were the first to examine the
role of OL in stimulus class formation. They recruited four adults with
moderate and severe intellectual disability and taught them to match arbitrary
visual stimuli. Each participant was taught one set of sample-comparison
stimulus relations (e.g., B1-A1, B2-A2) until they demonstrated a mastery
criterion. The participant then observed another participant perform a second
set of sample-comparison stimulus relations (e.g., C1-A1, C2-A2) such that
each observed relation had one stimulus in common with a relation that had
been directly taught to participants. During tests for derived relations, all
participants demonstrated at least one such relation between the modeled
stimuli.
Rehfeldt et al. (2003) replicated and extended this work with individuals
with autism and other developmental disabilities. Participants were taught to
conditionally relate a set of stimuli and were also required to observe a peer
learning conditional discriminations with a different set of related stimuli.
The experimenter verbally prompted the participants to observe the model by
saying, “Watch [name],” and provided verbal praise when the participant
observed the peers. This prompt was effective in establishing attending for all
participants in the study (Rehfeldt et al., 2003). The experimenter praised the
model while the observer watched the model perform conditional
discriminations. All participants demonstrated stimulus class formation for
the stimuli that were directly trained and improved posttest scores relative to
pretest scores for the stimuli that they observed a peer learning. A second
experiment replicated these results using stimulus sets from similar
overarching categories and words that could not be discriminated easily by
form or shape. Results showed that participants demonstrated emergent
performance for all stimulus sets they were trained on, and also displayed
some of the relations that they had observed a peer learning.
Finally, Ramirez and Rehfeldt (2009) explored the impact of OL to
teach Spanish vocabulary words to typically developing children. In their
experiment, one child was taught conditional discriminations between
dictated names and their corresponding pictures across three stimulus sets
while a second child observed. Results showed the emergence of symmetrical
relations in the form of labeling the pictures used in training by both children.
Collectively, the results of these studies show support for the use of arranging
observation of peers as an attractive approach to teaching conditional
discriminations that may result in derived stimulus relations. All three studies
used an MTS procedure, and each reported mixed success in establishment of
derived stimulus relations. These results suggest that the MTS training
procedure may not be effective for all learners and that alternative procedures
should be evaluated.

Stimulus-Pairing Procedure
One alternative to MTS training that has received empirical support in
recent years is a stimulus-pairing observation (SPO) procedure (also
termed stimulus-pairing training or respondent-type training; Rosales,
Rehfeldt, & Huffman, 2012; Smeets, Leader, & Barnes, 1997; Smyth,
Barnes-Holmes, & Forsyth, 2006). This procedure involves the observation
of a stimulus briefly presented in isolation, followed by an inter-stimulus
interval (ISI) and then presentation of a second stimulus. Thus, the
procedure involves pairing of stimuli with no overt response requirement
from the participant other than attending to the stimuli presented by the
experimenter. Presentations of stimuli may occur on a computer screen or in
vivo. These stimulus pairings are followed by tests for derived stimulus
relations (Leader & Barnes-Holmes, 2001a, 2001b). Since there is no mastery
criterion requirement in the SPO procedure, presentation of stimulus pairs
typically occurs for a predetermined number of trials before tests for derived
stimulus relations are presented.
The SPO procedure may be more efficient than MTS in terms of time
required for learning stimulus relations. It is possible that in MTS tasks,
responses are under various types of stimulus control, including irrelevant
features of comparison stimuli or varied position of stimuli (Fields, Garruto,
& Watanabe, 2010). If participants learn stimulus relations through an SPO
procedure, instructors can potentially decrease some of these errors.
However, studies that have directly compared the efficacy of MTS and SPO
have reported mixed results, with neither procedure showing facilitation of
derived stimulus relations for all participants (Clayton & Hayes, 2004;
Fields, Reeve, Valeras, Rosen, & Belanich, 1997; Kinloch, McEwan, &
Foster, 2013).
Leader and Barnes-Holmes (2001b) demonstrated the efficacy of SPO to
teach typically developing five-year-old children academically relevant
stimulus relations. The visual stimuli paired were fractions (A stimuli), the
corresponding pictorial representations for each numerical property (B
stimuli), and the corresponding decimals (C stimuli). The fraction-pictorial
representation pairing (A1-B1, A2-B2) was conducted first, followed by the
decimal-pictorial representation pairing (C1-B1, C2-B2). Tests for derived
stimulus relations subsequently demonstrated that all participants were
successful in matching the fractions to their corresponding decimals (A1-C1,
A2-C2) and the decimals to fractions (C1-A1, C2-A2). Two follow-up
experiments demonstrated that participants also established relations between
pictures that were not specifically targeted in training.
 Rosales et al. (2012) also recruited typically developing children
learning English as a second language and examined the effectiveness of the
SPO procedure to teach a small vocabulary set. Participants were required to
attend to the experimenter presenting an auditory stimulus paired with a
visual (object) stimulus in isolation for a predetermined number of trials.
Tests for emergence of stimulus relations demonstrated that the participants
acquired most of the listener relations targeted (i.e., selecting a stimulus
when named in an MTS format from an array of three items) and some
speaker relations (i.e., naming the stimulus when presented in isolation). The
experimenters then conducted multiple exemplar instruction (MEI;
involving direct teaching of tact relations with novel stimulus sets, followed
by probes with the original training set) in an attempt to demonstrate the
emergence of the untrained relations and reported moderate success.
The findings of the SPO literature reviewed thus far are exciting and
demonstrate support for the SPO procedure to establish derived stimulus
relations in typically developing learners. The utility of the SPO procedure
may be most evident for learners with autism and other developmental
disabilities due to the potential savings in training time. However, only a
handful of studies have examined the possibility of the SPO as an
instructional procedure for this population. For example, Takahashi,
Yamamoto, and Noro (2011) used SPO to teach relations among names
(printed words), faces, and Kanji symbols to two boys with autism. The
experimenters paired two visual stimuli (faces and printed words) in
succession on a computer screen for one participant, and an auditory and
visual stimulus (Kanji symbol plus corresponding auditory stimulus) for a
second participant. Following SPO, the experimenters administered
subsequent MTS tests to assess emergent relations among these stimuli for
each participant. Results indicated the SPO procedure was effective in
producing increases in correct responses across all participants, but the study
did not directly compare the efficacy of SPO with MTS procedures.
Vallinger-Brown and Rosales (2014) extended the applied research on
the SPO procedure for learners with autism by directly comparing the
effectiveness of MTS and SPO to facilitate the emergence of untrained
responses in children with autism. All participants were exposed to both
training conditions, with different sets of stimuli. During the MTS procedure,
participants were presented with a sample auditory stimulus and were
required to select the correct corresponding visual stimulus from an array of
six images presented on an iPad. During the SPO procedure, participants
were presented simultaneously with an auditory and visual stimulus and
required to observe the screen throughout the training. Results of this study
showed mixed results, with one participant demonstrating untrained stimulus
relations following training in both procedures and two participants
demonstrating minimal increase in performance relative to baseline following
the MTS procedure. Finally, a transfer of stimulus control procedure was
used to directly teach the target relations to all participants. These authors
noted that a possible prerequisite skill for success with the SPO procedure
may be overt or covert echoic responding, but this was not directly evaluated
in the study.
A similar study by Byrne, Rehfeldt, and Aguirre (2014) further
confirmed the effectiveness of an SPO procedure to teach speaker and
listener responses (i.e., selection-based responding) to learners with autism.
This study also demonstrated the value of MEI to help establish some of
these trained and untrained relations. However, these results counter the
notion that echoic responding may play a role in levels of accuracy following
training with the SPO procedure because the participant who emitted the
fewest number of echoic responses during training performed at optimal
levels during tests for untrained relations. The authors acknowledge the
possibility that this participant may have been engaging in covert echoic
responding, but further investigation is needed on this topic.
Future research on the effectiveness of the SPO procedure should
systematically examine the prerequisite skills needed for this procedure to be
effective. Some research has demonstrated prerequisite skills for learners to
benefit from OL more generally (MacDonald & Ahearn, 2015), and it would
be interesting to determine if these skills are similar to those that may be
required for the formation of derived stimulus relations from observation.
Another alternative for learners with more advanced language repertoires
may involve indirect measures such as posttraining interviews with
participants to inquire directly about procedures used to learn the relations.
Given the relative ease with which the SPO procedure can be implemented
and the parallels it has to naturalistic instruction, further research in this area
seems warranted. If derived stimulus relations may develop through
sequential or simultaneous presentation and observation of related stimuli,
there are broad implications for its use in practice.
 Extended Implications of Observational
Learning
Aside from derived stimulus relations, research in OL and specifically
on the use of the SPO procedure has demonstrated that transfer of stimulus
functions can also occur (Smyth et al., 2006). Smyth and colleagues
demonstrated this phenomenon using the SPO procedure in an experiment
with participants who self-identified as spider-fearful. The participants were
exposed to four stimulus pairings and subsequently tested for a transfer of
function for reports of self-arousal in the presence of stimuli that were never
directly paired. A second experiment demonstrated differential responding on
self-reported levels of arousal when videos of spiders where shown to spider-
fearful and non-spider-fearful participants. These results demonstrate the
utility of the SPO procedure and offer support for a derived-relations model
of the acquisition of anxiety responses (Smyth et al., 2006). Related studies
also show support for the notion that conditioned fear or excessive avoidance
behavior, which is a defining feature of anxiety disorders, may emerge
through observation (Cameron, Roche, Schlund, & Dymond, 2016; Cameron,
Schlund, & Dymond, 2015). Clearly, OL has implications for understanding
a wide range of psychological phenomena.

Summary of OL Research in Derived Stimulus

Relations
While more research is needed, the existing research demonstrates that
learners can acquire more skills via OL and derived stimulus relations
interventions than by directly observing a model performing those skills
(MacDonald et al., 1986; Ramirez & Rehfeldt, 2009; Rehfeldt et al., 2003).
That is, OL can also lead to the formation of derived stimulus relations and
transfers of stimulus functions (Smyth et al., 2006). In addition, the basic and
applied research on the SPO procedure provides some initial support for its
use to teach conditional discriminations and establish untrained relations
among stimuli for typically developing populations (Leader & Barnes-
Holmes, 2001b; Rosales et al., 2012) as well as learners with developmental
disabilities such as autism (Byrne et al., 2014; Takahashi et al., 2011;
Vallinger-Brown & Rosales, 2014). Moreover, there is growing support for
the participation of OL across a wide range of psychological happenings.
Implications and Future Directions
Behavior analytic work in the area of OL has a number of exciting
implications for both research and practice. An extensive amount of
learning occurs from observation; thus, the development of an OL
repertoire is of great importance. Moreover, developing the skills to learn
from observation permits individuals to learn faster, and therefore to learn
more, and to adapt to changing circumstances more quickly. Given this, we
encourage practitioners to become familiar with the suggested prerequisite
skills described in the research literature, assess for their presence, and
program for their development when deficits are identified. We also
recommend that generalization be assessed regularly, both across variations
of particular skills and across skill domains, with the ultimate goal being the
development of a highly generalized OL repertoire (e.g., MacDonald &
Ahearn, 2015). Behavior analytic researchers may contribute to this
important area by evaluating prerequisites more systematically in research
studies and by studying various interventions to promote OL with specific
learner profiles. Given the importance of developing an OL repertoire, this
area demands considerable attention from behavior analysts.
Practitioners should also consider the use of observational interventions
in the development of derived stimulus relations. A growing body of research
has shown how OL can lead to the development of derived stimulus relations,
with multiple skills being learned in a highly efficient manner. As with OL
more generally, more specific attention to prerequisite skills is needed in this
area, so that both practitioners and researchers develop a better understanding
of when to use such programs, and further, how to better prepare clients for
participation in such programs. This, in addition to the results of studies on
the development of conditioned reinforcers through observation, highlights
that OL may involve much more than observers’ engaging in modeled
behavior at a later time.
It seems likely that some type of OL participates in a great range of
common psychological events. As mentioned in this chapter, observational
processes have been shown to play a role in the development of fears, for
example. It is possible that observational processes contribute to the benefits
of mindfulness-based interventions as well, as such interventions often
involve the deliberate focus on particular features of one’s environment. To
the extent that some psychological problems are a product of a lack of
attention to and observation of stimulus relations in one’s environment,
interventions may be aimed at promoting such observation. Importantly,
however, Greer et al. (2006) noted that several terms within the OL literature
seem to be used in various ways, and additional conceptual clarity may be
needed as the research in this area continues to evolve.

Conclusion
Observational learning is a conceptually interesting topic with a number of
clinically important implications. Our aim in writing this chapter was to
provide readers with an overview of the behavior analytic work in OL,
including theoretical foundations, an overview of research in the area of
OL, a consideration of OL’s role in derived stimulus relations, and
implications for future research and practice. We hope that this chapter
impacts clinical work and stimulates further research in the area of OL.

Study Questions
1. Why is observational learning a socially important topic for
behavior analysts to be familiar with?
2. How does Bandura distinguish between learning and performance?
3. How does reinforcement history explain some of the findings in the
observational learning literature?
4. How might “atomic behavior” explain learning from observation?
5. Describe two of the processes outlined by Masia and Chase (1997)
to explain observational learning.
6. What are two important prerequisite skills for observational
learning?
7. Describe one of the peer-monitoring interventions reviewed in the
chapter.
8. How is the literature on conditioning reinforcers through
observation different from other research in the area of
observational learning?
 9. Describe the stimulus-pairing observation procedure. What makes it
different from match-to-sample training procedures?
10. Comment on the “extended implications” of observational learning.
11. What is one theme for practitioners to take away from the chapter?
Comment on why this is an important topic in clinical practice.
 Chapter 9:

Generative Responding Through

Contingency Adduction6

Kent Johnson
Morningside Academy
Elizabeth M. Street
Central Washington University

Link and Preview

The present chapter focuses on a topic that is central to the
behavior analysis of language and cognition: generative
responding. The chapter presents an overview of conceptual
foundations, provides references to seminal work in the area, and
addresses numerous implications for the design of instructional
programs. While much of the content focuses on the topic of
education specifically, the processes involved relate to the
consideration of language and cognition in behavior analysis more
generally.

Generativity is the study of the conditions that produce novel, untaught

responding in new circumstances, without directly programing them. Not
only is the study of generativity important for a scientific understanding of
complex human behavior, it is also important from a practical perspective.
Educators cannot possibly teach everything a student needs to learn in order
to be an effective independent adult. Even full mastery of the K–12
curriculum will not do the trick. Successful adults engage in untaught
blends and recombinations of behavior they previously were taught in order
to meet the requirements of contexts not previously presented in instruction.
In order to engage in untaught blends and recombinations that meet current
contingency demands, people must have component repertoires relevant to
those new situations that can be blended and recombined. The new
situations must function as an effective teacher. As Ferster (1965)7 noted,
It may seem a paradox that the listener (student) needs essentially
the same repertoire as the speaker (teacher) if communication is to
be effective. What, then, was communicated? Actually an
instruction was communicated, a rearrangement of existing verbal
behavior so that new combinations can occur.
This analysis is also consistent with Skinner’s concept of composition,
discussed in Verbal Behavior (Skinner, 1957, p. 123, pp. 344–352).
This conceptualization has driven our generative instruction model of
teaching and learning. The thrust of generative instruction is to arrange
instruction of key component skills, facts, concepts, and principles in such a
way that students will engage more frequently in more complex academic
behavior without direct teaching. We have discovered that complex
behavioral repertoires emerge without explicit instruction when well-selected
component repertoires are appropriately sequenced, carefully instructed, and
well rehearsed.
Our model was developed at Morningside Academy in Seattle, a
laboratory school for typical and near-typical elementary and middle school
students who struggle to acquire foundational reading, writing, math,
thinking, reasoning, and problem-solving repertoires. Our students have not
previously reached their potential; many have been diagnosed with learning
disabilities such as dyslexia, dysgraphia, dyscalculia, or ADHD. All have
average to above average intelligence test scores. Morningside is not a school
for children with significant emotional problems, behavioral problems,
developmental delays, or autism spectrum disorders. One hundred thirty
public partner schools and agencies in the United States, Canada, and Europe
have adopted our generative instruction model to date (Johnson & Street,
2004, 2012, 2013, Street & Johnson, 2014).
We will report our discoveries and investigations of generative
responding, or untaught responding, in academic skill development as well
as thinking, reasoning, and problem-solving development. The data that we
will share have come from many classrooms across the United States, as well
as an associated instructional design company. Both of us authors are
educational practitioners, so neither of us has conducted detailed or
exhaustive controlled studies.8 Yet our descriptive data show such consistent
patterns that we want to share them with the wider behavioral community in
the hopes that other practitioners will join us in our inductive explorations
and that researchers will join us by conducting controlled studies of the
phenomena that we have confirmed in our multiple settings.

Generativity and Contingency Adduction

The work of three groups of behavior analysis investigators has influenced
our work in generative instructional design: Epstein, Skinner, and their
colleagues’ studies of generativity; Andronis, Goldiamond, and their
colleagues’ research on contingency adduction; and Haughton, Barrett,
Binder, and other precision teachers’ work in behavioral fluency. Here, we
will briefly describe the first two groups’ influences, and later in the chapter
we will describe the third group’s influence.
Dr. Robert Epstein, a student of B. F. Skinner’s, described a series of
animal laboratory experiments that were intended to provide a behavioral
interpretation of “insight,” akin to Kohler’s gestalt psychology of the mental
life of other animals (e.g., Epstein, 1985; Epstein, Kirshnit, Lanza, & Rubin,
1984). Epstein and his coworkers found that pigeons that were individually
taught to push a small box around an experimental chamber, to step on the
box, and to peck at an object—all in separate training sessions—would solve
the problem of pecking at the object when it was out of reach. When
presented with the problem, the birds would demonstrate a series of
“insightful” maneuvers, much like Kohler’s apes, revealing that behavior
analysis can account for such behavior in other animals. In remarkably
similar behavior patterns, birds would observe the experimental chamber,
move the box under a banana, step on the box, and peck the banana. Only
birds that received instruction in all three of the component skills
successfully solved the problem. Epstein (1985) described the phenomenon
as a “spontaneous” (unprompted, not arranged) interconnection of existing
repertoires to solve a problem. The existing repertoires were learned
separately, under the same contingencies that prevailed in the problem
situation. He later expanded this work to show the emergence of novel
behaviors in humans under similar circumstances and used the term
“generativity” to name the phenomena. Epstein (1993, 1999) also developed
a computer-based algorithm for simulating the shifting probabilities of
competing behaviors as the interconnection process unfolds. If Epstein’s
pigeons were humans, most people would say that they “figured out” how to
solve their problem, that they showed “insight,” and that they were “very
creative” or “clever.” From this colloquial perspective, generative instruction
is meant to produce students who are clever, creative, insightful problem
solvers.
In contrast to Epstein’s studies, which showed how component teaching
can be added to an existing repertoire to solve a specific problem, Andronis,
Layng, and Goldiamond (1997) used the term contingency adduction to
describe a related but different phenomenon: Behaviors learned under one set
of contingencies can be selected or recruited by new, very different
contingencies that present a new and different problem. The performance is
reinforced in the new situation by a different effect on its environment than
the effects achieved when the originally established behaviors were emitted.
That is, the recruited behaviors form new combinations or blends that serve a
new or different function in a new context, solve a new problem, and meet
the requirements for success in a new situation, different from the
requirements for success under which those recruited behaviors were
previously learned. According to Andronis et al. (1997), the moment of
reinforcement of a behavior established with one set of conditions by an
entirely different set of conditions marks the “Aha!” moment of contingency
adduction. No direct instruction or programing of the adduced behavior or
behavioral repertoire occurs. Adduction can occur only once, after which the
performance becomes part of a student’s behavioral history, available as
needed to meet reinforcement requirements, and available for further
recombining and blending with other components during contingency
adduction in future problem solving and discovery learning.
Novel behavior produced by recruitment is different from novel
behavior produced by shaping of successive approximations. The new,
adduced behavior “comes pre-shaped, and is selected or recruited by different
contingencies than those responsible for the initial shaping of the behavior”
(Layng, Twyman, & Stikeleather, 2004a, p. 99). Andronis et al. (1997) used
the term “contingency adduction” rather than “adduction” to emphasize that a
student does not adduce the performance, the new contingency does. The
new contingency shares common features with the original contingencies that
produced the performance. The term “adduction” is applied in its definitional
sense: adduct means “to bring together similar parts,” and adduction means
“the act of adducing or bringing forward” (Webster’s New Collegiate
Dictionary, s.v. “adduct” and “adduction,” respectively). In the context of
academic skill development, contingency adduction allows a teacher to skip
one or more subsequent instructional steps on a curriculum ladder because
the student can already engage in the performance required by an
instructional objective.

Examples of Contingency Adduction: Five

Variations We Have Programmed
Johnson and Layng (1992) first provided empirical evidence for
contingency adduction in fractions problem solving, eliminating the need for
instruction. Four students enrolled in a Morningside summer school program
for neighborhood children at Malcolm X College in Chicago were assessed in
solving word problems with fractions. The number of correct responses
ranged from three to seven out of fourteen assessment problems. Other
assessments showed similar deficits in lower-level, whole number word
problems and fractions computation skills. The students were placed in
instructional sequences in which they were taught the skills required to solve
whole number word problems and to add, subtract, multiply, and divide
fractions and mixed numbers (i.e., combinations of whole numbers and
fractions). Both the whole number problem solving and fractions
computation skills were then shaped to accuracy and fluency criteria with
separate instructional sequences. After the students had practiced the lower-
level skills until they had reached high frequencies of correct responding,
they were given a contingency adduction probe: Could they now solve a
similar set of fraction word problems correctly when simply given the
instruction, “Can you do these problems now?” with no instruction in
fractions problem solving? Three of the four students correctly solved all
fourteen word problems and one solved thirteen of the fourteen problems
correctly. The same test environment in which their previous performance
had failed now resulted in a highly successful adduction of a new repertoire,
“for free” as we say colloquially. Adduction was not a product of gradual
shaping but appeared fully established as a function of establishing its
 component parts and providing a new instruction. With no fraction problem
solving instruction necessary, fluency building was prescribed to assure
retention of this adduced repertoire.
Let us examine this example closely to see how it meets the
requirements of contingency adduction. Applying Tiemann and Markle’s
(1990) method of concept analysis to contingency adduction reveals three
critical features, listed in table 9.1. Two relevant behaviors were in the
students’ behavioral histories—whole number word problem solving and
fractions computation—thus meeting the first requirement for contingency
adduction. A new, different contingency was presented—fractions word
problems—with the instruction, “Can you do these problems now?”
illustrating the second criterion. Finally, the previously established behaviors,
whole number word problem solving and fractions computation, formed a
new behavioral sequence that served a new and different mathematical
function, solving word problems involving fractions. Solving fractions word
problems meets the new reinforcement requirement. Neither of the originally
established behaviors was sufficient to meet the new contingency
requirement. Solving fractions word problems under these conditions is
adduced behavior.

Table 9.1. Three Critical Features of Contingency

Adduction

Contingency Adduction Criteria Notes

1. One or more behaviors are reinforced under one It/they become(s) part of
set of contingencies (conditions, circumstances). the behavioral history of
the learner.

2. A different set of current contingencies Recruitment is the

(conditions, circumstances) recruits one or more occurrence and
behaviors in a person’s history (#1). reinforcement or
selection of that
behavior in the new
circumstance.

3. A novel stimulus control relation is formed from The moment of

the behaviors in the person’s history, one that serves recruitment and
a new and different function or purpose than the reinforcement (#2) is the
function served when the behaviors were first moment of contingency
learned and reinforced. adduction.9

A complete concept analysis (Tiemann & Markle, 1990) includes both

critical features and variable features. Each example of a concept contains
features that vary from instance to instance while maintaining membership in
the concept class. We have so far discovered five variations of contingency
adduction that meet the requirements for membership in the concept class.
Table 9.2 describes these five variations, and table 9.3 shows three methods
for designing instruction to produce contingency adduction that we have
engineered so far. Space allows us to give only one or two examples of
academic skill development for each variation, but we encourage readers to
generate many more examples.

Table 9.2. Some Variations of Contingency Adduction

Variation New What Occurs Text Example

Antecedent
Condition

1. A combination A new sequence of Fractions word

Adduction of stimuli, each already established problem solving
of a new of which component behaviors without instruction, as
sequence of controls occurs. a function of new
previously different antecedent
learned behaviors requirements, and
behavior previous instruction in
whole number word
problem solving and
fractions computation

2. A combination A new blend (a new Saying consonant

Adduction of stimuli, each topography) of already blends without
of a new of which established behaviors instruction, as a
blend of controls occurs. function of new
previously different antecedent
learned behaviors requirements and
behavior previous instruction in
consonant sounds

3. A component of A component of already Saying consonant

Adduction a previously established behavior sounds without
of a encountered occurs. instruction, as a
component (“familiar”) function of new
of stimulus antecedent
previously requirements and
learned previous instruction in
behavior consonant blends
from
context

4. A new stimulus, (a) A new response Saying new

Adduction presented occurs from the set of vocabulary words,
of new together with new and previously pictures, and
vocabulary previously encountered stimuli. definitions, as a
from encountered function of new
(b) The sequence of
context, stimuli that antecedent stimuli and
training restricts
adduction serve as prior instruction in
response alternatives to
by process nonexamples other vocabulary
the one remaining
of based upon a words, pictures, and
stimulus, correlating the
elimination prior history of definitions
new stimulus to the
instruction
requirements for
responding contained in
the instruction.

5. A new (a) Already established (1) Saying untaught

Adduction arrangement of behavior not previously fractions, decimals,
of new previously occasioned by the new and percent
equivalence encountered antecedent stimulus equivalence relations,
relations stimulus- occurs. as a function of prior
response instruction in some
(b) The sequence of
relations fraction, decimal, and
training events restricts
 response alternatives, percent equivalence
linking a stimulus from relations.
one previously
(2) Selecting untaught
established stimulus-
picture-definition
response relation to a
equivalents and word-
behavior established in a
picture equivalents, as
different stimulus-
a function of prior
response relation.
instruction in word-
definition equivalents.
(3) All relational
operants?

Table 9.3. Three Procedures for Designing Instruction to

Produce Contingency Adduction10

Procedure Text Examples

Combined stimulus (1) Sequence adduction of fractions problem

procedure solving
(2) Blended topographies adduction of
consonant blends

Exclusion procedure (1) Component adduction of consonant sounds

(oddity-from-sample
(2) Adduction of vocabulary words, pictures,
procedure)
and definitions

Equivalence procedure (1) Saying fraction, decimal, or percent

equivalent of another fraction, decimal, or
percent
(2) Selecting definition-picture equivalents and
word-picture equivalents

VARIATION 1
 Returning to our example of fractions word problem solving, we see in
Variation 1 (see table 9.2) that a new sequence of previously established
behaviors was adduced by the contingencies of presenting word problems
with fractions. Consulting table 9.3, we see that the adduction of the fraction
word problem-solving sequence was produced by a combined stimulus
procedure. The combined stimulus process occurs naturally when students
read text and encounter new words that are recombinations of sounds they
have mastered. Students regularly decode thousands of new words after
learning the forty-six sounds of letters and key letter combinations (Alessi,
1987). These new math and reading performance sequences allow the
instructor to skip instruction for solving fractions word problems and sight
word reading and move on to the next more advanced instructional objectives
in the math and reading curriculum ladders, saving lots of time and effort.

VARIATION 2
Layng and colleagues (2004a) reported a second variation of novel
behavior that meets the three requirements for membership in the concept
class contingency adduction. It occurs in our interactive, animated Internet
reading program, Headsprout Early Reading (http://www.headsprout.com).
The program explicitly teaches consonant sounds, such as /c/, /r/, /f,/ /l/, /s/,
/p/, /t/, and /n/. Later, the program says, “I bet you can figure out new sounds
all by yourself!” Students are then presented combinations of the directly
taught sounds, such as /sn/, /cr/, /sl/, /fl/, /pl/, /sp/, /tr/, and /st/ and asked to
listen to and select the correct sound combinations.
Data collected on over 11,000 students showed that between 80% and
95% of students correctly identified the new sound combinations in the
absence of instruction (i.e., “for free”).11 Let us again examine table 9.1
closely to see how this example meets the requirements of contingency
adduction. Isolated consonant sounds were in the students’ behavioral
histories, thus meeting the first requirement. A new, different contingency
was presented—consonant blends with the instruction, “I bet you can figure
out new sounds all by yourself!”—illustrating the second criterion. Finally,
the students formed new consonant blends from the previously established
isolated consonant sounds, serving a new or different reading function. To
meet the new reinforcement requirement, students had to select the correct
consonant blends when spoken by the program. None of the originally
established consonant sounds were sufficient to meet the new contingency
requirement. Correctly selecting consonant blends when heard under these
conditions is adduced behavior. In this example we see that a new blend of
previously established behaviors was adduced by the contingencies of
presenting consonant sounds together, the second variation listed in table 9.2.
Consulting table 9.3, we see that adduction of consonant blends from
consonant sounds was produced by a combined stimulus procedure.
One student in the fractions problem-solving example made an error in
the contingency adduction probe, and 5% to 20% of the 11,000 students in
the consonant blends example did not successfully adduce the correct
behavior in the first adduction probe. One or more component repertoires
may emerge during an adduction probe. We call each response to an
adduction probe a candidate response. Sometimes a candidate response is
reinforced and becomes adduced behavior. When a candidate response is not
reinforced, it is not adduced and does not become part of the selecting
environment for further new repertoires.

VARIATION 3
We have engineered a third variation of novel behavior that meets the
three requirements for membership in the concept class, contingency
adduction. It occurs when a component of a previously learned behavior, but
not the composite behavior itself, meets the requirement for reinforcement.
Another sequence from instruction in Headsprout Early Reading illustrates
this process (Layng et al., 2004a). Some individual consonant sounds are
adduced by segmenting previously taught consonant blends. For example, the
sound /n/ is adduced from the previously taught sight word “an.” As the
program narrator says, “Some sounds have other sounds inside them!” “An”
moves to the left side of the screen and “n” is presented on the right side. The
narrator then says, “Click on the sound that does not say “an.” The student
responds “away” from the previously learned sight word “an,” and clicks on
“n.” The narrator then says, “Yes, n.” Then “n” is presented alongside of “an”
and “a.” The program says, “Click on ‘n.’” When the student clicks on “n,”
the program says, “Yes, n.” This is the moment of contingency adduction. Six
consonant sounds are taught this way: /n/, /c/, /l/, /f/, /r/, and /p/. Four dyads,
/an/, /cl/, /fr/, and /ip/, were taught earlier, from which these six consonant
sounds were segmented. Data collected on almost 33,000 students showed
that they correctly segmented between 86% and 97% of the new isolated
consonant sounds from the sound combinations “for free” (without
instruction). Eighty-nine percent to 98% of the sounds were isolated on the
first attempt, depending upon the sound.
Four sound combinations were in the students’ behavioral histories, thus
meeting the first requirement for contingency adduction in table 9.1. A new,
different contingency was presented, isolated consonant sounds, with the
instructions “Some sounds have other sounds inside them!” and “Click on the
sound that does not say [the letter combination],” illustrating the second
criterion. Finally, the previously established behaviors, four sound
combinations, were segmented into six isolated consonant sounds, serving a
new or different reading function. Selecting the correct isolated consonant
sounds from the sound combinations and other, incorrect isolated sounds met
the new reinforcement requirement. None of the originally established sound
combinations were sufficient to meet the new contingency requirement.
Correctly selecting isolated consonant sounds when heard under these
conditions is adduced behavior.
Adduction of a component of previously established behavior is the
third variation listed in table 9.2. Consulting table 9.3, adduction of these six
consonant sounds from previously established consonant blends was
produced by an exclusion procedure, specifically an oddity-from-sample
procedure (Layng et al., 2004a). The student “responds away” from a
previously learned stimulus, allowing a new stimulus, the only remaining
stimulus available, to recruit the novel behavior that meets the requirement
for reinforcement.

VARIATION 4
A fourth variation of contingency adduction is illustrated in the second
Internet program in the Headsprout system, Headsprout Reading
Comprehension (http://www.headsprout.com). One feature of the program
teaches students new vocabulary words in groups of four. The program
teaches thirty-one groups of four words, or 124 words, in total. Each word
has a correlated definitional word or phrase already in the student’s
repertoire, plus a picture illustrating the word. The fourth word in each group
is learned through an exclusion training procedure in the following manner.
After learning the first three vocabulary words with a different procedure,
described below, the fourth vocabulary word, definition, and picture are
introduced. The new vocabulary word appears in a row along with two of the
previously learned words, and the new definition appears below them. The
student must “respond away” from the two previously learned words and
select the remaining new word that correctly corresponds to the new
definition presented below the words. In subsequent instructional frames, the
new definition appears in a row along with two of the previously learned
definitions, and the new word or the new picture appears below them. The
student must “respond away” from the two previously learned definitions and
select the remaining new definition that correctly corresponds to the new
word or picture presented below the definitions. Likewise, other frames
present the new picture in a row along with two of the previously learned
pictures, and the new word or definition appears below them. The student
must “respond away” from the two previously learned pictures and select the
remaining new picture that correctly corresponds to the new word or
definition presented below the pictures. Whenever the student clicks on the
unfamiliar word, definition, or picture, the program narrator says, “Right!
That’s the word/definition/picture that goes with that
word/definition/picture!”; this is the moment of contingency adduction. Data
from twenty students show that just over 96% adduced correct answers on
the first trial through exclusion training (Joe Layng, personal communication,
August 14, 2017).
This example embodies the three critical features required for
membership in the concept class contingency adduction described in table
9.1. Table 9.2 lists this fourth variation, and table 9.3 identifies it as
adduction produced by an exclusion procedure. A large number of studies
have systematically replicated the exclusion effect with typically developing
children, children with disabilities, and undergraduate students learning
word-object relations. Most recently, exclusion procedures were used with
two-year-old children learning adjectives (Ribeiro, Gallano, Souza, & de
Souza, 2017). Our work also helps to extend the generality of using an
exclusion procedure.

VARIATION 5
A fifth variation of contingency adduction is familiar to behavior
analysts: stimulus equivalence. Much research exists to support
equivalence-based instruction, reviewed in chapter 10 in this book (e.g.,
Fields et al., 2009; O’Neill, Rehfeldt, Ninness, Munoz, & Mellor, 2015).
From a contingency adduction perspective, in stimulus equivalence
procedures, one component of a previously encountered stimulus-response
relation occurs with new instructions to relate it to a component of another
stimulus-response relation that has also already been learned. For example, at
Morningside we ask students to memorize the equivalences among certain
fractions, decimals, and percentages as a tool skill for learning algebra.
Students memorize the following equivalents among tenths, fifths, quarters,
thirds, and halves: 1/10 = .1 = 10%, 1/5 = .2 = 20%, 1/4 = .25 = 25%, 1/3 =
.33 = 33%, 2/5 = .4 = 40%, 1/2 = .5 = 50%, 3/5 = .6 = 60%, 2/3 = .67 = 67%,
3/4 = .75 = 75%, and 4/5 = .8 = 80%. Each of these ten sets of equivalence
relations contains three equivalents to master (i.e., fraction to decimal,
fraction to percent, decimal to percent), totaling thirty relations, which are
much more efficiently mastered in a stimulus equivalence paradigm. When
we teach students to memorize ten fraction-decimal equivalents and ten
decimal-percent equivalents, then present the ten fractions with appropriate
instructions (e.g., “Write the percent equivalents of these fractions”), the
adduced performance yields these ten relations “for free.”
This example of adduction by equivalence meets the requirements of
contingency adduction described in table 9.1. Adduction by equivalence is
the fifth variation listed in table 9.2. As Ferster said in 1965 (see the quote at
the beginning of this chapter, excerpted from Ferster, Culbertson, & Boren,
1975), the instruction rearranged the student’s existing repertoire. This time
the instruction, “Write the percent equivalents of these fractions,” rearranged
a component of students’ math repertoires, linking a stimulus from one
previously established stimulus-response relation to a behavior established in
a different stimulus-response relation. In other words, the sequence of
instructional events—memorizing fraction-decimal equivalents, then
memorizing decimal-percent equivalents—restricted response alternatives
when students were presented with fractions plus the instructions to provide
the percent equivalents. Knowing the decimal equivalents of fractions
restricted their responses to the instruction to the percent equivalents of the
decimals.
Stimulus equivalence is also incorporated in the procedures for teaching
vocabulary in Headsprout Reading Comprehension
(http://www.headsprout.com). To review, vocabulary words are taught in
groups of four. Each word has a correlated definitional synonym or phrase
already in the student’s repertoire, and a picture illustrating the word. Earlier
we described how the fourth word in the group is learned through an
exclusion procedure. The other three words in each group are learned through
exposure plus stimulus-equivalence practice, in the following manner.
Each word is visually presented with its definition in a sentence, along
with its picture. The narrator reads the sentence, then the student reads it.
After this exposure, stimulus equivalence is leveraged in a practice format
that adduces more formal word-definition equivalents, definition-picture
equivalents, and word-picture equivalents. Students are presented with a new
vocabulary word and, below it, the familiar definition equivalent in an array
of several other definitions. When they select the correct definition, the
narrator says, “Right! [New word] means the same thing as [familiar
definition].” Students also practice the reverse relation. When presented with
a familiar definition, students must select the correct new word from an array
of words. Likewise, students practice the other legs of the word-definition-
picture equivalents: definition-picture equivalents and their reverse, and
word-picture equivalents and their reverse. Thus, the stimulus-equivalence
procedure in these segments is exposure-initiated rather than contingency
shaped.
Data from twenty-four students from one of the thirty-one vocabulary
segments shows that they correctly selected the appropriate word when given
the definition, picture when given the definition, word when given its picture,
and the reverse of these relations 95% of the time on the first trial. Their
average number of errors was 1.29. Data from nineteen students from a
second vocabulary segment shows that they correctly selected the appropriate
answer 97% of the time on the first trial. Their average number of errors was
0.79
Table 9.3 lists adduction of new equivalence relations from previously
established equivalences as the third procedure that we have engineered so
far to produce contingency adduction. We are curious about the degree to
which contingency adduction analysis extends to other relational operant
investigations, which is why we have added “all relational operants” with a
question mark to table 9.2 as a third text example of the fifth variation of
contingency adduction.

Some Complex Classroom Examples of

Contingency Adduction
 The adduction examples we have presented so far clearly indicate
specific antecedent procedures responsible for their occurrence. Many
classroom examples are not so clear-cut, representing combinations of more
than one of the variations of adduction that we have discussed. For example,
after reading comprehension instruction in drawing a conclusion based on
information in a text, students may encounter an adduction activity that asks
them to make a skillful prediction at a certain point in a text, a new
comprehension skill that has not been taught. To be successful, students must
draw a conclusion from previously read text in a story—for example, that a
character is shy and timid because they fail to speak up when others make
false accusations about their past behavior. Then students must go beyond
their conclusion and use a new event in a story—another accusation by a
different character—to predict that when the character hears the false
accusation they will fail to object to it by providing an accurate account of
their own behavior. The combined stimuli of a new story event, a new
accusation, and the instruction to make a prediction adduce a prediction
about a future outcome in the story.
In writing instruction, after learning how to modify nouns with
adjectives and how to write dependent clauses, students may encounter an
adduction activity that asks them to write sentences with appositives, without
any additional instruction beyond a model sentence or two. For example,
after learning how to construct sentences such as “The surly, arrogant
candidate lost the election to a grassroots candidate,” the student may adduce
appositive sentence writing and write, “The candidate for state representative,
a surly and arrogant man, lost the election to a grassroots candidate.”
Excellent research-based methods for teaching grammar through sentence
combining such as those described by Linden and Whimbey (1990),
Whimbey, Johnson, Williams, and Linden (1993/2017), O’Hare (1973), and
Archer, Gleason, and Isaacson (2008) greatly enhances the power of
adducing sentence writing.
To illustrate the information described thus far in the chapter, below we
present an example of a teacher’s use of adduction in the classroom.

A Case Study in Adduction

Morningside teacher Marianne Delgado designed a generative process
to teach students how to adduce definitions of unfamiliar vocabulary words
by using four reading comprehension procedures: (a) using context clues to
infer the meaning of a word, (b) describing the connotation and mood that a
word evokes, (c) writing multiple forms of the word corresponding to various
parts of speech, and (d) writing the word in a complete sentence that
indicates its meaning (Delgado & Johnson, 2010). Using context clues to
determine the meaning of “writhing,” students copied the sentence in which
the unknown word occurred from the text, read the sentences that
immediately preceded and followed the sentence containing the unknown
word, and wrote their best guess of the word’s meaning. For example, when
students read, “LeGleo peered through the spyglass and viewed the writhing
mass, shuddering as he spied rats moving on every piece of the decrepit
ship,” one student copied the sentence, read the sentences before and after it,
and recorded this best guess: “I think writhing must mean wiggling as if the
rats are moving constantly. He shuddered so I know he didn’t like it.”
To describe the connotation and mood that “bedraggled” evokes,
Delgado instructed the students to decide whether the word had a positive or
negative connotation and give an instance in which they would use the word
in daily conversation. One student wrote, “I think bedraggled has a negative
connotation, because it was how (story character) Rikki Tikki Tavi appeared
after he had been in a flood. I would use bedraggled if I was describing how I
got caught in a rainstorm while waiting for a bus.” When writing multiple
grammatical forms of a word, one student wrote, “cache–noun, used with in.
He stored the dynamite in a cache. Cache–verb, cached, caching, caches. He
cached the code in secret files deep in the computer.” When writing sentences
that show they know the meaning of a word, students were instructed to
surround the word with words that might be contained in a definition and
with words that would describe a picture of the word in action. For example,
one student wrote, “On my trip to the farm I saw two oxen wearing a yoke
that was attached to the plow behind the animals,” and “On my trip I saw two
fledglings in a nest that hung over a masonry wall built with stones from the
riverbank.”
Delgado conducted multiple informal assessments of her procedures
with different subsets of the vocabulary words she taught throughout the
school year. For some new vocabulary words, students followed the
adduction process outlined above; for other words, students practiced
flashcards with new words on one side of the cards and teacher-given
definitions of the words on the other sides. When given a multiple-choice test
of words and definitions, students who adduced word definitions performed
as well as students given words and teacher definitions, averaging 4.3 correct
of five test items, compared with 4.1 correct of five items for the words they
learned by practicing flashcards. Later, when asked to use an accumulation of
twelve previously adduced words to write a story, and twelve words learned
from flashcards to write another story, students correctly used an average of
10.4 of the twelve adduced words, compared to only 8.9 correct of the twelve
words learned from flashcards.
The differences in the flashcard and adduction procedures became even
more obvious on two retention tests, using a different set of new vocabulary
words from those reported above. On one retention test of words memorized
from flashcards, students used 8.5 words correctly out of twelve words in
November, but six months later their performance declined to an average of
7.1 out of twelve correctly used words. However, on a retention test of
adduced vocabulary words, students wrote an average of 9.8 words correctly
in November, and maintained correct usage six months later, averaging 9.3
correct words. When asked to write two separate paragraphs employing six
different adduced words and six different flashcard words they had learned
six months earlier, students used an average of 5.2 of the six adduced words
correctly, but only 4.5 of the six flashcard words correctly.
Delgado and another Morningside teacher, Nicole Erickson (Delgado,
Erickson, & Johnson, 2015) replicated these results. Delgado and Johnson
(2010) had implemented the adduction procedure with students who
performed above grade level in reading and writing. Those students had
adduced vocabulary words in the course of preparing a presentation in a
project-based learning class. In the study that is reported in Delgado et al.
(2015), Erickson successfully used the adduction procedure with corrective
readers who performed below grade level in reading and writing. Her
students simply required more explicit instruction and correction procedures
to master the four reading comprehension procedures.
In addition to designing instruction to promote contingency adduction,
we can teach generative repertoires such as thinking, reasoning, and
problem solving that themselves will facilitate contingency adduction within
a sequence of components and composites.

Generative Repertoires
 Colloquially speaking, generative repertoires are at work when we
“figure out” how to do things we’ve never done before, draw new
conclusions, infer new knowledge, “have insights,” and so on. From a
behavior analytical perspective, current contingencies occasion generative
repertoires: supplementary verbal behavior from the problem solver that may
lead to repertoires not previously taught. The student’s generative repertoires
rearrange current circumstances to prompt generative responding that is
based upon new combinations and blends of behavior in our current relevant
repertoires. At Morningside, teaching generative repertoires is a high priority.
One important influence is the work of Arthur Whimbey and his
colleagues. Their book Blueprint for Educational Change: Improving
Reasoning, Literacies, and Science Achievement with Cooperative Learning
(Whimbey et al., 1993/2017) makes the case that much of novel responding
comes from the ability to reason or to engage in analytical thinking.
Whimbey et al. reason that many nonanalytic thinkers don’t engage in the
careful process of analyzing what they read or hear or think, a process that,
they assert, analytical thinkers use routinely. Specifically, Whimbey et al.
provide examples of how nonanalytical thinkers don’t read carefully or draw
clues from what they read. They say that nonanalytical thinkers don’t
question their first assertions or draw pictures to represent complex problems
as their more analytical classmates do; rather, they guess at answers that
seem, on the surface, to be correct. Whimbey and his colleagues show how
these habits lead to incorrect assumptions and ultimately incorrect answers.
Our observations confirm these findings. In addition we find that some
students don’t engage in self-questioning and that some have very
disorganized and nonsequential thinking.
Whimbey et al. (1993/2017) then build the case for making overt the
reasoning processes of good analytical thinkers that most often occur out of
view of an observer. They suggest that one of the reasons that some students
don’t acquire good reasoning skills is that they don’t get to see or hear what
the good analytical thinker is doing privately, in order to be able to learn from
or imitate it. It is this thesis that set the stage for Whimbey and his colleagues
to develop their seminal think aloud pair problem solving (TAPPS) method
of reasoning from a given problem. TAPPS builds on the notion that effective
analytical thinkers engage in a dialogue with themselves. Deciding what to
buy at a grocery store, solving a math word problem, and making predictions
while reading a mystery all may involve self-dialoguing. This notion aligns
perfectly with a radical behaviorist way of thinking. We can apply what we
know about observable behavior to events like thinking that we cannot see—
that’s the radical component of radical behavior analysis.
A significant amount of thinking is private verbal behavior: conversing
with oneself as a speaker and listener in the same skin, as Skinner (1953) put
it. Skinner called thinking precurrent behavior, a sequence of behaviors that
prime and prompt final solution behavior. The self as speaker prompts and
probes their own behavior and generates stimuli to supplement other
behavior already in their repertoire. The self as listener confirms, rejects, or
prompts new speaker behavior. The specific consequences of thinking may
advance problem solving closer to the solution, having a meaningful effect
on the thinker; the thinking behavior will thus be automatically reinforced.
Whimbey et al. (1993/2017) use a problem taken from the Whimbey
Analytical Skill Inventory, a test that Whimbey and colleagues constructed
and tested for its ability to separate analytical from nonanalytical thinkers.
They also provide a “script” of the think-aloud dialogue used by a
particularly analytical problem solver—a male lawyer, in this case—to solve
the problem, shown in the text box below. The lawyer’s dialogue illustrates
several of the characteristics of a good problem solver, including his concern
for accuracy and how he achieves it: writing a segment of the alphabet to
help him find the correct answer, pointing to letters as he counts them in the
partial alphabet he has written on the paper before him, making marginal
notes on what he has concluded at each point in his analytical thinking rather
than trusting his conclusions to memory, and, finally, because the problem
asks him to “Circle the consonant that is farthest to the right in the word,”
saying, “Which is my right hand? This is my right hand. So this direction is
right,” as he points to his right to be sure he is going in the right direction as
he counts off letters in the word. Similar examples are used to teach students
the value of drawing pictures and other similar routine skills of good problem
solvers.

An Expanded Version of a “Think Aloud” from

Whimbey et al. (1993/2017)
OK. The problem says, “If the word sentence contains fewer than
nine letters and more than three vowels, circle the first vowel.
Otherwise, circle the consonant that is farthest to the right in the
word.”

I’m going to start from the beginning.

If the word sentence contains fewer than nine letters…
Well, the first thing I need to do is to count the letters in the word
“sentence.” OK. 1, 2, 3, 4, 5, 6, 7, 8. I think eight is right, but just to
be sure, I’m going to count again, and this time, to make sure that I
don’t miss any letters, I’m going to touch the letters in order with
my pen as I say each number. 1, 2, 3, 4, 5, 6, 7, 8. OK, so the word
‘sentence’ has eight letters in it, which is fewer than nine. So, I’m
going to write the word “yes” above that part of the question so I
won’t forget.
But that’s not all. The problem says, “If the word sentence contains
fewer than nine letters and more than three vowels, circle the first
vowel.” I already know that the word sentence contains fewer than
nine letters, having determined that it actually contains eight
letters. That was the first test. So what is the second test? Oh,
here it is. It says “and more than three vowels.” Because the
instructions say and, it means that for me to circle the first vowel,
the word sentence must pass both tests: it must have fewer than
nine letters and more than three vowels. So the next thing I need
to do is to count the number of vowels in the word sentence.
But first, which letters are vowels? Oh yes. They are a, e, i, o, and
u.
Are there any of those letters—any vowels—in the word
“sentence?” Oh, I see three of them, here and here and here. OK. I
need to count them. I’ll use my pen. There are 1, 2, 3. OK. There
are three. Now, I want to read the part of the problem that deals
with this again. It says, “more than three vowels.” Well, “sentence”
does have three vowels, but it doesn’t have more than three. So I’ll
write a “no” above that part of the problem. And since I have
answered only one of the questions “yes,” I now know that my
answer will be wrong if I circle the first vowel.
So, now what do I do? Let me read the rest of it. “Otherwise, circle
the consonant that is farthest to the right in the word.” Okay, now I
know what to do. But wait, which direction is right? Oh yes, I know
that this is my right hand so this direction will be to the right. So the
letter that is farthest to the right is an “e.” But an “e” is a vowel, so I
need to keep moving back from the end until I come to a consonant.
Oh, here is one. It’s the “c” right before the final “e.” Now what was I
to do with it? Oh yes, I was to circle it. So that’s what I’ll do.

In addition to creating the inventory that successfully separates

analytical reasoners from nonanalytical ones, Whimbey has either authored
or coauthored at least four analytical reasoning workbooks: Analytical
Reading and Reasoning (2nd ed.; Whimbey, 1990), Mastering Reading
Through Reasoning (Whimbey, 1995), and Analyze. Organize. Write. (rev.
ed.; Whimbey & Jenkins, 1987), as well as Problem Solving and
Comprehension (6th ed.; Whimbey & Lochhead, 1999), the classic book, first
published in 1970, that outlines the TAPPS procedures and provides
hundreds of practice problems. Numerous programs have used Whimbey and
Lochhead (1999) along with Whimbey’s TAPPS program to improve their
students’ analytical reasoning. For example, Xavier University, a historically
black college in New Orleans, teaches a five-week summer program called
SOAR (Stress on Analytical Reasoning). Over its first dozen years, students
enrolled in SOAR made pre- to posttest gains averaging twelve points on the
Preliminary Scholastic Achievement Test (equivalent to gains of 120 points
on the Scholastic Achievement Test) and three grade levels on the Nelson-
Denny Reading Test. Students enrolled in SOAR during those first years
were more than twice as likely to pass science and math in the college
sequence than unenrolled students, and more African American students
were accepted in medical and dental schools from Xavier than from any other
US college (Whimbey et al., 1993/2017). Another example of the success of
TAPPS comes from the chemical engineering department at McMaster
University—a very selective engineering university in Canada—which has a
specific program devoted to teaching problem-solving processes using
TAPPS as a primary teaching method
(http://www.chemeng.mcmaster.ca/mcmaster-problem-solving-mps-
program). Both of these programs, which began before the turn of the
century, continue today. Other research shows that TAPPS improves reading
comprehension, vocabulary acquisition, writing and editing skills, math
performance, direction following, understanding assignments, standardized
achievement test scores, aptitude, and even intelligence quotients (Whimbey
et al., 1993/2017). Reading between the lines, it also holds promise for
setting the stage for students to adduce new repertoires.
Morningside Academy’s principal, Joanne Robbins, modified
Whimbey’s TAPPS program to make it appropriate for younger students at
Morningside learning to think aloud in their problem solving (Robbins,
2011). Robbins’s (2014) teacher-training program, which she has designed,
field-tested, and revised, is called Learn to Reason with TAPS: A Problem
Solving Approach. As Johnson (2015) reported in The Journal of
Organizational Behavior Management, to reduce students’ anxiety about the
procedure, Robbins introduces talk aloud problem solving (TAPS) in a
cooperative learning setting and engages students initially in content-free and
typically enjoyable exercises such as solving logic problems, puzzles, or
brainteasers. The program teaches students how to function in only one of the
problem-solving roles—the problem solver or the active listener—at a time.
The ultimate goal of the program is for each student to adopt the practices of
both a good problem solver and an active listener.
The five qualities of a problem solver include
1. shows a positive attitude,
2. works carefully,
3. breaks the problem into parts,
4. actively works the problem, and
5. never just guesses.

The five qualities of an active listener include

1. follows along,
2. gives encouragement,
3. catches mistakes,
4. leads and hints instead of giving an answer, and
5. reviews for accuracy.
 After roles are assigned, each partner in the team receives a card
displaying their role. Students learn specific dialoguing practices for each
quality. One student—the Problem Solver—performs aloud as the other
partner—the Active Listener—provides reinforcement, corrections, and
shaping. As students master both the speaker and listener roles, they receive a
variety of problems to solve independently, although, even then, they are
required to comment out loud on their own speaking, as a listener would. At
this stage, the student is practicing combining their speaker and listener
repertoires in the same skin into a controlled intraverbal chain. During this
phase of the teaching process, Morningside teachers provide reinforcement,
corrections, and shaping in a manner similar to what they did when students
played only one of the roles for a particular problem. According to Johnson
(2015), “Finally, each student practices covertizing the speaker and listener
routines while maintaining problem solving fluency. Of course, by
encouraging students to covertize the speaker and listener routines, teachers
can only see the outcomes of the private dialogue in a correctly solved
problem” (p. 143). As Morningside Academy has routinized its own TAPS
procedures, students have been providing evidence that they understand its
importance in working through many challenging problems, puzzles, logic
questions, and even everyday problems that arise. For example, we’ve
witnessed students—deciding where to go for a special lunch or what to do
on their break—say, “Let’s TAPS that!”
Morningside instructional designers are in the process of developing
several generative procedures, including routines for questioning teachers
and peers (e.g., Robbins, Layng, & Jackson, 1995), a constructivist approach
called “questioning the author” (Beck, McKeown, Hamilton, & Kucan,
1997), and pragmatic natural philosopher and educator John Dewey’s (1933)
method for reflective thinking, descriptions of which are beyond the confines
of this chapter. Besides strengthening our students’ problem-solving
repertoires, designing generative repertoire procedures reduces the need to
design instructional procedures that directly promote contingency adduction
and provides more opportunities for using commercially available
instructional materials that are not generative in nature.

Frequency Building of Component Skills and

Application to Learning Complex Skills
 The third group of behavior analysis investigators who influenced our
generative instruction model was Eric Haughton’s behavioral fluency group,
which included Beatrice Barrett, Carl Binder, Annie Desjardins, Michael
Maloney, and Clay Starlin, among others. Haughton in particular was
interested in the rather explicit nature of the relation between practice to
fluency and generative responding. The present authors eventually became
part of this group. Before the discoveries of contingency adduction, the
behavioral fluency group discovered a wealth of Standard Celeration Chart
evidence showing that accuracy of a response class is insufficient for smooth
progress through a curriculum sequence. Building component elementary or
“tool” skills such as letter sounds and math facts to certain frequencies made
it easier to develop useful frequencies of complex skills that require the
synthesis of these component skills, such as word and passage reading and
computational arithmetic. Eric Haughton (1972) described many Standard
Celeration Charts that illustrated these relations, discovered while he was
coaching teachers using precision teaching in elementary schools in Canada.
Importantly, he calculated the relation between the frequency of saying
sounds correctly and oral reading frequency across the Standard Celeration
Charts of many teachers representing many students. The correlations ranged
from .6 to 1.0. Notably, the correlation between number writing frequency
and computation averaged 0.9! Low tool skill frequencies clearly predicted
low frequencies and higher errors in more complex reading, math, and
writing performance.
Over time, Haughton and his teachers discovered three outcomes of
effective classroom practice: (1) when tool skill frequencies reach between
100 and 200 per minute, more complex skills will accelerate faster and
require less error correction, and students will progress through curriculum
steps at a faster rate; (2) errors are much easier to correct when overall
performance on a given skill is higher than fifty to sixty per minute; and (3)
students can successfully engage in independent practice without risk of
practicing errors when they can perform at approximately half the
proficiency level for a given skill (Haughton, 1980). A large number of
precision teaching investigations since 1980 have also supported the
conclusion that increased performance frequencies improve application (e.g.,
Evans & Evans, 1985; Johnson & Layng, 1992, 1994; Maloney, Desjardins,
& Broad, 1990; Mercer, Mercer, & Evans, 1982). By application, precision
teaching means “integration of component response classes into composite
response classes” (Binder, 1996, p. 178).12 One controlled study has
examined the relation between the frequency of prerequisite tool skills and
complex skills that incorporate them. Van Houten (1980) measured the
frequency of long multiplication (e.g., 42 x 26) and long division (e.g., 342 ÷
12) problems completed in ten minutes. During some conditions students also
practiced multiplication and division facts during another part of the school
day, and during other conditions they did not. Results showed that students’
frequencies of completing long multiplication and long division problems
increased only in the conditions in which they also practiced multiplication
and division math facts. When math facts practice did not occur, frequencies
of long multiplication and division remained the same. Math fact frequencies
rose from thirty to forty per minute at the beginning of the experiment to
sixty to seventy per minute at the end of the experiment, highly correlated
with increases in long multiplication and division frequencies.13
Since precision teaching investigations and data are focused on the
relation between growth in tool skill frequencies and growth in more
complex skill frequencies, it is not possible to draw conclusions about
whether tool skill frequency building produced adduction of more complex
skills in the absence of instruction. We can say that tool skill frequency
building made it appear much easier to learn subsequent more complex skills
that required the synthesis of these skills. By easier, precision teachers are
talking pragmatics: more complex skills will accelerate faster and require less
error correction, and students will progress through curriculum steps at a
faster rate. Tool skill frequency building enables composite behavior
instruction by ensuring that the frequencies of the components are sufficient
for them to be easily applied in larger composites, after at least some
instruction in the composite. Instruction is presumably still necessary but is
briefer and requires less error correction. From these data we can
conceptualize degrees of generativity in responding and a continuum of
instruction ranging from many demonstrations and opportunities for guided
practice and feedback to a few tips and quips, or no help at all. It appears that
the amount or dose of instruction required for students to build high
frequencies of complex skills with fewer errors was much lower when
students had higher frequencies of their component skills.

Collateral Emotional Effects of the Moment of

Contingency Adduction
 The moment of contingency adduction often produces collateral positive
emotional effects. The student’s adduced performance may be accompanied
by “Aha,” smiling, and other overt responses if the student tacts the
correspondence between their novel behavior and the criteria for
reinforcement in the new situation. Aha tacts usually follow sudden and
dramatic shifts in stimulus control where the student’s performance changes
from unsuccessful candidate responses to adduced responses that match the
teacher’s score sheet. If engineered well, inductive, discovery learning can be
enormously more fun for students and teachers than didactic, expository
learning.
One reason constructivist-style education dominates current educational
practices is likely the aesthetically pleasing outcomes discovery learning in
principal can produce. However, constructivist lessons are often not
successful for all students. Student outcomes often do not meet high
standards, continuing a sorting machine philosophy of education where the
best students rise to the top while others are left behind. Often these problems
are masked by cooperative learning arrangements in which a few students in
a small group do the majority of the learning while the others are left
puzzled. Discovery learning lessons also often proceed at a slow pace of
learning, bogged down with errors and collateral negative emotional effects.
Behavior analysis in education may gain a competitive foothold in current
thinking about education by investigating and proposing methods based on a
thorough analysis of the conditions under which generative responding
through the process of contingency adduction can be arranged.

Contingency Adduction and Behavioral Evolution

The behavioral process of contingency adduction (or comparable accounts
of generative responding) is essential to a plausible account of the origins of
the vast number of behaviors emitted by organisms, especially humans,
during the course of their lives. Behavioral accounts that incorporate only
the basic behavior principles of reinforcement, shaping, and stimulus
control are insufficient, as evidenced by the large majority of skeptics who
find behavioral accounts too elementary and too naive (e.g., Chomsky,
1959). Figure 9.1 depicts a more plausible account. In the course of an
organism’s behavioral evolution, various contingencies operate
simultaneously and continuously on the many behaviors in an organism’s
history. Novel behavior relations are the outcome of these dynamics. In
each new situation or circumstance, contingencies recruit current relevant
component repertoires from an organism’s history. Novel adduced
repertoires are selected from the combinations and recombinations of
component behaviors in an organism’s history to meet the requirements of
those current reinforcement contingencies. These adduced repertoires
become part of the selecting environment for further new repertoires
required in subsequent new circumstances, and so on, thus producing
increasingly complex behavior. When current contingencies are not at
sufficient strength, contingency adduction is often facilitated when the
student engages in generative repertoires, such as TAPS, best guess study
skills, and other analytical routines. Good instructional design assures that
students practice generative repertoires to heighten their occurrence in the
presence of novel contingencies.

Figure 9.1 Behavioral evolution.

The quote below, by Charles Ferster (1965)14, conveys the essence of

contingency adduction as it applies to instruction.
 The essential point is that the speaker (teacher) does not build the
repertoire in the student. He can only prompt, modify, or rearrange
verbal behavior that the listener (student) already has. Thus, the
interaction between speaker and listener (teacher and student) is
most effective when their repertoires are close but not identical. The
listener (student) is well prepared for the speaker’s (teacher’s)
influence when he would have said the same thing given another
ten minutes (italics added).

Implications and Future Directions

Our work with contingency adduction has substantially informed our
generative instruction model of academic instruction and our other clinical
work with children and youth (Johnson & Layng, 1992, 1994; Johnson &
Street, 2004, 2012, 2013; Layng, Twyman, & Stikeleather, 2004a, 2004b;
Street & Johnson, 2014). Generative instruction makes teaching and
learning more efficient. Eliminating the need for explicitly teaching
composite performances reduces time to mastery of a typical curriculum by
one-third or more. Generative instruction also provides explicit practice in
problem solving without instruction as required in the real world. Direct
and explicit teaching alone avoids teaching students how to think and
creates dependency upon teachers and instruction for learning. And if we
teach ourselves as well as constructivists how to engineer discovery
learning effectively, we can heighten the aesthetically reinforcing
effectiveness of discovery learning by making reinforcement denser and
more probable. Then maybe we’ll command a bigger role in American
general education—a role that constructivists now claim and that we once
claimed in the 1960s and 1970s.
In light of these advantages, we recommend that behavioral instructors
revise their typical behavioral instruction protocols such as discrete trial
training, direct instruction, and precision teaching in the following manner:
1. Focus instruction on component behaviors to facilitate untaught
mastery of complex behavior.
Deconstruct the current holistic activities and real-world
composite behaviors identified in today’s academic
curriculum frameworks and standards. Use
 component/composite analysis to break down the composite
behaviors (Binder, 1996; Haughton, 1980; Johnson & Layng,
1992). We have identified over forty foundational
component behaviors in reading, writing, and mathematics
(Johnson & Street, 2013).
Design and present explicit instruction for each component
behavior, using the demonstrate-guide-test model described
by Gilbert’s (1962a, 1962b) Mathetics and Engelmann and
Carnine’s direct instruction (Engelmann & Carnine, 1982;
Marchand-Martella, Slocum, & Martella, 2004) to build
highly accurate performance.
Design and present frequency-building activities for each
component, using Lindsley’s precision teaching (Binder,
1996; Johnson & Street, 2012, 2013; Kubina & Yurich,
2012) to build high frequencies of performance.
2. Build complex repertoires from the component behaviors thus
established by engineering discovery learning.
Sequence instruction so that composite activities follow
component instruction.
After teaching the sequence of component behaviors, design
and present a contingency adduction moment: one or more
composite behavior activities that incorporate the component
behaviors that you have explicitly taught. When presenting
the composite activity, provide a blend of brief context-
setting instruction and connections to natural, real-world
problem solving to assure that the activity functions as a
motivative operation.
Use a delayed prompting procedure to guarantee that the
composite activity will recruit the components and adduce
the complex untaught composite, instead of directly teaching
it.
3. Delayed prompting is essentially an upside-down version of explicit
instruction; begin with a test phase (“Can you do this now?) and
proceed to prompting and guiding, then full-blown demonstrations,
as needed.
 Reinforce the new performance to produce adduced
behavior.
Teach generative repertoires such as TAPS, questioning, and
reflective thinking to facilitate contingency adduction and
discovery learning.
4. Use these procedures for corrective, remedial instruction as well as
instruction for new learning. Trouble in algebra does not mean we
should intensify instruction and practice in algebra.
Component/composite analysis and student performance assessment
will reveal the components of algebra that are absent, error-prone, or
at frequencies too low to be useful in solving algebra problems.
Such nonlinear analyses reveal that the problem that a student
presents—in this case, trouble in algebra—is not the main problem
to solve.

Conclusion
Generative behavior is behavior that occurs in novel forms under novel
circumstances without direct instruction. That behavior analysts can
program for generative responding is important, because educators cannot
possibly teach every academic skill that a student needs to learn. The
purpose of generative instruction is to arrange for the teaching of a number
of component skills so that students will demonstrate new forms of complex
academic skills that combine several component skills. Research on
generativity, contingency adduction, and behavioral fluency have all
contributed to the development of the generative instructional model. Five
variations of contingency adduction have been programmed in classroom
settings and are incorporated into the Headsprout Internet reading program.
Results for the Headsprout Early Reading program have shown that a large
proportion of students identified new sound combinations in the absence of
instruction and formed new consonant blends. Likewise, many students
showed the emergence of new vocabulary skills following their completion
of the Headsprout Reading Comprehension program. Whimbey and
colleagues have provided a framework for analyzing novel behavior in
terms of analytic thinking and problem solving. Importantly, for all forms of
generative repertoires, the speaker or teacher does not build the repertoire
but prompts, modifies, or rearranges verbal behavior that the listener or
student already has in the repertoire.

Study Questions
1. What is generativity? Explain in your own words.
2. What are the practical benefits of generative responding? Why is it
socially important?
3. Describe two areas of research that have influenced the authors’
generative instruction model.
4. Explain “contingency adduction.”
5. What are the three critical features of the concept class contingency
adduction?
6. How does exclusion learning facilitate contingency adduction? Give
an example.
7. Describe two of the examples of contingency adduction reviewed in
the chapter.
8. What is the main idea behind the development of TAPPS/TAPS
methods?
9. What are the five qualities of problem solvers and active listeners as
described in the chapter?
10. What is one of the main conclusions from the precision teaching
research described in this chapter?
11. Describe the implications for teaching provided by the authors at the
end of the chapter.
12. What is a “contingency adduction moment”? Give an example.
13. The authors say that trouble in algebra doesn’t necessarily mean that
more practice in algebra is needed. What point are the authors
making with this example?
 Chapter 10:

Equivalence-Based Instruction: Designing

Instruction Using Stimulus Equivalence

Daniel M. Fienup15
Teachers College, Columbia University
Julia Brodsky
The Graduate Center of the City University of New York

Link and Preview

The discovery of a behavior process known as stimulus
equivalence was important for behavior analysts because it set the
stage for an analysis of conditions under which novel patterns of
responses may emerge in the absence of explicit instruction. The
present chapter introduces the concepts of stimulus equivalence
classes and elaborates upon how the formation of equivalence
classes relates to the concepts of stimulus and response
generalization. The chapter also describes instructional
procedures that have been used in both the laboratory and
educational settings for producing equivalence relations.

The challenge of education is to engineer antecedents and consequences to

produce accurate performance under appropriate stimulus control
conditions. An educator must begin the design of instruction by determining
the stimuli that should evoke the desired academic response(s). At the
simplest level, an educator may wish to establish responding in the presence
of a discriminative stimulus such that the response occurs in the presence
of certain stimuli—such as saying “five” in the presence of the numeral five
(5) but not in the presence of other stimuli, such as other numerals—thereby
demonstrating stimulus control (Herrnstein, 1990). At a higher level, an
educator may want to develop responding to an open-ended category of
stimuli, such as treating the numeral five displayed in different sizes and
fonts, as functionally interchangeable (see top portion of figure 10.1). The
formation of open-ended categories requires stimulus generalization, or
the spread of effect across topographical stimulus dimensions (Herrnstein,
1990; Stokes & Baer, 1977). A more complex type of stimulus control is
involved in concept formation, which entails responding to a range of
nonidentical stimuli that may vary across perceptual modalities—such as
the numeral five (5), the letters f-i-v-e, the spoken word “five,” and five
objects—and treating them as functionally interchangeable (see bottom
portion of figure 10.1). Concept formation is quite a different phenomenon
from stimulus generalization because of the range of stimuli involved. At
the level of concept formation, educators may look to the long history of
basic and applied studies dealing with the formation of classes of physically
disparate stimuli for guidance in teaching such repertoires. This area of
research is known as stimulus equivalence (Sidman, 1994).

Figure 10.1 A depiction of stimulus generalization and stimulus

equivalence.
 Equivalence-based instruction (EBI) is a generative instructional
design tool that incorporates the principles of stimulus equivalence, which
bears resemblance to logical deduction (Fienup, Covey, & Critchfield, 2010).
Typically conducted via match-to-sample (MTS) procedures, EBI teaches
classes (sets) of stimuli containing three or more stimuli each (e.g., A, B, and
C). To reduce the amount of direct teaching while maintaining high levels of
academic output, educators consider how to link a small subset of stimulus-
stimulus relations to expand to a new, larger set of stimulus-stimulus
relations. EBI is concerned with the design of instruction to teach functional
interchangeability between topographically dissimilar, but thematically
similar, stimuli.
The trajectory of stimulus equivalence research is atypical. Many
behavioral technologies are inspired by the findings of basic research. Initial
stimulus equivalence research was conducted in the applied field, where the
phenomenon was discovered and studied by Murray Sidman and his
colleagues (Sidman, 1971; Sidman & Cresson, 1973), who taught reading
comprehension to learners with an intellectual disability (ID) and Down
syndrome. Questions generated in applied settings led researchers to further
study this phenomenon in laboratory settings, where the majority of research
in this area has been conducted. In the last few decades, while basic
researchers have continued conducting research in highly controlled settings,
parallel applied and translational research has studied the application of this
technology to the design of educationally relevant academic instruction
(Rehfeldt, 2011). Areas of instruction have included spelling (De Rose, de
Souza, & Hanna, 1996), matching names and faces (Cowley, Green, &
Braunling-McMorrow, 1992), geography (e.g., Hall, DeBernardis, & Reiss,
2006), and mathematics (Lynch & Cuvo, 1995). These skills have been
taught to participants with a variety of diagnoses, such as Down syndrome
(Sidman & Cresson, 1973), fragile X syndrome (Hall et al., 2006), ID
(Sidman, 1971), autism spectrum disorders (Stanley, Belisle, & Dixon, 2018),
brain injury (Cowley et al., 1992), and visual impairments (Toussaint &
Tiger, 2010).
Lynch and Cuvo (1995) conducted a study that typifies EBI. These
researchers taught fraction-decimal mathematics classes to general education
fifth and sixth graders. The stimuli in the classes included numerical
representations of fractions (A stimuli); pictorial analogues, which were grids
composed of 100 squares with a particular number of squares filled in black
(B stimuli); and numerical representations of decimals (C stimuli). There
were twelve separate fraction-decimal classes included in training. For
example, class one contained the stimuli 1/5 (A), a pictorial analogue with 20
of 100 squares filled in (B), and 0.20 (C). Figure 10.2 displays examples of
taught and derived stimulus-stimulus relations in this study. The researchers
used MTS procedures to teach AB and BC relations and test for the
emergence of derived relations.
During AB training, the researchers presented a sample stimulus and the
participant selected the sample (observing response) to produce four
comparison stimuli. In AB training, fractions (A) served as samples and
pictorial analogues (B) served as comparisons. Of the four comparison
stimuli presented during a trial, one was from the same class of stimuli, and
selecting this stimulus resulted in an audible “yes” generated by the computer
program. Three comparison stimuli were from classes other than the sample,
and selecting those comparison stimuli resulted in no effect—that is, the
stimuli remained until the participant clicked the correct comparison
stimulus. Following mastery of all twelve AB relations, the researchers
applied the same training procedure to BC relations where pictorial
analogues (B) served as sample stimuli and decimals (C) served as
comparison stimuli. Prior to and after the instruction, the researchers tested
stimulus-stimulus relations that were expected to emerge from the
instruction. Two of these relations were symmetry (BA, CB) and tested for
whether sample-comparison conditional discriminations resulted in
bidirectional transfer whereby the participant made accurate responses when
samples became comparisons and comparisons became samples. One derived
relation was transitivity (AC) and represented an association between two
stimuli not directly taught, but related through a mutual stimulus (in this
study, the B stimuli). A final derived relation was equivalence (CA), or
combined symmetry and transitivity, which involved both bidirectional
transfer and a novel association.
Prior to instruction, participants showed low levels of class-consistent
(accurate) responding. After learning the AB and BC conditional
discriminations, participants showed the emergence of equivalence classes by
responding with high accuracy to symmetry, transitivity, and equivalence
probe trials, thus demonstrating the ability to accurately relate fractions and
decimals (see figure 10.2). This is an important educational outcome because
participants treated three topographically different stimuli as interchangeable,
and this socially significant repertoire resulted from teaching just two
overlapping relations. Thus, the teaching strategy appeared to be time
efficient, in that several skills emerged after only a few were taught.

Figure 10.2 Trained relations (top row) and tested derived relations
(bottom row) based on Lynch and Cuvo (1995).

Equivalence Procedures as Instructional Design

Choices
In this section we will review different components of EBI and the
supporting research base. Many instructional design features of the study
conducted by Lynch and Cuvo (1995) are worth noting. The first is that by
teaching two relations (AB, BC), the researchers observed the emergence of
four derived relations (BA, CB, AC, CA). The number of untaught
emergent relations shows the power of stimulus equivalence and why
educators and clinicians may be interested in adopting this technology. For
example, Lynch and Cuvo (1995) observed the emergence of relations
between fractions and decimals that were never directly taught. The effect
of programming for and observing the emergence of derived relations has
been replicated across a large number of different content areas and
populations, including college students learning about single-subject
research designs (Lovett, Rehfeldt, Garcia, & Dunning, 2011) and
neuroanatomy (Fienup, Mylan, Brodsky, & Pytte, 2016), and learners with
disabilities mastering reading and spelling (de Rose et al., 1996) and coin
equivalences (Keintz, Miguel, Kao, & Finn, 2011). Broad reviews of
equivalence applications (Rehfeldt, 2011), and those specifically with
college students (Fienup, Hamelin, Reyes-Giordano, & Falcomata, 2011;
Brodsky & Fienup, 2018) and individuals with autism (McLay, Sutherland,
Church, & Tyler-Merrick, 2013), summarize the variety of concepts
researchers have taught using EBI.
The derived relations reported in EBI studies represent properties of
equivalence classes, and the presence of these properties establish that a class
of stimuli function interchangeably. The properties established in the Lynch
and Cuvo (1995) study were symmetry, transitivity, and equivalence. One
additional property is reflexivity, whereby a stimulus bears a relation to
itself, akin to identity matching (see Sidman, 1994). Lynch and Cuvo (1995)
did not test for the property of reflexivity, nor do most studies involving
typically developing participants, but this property is important to establish
and test in applications with individuals with IDs (e.g., Sidman & Cresson,
1973).

Match-to-Sample Procedures
Lynch and Cuvo (1995) made an instructional design choice by
implementing simultaneous MTS procedures (see top portion of figure 10.3)
to conduct AB and BC conditional discrimination training. In simultaneous
MTS procedures, the sample and comparison stimuli are present at the same
time. Lynch and Cuvo required an observing response to the sample stimulus
to ensure a participant was attending to the sample, and then the comparison
stimuli appeared while the sample remained on the screen. A variation of this
procedure is delayed MTS (see bottom portion of figure 10.3), in which the
sample disappears once clicked, and the comparison stimuli appear
immediately (zero-second delay) or after some time (Arntzen, 2006). Indeed,
most of the EBI applications we are aware of utilize a simultaneous MTS
procedure. This is interesting because basic research has shown that delayed
MTS results in higher levels of equivalence class formation. Arntzen (2006)
conducted a series of experiments with adult learners comparing
simultaneous MTS and delayed MTS arrays with delays between zero and
nine seconds. Results showed that longer delays between the offset of the
sample stimulus and the onset of comparison stimuli resulted in higher
numbers of participants demonstrating equivalence classes. To the best of our
knowledge, this phenomenon has not explicitly been studied in the context of
EBI and teaching academically relevant equivalence classes, although there
are examples of delayed MTS procedures used in EBI research (e.g., Lane &
Critchfield, 1998; Stromer & Mackay, 1992).

Figure 10.3 Simultaneous MTS (top row) and delayed MTS (bottom
row).

Training Structure
Lynch and Cuvo (1995) made an instructional design choice to arrange
the training structure of the fraction-decimal classes as a linear series (LS),
whereby A sample stimuli were conditionally related to B comparison
stimuli, and B sample stimuli were conditionally related to C comparison
stimuli (see left portion of figure 10.4). The B stimulus was the node, or
stimulus that connected the other two, which served as both a comparison
stimulus and sample stimulus in different training phases. Basic researchers
have rigorously tested other training structures. The other training structures
are colloquially described as one-to-many (OTM; middle portion of figure
10.4) and many-to-one (MTO; right portion of figure 10.4), or technically
named sample-as-node and comparison-as-node, respectively.
In an OTM training structure, given stimuli ABC, an instructor teaches
AB and AC relations and tests for the emergence of (1) BA and CA
symmetry relations and (2) BC and CB equivalence relations. In this case, A
stimuli serve as samples on all training trials and A stimuli are the node,
hence the technical name sample-as-node. In an MTO training structure, an
instructor teaches BA and CA relations and tests for the emergence of (1) AB
and AC symmetry relations and (2) BC and CB equivalence relations. The
naming of a particular relation depends upon the training structure used.
You may notice that the BC relations mentioned here represent a trained
relation in the LS example and an equivalence relation in OTM and MTO
examples. The way to distinguish between these relations is to first note the
trained relations (see black lines in figure 10.4) and then consider whether a
derived relation (red lines) entails (1) bidirectionality (symmetry), (2)
associations between novel stimuli without bidirectionality (transitivity), or
(3) associations between novel stimuli with bidirectionality (equivalence).
Researchers have directly compared training structures and their effect on the
probability of forming equivalence classes. With typically developing adults,
the OTM training structure produces the highest levels of derived relations,
followed by the MTO training structure (Arntzen, Grondahl, & Eilifsen,
2010; Arntzen & Holth, 1997, 2000). With children and individuals with IDs,
the MTO training structure is the most successful arrangement, followed by
OTM (Saunders, Drake, & Spradlin, 1999; Saunders, Wachter, & Spradlin,
1988). In basic research studies that directly compared training structures, the
LS structure has reliably produced the lowest level of class-consistent
responding; however, several applied EBI studies have used the LS training
structure with success (e.g., Lynch & Cuvo, 1995; Fields et al., 2009).
Figure 10.4 The linear series, one-to-many, and many-to-one training
structures.

Training Protocol
Another instructional design choice by Lynch and Cuvo (1995) was how
to order training and testing phases, referred to as training protocol (see
figure 10.5). Lynch and Cuvo programmed all training phases (AB and BC)
consecutively and prior to conducting tests of derived relations. During
testing, probes for symmetry and transitivity were conducted first, followed
by equivalence probes. The training protocol used in Lynch and Cuvo mimics
a simultaneous (SIM) protocol (see top portion of figure 10.5). In this
protocol, all training phases are grouped and all testing phases are grouped,
with no overlap between (Fields, Reeve, Adams, Brown, & Verhave, 1997).
Another training protocol is the simple-to-complex (STC) protocol (see
middle portion of figure 10.5), which intersperses training phases and derived
relations probes. Across the whole program, derived relations are tested
individually, and tests are arranged from the simplest type of derived
relations (i.e., symmetry) to the most difficult (i.e., equivalence; Adams,
Fields, & Verhave, 1993). Lynch and Cuvo’s (1995) protocol would have
looked a bit different if conducted as an STC protocol. The authors taught
AB, then BC, and then tested for BA, CB, AC, and BC, respectively. In an
STC protocol, the test for symmetry of AB would have directly followed AB
training. With three-member classes, performance differences produced by
the SIM and STC protocols is minimal, but the difference becomes more
obvious as class sizes increase (see Fienup, Wright, & Fields, 2015).
The SIM and STC protocols are the most commonly used in basic and
applied studies, although other protocols have been discussed. The complex-
to-simple (CTS) protocol completes training phases consecutively and then
tests for the emergence of derived relations individually, starting with the
most complex relations (for a description see Adams et al., 1993; for an
applied example see LeBlanc, Miguel, Cummings, Goldsmith, & Carr, 2003).
The hybrid protocol (Imam & Warner, 2014) includes training in
consecutive phases, and tests are ordered from simple to complex (see
bottom portion of figure 10.5). In other words, the hybrid protocol includes
SIM training and STC testing. The first work on training protocols was
conducted by Adams et al. (1993) and showed that when instructors taught
relations between arbitrary stimuli (e.g., nonsense syllables), the STC
protocol produced a very high number of participants who demonstrated
equivalence classes. This outcome was considerably better than previous
studies utilizing SIM and CTS protocols. Fienup et al. (2015) directly
compared the SIM and STC protocols while teaching college students
neuroanatomy classes. The researchers found that when teaching three-
member (ABC) and four-member (ABCD) classes, participants who
completed the STC protocol always formed classes during the first
administration of the final test of equivalence. However, fewer participants
who completed the SIM protocol formed three-member classes (75%), and
this number dropped further with four-member classes (42%). Additional
research has demonstrated that the breakdown of testing is potentially more
important than whether training and testing are interspersed. Imam and
Warner (2014) compared the SIM protocol to the hybrid protocol and found
that ordering test phases from simple to complex dramatically influenced
whether an individual formed equivalence classes. This procedure mimics the
testing order used by Lynch and Cuvo (1995) and may explain why Lynch
and Cuvo found high performance during tests of equivalence class
formation.
 Figure 10.5 The SIM, STC, and hybrid training protocols. Dashed
boxes represent testing probes; solid boxes represent training
phases.

Generalization of Class-Consistent Responding

Lynch and Cuvo (1995) conducted generalization trials and paper-based
generalization tests using novel fraction-decimal classes. Frequently, EBI
researchers test for generalization across stimuli, responses, and time. For
example, Trucil, Vladescu, Reeve, DeBar, and Schnell (2015) taught portion
size estimation using EBI, and tested for generalization by asking participants
to estimate portions of novel foods, demonstrating stimulus generalization.
Such tests underscore the educationally relevant outcomes produced by EBI.
Reyes-Giordano and Fienup (2015) conducted MTS EBI and tested derived
relations in computerized MTS, multiple-choice, and fill in the blank formats.
The outcomes suggested response generalization. Albright, Reeve, Reeve,
and Kisamore (2015) taught participants statistics concepts and demonstrated
maintenance one week after equivalence training. Fienup et al. (2016)
demonstrated that learning neuroanatomy classes via computerized EBI
resulted in higher course exam performance a few weeks later as compared to
no equivalence training, demonstrating a combination of stimulus and
response generalization and maintenance.
Successful performance on generalization measures demonstrates that
instructors may choose to use EBI teaching procedures even when the target
response topography may differ from the training topography. This is
educationally important because many curricular goals include writing and
talking about stimuli. However, teaching in this modality may be difficult,
and the results of the EBI literature suggest the MTS training often results in
improved writing and discussion of educational stimuli (e.g., Walker,
Rehfeldt, & Ninness, 2010). Additionally, the research has shown that when
maintenance tests are administered, participants continue to respond in a
class-consistent manner in both the experimental settings (Albright, Schnell,
Reeve, & Sidener, 2016; Trucil et al., 2015) and the naturalistic educational
settings (e.g., Fienup et al., 2016); however, some booster training may be
required to maintain responding (Walker & Rehfeldt, 2012). Such
generalization outcomes may be beneficial in clinical work, where a therapist
can train concepts such as classes of animals using EBI in a client’s home,
but the learner may generalize to writing about classes of animals at school or
speaking about classes of animals with peers on the playground.

EBI Applications
Researchers have demonstrated the utility of EBI in teaching a wide variety
of socially significant behaviors across a number of individuals of varying
skill levels. Sidman (1971) first demonstrated equivalence relations in a
learner with ID. The learner had a preexisting repertoire of matching
spoken words to pictures. After Sidman taught the learner to match spoken
and printed words, the learner was then able to match words to their
corresponding pictures, thereby demonstrating reading comprehension. In a
follow-up study, Sidman and Cresson (1973) taught two young adults with
Down syndrome reading comprehension, but unlike in the 1971 study, the
participants did not have a preexisting spoken word-picture repertoire. The
researchers taught the participants to choose pictures and written words
when presented with dictated words and observed the emergence of reading
comprehension—matching pictures and written words. This study
replicated and extended the original 1971 study and experimentally
demonstrated the emergence of transitive relations and equivalence classes
contingent upon teaching specific, linked conditional discriminations.
Much of the applied literature represents a replication and extension of
the Sidman and Cresson (1973) procedures to teach different repertoires. For
example, Keintz et al. (2011) taught coin equivalences to children with
autism, which is an important step in teaching functional use of money. Coin
names served as A stimuli, physical coins served as B stimuli, printed prices
served as C stimuli, and dictated prices served as D stimuli. Following
training, participants were able to engage in a number of derived skills, such
as matching coin names with prices. Such a procedure could also be used to
teach equivalences with paper-based currency in addition to coin-based
currency in clinical practice.
A number of studies have involved utilizing EBI with children and
adults with disabilities. LeBlanc et al. (2003) taught classes that involved US
state names (A), pictures of states (B), and state capital names (C) to children
with autism. Hall et al. (2006) replicated and extended the work of LeBlanc
et al. (2003) and Lynch and Cuvo (1995) by teaching geography and
mathematics skills to children with fragile X syndrome. Stanley et al. (2018)
taught adolescents with autism a number of academic content areas,
including chemical compounds (technical names, compound symbols,
common names), units of measurement (yards, feet, inches), and historical
figures (names, pictures, facts, country of origin). EBI procedures have also
been used to help adults with brain injury match names and faces (Cowley et
al., 1992) and recognize emotions (Guercio, Podolska-Schroeder, & Rehfeldt,
2004), as well as teach Braille to children with degenerative visual
impairments (Toussaint & Tiger, 2010). Using EBI to address these socially
significant repertoires benefits both clients and clinicians by saving valuable
instructional time, allowing the clinician to progress to teaching other skills
with the saved time.
Several studies have focused on language and language arts skills,
including the initial studies conducted by Sidman (1971) and Sidman and
Cresson (1973) of reading competencies. Additionally, researchers have
taught competencies such as foreign languages, spelling, and prereading
skills. Equivalence researchers have trained other language arts skills in
similar ways. For example, Connell and Witt (2004) trained kindergarteners
to match uppercase and lowercase letters to the spoken name of the letter and
its sound, de Rose et al. (1996) taught reading and spelling skills to first
grade children, and Lane and Critchfield (1998) taught letter sounds to
individuals with intellectual disabilities.
Much of the basic research related to stimulus equivalence has focused
on demonstrating properties of classes or manipulating a parameter of
training to observe how it affects the formation of classes. One important
feature of equivalence classes is the transfer of stimulus functions across
stimuli related in an equivalence class (for a review, see Dymond & Rehfeldt,
2000). Researchers have found that once a set of stimuli function
interchangeably (as an equivalence class), functions associated with one class
member transfer to other class members. For example, Hayes, Kohlenberg,
and Hayes (1991) conditioned stimuli to serve reinforcing and punishing
consequences by pairing them with verbal feedback. Then, they included the
stimuli in a training procedure to form equivalence classes and taught a
conditioned reinforcer in one class and a conditioned punisher in another.
After equivalence classes were formed, the researchers found that stimuli in a
class with a conditioned reinforcer took on reinforcing functions and stimuli
in a class with a conditioned punisher took on punishing functions.
Unfortunately, there have not been enough applications that explore the use
of transfer of function to solve socially important skill acquisition deficits.
An application of these findings in clinical work might involve establishing
classes of “safe” (e.g., toys, books) and “dangerous” (e.g., knives, matches,
guns) stimuli and corresponding approach and avoidance responses to a
stimulus from each class, which should then transfer to other members of the
safe and dangerous classes.
Two important applications of transfer of function are reported in the
literature. One exemplar was conducted by Taylor and O’Reilly (2000). In
this study, the researchers taught individuals with disabilities to engage in a
behavior chain at a grocery store (e.g., get basket, collect food items, and
check out). After this, the participants engaged in an MTS task that matched
various grocery store stimuli—the written and spoken word
“SUPERMARKET,” pictures from the inside of a familiar supermarket, and
pictures from the inside of a novel supermarket. In effect, the researchers
developed a grocery store class using an MTS procedure after establishing a
discriminative function (i.e., how to shop) to some members of the class
(e.g., images of the inside of the trained/familiar supermarket). When the
participants were then taken to a novel grocery store, they were able to
engage in the behavior chain with high levels of accuracy similar to
individuals taught using multiple exemplar training (i.e., taught the behavior
chain in multiple grocery stores) and with higher levels of accuracy than
participants who completed single-exemplar training (i.e., taught the behavior
chain in one grocery store). Training in one supermarket and teaching
supplementary classes may have saved instructional time and resources. Such
a procedure could be expanded to teach other activities of daily living to
clients, such as borrowing books from the library, calling others on a phone,
ordering at a restaurant, and using public restrooms. In another example of
transfer of function, Rosales and Rehfeldt (2007; also see Rehfeldt & Root,
2005) began by teaching participants to mand for preferred items using
pictures of the desired item. The researchers taught participants to mand by
engaging the participant in a behavior chain that led to a reinforcing outcome
and withholding a step in that chain to create an opportunity to mand. For
example, one behavior chain included listening to music. The participant had
to mand for headphones by handing a picture of headphones to the researcher
in order for the chain to continue and result in the opportunity to listen to
music. After manding was established in the participant’s repertoire, the
researchers used EBI to teach the participant three-member equivalence
classes that included pictures, spoken names, and written words of those
same desired items. Next, during the behavior chain, the researchers omitted
the pictures but made the written words available to participants. The
researchers observed that participants were readily able to mand using both
vocal and written words. In other words, once Rosales and Rehfeldt (2007)
established pictures as discriminative stimuli for manding, the function
transferred to other members of the class (the corresponding text).
Much of the recent applied EBI research has focused on college-level
content. This phenomenon has been studied in a number of settings, such as
the laboratory (e.g., Albright et al., 2015), as part of a course requirement
(Critchfield, 2014), as a method of lecture (Pytte & Fienup, 2012), and as an
online supplement to in-class instruction (Walker & Rehfeldt, 2012). Several
studies have shown that EBI is more effective at concept formation than no
instruction. For example, Fields et al. (2009) taught college students four
four-member statistics classes using an LS training structure and an STC
training protocol. Each class contained a graph displaying an interaction (A),
a textual example of an interaction (B), a name of a statistical interaction (C;
e.g., crossover interaction), and a definition of that interaction (D). The
researchers compared the effects of EBI to a no-instruction control group that
completed pre and post assessments with a break in between. Fields et al.
(2009) found that participants had comparable levels of performance at
pretest, but the EBI group scored significantly higher on the posttest and
demonstrated the formation of statistical interaction equivalence classes.
EBI has been shown to be more effective than instructional procedures
typically used in the classroom, such as reading a textbook (O’Neill,
Rehfeldt, Ninness, Munoz, & Mellor, 2015) and video-taped lecture (Lovett
et al., 2011). It is also more efficient than directly teaching all stimulus-
stimulus relations in a class (Fienup & Critchfield, 2011). Researchers have
had success implementing EBI through various delivery methods. While
many researchers deliver this instruction via automated computer tutorials
(e.g., Albright et al., 2015), some have had success using paper-and-pencil
formats (Walker et al., 2010) or distance education formats, such as
Blackboard (Walker & Rehfeldt, 2012) and Adobe Connect (Sella, Ribeiro,
& White, 2014), which demonstrates the potential of EBI technology,
irrespective of instructional setting or format. Participants have demonstrated
equivalence class formation in a variety of response topographies, including
computer-based MTS (e.g., Fienup & Critchfield, 2011), paper-based MTS
(Fields et al., 2009), clicker-based responding (Varelas & Fields, 2017),
selection-based intraverbals (e.g., O’Neill et al., 2015), vocal intraverbals
(e.g., Walker et al., 2010), and estimating portions of food (Hausman,
Borrero, Fisher, & Kahng, 2014). EBI researchers have taught college
students a wide range of socially significant topics, such as neuroanatomy
(e.g., Reyes-Giordano & Fienup, 2015), research design and analysis (Fields
et al., 2009; Fienup & Critchfield, 2011; Sandoz & Hebert, 2016; Sella et al.,
2014), functions of behavior (Albright et al., 2016), portion size estimation
(e.g., Trucil et al., 2015), disability categorization (Walker et al., 2010), and
advanced algebra and trigonometry (Ninness et al., 2006, 2009).

Implications and Future Directions

There is robust experimental evidence for the effectiveness of EBI (Brodsky
& Fienup, 2018; Fienup et al., 2011; Rehfeldt, 2011). However, this claim
comes with a caveat that should shape future research and application: the
vast majority of experimental evidence comes from teaching individuals a
small set of discrete skills in highly controlled settings. Instruction ranges
from teaching children with disabilities three three-member geography
classes (LeBlanc et al., 2003), to teaching typically developing children
twelve three-member fraction-decimal classes (Lynch & Cuvo, 1995), to
teaching college students sixteen three-member neuroanatomy classes
(Fienup et al., 2016). These skill sets were taught in settings analogous to
controlled laboratory settings, such as a quiet area of a classroom
(participant 1; LeBlanc et al., 2003), separate classrooms (participant 2;
LeBlanc et al. 2003; Lynch & Cuvo, 1995), and traditional university
laboratory space (Fienup et al., 2016). Collectively, the research suggests
that EBI is promising for educational applications, but research is needed
that is conducted in traditional educational settings and scales this type of
instructional design to address the many educational needs of individuals
with and without disabilities.
Many educational content areas have yet to be addressed by EBI
researchers. Most individuals are in school from ages five to eighteen and
increasingly contact higher educational schooling for at least an additional
four years. The sum of the EBI research base touches upon a few academic
content areas students may come across during this long time of formal
education. Two recent studies demonstrate progress toward the goal of
scaling EBI. Greville, Dymond, and Newton (2016) trained four five-member
(name, image, function, pathology, anatomy) and eight four-member (name,
image, function, pathology) neuroanatomy classes to students in their first
and second year of medical school. Fienup et al. (2016) taught sixteen four-
member neuroanatomy classes that covered one-third of the neuroanatomy
content covered in an undergraduate behavioral neuroscience course. These
studies represent movement toward broader curricular applications of EBI.
We hope to see researchers designing EBI applications that span content
across entire courses.
Another glimmer of hope in scaling EBI comes from the publication of
Dixon’s (2015) PEAK relational training system: Equivalence module
(PEAK-E; see chapter 11). The PEAK-E module was designed for children
with disabilities and contains 184 different educational goals, spanning
repertoires such as demonstrating reflexivity with objects, pictures, and
stimuli from different sensory modalities (e.g., gustatory); demonstrating
symmetry with pictures and objects or written words and objects; the
formation of planetary equivalence classes (with pictures, spoken names, and
written names of planets); reading and mathematics classes; and even
perspective taking. For a particular skill, the curriculum specifies the training
procedures and specific derived relations to be evaluated. This curriculum
emphasizes tests of reflexivity, symmetry, transitivity, and equivalence
toward the development of higher-order concept learning—something that is
omitted from common curricula for children with disabilities.
There is a sizable evidence base demonstrating the efficacy of
equivalence work with typically developing college students (thirty-one
experiments across twenty-eight papers; Brodsky & Fienup, 2018) and a
number of studies evaluating EBI in individuals with developmental
disabilities (McLay et al., 2013). Very few studies address EBI as an
instructional tool for typically developing children. This problem is related to
the scaling issue mentioned above in that many general education topics have
not been examined by EBI researchers because the participants included in
studies determine the educational content of the studies. Thus, many of the
goals listed in national school curricula such as the Common Core (National
Governors Association Center for Best Practices & Council of Chief State
School Officers, 2010), which is used by forty-two states in the US, have not
been evaluated for whether or not they are amenable to EBI. The work
conducted by Albright et al. (2015), Lynch and Cuvo (1995), and Ninness et
al. (2006, 2009) dealing with mathematics EBI applications strongly suggests
this technology could be applied to a wide array of Common Core
mathematics competencies. For example, a first grade goal is “tell and write
time in hours and half-hours using analog and digital clocks” (see 1.MD16).
In an EBI application, classes could include spoken times (A), written times
(B), time displayed on an analog clock (C), and time displayed on a digital
clock (D). This set of stimuli is amenable to an OTM training structure where
spoken times serve as sample stimuli. One may find that certain relations,
like matching spoken and written times, are already in the participant’s
repertoire and, thus, training could focus on teaching AC and AD relations
and testing for the emergence of the remaining relations. Lynch and Cuvo’s
(1995) procedures could be used to address the third grade goal to
“understand two fractions as equivalent (equal) if they are the same size, or
the same point on a number line” (see 3.NF). This training could serve as a
prerequisite for addressing the fifth grade goal to “use equivalent fractions as
a strategy to add and subtract fractions” (5.NF). As a last example, Common
Core specifies a goal to “classify two-dimensional figures into categories
based on their properties” and provides the following subgoal: “Understand
that attributes belonging to a category of two-dimensional figures also belong
to all subcategories of that category. For example, all rectangles have four
right angles and squares are rectangles, so all squares have four right
angles” (5.G). Indeed, the educational goals proposed by the Common Core
curriculum could benefit from EBI.
Additional research is needed to identify training and testing parameters
that promote equivalence class formation toward building robust EBI
applications. Currently, a number of instructional design components have
empirical support. With adult learners, data exists suggesting that participants
should respond with mouse clicks (Kato, de Rose, & Faleiros, 2008) and
undergo OTM training structures (Arntzen & Holth, 1997) within the context
of the STC training protocol (Fienup et al., 2015), using delayed MTS
training procedures (Arntzen, 2006) and a high mastery criterion (Fienup &
Brodsky, 2017). With young learners and individuals with disabilities, the
only research comparing different EBI components suggests the MTO
training structure is best (Saunders et al., 1988, 1999). Clearly, many
components of EBI have not been tested, especially with young learners and
individuals with disabilities (McLay et al., 2013), which are frequently
reported populations in behavior analytic research. Additionally, an astute
reader will note that applied researchers have used some components of EBI
with success, despite negative findings for that component in the basic
research. For example, the LS training structure is reliably found to be
simultaneously ineffective in basic research (e.g., Arntzen & Holth, 1997)
and effective in applied work (e.g., Fields et al., 2009). What is not
communicated in the research is the number of pilot studies conducted by
researchers, including those conducted in our own lab, to optimize a specific
EBI tutorial. This process may hamper adoption by educators, and thus, we
encourage researchers to investigate optimal instructional design components
of EBI. Specific guidelines for designing EBI applications with different
populations would promote mainstream use of EBI.
Beyond the design of EBI components lies a challenge in disseminating
actual programs. Many EBI applications are computer based and require the
developer to have skills in software design. This may prevent educators from
adopting EBI. One solution lies in the development of commercially
available EBI applications. A platform where instructors could specify
instructional content and disseminate the program to students and clinicians
for use on smartphones and tablets would allow for considerable flexibility
for learners. Another avenue researchers can pursue is evaluating low-tech
procedures or procedures within already-existing educational systems. In
terms of low-tech applications, Walker et al. (2010) used a paper-and-pencil-
based procedure that entailed completing multiple-choice worksheets for
training. Many EBI applications with individuals with disabilities entail MTS
discrete trials similar to educational strategies already in place for the
individual. In this context EBI adds the programming for and testing of
derived relations. Regarding already existing educational systems, Walker
and Rehfeldt (2012) demonstrated that the popular course-management
system Blackboard could be used to administer quizzes for EBI of single-
subject research design. Accompanying task analyses that specify how to
create individual quizzes and quiz contingencies for systems like Blackboard
and Quizlet (see https://www.quizlet.com) could help speed the
dissemination of EBI technology.

Conclusion
This chapter provided an overview of processes involved in stimulus
equivalence and a review of instructional research that utilizes stimulus
equivalence (EBI). In addition, a number of areas in need of additional
research were highlighted. Collectively, EBI research findings show that it
is an effective technology for teaching a number of socially significant skill
sets to learners on a spectrum of age, skill level, and disability status
(Brodsky & Fienup, 2018; Fienup et al., 2011; Rehfeldt, 2011). As research
on effective EBI design choices grows, instructors will be able to optimize
their equivalence-based teaching procedures. This knowledge base,
combined with evidence of scaling and greater accessibility of the
technology, has the potential to bring behavior analytic instructional design
one step closer to mainstream adoption.

Study Questions
1. How is stimulus equivalence different from generalization?
2. Suppose you are teaching a child letter-name-sound classes. Letters
serve as the A stimuli, names of letters serve as the B stimuli, and
 letter sounds serve as the C stimuli. You teach AB and AC relations
(one-to-many training structure). In this scenario, which relations
represent symmetry and which relations represent equivalence?
3. If you were teaching a child math quantities (A: spoken name of
number, B: numeral, C: quantities of dots), how could you mimic
the training procedures of Lynch and Cuvo (1995) to teach math
quantity classes? What relations would you directly teach and what
specific relations would you expect to emerge as a function of that
training? (See figure 10.2 for assistance.)
4. What is training structure? How would you differentiate instruction
based on this variable and the clientele you work with?
5. What is a training protocol? What evidence suggests one should use
one training protocol over another?
6. How can you incorporate tests for generalization within an
equivalence-based instruction protocol?
7. What is transfer of function? Describe one application that has
incorporated transfer of function.
8. How have educators applied equivalence-based instruction to teach
college-level content?
9. How could the content of this chapter be applied to teaching classes
of stimuli that are not related by sameness (e.g., frames of
opposition or comparative frames)?
 Chapter 11:

Relational Frame Theory: Basic Relational

Operants

Mark R. Dixon
Southern Illinois University
Caleb R. Stanley
Southern Illinois University

Link and Preview

Chapter 10 provided an overview of the phenomenon known as
stimulus equivalence. Chapter 11 builds upon that chapter by
delineating arbitrary relations other than stimulus equivalence, all
of which play an important role in language and cognition. The
concepts of relational framing and derived relational responding
are introduced in this chapter, as well as the relevance of both
concepts for the understanding of human language. The chapter
surveys the types of additional relational frame families, including
research in support of each type and educational implications of
building repertoires of each.

On the Origin of Relational Frames

An emergent, or derived, relation is any relation between two stimuli that
develops without a direct history of reinforcement. Derived relational
responding has appeared in the behavior analytic literature for over forty
years, with the first demonstration of an emergent relation being traced back
to the work of Murray Sidman. In an attempt to teach reading skills to a boy
with intellectual disabilities, Sidman (1971) found that when he taught a
specific set of relations, new relations would emerge in the absence of
direct training. For example, he taught the boy that spoken words (A) were
the same as pictures (B) (A-B), and the pictures (B) were the same as the
printed words (C) (B-C). Following this training, the boy not only was able
to respond to the directly trained relations (i.e., A-B and B-C), but also
responded correctly to relations that had not been specifically taught. For
instance, he responded correctly by saying the spoken word (A) in the
presence of printed words (C) (C-A) and vice versa (see chapter 10). While
it is not surprising that the participant made this response, as most
individuals would with sufficient training, the most surprising finding was
that he made this response in the absence of reinforcement for responding
in accordance with that specific relation. Novel responding in the absence
of reinforcement poses an interesting question for behavioral science.
Sidman theorized the boy responded correctly to these untrained relations
because the stimuli had become equivalent, and he coined this phenomenon
stimulus equivalence. As a result, many behavioral scientists began
researching this phenomenon in an attempt to explain it. Over the years,
numerous conceptual and theoretical debates ensued, each offering an
account that attempts to explain the phenomenon. Relational frame theory
(RFT; Hayes, Barnes-Holmes, & Roche, 2001) is one such attempt.
Since stimulus equivalence was first popularized, the connection
between derived relational responding and language has been obvious. One
of the first studies to evaluate this connection was conducted by Devany,
Hayes, and Nelson (1986), who compared the performance of three groups of
children on tests of stimulus equivalence. Groups consisted of typically
developing children with full language, children with an intellectual
disability who demonstrated expressive language, and children with an
intellectual disability who had more severe language deficits. The researchers
exposed the children to a series of trainings in the form of conditional
discriminations, then tested them on their ability to form equivalence
relations. The findings showed that all children who formed equivalence
classes also demonstrated language, whereas those who did not form
equivalence classes were without language. Additionally, since all children
were exposed to the same trainings and reached the same mastery criteria
prior to the equivalence tests, failure to form equivalence was not due to the
inability to learn the conditional discriminations. Instead, the evidence
pointed to a relationship between language and stimulus equivalence.
Most early literature on derived relational responding consisted of
establishing relations among stimuli based on their physical features, referred
to as nonarbitrary relational responding. Lowenkron (1989) provided one
of the first demonstrations that this type of relational responding could be
brought under contextual control. In this study, Lowenkron required
participants to select between sets of stimuli dependent on background color.
For example, selecting the comparison stimulus that was the same as the
sample in the presence of a blue background would result in reinforcement,
and selecting a comparison stimulus that was larger or longer than the sample
in the presence of a green background would also result in reinforcement.
The background color eventually came to exert contextual control over the
participants’ responses, even when the participants were presented with novel
stimuli.
Steele and Hayes (1991) expanded on the findings of Lowenkron’s
(1989) study by examining if relational responding to arbitrary stimuli could
be brought under contextual control. In their study, the authors exposed
participants to a series of trainings in which participants were taught relations
between arbitrary stimuli using contextual cues of same, opposite, and
different. The results showed that the relational responses were brought to
bear on the stimuli due to the contextual cues. This suggested that humans
not only demonstrate relational responding to stimuli based on their physical
properties, but also demonstrate relational responding to arbitrary stimuli
based on the presence of a contextual cue. This had implications for language
development since this pattern of responding may explain the process by
which words acquire their meaning—through their relation with other stimuli
in the presence of a contextual cue. For example, imagine that a young child
with limited verbal capabilities is taught to select a picture of a fox from an
array when asked “Show me the same as fox.” In this example, the word
“same” serves as a contextual cue that specifies the relational response that
would be reinforced in the presence of the word “fox”—selecting the picture
of the fox. When this relational response is made, the functions of the stimuli
are altered depending on the contextual cue. In this example, the word “fox”
acquires the same stimulus functions as the picture of the fox, and the child is
likely to respond to the word “fox” as though it is the same as the picture,
even when this visual stimulus is absent. In this way, the word “fox”
essentially acquires the meaning of the picture and vice versa, which
provides an explanation for the process in which words can acquire meaning.
Further support for the relationship between language and relational
stimulus equivalence was found when experimenters followed a sixteen-
month-old infant over the course of an eight-month period. In this study,
Lipkens, Hayes, and Hayes (1993) documented the development of language
as well as the infant’s performance on tests of stimulus equivalence and
exclusion when presented a series of relational tests. The data showed that
the child was able to respond correctly to both tests of equivalence and
exclusion following the various trainings of the study, despite the limited
verbal repertoire of the child at the time. These findings yielded two
significant implications. First, because the child had little to no verbal
repertoire, it suggested that derived relational responding was not dependent
on language mediation, and therefore, language was not necessary for
derived relational responding to occur. Instead, although derived relational
responding is not dependent on language, language may be dependent, at
least in part, on derived relational responding. The second major implication
was that the child responded correctly to the tests of exclusion even though
no training had been conducted to that specific test. This could not be readily
accounted for by the stimulus equivalence model and suggested other
explanations may be necessary to account for this pattern of responding.
The prevailing notion at the time the studies described above were
conducted was that language gave rise to derived relational responding, and
derived relational responding was dependent on language. Relational frame
theorists believed that language was dependent on and resulted from a form
of relational responding called arbitrarily applicable relational responding
(AARR). Steele and Hayes (1991) provided this evidence in their study when
they demonstrated that AARR could be brought under contextual control.
These studies served to provide the first empirical support for the core
assumptions behind RFT, and, while the origins can be traced back to
stimulus equivalence, not everything is equal.

Fundamental Concepts in Relational Frame

Theory
RFT holds that emergent relations result from the behavior of relating.
Relating means responding to one event in terms of another event (Hayes,
Barnes-Holmes & Roche, 2001). For instance, if an individual was taught to
select a black square in the presence of an identical black square and a
triangle when presented with the statement “Which is the same?” they are
relating, or making a relational response. In this example, the individual was
responding to a stimulus in terms of its physical or formal similarities (i.e.,
the shape of the stimulus) to another stimulus. This type of relational
responding is referred to as nonarbitrary relating. In nonarbitrary relational
responding, stimuli are related in terms of their physical properties, such as
color, shape, size, weight, texture, smell, and so on. An example of
nonarbitrary relational responding involving size would be if a clinician is
teaching a child the concept of bigger and smaller, whereby the child is
taught to select the largest block from an array of three blocks of varying size
when presented with the statement “Which is the biggest?” In this example,
the child is responding to each of the blocks in terms of their size to one
another. An interesting feature of nonarbitrary relational responding is that is
it not limited to humans, as many other species have been trained to respond
relationally to the physical properties of stimuli. For instance, nonarbitrary
relational responding has been observed in birds (Towe, 1954), fish (Perkins,
1931), and mammals (Harmon, Strong, & Pasnak, 1982).
RFT posits that relational responding is much more than responding to
the physical or formal characteristics of a stimulus. Rather, relational
responding can come under the control of cues that can be modified based on
social convention, known as arbitrarily applicable relational responding
(AARR). In contrast to nonarbitrary relational responding, AARR is not
dependent on the physical or formal properties of a stimulus to make a
relational response, but instead relies on a contextual cue that specifies the
relation between the stimuli. For example, imagine a child is being taught
differences in coin values. During training, the child is verbally told that a
dime (A) is more than a nickel (B) (A-B), and a nickel (B) is more than a
penny (C) (B-C). Without having ever been directly taught the relation
between the dime (A) and penny (C), the child will derive that the penny is
less than the dime (C-A) and the dime is more than the penny (A-C). Without
ever having seen the stimuli, the child can make this response because the
contextual cue (i.e., more than) is controlling this pattern of responding. In
fact, the nonarbitrary stimulus properties seem to compete with this pattern of
responding, as the physical size of the dime is less than that of the penny. In
this example, the individual is responding to the stimuli based solely on the
contextual cue.
 The above example highlights the notion that in the presence of a
contextual cue, individuals can respond relationally to any stimuli, regardless
of the actual physical properties of those specific stimuli. Additionally, unlike
nonarbitrary relational responding, AARR is a pattern of responding
belonging only to humans, with no other animal species having exhibited this
type of relational responding. Several studies have attempted to establish
equivalence classes in nonhuman animals (e.g., Murayama & Tobayama
1997; Yamamoto & Asano, 1995), but with little success. It has been
documented that with extensive training, certain species of animals have
occasionally responded with high accuracy in accordance with reflexive and
symmetrical relations (Oden, Thompson, & Premack, 1988; Pack, Herman,
& Roitblat, 1991), few have demonstrated responding in accordance with
transitivity (D’Amato & Colombo, 1985; Tomonaga, Matzusawa, Fujita, &
Yamamoto, 1991), and none have demonstrated high levels of correct
responding in accordance with stimulus equivalence. Because AARR is a
pattern of responding unique to humans, RFT rests on the assumption that
AARR is the core of complex human language and cognition, with our
advanced verbal repertoires arising from arbitrarily applicable relational
responses. As such, an understanding of the principles and concepts
underlying RFT can aid in the identification of the specific environmental
variables responsible for these repertoires, which has applied utility in that
interventions can be developed to establish and strengthen these behaviors.

PROPERTIES OF A RELATIONAL FRAME

Most literature on derived relational responding refers to a pattern of
responding relationally as a “relational frame.” This term is often used as a
noun, but it refers to the action of framing events relationally, or responding
to one event in terms of another (Hayes & Hayes, 1989). Specific patterns of
relational responding constitute different types of relational frames. For
example, responding to one stimulus in terms of its similarities to another
stimulus is one pattern of relational responding, while responding to one
stimulus in terms of its differences from another stimulus is another pattern
of relational responding, both of which represent distinct types of relational
frames. A relational frame consists of three defining properties: (1) mutual
entailment, (2) combinatorial entailment, and (3) transformation of stimulus
function. Mutual entailment occurs when an individual responds to one
stimulus in terms of another stimulus. For example, imagine a child is
presented with a block and a ball and is taught to select the ball (B) when
asked, “Which one is the ball?” (A), and is provided reinforcement for
making the correct response (A-B). As depicted in figure 11.1, mutual
entailment occurs when the individual responds by saying “ball” (A) when
shown the actual physical ball (B), without ever having received
reinforcement for producing the vocal utterance in the presence of that
stimulus (B-A).

Figure 11.1 Mutual entailment.

Combinatorial entailment adds another level of complexity and refers

to a derived stimulus relation in which two or more stimulus relations
mutually combine (Hayes & Chase, 1991). Stated another way, combinatorial
entailment occurs when an individual responds to a stimulus in terms of
another stimulus based on mutually entailed relations with other stimuli.
Using the previous example, imagine that the mutually entailed relation
between the physical ball (A) and the vocal utterance “ball” (B) has already
been established. An additional mutually entailed relation can develop if the
individual is directly taught to respond with the vocal utterance “round toy”
(C) when presented with the question “What is a ball?” (B) (B-C). In the
absence of direct training, they will demonstrate mutual entailment by saying
“ball” (B) when asked, “What toy is round?” (C) (C-B). If the child has a
sufficient repertoire of relational responding, they will additionally
demonstrate a combinatorially entailed response. In this example, the child
would respond with the vocal utterance of “round toy” (C) when presented a
ball (A) and asked, “What shape is this?” (A-C), and by selecting the ball (C)
when presented an array containing a ball and a block and asked, “Which one
is the round toy?” (A) (C-A; see figure 11.2).

Figure 11.2 Combinatorial entailment.

The final property of a relational frame is the transformation of stimulus

function. The transformation of stimulus function (see also chapter 10) is
an ongoing process that occurs contingent on the development of mutually
entailed and/or combinatorially entailed relations. The transformation of
stimulus function occurs when a certain stimulus in a relational network has
certain functions and serves to modify the functions of other stimuli in that
relational network in accordance with the derived relation (Hayes, Barnes-
Holmes, & Roche, 2001). More simply stated, the transformation of stimulus
function is demonstrated when the function of a stimulus is changed due to a
mutually and combinatorially entailed relationship with other stimuli. As a
result, this stimulus will not only evoke new behavior, but also evoke the
behavior in novel contexts because of its relational history with other events.
Building on the previous example, imagine that the mutually entailed and
combinatorially entailed relations between a ball and the vocal words “ball”
and “round” are established (A-B, B-C, A-C, and C-A). The child may be
able to request “ball” or a “round toy” when asked, “What do you want to
play with?” or ask for a ball in the absence of prompts. This pattern of
responding represents a transformation of stimulus function because neither
the question nor wanting to play with the toy would have evoked the request
for the ball prior to the establishment of a relational history.
Now, imagine the child from the previous example is presented with an
entirely novel item that they have never interacted with before. They are
taught that this new item is called a JES. If the child was asked, “Would you
like to play with the JES?” it is likely that the child will not respond to this
question. If the child was taught a JES is similar to, but also better than, a
ball, the child may now be more likely to engage with the item because it has
been related to an already preferred stimulus, and the new stimulus is better
than the already preferred item. Although this example contains an arbitrary
word with no actual meaning, it is no more arbitrary than any other novel
words that we are exposed to each day. Every word ever encountered
throughout the duration of an individual’s life was at one time an arbitrary
stimulus, void of any function or true meaning. It was not until that arbitrary
stimulus participated in relational networks with other stimuli that the
functions transformed. The functions of those once arbitrary stimuli acquired
meaning based on their relationships with known stimuli, without a direct
training history. The transformation of stimulus function gives rise to a true
understanding of the stimuli encountered in our daily lives. In the above
example, the resulting transformation of stimulus function allowed for the
reinforcing experience of interacting with a potentially preferred tangible
stimulus. The transformation of stimulus function may also allow for an
individual to solve problems or respond in terms of abstract relational events.
Once relational responding has been firmly established in an individual’s
repertoire, stimuli can be related to almost any other stimuli, resulting in a
near limitless amount of transformations of stimulus function, depending on
whatever complex combination of contextual cues happen to be present at
any given moment.

RELATIONAL FRAMING AS A GENERALIZED OPERANT

Demonstrating derived relational responses to stimuli in the absence of a
history of reinforcement is a hallmark characteristic of mutual and
combinatorial entailment. Emitting derived responses, however, first relies
entirely on a history of differential reinforcement for relating events to other
events in this way. Responding in accordance with mutual entailment, along
with combinatorial entailment and the transformation of stimulus function,
comes about due to a history of differential consequences and multiple
exemplars. In this way, relational framing is probably best conceptualized as
generalized operant behavior. When an individual is provided a sufficient
number of exemplars and receives reinforcement for making mutually and
combinatorially entailed relations, they will then begin to make these
responses seemingly naturally, and in contexts in which they were not
initially trained. Furthermore, as the individual is taught new stimulus
relations, more untrained stimulus relations will begin to emerge in the
absence of direct training. This pattern of responding allows for faster
learning and facilitates the development of more complex relational
repertoires.
Initial research on derived relational responding as a generalized operant
was scant, with most articles consisting of conceptual and theoretical
discussions on the topic (e.g., Barnes-Holmes & Barnes-Holmes, 2000) or
demonstrating the emergence of combinatorial relations in individuals with
existing relational repertoires (e.g., Healy, Barnes, & Smeets, 1998; Healy,
Barnes-Holmes, & Smeets, 2000). Recently, however, numerous empirical
articles have emerged supporting derived relational responding as operant
behavior. Some researchers have even established derived relational
responding repertoires when they were previously nonexistent (Barnes
Holmes, Barnes Holmes, Smeets, Strand, & Friman, 2004; Berens & Hayes,
2007).
A more recent study conducted by Belisle, Dixon, Stanley, Munoz, and
Daar (2016), for example, sought to teach perspective-taking skills (i.e.,
deictic relational responding) to three children with autism and evaluate if
this pattern of responding would generalize to novel stimuli. The researchers
used a two-sided picture card, which depicted a different picture on each side,
to test the children’s ability to make different types of perspective-taking
responses. To evaluate perspective-taking abilities, the child was first shown
both sides of the picture card and the card was positioned so that the
participant and the assessor could see opposite sides of the picture card; then
the assessor would ask a question that required the participant to perform
perspective taking. Baseline performance indicated that the participants failed
to respond correctly when presented with certain perspective-taking tasks.
The researchers then conducted training for one set of picture cards until the
participants could respond accurately to the required tasks. Following
training, the participants were exposed to a set of picture cards containing
novel stimuli (i.e., stimuli that differed from those taught in training) to
evaluate whether they demonstrated this pattern of responding with this new
set of cards. Two of the three participants demonstrated high levels of
responding to the novel set of picture cards, while the third participant
required additional training to reach the same criterion. Because the stimuli
in the second set of picture cards were not formally similar to those used in
the first set, correct responding to the second set cannot be interpreted as
stimulus generalization, but rather suggests this pattern of relational
responding as a generalized type of operant behavior. These findings
contribute to the growing body of literature supporting derived relational
responding as generalized operant behavior, which develops from and is
shaped by a history of differential consequences.

An RFT Account of Language and the Response

from the Behavior Analytic Community
The first behavior analytic account of language was provided by B. F.
Skinner (1957) in his book Verbal Behavior. In this analysis, Skinner
proposed that language arises in the same way as any other operant behavior
—it is reinforced. Skinner defined verbal behavior as behavior that is
reinforced through the mediation of a listener who has been trained to
provide reinforcement to the speaker. Therefore, a speaker’s verbal responses
are dependent on the consequences the speaker contacts from the listener.
While this account of verbal behavior served the behavior analytic
community for over fifty years, some believe there are challenges with this
definition. First, Skinner’s definition of verbal behavior may not be entirely
functional. A functional definition requires that the definition be stated in
terms of the history of the individual organism and the current environment.
Skinner’s definition of verbal behavior incorporated the history of
conditioning of a listener; therefore, the definition no longer depends solely
on the history of the speaker. Furthermore, this definition focuses only on the
role of the speaker and considers the listener’s behavior nonverbal. In
addition, Skinner’s definition of verbal behavior may be too broad. With
Skinner’s definition it is not hard to classify almost any behavior as verbal
behavior. For example, imagine a rat that has been trained to press a lever
and, contingent on a lever press, is provided food by a researcher. Using
Skinner’s definition, the rat’s behavior would meet the criterion for verbal
behavior. The speaker’s (i.e., rat’s) behavior (i.e., lever press) is reinforced
through the mediation of the listener (i.e., researcher). Given this example,
most basic operant experiments where an animal’s behavior is reinforced by
an experimenter would also meet this criterion. In other words, Skinner’s
definition of verbal behavior provides no meaningful way to discriminate
between all basic operant research ever done with rats and pigeons and the
complexity, generativity, and variability of human language.
An alternative account of language came in the form of a collection of
works on derived relational responding titled Relational Frame Theory: A
Post-Skinnerian Account of Human Language and Cognition (Hayes, Barnes-
Holmes, & Roche, 2001). In this book, the authors present the foundational
concepts of RFT along with the accumulation of publications supporting it
over years of research. In addition to the fundamental concepts of RFT, the
book offered an alternative account to human language and cognition.
According to RFT, language arises due to framing events relationally, and
this pattern of responding is the core of human language and cognition. Our
complex verbal repertoires are the result of relational responding in the
presence of a contextual cue that comes to control this pattern of responding
to stimuli, regardless of the physical or formal properties of the stimuli. In
short, RFT holds that AARR gives rise to language. This definition of
language addresses some of the concerns surrounding Skinner’s definition
described above. For example, RFT provides a functional definition of
language in that it is no longer dependent on an understanding of a second
person’s history of reinforcement in order to determine whether a first
person’s behavior is verbal. In RFT the behavior of both the speaker and
listener is considered verbal. The speaker produces verbal stimuli that are a
product of framing events relationally, and the listener responds in
accordance with these relationally framed events.

The Hope Relational Frame Theory Provides

A core foundational assumption of RFT is that relational training can
give rise to complex networks of untrained stimulus relations. For example,
imagine that a three-member stimulus equivalence class has been established
between a set of stimuli (e.g., ABC). From training only two relations (A-
B and B-C), another four relations (B-A, C-B, A-C, & C-A) emerge without
the need of explicit training. Now, imagine that stimulus A is then established
as being different from stimulus D (A-D). By establishing this single relation
between stimulus A and stimulus D, all other stimuli within the same
equivalence class as A are then also related to stimulus D, which results in a
total of nine untrained stimulus relations. Now, if stimulus D was then related
to a completely different set of stimuli, an even greater number of relations
would be made between the new stimuli and all other previously related
stimuli. This example demonstrates that with only a few directly trained
relations, the number of relations that emerge is exponential. With every
stimulus we have ever encountered being related to every other stimulus we
have and will encounter, it becomes clear just how generative relational
responding can be.
Because relational responding is so generative, it has particular utility in
terms of skill acquisition. Rather than directly training every single response
to an individual, trainers can set up programming so that untrained relational
responses can be derived. For example, imagine that you are trying to teach a
child to tact a novel stimulus, and you want to train vocal and printed words
to this stimulus. Rather than teaching every single relation among the stimuli,
which would be undoubtedly time consuming, you could set up the stimulus
relations so that only two relations would need to be trained for the other
relations to emerge. The generative nature of relational responding could be
particularly useful in situations where limited resources are afforded or time
constraints exist. For instance, a study conducted by Stanley, Belisle, and
Dixon (2018) sought to use procedures based on RFT to teach academically
relevant material to individuals with autism. One of the skills taught involved
unit conversion between yards, feet, and inches. Rather than teaching every
single relation between yards, feet, and inches, the researchers arranged the
stimuli so that training consisted of teaching conversion of inches to yards
and feet to yards. Once these were taught to criteria, the researchers tested for
the emergence of the unit conversion of inches to feet and feet to inches. The
participants responded with high accuracy to these specific relations, despite
the fact that they had not been directly trained. This study demonstrates the
generative nature of derived relational responding. Rather than teaching
every single relation among the stimuli (i.e., a total of six relations), the
researchers arranged the stimuli so that only two relations were directly
trained and the remaining relations were derived. Although this represents
only a single demonstration of this phenomenon, a clinician could replace the
stimuli with stimuli individualized to their client, while retaining the same
arrangement, in order to facilitate the emergence of derived relations.
 Types of Relational Frame Families
There are a number of different ways in which stimuli can be related to
one another. Take, for example, an apple and a cat. Initially, it may appear as
though these two stimuli are not related to one another at all. But upon
further inspection, these two seemingly unrelated stimuli can be related to
one another in countless ways. For example, it might be easy to think of ways
in which these two stimuli are different from each other. Perhaps the color of
the two stimuli is different (e.g., the cat is black and the apple is red), or
maybe the cat has hair, whereas the apple does not. The cat may be bigger
and weigh more than the apple. Both the cat and the apple are the same
because they are living organisms; however, the apple is classified as a fruit,
whereas the cat is classified as an animal. Relations such as same, different,
more or less than, opposite, temporal relations, and many more resulted from
only a single example. Within the RFT literature, stimulus relations have
typically been categorized into nine distinct families of relations: (1)
coordination, (2) opposition, (3) distinction, (4) comparison, (5) hierarchical,
(6) temporal, (7) spatial, (8) conditionality/causality, and (9) deictic. In this
chapter, relational frames are categorized into six families, with the frame
families of temporal, spatial, and conditionality/causality grouped into the
frame family of comparison, since these topographies are compared relative
to each other across a wide variety of dimensions (i.e., time, space, size).

COORDINATION
The first type of relational frame family is the frame of coordination
(see figure 11.3). This relational frame has been considered to be the simplest
relation and the first to emerge in an individual’s repertoire. The frame of
coordination appears to serve as the foundation on which other frames are
built (Hayes, Barnes-Holmes, & Roche, 2001). The frame of coordination
deals exclusively with sameness and similarity. The examples from
equivalence research described in the beginning of the chapter are examples
of frames of coordination. An example of similarity (but not equivalence)
would be if an individual was taught that a pen is similar to a paint brush. In
this instance, the pen and paint brush are not exactly the same physically but
are the same in terms of their function—they both are utensils used to
produce markings on a page.
 The relational frame of coordination is the most researched type of
stimulus relation, encompassing approximately 73% of the literature on
derived stimulus relations (Rehfeldt, 2011). Research for this type of frame
dates back over forty years, beginning with Sidman (1971). Since then,
abundant basic and applied studies have been published, a list far too
exhaustive to outline in this chapter. However, basic research on this
relational frame includes demonstrations of contextual control over
transformations of stimulus functions in frames of coordination (Dougher,
Perkins, Greenway, Koons, & Chiasson, 2002), relating equivalence classes
to other equivalence classes (Barnes, Hegarty, & Smeets, 1997), the
transformation of respondently conditioned stimulus functions in accordance
with arbitrarily applicable relations (Roche & Barnes, 1997), and many
others (e.g., Barnes-Holmes, Barnes-Holmes, Roche, & Smeets, 2001a;
Healy et al., 2000; Valverde, Luciano, & Barnes-Holmes, 2009). Researchers
have used instructional procedures grounded on the frame of coordination in
a variety of areas, including teaching fraction-decimal relations (Lynch &
Cuvo, 1995), second languages (Haegele, McComas, Dixon, & Burns, 2011;
Joyce, Joyce, & Wellington, 1993), and accurate portion size estimates
(Hausman, Borrero, Fisher, & Kahng, 2017) to children and adolescents, as
well as to college students (e.g., Critchfield & Fienup, 2010; Fienup, Covey,
& Critchfield, 2010; Sella, Ribeiro, & White, 2014; Walker, Rehfeldt, &
Ninness, 2010). In addition, coordination has demonstrated particular utility
for skill acquisition in individuals with disabilities and has been used to teach
name-face matching to adults with brain injuries (Cowley, Green, &
Braunling-McMorrow, 1992), as well as to teach US geography (LeBlanc,
Miguel, Cummings, Goldsmith, & Carr, 2003), music skills (Arntzen,
Halstadtro, Bjerke, & Halstadtro, 2010), monetary values (Keintz, Miguel,
Kao, & Finn, 2011), academic skills (Stanley et al., 2018), and emotions
(NoRo, 2005) to individuals with autism. (See chapter 10 of this volume for
more on frames of coordination as it pertains to equivalence-based
instruction.)
 Figure 11.3 Frame of coordination.

DISTINCTION
The relational frame family of distinction refers to relating one stimulus
in terms of its differences to another stimulus. For example, if an individual
is taught that the text C-U-P is the same as the vocal utterance “cup” and then
taught that this vocal utterance is different from the vocal utterance “bowl,”
the individual will then respond not only as though the vocal utterance
“bowl” is different from the vocal utterance “cup,” but also that it is different
from the text C-U-P and vice-versa (see figure 11.4). Again, notice in this
example that the relation of sameness must first be established or exist in the
individual’s repertoire before the relation of difference develops. An
additional component of the frame of distinction worth noting is that the
relevant dimension by which the stimuli are being differentiated is usually
not specified. We do not know exactly which specific features of the stimuli
differ, only that they differ in some way. Using the previous example, it is not
clear how exactly a bowl and a cup are different, since they are different
along several properties (e.g., shape, color, size, texture). Successful AARR
does not require this specification; it only requires responding accurately to
the contextual cue “different.” As a more clinically oriented example, a
clinician might target this pattern of responding if attempting to teach a child
olfactory discrimination (i.e., discrimination of different scents). The
clinician could present a sample scent (e.g., vanilla), followed by a
sequentially presented identical scent and a nonidentical scent (e.g.,
cinnamon) and the discriminative stimulus “Which is different?” If the child
received reinforcement following a correct relational response, it would
strengthen this pattern of nonarbitrary relational responding. Although this
may appear as a trivial skill not worth targeting, it has potential applied
utility for individuals with limited verbal repertoires who have difficulty
responding accurately to problems involving the contextual cue of different.

Figure 11.4 Frame of distinction.

Research on the frame of distinction dates back to the first

demonstration of AARR, in which participants’ responses to arbitrary stimuli
were brought under contextual control (Steele & Hayes, 1991). Several basic
studies on the frame of distinction have been conducted since then (e.g.,
Roche & Barnes, 1997), most of which have been demonstrations of basic
behavioral phenomena that support underlying theory of RFT, lending further
empirical support for the theory. For instance, a study conducted by Dixon
and Zlomke (2005) sought to evaluate relational frames of coordination,
opposition, and distinction using computerized training procedures consisting
of a series of conditional discriminations. During the first part of their study,
participants were shown two stimuli on the screen and asked to choose
between two arbitrary colored response options that served to describe the
stimuli as same, opposite, or different. This procedure established relations of
same, opposite, and different between these stimuli. Participants were then
exposed to a series of tests to determine if they were able to correctly respond
to the derived, symmetrical relations they were initially trained on. All
participants demonstrated high levels of correct responding to the tests, as
well as when presented with novel stimuli. This study demonstrated that
procedures other than match-to-sample can be utilized as a means of
establishing relational responding. Although there are several basic and
translational studies on the frame of distinction, most demonstrations were
conducted with typically developing participants who likely could already
respond in accordance with the frame of distinction. Currently, there have
been no studies that have successfully established a frame of distinction in
individuals previously lacking this pattern of responding. Therefore,
substantial applied research is necessary to provide empirically based
methodology for teaching these behaviors. A particular population that would
greatly benefit from applied procedures for establishing relational framing of
distinction is individuals with autism, given that these individuals often have
deficits in their relational repertoires.

OPPOSITION
The relational frame family of opposition deals with relating one
stimulus in terms of another stimulus based on the contextual cue of opposite.
For example, if a child was taught an elephant is big and then taught big is
the opposite of small, the child will subsequently respond to the elephant as
the opposite of small (see figure 11.5). The relational frame families of
opposition and distinction are similar to one another, yet differ based on the
relevant dimensions of the stimulus (Rehfeldt & Barnes-Holmes, 2009). With
the frame of opposition, the stimuli are being differentiated based on some
reference point along a continuum of a relevant dimension, with each event at
opposing ends of the continuum. In a frame of distinction, the stimuli are also
differentiated based on a relevant dimension, but the dimension remains
unspecified and is not often implied. For example, an elephant is the opposite
of a mouse because each are at differing ends on the continuum of the
specified dimension of size. In contrast, the statement “the elephant is
different from the mouse” does not indicate which dimension they are being
differentiated upon (e.g., size, shape, color).

Figure 11.5 Frame of opposition.

Numerous studies have been published on the frame of opposition, both

basic and applied. For example, Dymond, Roche, Forsyth, Whelan, and
Rhoden (2007) conducted a study demonstrating the transformation of
avoidance response functions in accordance with the relational frames of
same and opposite. In this experiment, participants were first exposed to a
relational training procedure to establish same and opposite relations among
arbitrary stimuli. Once relational networks were established, one stimulus
from the network was established as discriminative for an avoidance
response that canceled a scheduled presentation of an aversive image/sound.
Results showed that for participants exposed to the relational training
procedure, other stimuli in the network acquired discriminative functions in
accordance with the relations of sameness and opposition. In other words,
stimuli that were related in a frame of sameness or coordination were
discriminative for the previously taught avoidance responses, whereas those
that were related in a frame of opposition were not.
 Other basic research on the frame of opposition has demonstrated the
transformation of consequential function in accordance with the frame of
opposition (Whelan & Barnes-Holmes, 2004), as well as generalized
responding to the contextual cue of opposite (Healy et al., 2000).
Furthermore, Barnes-Holmes, Barnes-Holmes, and Smeets (2004) conducted
a study in which children who were lacking the frame of opposition were
exposed to a set of relational trainings that resulted in the emergence of
generalized responding in accordance with the frame of opposition. The
authors of this study accomplished this by first exposing the participants to a
series of training and testing phases, which involved establishing opposite
relations among several paper coins and presenting a problem-solving task to
the participant. The problem-solving task required the participant to select the
coin that could be used to buy as many or as few sweets as possible, which
the participant could only accomplish if they could respond in accordance
with relational frames of opposition. After a series of trainings, the
experimenters were successful in establishing relational responding in
accordance with opposition, and this responding generalized to presentations
involving novel stimuli and to novel experimenters. These findings support
the assumption that relational responding is a generalized operant behavior.
Similar studies have been conducted with children with disabilities, and the
results suggest that relational training procedures are effective in establishing
relational responding of opposition with this population as well (e.g., Dunne,
Foody, Barnes-Holmes, Barnes-Holmes, & Murphy, 2014).

COMPARISON
The relational frame of comparison deals with relating one stimulus in
terms of another based a specific quantitative or qualitative dimension.
Common examples of responding in accordance with this relation include
more/less, better/worse, bigger/smaller, and faster/slower. Countless other
comparative relations among stimuli can exist. It is important to note that the
relevant controlling variable for any particular response is relative to the
comparison being made; it is not inherent in any property of a particular
stimulus. For example, if you showed an individual a large book and a
medium-sized book and asked, “Which is bigger?” they would select the
large book rather than the medium book. Now, if you presented the medium-
sized book alongside a small book, the person would select the medium-sized
book. In this example, the medium book is only smaller when it is being
compared to a larger stimulus, thus illustrating how a direct relational
comparison must be made between the stimuli (see figure 11.6). A more
clinically relevant example might involve teaching monetary values of coins
to a child, such as the relations among the values of a quarter, a penny, and a
dime. To accomplish this, a clinician could present the dime as a sample
stimulus along with an array containing a penny and a quarter, while
reinforcing the learner’s selection of the quarter when asked, “Which is
worth more?” The clinician could then repeat the stimulus arrangement
except reinforce the learner’s selection of the penny when asked, “Which is
worth less?” This arrangement would effectively establish the quarter as
more than the dime and the penny as less than the dime, while
simultaneously allowing for the derivation of the relationship between the
quarter and the penny.

Figure 11.6 Frame of comparison.

Like the relational frame of coordination, the relational frame of

comparison has a well-established empirical base. Several basic studies have
been conducted demonstrating the emergence of AARR in terms of this
frame and the resulting transformation of stimulus functions (e.g., O’Hora,
Roche, Barnes-Holmes, & Smeets, 2002; Vitale, Barnes-Holmes, Barnes-
Holmes, & Campbell, 2008; Whelan, Barnes-Holmes, & Dymond, 2006). In
a study conducted by Dougher, Hamilton, Fink, and Harrington (2007), for
instance, the experimenters exposed college students to a series of trainings
that established a frame of comparison using arbitrary stimuli. They then
taught the participants that stimulus “A” was smaller than stimulus “B,” and
stimulus “C” was the largest stimulus. The experimenters then paired
stimulus “B” with mild shock using respondent conditioning. Next, the
experimenters presented the participants with each of the three stimuli and
measured changes in their skin conductance. The results showed changes in
skin conductance were greater for stimulus “C” than for stimulus “B,” and
were smaller for stimulus “A” than for stimulus “B” for six out of eight
participants. This exemplifies the transformation of stimulus function for the
“A” and “C” stimuli, since only “B” was paired with shock. Recent applied
studies on this frame have demonstrated that the frame of comparison can be
established in typically developed children’s repertoires when it is lacking
(Berens & Hayes, 2007), and this pattern of responding can generalize to
novel stimuli and trainers (Barnes-Holmes, Barnes-Holmes, Smeets, et al.,
2004). Data from these studies support AARR as an operant behavior and
demonstrate the utility of multiple exemplar training in establishing relational
frames. Other studies on the frame of comparison have demonstrated utility
in populations of those with disabilities, such as autism (e.g., Dunne et al.,
2014).

HIERARCHICAL
Hierarchical relations consist of responding to a stimulus in terms of its
membership to or as being part of another stimulus (see chapter 13 in this
volume). In other words, hierarchical relations denote belongingness between
a group of stimuli and a common categorical relation, and they are often
stated in the form of “stimulus A is an attribute of stimulus B” or “stimulus A
is contained by stimulus B.” For example, if an individual was taught that
cats and dogs belong to mammals, and snakes and lizards belong to reptiles,
they would respond to mammals as containing cats and dogs, and reptiles as
containing snakes and lizards (see figure 11.7). Similarly, if the individual
was taught that mammals and reptiles belong to animals, the individual
would respond as though animals contained dogs, cats (both of the
mammals), snakes, and lizards (both of the reptiles). Notice, however, that
although all dogs are animals, not all animals are dogs. The functions of
belongingness flow top down, not bottom up, which is a defining
characteristic of hierarchical frames. This type of responding may be the
closest to providing an understanding of complex topics such as abstraction
and concept formation.
Research on the relational frame of hierarchy is less established than
the other relational frames, with few empirical demonstrations reported in the
literature (see chapter 13). One of the first demonstrations of a hierarchical
relation was conducted by Griffee and Dougher (2002), who taught college
students a series of relations to arbitrary stimuli and conducted subsequent
tests to determine the extent to which the participants formed hierarchical
relations. The findings suggested that relational responding could be brought
under contextual control consistent with that of hierarchical relations.
Furthermore, results showed all stimuli at the bottom of the hierarchy shared
functions with stimuli at the top of the hierarchy, but the stimuli at the top of
the hierarchy did not share the functions of those at the bottom.
Additional studies have emerged demonstrating the utility of multiple
exemplar training to teach hierarchical relations (e.g., Gil, Luciano, Ruiz, &
Valdivia-Salas, 2012, 2014). A recent study conducted by Mulhern, Stewart,
and Elwee (2017) found a correlation between age, intellectual performance,
and hierarchical responding. These data suggest hierarchical responding
seems highly correlated with intellectual performance, and hierarchical
relational frames tend to develop over the course of several years as
individuals age. Although the previously mentioned studies add empirical
support for hierarchical relations, further research is necessary on this frame.
With the exception of the study by Mulhern et al. (2017), all the previously
mentioned studies were conducted with typically developed adults, which
limits the generalizability of the data to other populations. Future research
can evaluate different procedures for establishing hierarchical relations in
children, since children likely have simpler verbal repertoires than adults.
Further research is needed on this relational frame in terms of applied
demonstrations in populations with disabilities. Research has yet to be
conducted that evaluates the effectiveness of procedures in establishing
hierarchical frames in individuals lacking this repertoire (see chapter 12 for
more on hierarchical frames).
 Figure 11.7 Frame of hierarchy.

DEICTIC
Deictic relational frames involve responding to a stimulus in terms of its
perspective to the speaker themself. Deictic responding appears to be central
to perspective taking and is often under the contextual control of pronouns,
such as I/you, me/she, us/them, and so on. Perspective taking can also
include spatial relations of here/there and temporal relations of now/then,
since one is always speaking here and now and others are always listening
there and then. Given that deictic relational responding requires some form of
perspective taking by the speaker, this pattern of relational framing is
essential for the development of an understanding that different people see
different things. Although this seems like a relatively simple skill, it is often
deficient in individuals with disabilities such as autism, Down syndrome, and
other developmental and intellectual disabilities. Therefore, clinicians can
teach deictic relational responding to these individuals in an effort to
facilitate the development of many perspective-taking skills.
Studies have reported a relationship between an individual’s
perspective-taking abilities and intellectual functioning (Gore, Barnes-
Holmes, & Murphy, 2010). Individuals who respond with low accuracy to
perspective-taking tasks typically have lower intellectual functioning and
verbal ability. Other studies have found that those with schizophrenia
(Villatte, Monestès, McHugh, Freixa i Baqué, & Loas, 2010) and social
anhedonia (Vilardaga, Estévez, Levin, & Hayes, 2012; Villatte, Monestès,
McHugh, Freixa i Baqué, & Loas, 2008) demonstrate difficulty responding in
accordance with deictic relations, suggesting a potential connection between
the two. Several studies have reported successful results in terms of
establishing this type of responding in typically developed individuals (e.g.,
Barnes-Holmes, Barnes-Holmes, & McHugh, 2004; Heagle & Rehfeldt,
2006; Weil, Hayes, & Capurro, 2011) as well as those with
neurodevelopmental disabilities (e.g., Belisle et al., 2016; Jackson, Mendoza,
& Adams, 2014; Rehfeldt, Dillen, Ziomek, & Kowalchuk, 2007) and
schizophrenia (O’Neill & Weil, 2014). Taken together, results from these
studies suggest that deictic relating is amenable to training via multiple
exemplar training. Research on deictic relations could benefit from additional
studies establishing single and double reversals in nontypically developed
individuals, especially in terms of here/there and then/now relations. (See
chapter 17 for more on perspective taking.)

Implications and Future Directions

Within only a few decades, RFT has generated a significant amount of
research, and the number of studies published on derived relational
responding and RFT continues to increase (Dymond, May, Munnelly, &
Hoon, 2010). There are areas, however, in which additional research would
be beneficial, one being cross-sensory stimulus relations (i.e., stimulus
relations encompassing more than one sensory modality). Much of the
literature on derived relational responding focuses primarily on establishing
stimulus relations between visual and vocal stimuli, with little attention
given to the olfactory, gustatory, and tactile sense modes. While studies
demonstrating cross-modal relational frames exist (e.g., Belanich & Fields,
1999; Dixon, Belisle, Stanley, Munoz, & Speelman, 2017; Toussaint &
Tiger, 2010), the limited research does little in providing a well-documented
framework for program development by clinicians. Despite the limited
demonstrations, movement toward broadening RFT’s applications—such as
the incorporation of cross-sensory stimulus relations—can be seen with the
development of a curricular application based on the theory and concepts
underlying RFT, known as the Promoting the Emergence of Advanced
Knowledge (PEAK) Relational Training System. PEAK consists of four
modules, each designed to target a specific type of learning. The final two
modules, the PEAK-Equivalence (Dixon, 2015) and PEAK-Transformation
(Dixon, 2016) modules, are directly based on the concepts and principles
underlying stimulus equivalence and RFT. These modules incorporate these
principles and concepts into (1) an assessment designed to identify an
individual’s potential relational framing deficits and (2) a curriculum
designed to teach the skills identified as deficits by the assessment. There
are 184 different programs within the curricula for each of the modules that
clinicians can use to teach relational framing skills to their clients. With the
programs encompassing a broad range of relational skills, including cross-
sensory relational responding, the overarching goal of the PEAK is to
establish relational framing as a generalized operant behavior. Clinicians
can utilize this technology to facilitate relational framing repertoires, and
researchers can use the programs as a framework to answer empirical
questions.
A second avenue for future RFT research would be to conduct
longitudinal studies. Longitudinal studies would be beneficial for, and add to
the existing empirical support for, RFT in a number of ways. First, having
longitudinal studies that document the development of relational frames and
language would allow for a better analysis of the contribution of relational
framing to language development. RFT holds that relational framing emerges
first and gives rise to language, or more precisely, that language consists of
relational framing. Although Lipkens et al. (1993) conducted such an
analysis and obtained results supporting this perspective, their data is limited
to only a single child. Longitudinal studies would yield meaningful data in
terms of whether language or relational framing emerges first. Second,
longitudinal research could be afforded to populations that are often lacking
complex verbal repertoires. Having longitudinal studies conducted in this
area would allow researchers to evaluate the effect relational training may
have in producing more complex verbal responses in these individuals.
Researchers may also conduct randomized controlled trials (RCTs) of
various RFT-based procedures. While most studies in behavior analysis
feature the use of single-subject designs, RCTs would add an additional level
of scientific credibility outside of the behavior analytic community. RCTs
would also allow researchers to compare the relative effectiveness of RFT
and other approaches for producing changes in larger and longer-term
repertoires. For example, researchers may compare RFT-based procedures
with traditional verbal behavior therapy and control groups to evaluate which
form of training is most effective and efficient in producing sophisticated
verbal repertoires. RCTs may also be utilized to compare groups of people
who receive varying amounts of relational training to evaluate if a dosage
effect is present.
Finally, future research can be conducted evaluating the relationship
between neurological data, derived relational responding, and language.
Numerous studies have evaluated the relationship between derived relational
responding and language using functional MRIs (fMRIs; Hinton, Dymond,
Von Hecker, & Evans, 2010; Ogawa, Yamazaki, Ueno, Cheng, & Iriki, 2010;
Schlund, Cataldo, & Hoehn-Saric, 2008; Schlund, Hoehn-Saric, & Cataldo,
2007) and electroencephalograms (EEGs; e.g., Barnes-Holmes et al., 2005;
Roche, Linehan, Ward, Dymond, & Rehfeldt, 2004; Yorio, Tabullo,
Wainselboim, Barttfeld, & Segura, 2008). Most of these studies involve
exposing participants to a series of relational tests while documenting which
areas of the brain are activated during exposure to these tests. Rather than
documenting which areas of the brain are activated when exposed to
relational procedures, future research could attempt to link neurological data
with observed behavioral performances.

Conclusion
Sixty years ago, Skinner proposed the first behavior analytic account of
human language. This analysis has served as the primary account of
language for the behavior analytic community since that time; however, the
definition of verbal behavior and the analysis itself was not without
limitations. RFT offers an extension that addresses some of the potential
limitations of Skinner’s analysis. An RFT account of language holds that
AARR is the core of language. It becomes increasingly hard to dismiss RFT
as the evidence supporting it continues to grow. If the evidence supports the
utility of an RFT approach for producing analyses and outcomes that are
not possible using Skinner’s verbal operants alone, then the utility of
continued work using only Skinner’s verbal operants may require
reexamination.
 Since the emergence of RFT in the 1980s, a substantial body of
evidence has been established that supports the underlying theoretical and
conceptual foundations of RFT. Additionally, RFT has demonstrated
ubiquitous utility in terms of teaching skills to individuals and promoting
responding that is generative of numerous other behaviors. Although RFT
has generated a significant amount of empirical support, many areas could
benefit from further empirical exploration. As we move forward in our
pursuit of understanding language and cognition, it is imperative that
behavior analysis continue to research and scientifically evaluate the
underpinnings of language and cognition, while making data-driven decisions
rather than appealing to dogma.

Study Questions
1. What characterizes an emergent or derived relation?
2. Distinguish between nonarbitrary relational responding and
arbitrarily applicable relational responding.
3. From an RFT perspective, what gives rise to complex human
language and cognition?
4. What is a relational frame?
5. What are the three defining properties of a relational frame and how
do they differ from one another?
6. According to RFT advocates, what are some potential limitations of
a purely Skinnerian approach to language?
7. What are the different types of relational frame families discussed in
this chapter?
8. In terms of applied utility, what is the primary advantage offered by
programming for the emergence of derived relations?
 Chapter 12:

Assessing and Teaching Complex Relational

Operants: Analogy and Hierarchy

Ian Stewart
National University of Ireland, Galway
Shane McLoughlin
University of Chichester, UK
Teresa Mulhern
National University of Ireland, Galway
Siri Ming
Private Practice
E. B. Kirsten
National University of Ireland, Galway

Link and Preview

The preceding chapter surveyed the different types of relational
operants. To supplement that discussion, the current chapter
focuses exclusively on two in particular, analogy and hierarchy. As
was the case with the operants described in chapter 11, the
discussion in this chapter is inspired by relational frame theory
(Hayes, Barnes-Holmes, & Roche, 2001). Regardless of
theoretical focus, the implications for behavior analytic
instructional procedures focused upon analogical and hierarchical
reasoning are clear.
As has been discussed previously, relational frame theory (RFT) has
introduced the idea that derived arbitrarily applicable relational responding,
or relational framing, is an operant repertoire that underlies human language
and cognition. Substantial research findings now support this thesis,
including studies on diverse patterns of derived relational responding that
show a consistent correlation with complex human performance (e.g.,
Dymond & Roche, 2013; Zettle, Hayes, Barnes-Holmes, & Biglan, 2016).
Much of the work has been focused on relatively simple relational operants
(primarily equivalence but also others such as opposition and comparison).
However, RFT research has also modeled and investigated more complex
derived relations, including analogy and hierarchy, in both children and
adults as well as in populations of both typically developing individuals and
those with developmental delay. The aim of this chapter is to consider the
research on these two particular relatively complex operants.

Analogy
A critical feature of analogy (“A is to B as C is to D”) is the transfer of
existing knowledge from one context to another through an assessment of
the relationship or overlap between them. It is a core component of higher-
order language and cognition, including scientific and mathematical skills
(e.g., Polya, 1954) as well as problem solving more generally (e.g., Brown,
1989) and is commonly used as a metric of intellectual potential (e.g.,
Sternberg, 1977). In more basic, functional analytic terms, analogy may
underpin much of our untaught behavior in novel contexts because it allows
us to use existing knowledge of familiar contexts to guide behavior in new
or unfamiliar ones. Examples of this abound in educational and scientific
domains. Consider for instance “electron (A) is to nucleus (B) as earth (C)
is to sun (D).” This shows the basic process involved in analogy whereby
an individual’s familiarity with a known domain (e.g., the solar system) can
be used to teach them about important aspects of a second unknown domain
(e.g., atomic structure). Furthermore, analogy is not just confined to
education and science but is evident in diverse situations in everything from
the understanding of figurative speech (e.g., “arguing with him is fighting a
losing battle”) to dealing with new procedures or technologies (e.g., “using
that new application is like reading a map”).
 Based on its perceived importance, the study of analogy has long been a
key area within mainstream cognitive research (e.g., Bod, 2009; Gentner,
1983; Vosniadou & Ortony, 1989). In this field, the general consensus is that
analogy involves a transfer of relational information from a domain that
already exists in memory (usually termed the “source,” the “base,” or the
“vehicle”) to a domain to be explained (termed the “target,” or the “topic”;
see, for example, Vosniadou & Ortony, 1989). Furthermore, over the years a
variety of theories have been produced in accordance with this
conceptualization (e.g., Christie & Gentner, 2014; Corral & Jones, 2014;
Gentner, 1983; Gentner, Holyoak, & Kokinov, 2001; Holyoak & Thagard,
1989; Keane, 1997; Lu, Chen, & Holyoak, 2012; Sullivan & Barner, 2014).
Until the 1990s, behavioral psychologists showed little or no interest in
analogical reasoning. The advent of RFT has changed this. RFT posits
derived relational responding as the core operant pattern underlying human
language and cognition, including all of the complex performances of which
we are capable, such as analogy. Equivalence (or coordination) relations are
seen as the simplest relational pattern that can be derived and most likely the
first to be acquired (e.g., Lipkens, Hayes, & Hayes 1993). In this case, if a
learner is told that A is the same as B, then they will readily determine that B
is the same as A, even though this has not been directly taught. In this case,
the word “same” cues a coordination relation between A and B and enables
the learner to derive the mutually entailed B-A coordination relation,
assuming an appropriate training history. Furthermore, if this learner is then
taught to match B and C, then a more complex mutually entailed A-C
equivalence relation can also be derived. Multiple other derived relations
have also been empirically shown, including for example, comparison (e.g.,
“if A is more than B, then B is less than A”), opposition (e.g., “if A is
opposite of B, then B is opposite of A”), temporality (e.g., “if A is before B,
then B is after A”), and so on. Of key importance from an RFT perspective is
the fact that derived relations are arbitrary, and thus the relations need not be
based on formal properties of the stimuli. For example, the C and A terms in
derived equivalence don’t go together because they look like each other in
some way (just as the word “cat” does not resemble an actual cat) but based
on a learned generalized relational pattern.
A key feature of analogies is that they involve the derivation of an
equivalence relation between derived relations themselves. Hence in the
analogy “A : B :: C : D,” the task is to derive a relation between A and B and
between C and D, and also to derive an equivalence relation between these
two relations themselves based on the fact that they are the same type of
relation. Consider the analogy presented in figure 12.1, denoted as “chair :
stool :: glove : sock.”

Figure 12.1 A representation of the analogy denoted as “chair : stool

:: glove : sock” showing the relations between the elements.

From an RFT perspective, “chair” and “stool” participate in a relation of

equivalence (because both are furniture), while “glove” and “sock” also
participate in a relation of equivalence (because both are items of clothing).
A correct analogical response involves the derivation of these two
equivalence relations and the derivation of a further equivalence relation
between the relations (i.e., “chair” is equivalent to “stool” in the same way as
“glove” is equivalent to “sock” because each are members of the same
respective class).
In accordance with this conception, Barnes, Hegarty, and Smeets (1997)
reported the first RFT model of analogical reasoning as the derivation of
equivalence relations between equivalence relations, or equivalence-
equivalence responding. In this study, the experimenters first gave
participants match-to-sample (MTS) training and testing for the induction of
several equivalence relations between arbitrary (nonsense syllable) stimuli. In
a subsequent, critical stage, they exposed the participants to equivalence-
equivalence (analogical) testing, which was similar to the equivalence testing
but comprised two different types of trials, referred to as similar–similar and
different–different trials, respectively. During the former (similar-similar)
trial, the sample stimulus was always a compound of a combinatorially
entailed relation of sameness (e.g., B1C1), and the two comparisons included
both a compound stimulus formed by a combinatorially entailed relation of
sameness (e.g., B3C3) and one formed by a combinatorially entailed relation
of distinction (e.g., B3C4). Different–different trials entailed the same
comparisons, but this time the sample stimulus was a combinatorially
entailed relation of distinction, not sameness. In the language of RFT, this
equivalence-equivalence test was designed to see if participants would relate
two derived coordination or distinction relations to each other in accordance
with a derived frame of coordination (i.e., relationally frame one relation as
being coordinated with another relation). Findings were that a range of
participants, including adults as well as nine- and twelve-year-old typically
developing children, readily passed this analogical-type test.
A considerable quantity of studies has now extended this work. One
aspect on which several follow-up studies have focused is the importance of
formal, or nonarbitrary, properties in analogy. Consider again the analogy
from figure 12.1: “chair: stool :: glove : sock.” In this example, the arbitrary
equivalence relation between the words “chair” and “stool” is based, to some
extent, on the coordination of physical properties shared by the actual stimuli
with which the words are coordinated (i.e., both actual items have legs and a
flat platform for sitting on). Similarly, the arbitrary equivalence relation
between the words “glove” and “sock” is based on the shared physical
properties between the clothing items (i.e., both are made of soft material and
fit on parts of the body). Thus, although the two equivalence relations and the
equivalence-equivalence relation are entirely arbitrary (there are no formal
similarities shared by the words or shared between the words and the actual
objects), they can readily be traced back to shared nonarbitrary features. In
simple terms, the nonarbitrary features are brought to bear automatically in
the derivation of the coordination relations.
A study by Stewart, Barnes-Holmes, Roche, and Smeets (2001) was the
first to capture this dimension of analogy, by showing equivalence-
equivalence relations based on stimuli that differed along various physical
dimensions including, for example, color. Stewart, Barnes-Holmes, Roche,
and Smeets (2002) subsequently demonstrated that relating derived relations
could allow the discrimination of a common physical similarity between the
relations and that this in turn could produce transformation of functions in a
separate task. For example, the researchers exposed participants who had
previously sorted a series of wooden blocks according to color instead of
shape to an analogical protocol in which all trials involved the discrimination
of common shapes across the equivalence-equivalence network. After that,
participants changed the way they sorted the blocks so that they sorted
according to shape. This arguably modeled the experience of “insight” via
analogy, whereby analogical responding facilitates a new more effective
response to the environment. One more recent study by Ruiz and Luciano
(2015) showed that participants would choose equivalence relations
involving common physical properties as more closely related and thus more
suited to form an equivalence-equivalence relation than equivalence relations
involving different physical properties.
RFT research has extended the equivalence-equivalence model in a
number of other ways also. For example, a number of researchers have used a
method known as the relational evaluation procedure (REP) to demonstrate
a much more generative model of analogy than in the original (i.e., Barnes et
al., 1997) study (Stewart, Barnes-Holmes, & Roche, 2004). Other researchers
have measured relating derived relations, as a model of analogy, using
reaction times and event-related potentials (ERPs). This work showed that,
based on presumed lower levels of complexity, derived similar-similar
relations were emitted with greater speed than derived different-different
relations and were underpinned by different patterns of neural activity
(Barnes-Holmes et al., 2005). This outcome is similar to findings seen
elsewhere in the neurocognitive literature (Luo et al., 2003) and suggests that
the derivation of relations between relations recruits similar neural resources
to solving analogies in conventional format. Subsequently, Lipkens and
Hayes (2009) demonstrated the derived relating of opposition and
comparative relations and showed that directly teaching an analogy between
two relations allowed participants to derive a number of untaught analogies
between novel stimuli. Finally, Ruiz and Luciano (2011) extended the RFT
model of analogy by teaching and testing cross-domain analogies, defined
as the derived relating of relations across separate relational networks.
Whereas the research reviewed thus far focused exclusively on within-
domain analogies (e.g., “overcoming a difficult personal problem is like
overcoming a difficult work problem”), cross-domain analogies involve the
transfer of knowledge from one domain to a completely unrelated domain
(e.g., “overcoming a difficult personal problem is like reaching the top of a
mountain”). Ruiz and Luciano (2011) found that such analogies could be
established via a history of multiple exemplar training (MET) and that
performance on the model strongly correlated with that on a standard analogy
test.
 For the present purposes, perhaps the most relevant extension of Barnes
et al. (1997) is research assessing derived relations between relations in
young children. If RFT researchers are correct that equivalence-equivalence
responding represents a basic empirically controlled form of analogy, then it
might be expected that performances based on equivalence-equivalence
responding would coincide with those recorded on more traditional analogy
tests used by mainstream researchers.
For cognitive developmental psychologists, one key issue has always
been at what age children become able to engage in analogical reasoning.
Because of the view that the relations contained within analogies are complex
(often referred to as “higher-order”), it was traditionally believed that
children under twelve could not solve analogies (e.g., Piaget, Montangero, &
Billeter, 1977). According to Piaget, higher-order relations were present only
in formal operational thinking, which is not fully developed until
adolescence. Apart from providing empirical evidence supporting this claim,
Piaget also tried to identify the incorrect strategies that younger children in
his studies may have used to try to solve the analogies. One type of incorrect
answer was indicative of simple associations rather than higher-order
relations. For example, consider the task “Bicycle is to handlebars as ship is
to ?” In this case the relationship between A and B refers to
“steering mechanism,” so the correct answer is “rudder.” However, children
would often select “bird” based on the logic that “both birds and ships are
found on lakes.” In doing so, they were focused only on the “lower-order”
relation in the incomplete pair (C and D); they ignored the first pair (A and
B) and thus also missed the higher-order relation between relations.
 Goswami and Brown (1990) have also closely examined the types of
simple associations to which children incorrectly default in the context of
analogies. Their research deployed pictorial stimuli, presenting three of the
stimuli from an A : B :: C : D analogical pattern as an incomplete pattern and
an additional four stimuli as response options. Consider, for example, the
following sequence: “Bird is to nest as dog is to ” with the options
“doghouse,” “cat,” “another dog,” and “a bone.” The correct (analogical)
choice involves selecting the doghouse, based on the common relation of
habitat. The remaining, incorrect options in this case include the bone as a
simple associative choice (as dogs and bones frequently go together), the
other dog as a match on the basis of “surface similarity” derived primarily
from perceptual features, and the cat as a “category match.” Findings showed
that children between ages four and nine performed at levels significantly
above chance, with accuracy positively correlated with age. On this basis,
Goswami and Brown (1990) concluded that children as young as four or five
can show analogical reasoning, and that increasing age allows them to
respond less on the basis of simple associations and more in accordance with
higher-order relations.
One possible confound that may have influenced outcomes in both the
Piaget et al. (1977) and Goswami and Brown (1990) studies, however,
concerns the level of the children’s familiarity with the target pairs and the
relations between them. In Piaget’s research, the children’s analogical
reasoning skills may have been underestimated because they did not have
adequate experience or knowledge of concepts such as steering mechanisms,
for example. Regarding the findings of Goswami and Brown, it is possible
that age differences may have reflected differences in levels of experience
with birds, nests, and so on, rather than in relational competence.
 As it happens, data from Goswami and Brown (1989) shed light in this
regard. This study involved a series of item analogies based on causal
relations (e.g., cutting, wetting, and melting), with which most three- or four-
year-olds are familiar. Consider the task “Playdoh is to cut playdoh as apple
is to .” The experimenters presented tasks such as this to children in
conjunction with other tasks that probed their familiarity with the target
causal relations. For example, they showed the children three pictures of
items that had been causally transformed (e.g., cut playdoh, cut bread, cut
apple) and asked them to select the causal agent responsible for the
transformation (e.g., a knife, water, the sun). Results showed a strong
correlation between scores on the analogies and on the control tasks, thus
suggesting the importance of familiarity with target relations for analogical
reasoning. This implies, first, that relational knowledge and analogical
competence codevelop, and second, that where relational knowledge is high,
children even as young as four may derive the target relations and respond in
accordance with analogy. Conversely, where relational knowledge is
inadequate, associative, and likely incorrect, strategies will dominate.
Carpentier, Smeets, and Barnes-Holmes (2002) used the equivalence-
equivalence paradigm to provide a functional analytic test of the emergence
of analogical responding in young children. They assessed a range of age
groups including adults, nine-year-olds, and five-year-olds. Across several
experiments, adults and nine-year-olds successfully demonstrated
equivalence-equivalence relations, but the five-year-olds did not. Only when
the latter were exposed to equivalence-equivalence type tests in which the
composite equivalence relations were directly taught as opposed to being
derived did they subsequently demonstrate the target performance. Further
empirical support for these findings was reported in follow-up research by
Carpentier, Smeets, and Barnes-Holmes (2003).
The fact that five-year-olds in Carpentier et al. (2002) did demonstrate
derived analogical relations seems consistent with Goswami and Brown’s
(1990) research; however, in contrast to the outcomes of the latter, extensive
relational pretraining was required. As such, the issue of whether five-year-
olds can genuinely show analogical reasoning based on the core processes
described here (i.e., the derivation of relations between derived relations)
seems as yet open. From the RFT perspective, derived relational abilities
should be relatively well honed by age five, and five-year-olds should also
have a robust general relational knowledge base. As such, it might be
predicted that they could solve at least some analogy tasks, though it would
need to be ascertained that such performances are based on derived
equivalence-equivalence rather than on some simpler associative basis. In
any event, RFT would predict that explicit exemplar training in equivalence-
equivalence responding should greatly boost performance on traditional
analogy tasks.
As an example of the potential application of the RFT approach to
analogy in an applied arena, one relatively recent study used MET to teach
children with autism to solve novel metaphors (Persicke, Tarbox, Ranick, &
St. Clair, 2012). Metaphor is a subtype of analogy in which a primary
function of the relational process is to highlight particular nonarbitrary
properties of the analogical target that are shared with the vehicle. For
example, the aim of a metaphor such as “He’s as busy as a bee” is to
highlight how busy the target is by comparing it (he, in this case) to an insect
known for being fast moving and productive. Persicke et al. (2012) taught
children with autism to solve novel metaphors (e.g., “This apple is candy”)
through induction of the following steps: (1) stating properties of the target
(i.e., apple—it’s fruit, grows on trees, tastes sweet), (2) stating properties of
the vehicle (i.e., candy—it’s food, tastes sweet, rots your teeth), and (3)
identifying the property shared between the two, thus identifying the
metaphor’s meaning (in this case, that the apple tastes sweet). RFT would
suggest that a number of relational performances were involved. Teaching the
children to relate each stimulus in the metaphor to its respective properties
might be seen as instances of equivalence or possibly hierarchical relating,
while comparing the inferred properties of the two stimuli in the metaphor
arguably consisted of both distinction relating (for nonidentical properties)
and coordinate relating (when the shared property was identified). All
participants learned to solve novel metaphors, and their responding
generalized to novel, untaught metaphors.
In conclusion of the discussion of analogy, in spite of the core
importance of analogical reasoning within higher cognition, there remain
limited instructional protocols to help establish the necessary skills. From an
RFT point of view, the core repertoire of analogy is the capacity to derive
relations between relations. As such, a protocol such as that implemented by
Carpentier et al. (2002) that targets equivalence-equivalence responding and
related composite skills may be useful in establishing the basis of analogical
abilities.
Hierarchy
The second complex repertoire under discussion in this chapter is
hierarchical relational responding, which RFT sees as underlying
responding in accordance with conceptual hierarchies (see figure 12.2). One
key example of conceptual hierarchies is hierarchical classification, in
which classes of stimuli are treated as members of larger classes (e.g.,
“poodle” is classified as a member of the category “dog,” while “dog” is
classified as a member of the category “animal,” and so on; see e.g., Griffee
& Dougher, 2002; Slattery, Stewart, & O’Hora, 2011). Another example is
hierarchical part-whole analysis, in which elements are treated as parts of
larger, more inclusive “wholes” (e.g., “nail” is categorized as part of
“finger,” while “finger” is categorized as part of “hand,” and so on; e.g.,
Slattery & Stewart, 2014; Stewart, Slattery, Chambers, & Dymond, 2018).
Responding in accordance with conceptual hierarchies is a
fundamentally important repertoire with regard to understanding and
navigating the world (Bornstein & Mash, 2010; Furrer & Younger, 2008;
Gelman, 1988; Kalish & Gelman, 1992; Lin & Murphy, 2001; Markman,
1989; Proffitt, Coley, & Medin, 2000). For example, a young child
sufficiently skilled in this repertoire who learns the category of a new person
in her environment can derive likely properties of that person, and compare
and contrast with other already known people and make predictions about
what to expect (e.g., “Jenny is a doctor like my father; that means she treats
people who are ill and probably works in a hospital, like he does; she
probably knows a lot about medicine and how to treat sick people”). This
type of responding can thus facilitate rapid learning and understanding.
Furthermore, it is critical for promulgation and promotion of key skills
needed in the educational arena, including, for example, logical and scientific
thinking. For instance, the understanding of taxonomies (e.g., within biology
or chemistry) is heavily dependent on hierarchical organization. As such,
hierarchical conceptual responding is a very valuable repertoire in
educational terms.
 Figure 12.2 Examples of conceptual hierarchies: left, classification
hierarchy; right, analytic hierarchy.

A substantial quantity of research on conceptual hierarchies has

previously been provided by mainstream cognitive-developmental
researchers (e.g., Blewitt, 1994; Deneault & Ricard, 2006; Greene, 1994;
Inhelder & Piaget, 1964). The main focus of such work has been hierarchical
classification, in which stimuli are organized into a hierarchy of classes
within classes. Theorists from this tradition see this phenomenon as involving
three core features, namely transitive class containment, asymmetrical class
containment, and unilateral property induction.
Transitive class containment refers to classifying a stimulus (A) as a
member of a higher-order class (C) on the basis that it is a member of a
subclass (B) that is a member of that higher-order class. For example, if a
child is taught that “poodle” (A) is a type of “dog” (B), then they may also
classify “poodle” as an “animal” (C) on the basis that “dog” (B) is a type of
animal. Asymmetrical class containment refers to the fact that a higher-
order class (e.g., “animal”) contains a lower-order class (e.g., “dog”) but not
vice versa (i.e., “dog” does not contain “animal”). Unilateral property
induction refers to the concept that properties or features of a higher-order
class (e.g., “animal”) will also be found in a lower-order class (e.g., “dog”)
but not vice versa. For example, all animals breathe and thus dogs breathe;
however, while dogs have four legs, not all animals do (e.g., Halford,
Andrews, & Jensen, 2002). Learning to respond in accordance with these
properties is seen as critical for being able to think and derive effectively,
especially in the educational arena. At the same time, while important,
hierarchical classification may be just one variety of conceptual hierarchical
responding. Markman and colleagues (e.g., Markman & Seibert, 1976), for
example, compared classification (class/member) type hierarchy with
analysis (part/whole) type hierarchy (involving the analysis of a stimulus
[e.g., a face] into its parts [e.g., eyes, nose, mouth]). They found that younger
children responded with higher accuracy to questions based on analysis
hierarchy than to those based on classification hierarchy; accordingly, the
authors argued that analysis hierarchy develops earlier than classification
hierarchy.
Until recently there has been relatively little research on hierarchical
responding by behavior analysts. The first such study was one on hierarchical
classification by Griffee and Dougher (2002), who modeled features of
responding in accordance with a natural language hierarchy (e.g., robin–bird–
animal). This work was an important first step in providing a bottom-up
model of hierarchical classification. However, Griffee and Dougher did not
explicitly test for the three features that cognitive developmental theorists
had suggested were important in hierarchical classification. A more recent
study by Slattery et al. (2011) tested the Griffee and Dougher (2002) model
for the presence of one of these features, namely, transitive class
containment, and found that only two out of eight participants showed this
pattern. Though Slattery et al. subsequently adapted the protocol such that
participants did begin to show behavior consistent with transitive class
containment, they recommended an alternative approach to modeling
hierarchical responding including classification using RFT.
Within RFT, hierarchical conceptual responding can be conceptualized
and modeled as a pattern of hierarchical relational framing. As with other
frames, the latter might be hypothesized to originate in the teaching of
nonarbitrary relations. For example, one such nonarbitrary relational pattern
that might be important is containment: A child might learn, in one context,
to describe things as being physically inside other things (e.g., “my hand is in
my glove”) and in another, to describe things as containing other things (e.g.,
“the house contains the doll”). This repertoire might then come under
contextual control (i.e., of cues such as the words “in” and “contains”) and
generalize, developing into more abstract patterns such as classification (i.e.,
responding to “members” as being contained in “classes”) or analysis (i.e.,
responding to “parts” as being contained within “wholes”). Accordingly,
hierarchical responding can be investigated by establishing arbitrary stimuli
as cues using nonarbitrary training and then using those cues to establish
hierarchical relations (e.g., “member/class”) between arbitrary stimuli and
gauging additional derived relations and/or properties.
To date, a number of studies have modeled hierarchical responding as
hierarchical relational framing. The first of these was conducted by Gil,
Luciano, Ruiz, and Valdivia-Salas (2012). In this study, nonarbitrary
relational training was used to establish contextual cues for both hierarchical
(“includes,” “belongs to”) as well as “same” and “difference” relations. In the
case of the hierarchical cues, for example, participants were exposed to two-
dimensional picture stimuli inside two-dimensional outline stimuli and were
taught to choose the former in the presence of the “belongs to” cue and to
select the latter in the presence of the “includes” cue. In later stages, these
cues were used to test expected transformation of functions of arbitrary
stimuli via hierarchical relations. As an example, the researchers taught
participants that stimulus X.1 had the property of being cold and that it
contained a number of other arbitrary stimuli including A1 and B1. They then
tested them for whether a stimulus C1, previously derived as being in the
same class as A1 and B1, was also cold, and the participants showed this
predicted transformation of functions. This study demonstrated a number of
empirical and methodological innovations, including (1) the establishment of
contextual cues for containment relations and (2) the demonstration of a
format in which instructors require learners to select contextual cues for
particular frames in order to probe the learners’ responding in accordance
with multiple stimulus relations (same, different, belongs to, includes). More
recently, Gil, Luciano, Ruiz, and Valdivia-Salas (2014) extended this initial
study by showing additional patterns of derived hierarchical relations and by
providing an improved set of controls over participants’ performance.
Both Gil et al. (2012) and Gil et al. (2014) modeled hierarchical
responding as broadly conceptualized. As indicated above, however,
mainstream research has suggested a number of varieties of hierarchical
responding, including hierarchical classification and hierarchical analysis,
and these varieties may have functionally important differences. Accordingly,
a number of more recent studies have focused on using an RFT approach to
investigate particular varieties of conceptual hierarchy.
Slattery and Stewart (2014) used relational framing to model
hierarchical classification specifically and to probe for the three core features
said to characterize it, namely, transitive class containment, asymmetrical
class containment, and unilateral property induction. In Phase 1 of their
study, they established arbitrary shapes as contextual cues for hierarchical
relational responding, similar to Gil et al. (2012, 2014). In contrast to the Gil
et al. studies, however, the core of this training involved a very simple set of
colored shape stimuli that might be grouped together only along particular
physical dimensions. This aimed to facilitate tight control over the
nonarbitrary relational pattern supporting the establishing of the contextual
cues, so as to facilitate hierarchical classification in particular, and more
specifically, by inducing relating of abstracted common physical properties
(“classes”) with examples of shapes that included those particular properties
(“members”). In Phase 2, the cues thus established were used to teach and
test a hierarchical relational network of nonsense syllables. Nine out of ten
participants who completed the protocol exhibited predicted patterns of
derived relational responding and transformation of functions in accordance
with all three of the properties of hierarchical classification by showing
asymmetrical mutual entailment, transitive combinatorial entailment, and
unidirectional transformation of functions.
Another more recent study has extended Slattery and Stewart’s (2014)
approach by using a similar two-phase protocol to demonstrate and
investigate hierarchical analysis rather than hierarchical classification and to
examine whether the properties of this pattern of relational framing would
differ from those implicated in hierarchical classification (Stewart et al.,
2018). The key difference between this study and the previous one was in
Phase 1, which established contextual cues. In Slattery and Stewart’s (2014)
study, the relations that were relevant in teaching the contextual cues were
between concepts based on the abstraction of common physical properties
(“classes”) and examples of shapes that included those particular properties
(“members”)—for instance, between the concept “green” and particular
shapes that were green. In the Stewart et al. (2018) study, in contrast, the
relations that were relevant in teaching the contextual cues were between
shapes made up of a number of different parts (“wholes”) and examples of
the parts themselves (“parts”)—for example, between a shape made up of a
red triangle, a yellow star, and a green circle (the “whole”) and a yellow star
(a “part”). As in the previous study, Phase 1 established the functions of
contextual cues, and then in Phase 2, these cues were used to teach and test
for derived relational responding and transformation of functions. Findings
showed a similar pattern of mutual and combinatorial entailment but, in
contrast with Slattery and Stewart’s (2014) results, the absence of any
consistent pattern of transformation of functions. The authors argued that
although in part this was predictable—because, in contrast with classification
hierarchy, analytic hierarchy is not associated with any particularly clear-cut
pattern of transformation of functions—further research would be useful for
testing ways in which particular patterns of function transformation might be
induced.
These examples of RFT work on hierarchical relational responding
show that verbally competent individuals engage in hierarchical relational
framing and that this form of behavior underlies patterns of responding in
accordance with conceptual hierarchies, including both hierarchical
classification and analysis. Recent research has extended this work in
directions that are of direct relevance to the current focus on the early
development and establishment of these repertoires.
One example is a study by Mulhern, Stewart, and McElwee (2017), who
measured patterns of relational framing linked with categorization in a large
sample of young (three- to eight-year-old) typically developing children. The
protocol correlated framing performance, including nonarbitrary and
arbitrary containment and arbitrary hierarchy, with linguistic and cognitive
potential as measured by standardized instruments. To assess nonarbitrary
containment, the researchers presented children with differently colored
boxes in which smaller boxes were physically contained inside larger boxes,
and asked the children whether one particular box was inside another or
contained the other. To assess arbitrary containment, they presented children
with a number of differently colored circles that were physically the same
size as each other. The researchers then told the children about one or more
“containment” relations between the circles and probed for the derivation of
further relations in the absence of any physical containment relations being
demonstrated (hence, these tasks probed for arbitrary relations). For example,
for mutual entailment tasks, they told the children that one particular colored
circle (e.g., the green one) contained another particular colored circle (e.g.,
the red one) and then tested for predicted derived responding (in this case,
that the red one was inside the green one). Finally, to assess arbitrary
hierarchy, the researchers presented children with real and nonsense words on
a computer screen, told them about one or more “hierarchical class” type
relations between the stimuli, and probed to see whether they could correctly
derive further relations. For example, for mutual entailment tasks, they told
the children that one particular nonsense word (e.g., “tol”) was a type of
animal and then tested to see whether they could endorse the correct
entailment relation (in this case, that the class “animal” contained “tols” as
members). As for the previous task, this was an arbitrary relational task
because the stimuli representing classes and members did not show any
physical relationship of relevance to the task. Results showed expected
variation in level of relational categorization skill based on age, as well as
strong correlations between relational performance and scoring on
standardized measures.
Another recent study, following on from the latter, used an adapted
version of the protocol used in Mulhern et al. (2017) to assess and teach
relational framing of categorization in young typically developing children
and to assess the impact of training on relevant measures of language and
categorization (Mulhern, Stewart & McElwee, 2018). In two experiments, the
researchers assessed and instructed children in mutual and combinatorial
entailment and transformation of functions of hierarchical relations. In
Experiment 1, they assessed and instructed five-year-olds in arbitrary
containment, while in Experiment 2, they assessed and instructed six-year-
olds in arbitrary hierarchy. Both experiments employed a multiple baseline
design across responses and participants and compared children receiving
instruction with controls. In both experiments, correct responding increased
to criterion levels on introduction of teaching, and both generalization and
maintenance were observed. Additionally, in both cases the instructional
group showed better performance than the controls on measures of language
and categorization. Such results suggest that hierarchical framing can be
systematically taught and has potential intellectual benefits, though additional
research using larger samples and conventional analysis is needed to confirm
the latter.
Apart from this work, researchers have also recently applied RFT to the
assessment and teaching of class inclusion responding, a sub-repertoire that
falls under the umbrella of hierarchical classification skill. In a typical
example of such a task, the researcher first shows a child an array of stimuli
in a particular class that includes two different subclasses, with a greater
quantity of one subclass than the other. They then typically ask the child
whether there are more members of the more populous subclass or more
members of the class. For example, they might show the child an array of red
and blue flowers with more red flowers than blue flowers and ask, “Are there
more red flowers or are there more flowers?” The aim of the task is to
ascertain if children can respond to a stimulus as simultaneously belonging
both to a class as well as to a subclass contained in it.
Ming, Mulhern, Stewart, Moran, and Bynum (2018) used an RFT
paradigm to assess and train this ability in three typically developing
preschoolers (Experiment 1) and three individuals with autism spectrum
disorders (Experiment 2). In each of the two experiments, the researchers
first assessed the three participants, who failed to demonstrate class
inclusion. They then gave them nonarbitrary relational instruction using a
non-concurrent multiple baseline design. More specifically, they showed the
participants how the two sets of stimuli used on a particular trial (e.g.,
pictures of cats and pictures of dogs) might be arranged into separate
transparent boxes representing each of the two classes, and then how both of
the boxes might themselves be put into a larger transparent box representing
the overarching class (i.e., animals). The rationale was that this
demonstration of the physical containment relations involved could support
establishment of the appropriate relational repertoire. In both experiments,
relational teaching successfully established the target repertoire, and
subsequent testing demonstrated both maintenance and generalization to
untrained stimulus sets.
In conclusion, with regard to conceptual hierarchy, as is the case with
analogical reasoning, there remain relatively limited instructional protocols to
help establish the necessary skills. We have described a number of
approaches based on recent empirical studies that might be employed in this
context. However, much further exploration and refinement of these
protocols is needed in order to further this work.

Implications and Future Directions

There are a number of topics that require further research related to the
current chapter. With regard to the area of analogy, discussed earlier in this
chapter, further work is warranted to explore the utility of a protocol that
targets equivalence-equivalence responding and related composite skills
with young children and individuals with developmental delays. Research
exploring generalization and extension beyond core abilities is also needed.
For example, once a core repertoire is in place, then MET using more
traditional picture-based tasks such as those used by cognitive
developmental psychologists might be deployed. Research improving the
functional utility of such tasks is also warranted (see, for example,
Carpentier, Smeets, Barnes-Holmes, & Stewart, 2004). In addition, in the
case of both analogies and metaphors, instructors might query underlying
formal or nonarbitrary properties during assessment and highlight these
properties during instruction, with a particular emphasis on these in the
teaching of metaphor.
Turning to conceptual hierarchy, the Mulhern et al. (2017, 2018)
protocols, which evaluated and established containment and hierarchical
relational responding, focused on hierarchical classification only while not
exploring hierarchical analysis. Furthermore, even in the exploration of
hierarchical classification, much further investigation is needed. For example,
the protocols used focused on contextual cues for hierarchical relational
framing per se (e.g., type of, contains) exclusively. This is reasonable in the
case of simpler frames such as coordination or comparison. However,
hierarchical relational framing is more complex than these frames in that
derivation in accordance with the full scope of hierarchy should implicate
other simpler frames also. For example, in addition to framing appropriately
in response to the cues of “type of” and “contains,” other important aspects
of this pattern might also be tested. When two stimuli are framed as being
part of the same class, then they should also both be framed as being different
from other stimuli framed within a different class and equivalent to each
other, in at least some contexts, independent of their physical properties.
Such a pattern of responding might be expected from someone with a
sufficiently advanced repertoire of hierarchical classification. The Mulhern et
al. protocols did not test for such relations; adjusting it so as to do so might
be expected to improve its reliability and validity as a test of classification.
This would be one useful direction for future work. Despite the fact that
further work is needed, however, the protocols described in some of the
studies discussed, including especially those focused on assessment and
teaching of hierarchical relational frames in children and individuals with
developmental delay, do indicate the potential of RFT as an approach to this
important repertoire.
Conclusion
The repertoires on which this chapter has focused, namely analogical
reasoning and conceptual hierarchical responding, are simultaneously
complex but also of central importance in terms of the development of
sophisticated thinking and reasoning skills. This chapter indicates that there
is potential in adopting an RFT approach to the assessment and instruction
of these skills. Key aspects of RFT exemplified in these studies include
nonarbitrary relational support, MET of relational patterns, and consistency
of contextual control. This work is starting to allow sophisticated behavior
analytic investigation of these protocols, which can capture key aspects of
the features of these phenomena. Although relational framing paradigms
have been discussed and outlined in traditional cognitive developmental
literature, developing a functional analytic conception of these repertoires
can readily facilitate practically oriented behavior change.

Study Questions
1. What are the core features of analogy as conceptualized by
relational frame theory?
2. What part do nonarbitrary features play in analogy? Give an
example of an analogy and describe the nonarbitrary features
involved.
3. List and briefly describe three ways in which researchers have
extended the findings of Barnes, Hegarty, and Smeets (1997).
4. What is the significance of the findings of Carpentier, Smeets, and
Barnes-Holmes (2002) with regard to the debate concerning the age
at which children become capable of analogical reasoning?
5. What are the three defining properties of hierarchical classification
as explored by cognitive developmental researchers?
6. How did Stewart, Slattery, Chambers, and Dymond (2018) extend
Slattery and Stewart’s (2014) protocol?
7. How did Mulhern, Stewart, and McElwee (2017) assess hierarchical
relational framing of categorization?
8. What is class inclusion? Briefly describe how Ming, Mulhern,
Stewart, Moran, and Bynum (2018) have approached class inclusion
using an RFT paradigm.
 Chapter 13:

Rule-Governed Behavior and Verbal

Regulation

Jonathan Tarbox
University of Southern California and FirstSteps for Kids
Vincent Campbell
University of Southern California and FirstSteps for Kids
Savannah Pio
University of Southern California and FirstSteps for Kids

Link and Preview

The present chapter focuses on an important topic in the behavior
analysis of language and cognition, rule-governed behavior. While
it may seem obvious that much of the behavior of humans is
influenced by rules (or “instructions”), there are a number of
interesting issues and implications related to this topic. This
chapter provides an overview of conceptual foundations, reviews
basic and applied research, and suggests several areas for future
research and practice in the area of RGB.

The demonstration of the power of direct-acting contingencies of

reinforcement to influence behavior is among the greatest contributions of
the science of behavior analysis. Indeed, the basic principles of behavior
analysis can largely be described as a collection of the ways in which
immediate environmental contingencies directly impact the behavior of
organisms. However, for humans with complex verbal behavior, the picture
is somewhat more complicated. B. F. Skinner (1969) made the observation
that talking about the future can affect one’s behavior in the present. In
particular, rules that describe contingencies can control behavior as if the
behavior had contacted the contingencies described, even though it never
has in the past. For example, this book might provide you with a web
address that links you to study materials on the Internet, along with a
statement that, if you study those materials (behavior), you are likely to get
a better grade on the exam (consequence). You might then follow that rule
by going to that web address and studying those materials, although you
have never engaged in that behavior at that website in the past and so you
never could have received reinforcement for it. Put simply, verbally
competent humans, starting around the age of three to five (Bentall, Lowe,
& Beasty, 1985), can work toward a verbally constructed future that they
have never had—and may never have—direct contact with. This is the
essence of rule-governed behavior (RGB).
Skinner (1974a) pointed out that RGB is critical to human civilization.
Aside from cases of observational learning (see chapter 8), without RGB,
every individual human would have to contact virtually every direct
consequence in order to learn. In some cases, this would be merely
inefficient, as in learning to cook without recipes. In other cases, this would
be downright deadly, as in learning to avoid poisonous animals and foods,
and so on. The earliest developments in agriculture likely involved RGB
behavior of some sort because it necessarily involved behaving with respect
to a verbally constructed future (e.g., planting today to prepare for harvest
time several months away) that was too delayed and infrequent to control
behavior directly. Even the development of a behavior as rudimentary as
building a hut today because it might rain next week likely involved some
participation of RGB, and engaging in behaviors such as these undoubtedly
contributed to the survival of the cultures that induced their members to do
so. But aside from immediately practical matters of survival and
convenience, RGB is critical to creating and maintaining all of the higher-
order pillars of civilization. Virtually all laws, for example, are rules that
describe behaviors to be avoided and the aversive consequences of failing to
do so (Skinner, 1989). The laws of science, as Skinner pointed out, are
essentially “rules for effective action” with respect to the natural world
(1974a, p. 259). Skinner (1969) explained that rules are largely how
knowledge is passed down from one generation to another, so that each
generation can learn from its elders, rather than only through trial and error.
Without RGB, humanity would quite literally have to reinvent the wheel
every generation.
Despite the critical role that RGB plays in civilization, rules can also
come to influence behavior in maladaptive ways. Control by rules often
overrides control by prevailing contingencies of reinforcement (Baron &
Galizio, 1983). For example, a participant in a laboratory experiment may
follow the rule “I need to press fast,” even though they continue to fail to
earn reinforcers on a differential reinforcement of low rate (DRL)
schedule, or a gambler may continue to follow the rule “I’m going to win; I
just need to play one more spin,” while losing all their money in a slot
machine. Because of the power of rules to influence behavior and because of
the ubiquity of rules in daily life, a thorough understanding of what rules are
and how they control behavior is likely critical to a comprehensive applied
science of behavior. We now turn to a discussion of what, exactly, RGB is.

Defining Rule-Governed Behavior

Rule governed behavior is behavior that occurs due to contact with an
antecedent rule and not due to prior contact with the contingencies the rule
describes (Skinner, 1969). RGB can be contrasted to contingency-shaped
behavior, which is behavior that occurs due to prior contact with
contingencies, not contact with rules. For example, if you are told,
“Pressing the lever slowly will earn you points,” and you then press the
lever slowly, that slow rate of lever pressing is likely rule governed,
assuming you have not participated in a similar experiment in the past. If
you had been given no instructions and you tried pressing the lever quickly
and received no points and then tried pressing the lever slowly and then
received points, subsequent slow pressing might be primarily contingency
shaped. Note that exactly the same topography of behavior (i.e., slow
pressing) may be rule governed or contingency shaped purely on the basis
of the history and function of the behavior.
Outside of the laboratory, behavior is rarely 100% contingency shaped
or 100% rule governed. People talk to themselves frequently about what they
are doing and are in constant contact with contingencies, so most behavior is
likely partially controlled by both rules and contingencies. A laboratory
example might be if the experimenter told you, “Try pressing fast or slow and
see what works,” and you then follow the rule and come in contact with the
contingencies and then continue to press slowly. In this case your slow
pressing might be partially controlled by the word “slow” in the
experimenter’s instructions, partially controlled by the points you received
for slow pressing, and perhaps partially controlled by a rule you constructed
yourself during that experience with the contingencies (e.g., “Aha, it works
better when I press slowly; the experimenter was telling the truth”).

What Is a Rule?
If RGB is defined as behavior that occurs due to contact with a rule,
then a scientific account requires a technical definition of a rule. Skinner at
various times described a rule as a discriminative stimulus (1969, p. 148), as
a “contingency specifying stimulus” (1969, p. 169), and as an antecedent that
altered the function of other stimuli (1957, p. 359). A rule cannot be a
discriminative stimulus because a discriminative stimulus is defined as a
stimulus that was present in the past when a behavior was reinforced, and
RGB, by definition, has not occurred or been reinforced in the presence of
that rule in the past. So it is likely more appropriate to state that rules affect
behavior as though they were discriminative stimuli. Stating that a rule is a
contingency-specifying stimulus gets more to the heart of what rules seem to
do—they tell the listener what contingencies are likely to bear on their
behavior. But this definition was problematic for decades because we did not
yet have a technical definition, in terms of behavioral principles, of what is
meant by specification (Hayes, 1991).

Contemporary Accounts of Rules and Rule-

Governed Behavior
The crux of the challenge with rules is that they have the function of
discriminative stimuli, even though they have never been directly established
as such. For the first several decades of research in behavior analysis, little or
no research was done that showed that a stimulus can come to have a
function other than that which was directly taught. In cases of stimulus
generalization, a stimulus comes to function as though it were a
discriminative stimulus even though the behavior was never reinforced in its
presence (e.g., someone tacts all examples of apples as “apple” after learning
to tact just a few apples as “apple”), but this is due to the physical similarity
between the teaching stimuli and the generalization stimuli (e.g., all apples
are round). Rules do not need to be physically similar to other rules that have
had the same function in the past. To demonstrate this easily, we could
present to you any random combination of action words (e.g., clap, jump,
smile), any random combination of temporal words (e.g., before, after, first,
last), and any random sequence of the words “do” and “don’t,” and if you
were attending carefully, you could follow the rule, even though you
certainly have never done that same sequence of actions in response to a
physically similar instruction in the past. In lay terms, you understand the
meaning of the novel rule, apart from responding to how it is physically
similar to rules you have followed in the past.
A functional analysis of meaning, in terms of behavioral principles, was
lacking for decades in behavioral research (Parrott, 1984b). Skinner (1957)
was clear that the meaning of a word was to be found in the contingencies of
reinforcement for the behavior of the speaker (or listener), but what are those
contingencies? Starting in the 1970s, Dr. Murray Sidman began research on
reading comprehension that later came to be called stimulus equivalence
(see chapter 10 for in-depth explanation of equivalence). When words
participate in equivalence relations with objects, an analysis of these relations
may comprise a functional analysis of the “meaning” of those words
(Catania, 1997) or of speaking with meaning and listening with
understanding (Parrott, 1984b). For example, an English-speaking child
learns to tact cars as “car,” and then derives the response of pointing to a car
when hearing the word “car” (i.e., listener response).
Equivalence research provided a partial explanation for how rules can
control behavior as though the person had contacted the contingencies
described in the rule, because the research showed how words come to have
the functions of other stimuli, as though the words were those stimuli
themselves. For example, the word “clap” can have some of the same
functions as seeing someone clapping (e.g., you yourself can clap when you
hear the word “clap”). However, rules do not simply provide words that are
equivalent to action. Rules also provide words that specify contingent
relations between antecedents and behaviors (e.g., “If it’s raining, take an
umbrella”), between behaviors and consequences (e.g., “Clean your room
and then you can go out and play”), and between antecedents, behaviors, and
consequences (e.g., “If mom is already mad [antecedent], and you ask her for
something [behavior], then she is going to say no and you won’t get to have
it [consequence]”). Relational frame theory (RFT; see chapter 11) provides
an account of how humans learn to respond to the relations between stimuli
in multiple ways, including conditional or causal relations. According to
RFT, the ability to respond to one stimulus as conditional, or contingent upon
another, is itself generalized operant behavior learned through multiple
exemplar instruction. For example, over perhaps hundreds or thousands of
exemplars of a child’s parents saying conditional statements that involve two
stimuli (e.g., “First eat your veggies, then you get dessert”), the child then
responds to new such statements as though she has contacted those
statements in the past, even though she has not. The antecedent stimuli that
are present that cue the contingent relation (e.g., “First/then,” “If/then,” “If
and only if”) become functionally substitutable for a stimulus that is
discriminative for the actual contingent relation so that the child can respond
to those cues as though he has contacted the relation they describe in the past,
even though they have not (Tarbox, Tarbox, & O’Hora, 2009).
Figure 13.1 depicts relations involved in a rule that describes an
antecedent, a behavior, and a consequence. Such a rule likely involves, at a
minimum, (1) an equivalence (or coordinative, in RFT terminology) relation
between the word for the antecedent and the actual antecedent, (2) an
equivalence relation between the word for the behavior and the actual
behavior, and (3) an equivalence relation between the word for the
consequence and the actual consequence, plus (4) a conditional relation
between all of the above, cued by the words “if” and “then” (Barnes-Holmes,
O’Hara, et al., 2001). The complexity of the contingency described by the
rule can be increased by cueing additional relations. For example, rules might
describe what not to do (i.e., frames of distinction), when to do something
(i.e., temporal frames), that consequences will be bigger or smaller than in
the past (i.e., comparative frames), and so on.
 Figure 13.1 Schematic of the minimum relations constituting a
complete rule that describes an antecedent, behavior, and
consequence.

Dimensions and Characteristics of Rules

The defining feature of a rule would seem to be that it functionally
describes a contingency to the listener’s contacting it, but a virtually endless
range of complexity is possible, presumably limited only by the complexity
of the verbal and relational repertoire of the listener contacting the rule.
Peláez (2013) proposed a framework for analyzing some of the key
dimensions of rules that are likely to impact whether a listener will follow the
rule: (1) explicitness, (2) accuracy, (3) complexity, (4) source, and (5) time
(immediate, delayed, or remote).
The explicitness of a rule refers to how clearly and completely the rule
describes the contingency. For example, “Clean your room by five o’clock or
you lose your cell phone” clearly describes the behavior, when it needs to
occur (the antecedent), and the consequence to be avoided. A less explicit
version of a similar rule might be “Get some cleaning done, OR ELSE.”
More explicit rules are more likely to be followed than less explicit rules.
The accuracy of a rule is the degree to which the rule describes the
actual contingency the listener will experience if they follow the rule. All
other things being equal, more accurate rules are probably more likely to be
followed, particularly if the listener has a history of contacting and following
other accurate rules from similar sources in the past.
 The complexity of a rule refers to how many variables and relations
between variables the rule entails. In the example above, a more complex
rule might be “Clean your room more than you cleaned it last time, by five
o’clock, or you lose your cell phone.” That rule is more complex because it
also entails the comparative relation more between the behavior to be
executed now and a behavior executed in the past. All other things being
equal, a less complex rule is probably more likely to be followed than a more
complex rule.
The source of the rule refers to who the speaker is. All other things
being equal, rules are more likely to be followed when they are delivered by
speakers who have a larger history of mediating reinforcement for following
rules they have delivered in the past. Listeners can also derive and speak their
own rules, in which case they act as both the speaker who delivers the rule
and the listener who listens to it and then follows it (or fails to). Individuals
who have a larger history of reinforcement for deriving and following their
own rules are probably more likely to follow their own rules, relative to those
have less of such a history.
The element of time refers to temporal components of the rule. For
example, a rule can describe a contingency with a long-delayed or an
immediate consequence for behavior (e.g., “If you smoke cigarettes today,
you will get cancer in twenty years” versus “If you drink poison, you will die
right now”). All else being equal, rules that describe more immediate
consequences are probably more likely to be followed than rules that describe
more delayed consequences. Another temporal component that could affect
the likelihood that a listener will follow the rule is the duration of time that
passes from the time the listener contacts the rule to the time the listener has
the opportunity to follow the rule. For example, rules that describe what to
today or tomorrow are probably more likely to be followed than rules that
describe what to do next week, month, or year. Peláez (2013) notes that little
research has systematically analyzed these components of rules and that
much future research is needed. We return to a brief discussion of these
parameters in the “Practice Recommendations” section later in the chapter.

Functional Units of Rule Following

Zettle and Hayes (1986) and Hayes, Zettle, and Rosenfarb (1989)
proposed three functionally distinct types of rule following, based on the
history that established each repertoire as well as how it functions in the
present moment: (1) pliance, (2) tracking, and (3) augmenting.
Pliance is a repertoire of rule following that was acquired due to a
history of socially mediated arbitrary reinforcement for following rules. This
repertoire is likely established when parents reinforce a child’s behavior of
doing what they are told by meting out positive and negative reinforcement
that is not related in any meaningful way to the natural consequences of the
child’s behavior. For example, “First take a bath and then I will read you a
book,” or “Finish your chores or I’ll take away video games.” The reason for
the term pliance,” which comes from the word “compliance,” is that this
repertoire is acquired through arbitrary reinforcement of compliance, per se.
Tracking is a repertoire of rule following that is acquired due to a
history of following rules and then coming into contact with the
contingencies that the rules describe. Examples include “Take a bath so you
are not smelly and your friends don’t make fun of you,” “Clean your room so
that the next time you want to find your toy, you will be able to find it
easily,” or “Put your shoes on before you go outside and play, so your feet
don’t get hurt by stepping on sharp rocks.” The relation between the behavior
and the reinforcer is the primary distinction between pliance and tracking. In
pliance, the relation is arbitrary—the consequence is simply whatever the
authority figure chooses to use to modify the listener’s behavior. In tracking,
the consequence is a natural consequence of the behavior. In lay terms,
pliance is more like obedience, whereas tracking is more like following
accurate advice. Research has shown that individuals may be more likely to
form and follow rules that describe actual contingencies, as opposed to just
doing what authority figures tell them to do, when authority figures exert less
of a presence through monitoring compliance (Donadeli and Strapasson,
2015). To the extent that behavior analysts want to help clients form and
follow rules that accurately help them achieve valued goals, this research
implies that it may be beneficial to sometimes help clients “figure things out
themselves,” rather than always directly telling them what to do.
Augmenting is a repertoire of rule-following in which the rule
transforms the function of stimuli for the listener. In particular, contact with
rules can transform previously neutral stimuli to have novel reinforcing
and/or discriminative properties. For example, when being trained by a
behavior analyst, a parent of a child with autism might contact the rule
“Doing extinction for your child’s tantrums is moving him closer to making
friends and succeeding in school.” Contact with this rule may help mitigate
the aversive functions of hearing their child’s crying. In addition, the sound
of their child crying during a tantrum may have a transformed discriminative-
like function of evoking increased effort on the part of the parent, as opposed
to avoidance. Augmenting seems to be particularly relevant to behavior that
involves what is often referred to as “values” in lay terms (Hayes, Barnes-
Holmes, & Roche, 2001). For example, the function of losing money, which
is clearly aversive, may be transformed into a source of positive
reinforcement via contact with a rule such as “If you donate money to this
charity, you will help treat children who have cancer.” Rules that control
augmenting behavior are referred to as “augmentals.” It is worth noting that
augmentals have much the same effect as motivating operations (Laraway,
Snycerski, Michael, & Poling, 2003), so it may be reasonable to think of
augmentals as verbally motivating operations made possible by derived
relational responding.
It is important to note that, although repertoires of pliance, tracking, and
augmenting likely develop in that respective order in typical child
development, the distinction is not intended as implying that one is better or
more adaptive than another. In any given person, if one of the repertoires of
rule following is excessively strong or rigid, it could have negative
consequences. For example, if one’s pliance repertoire is excessively strong
or rigid, they might only do what they are told, which could severely limit
their creativity, career advancement, social relationships, and so on. If one’s
tracking repertoire is overly strong or rigid, they might never comply with
authority figures, even when to do so is critically important (e.g., police,
supervisor at work). Finally, if one follows augmentals too rigidly, it can lead
to negative life outcomes. For example, the augmental “No pain, no gain”
can be useful when followed to a reasonable degree (e.g., moderate exercise)
but can be damaging when applied too rigidly (excessive exercise that
produces injury) or under contextual circumstances that lead to harm (e.g., in
a physically abusive intimate relationship).

Basic Research on Rule-Governed Behavior

In this section, we describe some of the classic basic studies on RGB and
some of the major findings, thereby laying the groundwork for our later
discussion of applied research and practice recommendations.
 Sensitivity
Among the earliest findings in basic research on RGB was that RGB
tends to be less sensitive to changes in contingencies, compared with
contingency-shaped behavior. Kaufman, Baron, and Kopp (1966) provided
participants with instructions that occasioned behavior that was appropriate
to reinforcement schedules but found the behavior was insensitive to changes
in those contingencies. Shimoff, Catania, and Matthews (1981) studied
schedule sensitivity in two groups of college students by giving one group
clear instructions on how to respond by pressing a lever on a DRL schedule.
The other group was given no specific instructions; instead their lever-
pressing behavior was gradually shaped to occur at a low rate on the DRL
schedule. The researchers then changed the reinforcement schedule to
reinforce high-rate responding, without giving any new instructions. The
lever pressing of the group that was not given explicit rules (e.g., whose
behavior was shaped by contingencies) increased in rate when the DRL
contingency was removed, whereas the group whose behavior was directly
instructed by explicit rules describing DRL did not. The results suggested
behavior that is primarily rule governed tends to be less sensitive to changes
in contingencies than behavior that is primarily contingency shaped, all other
things being equal.

Instructed Versus Derived Rules

A small amount of research has found that the origin of a person’s rules
may affect how sensitive those rules are to changes in environmental
contingencies. For example, participants who were directly instructed on
what rules to follow were less sensitive to unsignaled changes in
reinforcement schedules than participants who developed their own rules to
follow based on past direct contact with contingencies (Catania, Shimoff, &
Mathews, 1989). Hayes, Brownstein, Haas, and Greenway (1986) compared
sensitivity to contingencies in groups of participants who had been given
accurate rules, minimal rules, or partially accurate rules. When confronted
with changes in contingencies, participants who had been given accurate
rules were less sensitive to these changes than those who were given minimal
or only partially accurate rules. Put simply, when participants were given
accurate rules describing reinforcement contingencies and they apparently
behaved with respect to those contingencies, they actually were not behaving
sensitively to the contingencies because their behavior did not change when
the contingencies changed. Participants whose behavior was either mostly
contingency shaped or controlled by self-derived rules were far more
sensitive to changes in contingencies.

Rules Describing Behavior Versus

Contingencies
Matthews, Catania, and Shimoff (1985) conducted an experiment with
undergraduate college students responding on multiple random ratio (RR)
and random interval (RI) schedules of reinforcement. The researchers
taught participants to derive rules that described either behavior (e.g., “press
fast” or “press slow”) or contingencies (e.g., “The button works after a
random number of presses”). They then reversed the actual contingencies,
and participants who had been taught to create rules that described behavior
continued to follow their old rules, even when the rule was inaccurate.
However, participants who had been taught to derive rules that described
contingencies were more likely to adjust to the new contingencies. The
results suggest that rules that encourage attending to the environment, rather
than merely executing specific behaviors, may produce behavior that is more
sensitive to changes in the environment.

Applied Research on Rule-Governed Behavior

Despite the tremendous importance of RGB to human civilization, there has
been surprisingly little research conducted on applying what we know about
RGB to solving problems of social significance. However, what little
research currently exists has addressed a rather broad range of topics, and
we describe some of this research below.

Gambling
Problem gambling is an interesting behavioral phenomenon because
most games, particularly slot machines, by mathematical necessity, do not
allow players to win in the long term. While reinforcement is occasionally
delivered, if a gambler plays for long enough, it is absolutely certain that they
will lose money overall. It is also apparent that, although reinforcement
occurs occasionally, the much more frequent direct consequence of gambling
is money loss, which would presumably function as extinction or
punishment. The fact that most people are not problem gamblers may support
this analysis. However, researchers estimate that between 1% and 2% of the
US population meets diagnostic criteria for pathological gambling disorder,
while up to approximately 14% of the population engages in problem
gambling behaviors (Cunningham-Williams et al., 2005). Furthermore, the
effects of problem gambling can be severe, so it would seem that, for these
individuals, some variables other than direct consequences are likely
contributing to gambling.
Dixon (2000) conducted one of the first studies to apply an analysis of
RGB to gambling behavior. The study involved manipulating the delivery of
rules in a simulated roulette game in which participants were sometimes
allowed to place their own bets and the experimenter sometimes dictated
where bets would be placed. The dependent variable of interest was the
amount of chips wagered. Participants consistently wagered more chips when
they had control over chip placement versus when the experimenter did,
despite there being no difference in the probability of winning. When the
experimenter provided the participants with accurate rules (e.g., “I cannot
pick numbers that make you lose”), the difference in wagering between
participant-controlled versus experimenter-controlled trials decreased
substantially. The implications of this study are that gambling behavior may
be largely influenced by inaccurate rules that gamblers follow, rather than
solely by the reinforcement schedule entailed in the game. In addition, the
results imply that interventions that target these rules have the potential to
affect gambling behavior.

Public Health
Public health is an area in which RGB is implicated, almost by
definition. It is very rarely the case that members of the public receive
adaptive, immediate, direct-acting consequences for their health- and safety-
related behaviors. Indeed, the immediate consequences of many health- and
safety-related behaviors are maladaptive. For example, the immediate
consequences of smoking, taking off a helmet, having intercourse without
condoms, drinking alcohol, consuming illegal drugs, consuming high-calorie
processed foods, and so on, can be highly reinforcing. And since the vast
majority of typically developing adults do not have professional staff
following them during all hours, it is not possible to deliver professionally
mediated direct consequences for most behaviors related to public health.
RGB would seem particularly relevant to the area of public health, then, and
a few researchers have studied rules and/or analyzed how rules might be
involved in strengthening and weakening safe and healthy behaviors.
Jonah and Grant (1985) evaluated the effects of a Selective Traffic
Enforcement Program (STEP) in Ottowa, Canada, that involved intensive
public education about the benefits of seat belt use and the law requiring it, as
well as stepped-up enforcement of seat belt laws via traffic citations. The
program was theorized to work via two mechanisms: (1) increased
probability of punishment for driving without a seat belt, and (2) increased
awareness of this contingency, as well as of the greater safety that seat belts
afford. The results of the evaluation showed that seat belt use increased from
64% to 84%, that driver casualties declined by 14%, and that these results
were maintained two years after the intervention. For individual drivers who
received a speeding citation and then drove more slowly, one might attribute
the effects of this intervention to direct-acting punishment, not RGB.
However, for anyone whose behavior changed before receiving a citation, the
changed behavior was likely largely rule governed. Rules such as “I have to
wear a seat belt so I don’t get a ticket” and/or “I have to wear a seat belt so I
don’t get killed in a crash” plausibly helped mediate the effects of the public
safety campaign.
Mathews and Dix (1992) hypothesized that RGB may be relevant to
influencing how cartoonists depicted characters wearing seat belts in
nationally syndicated newspaper cartoons. In a personal letter to eight
cartoonists, the experimenters described the importance of drawing cartoon
characters wearing seat belts, for the purpose of modeling safe behaviors to
the public. The percentage of cartoon characters depicted wearing seat belts
increased from 15% to 41%, but the effects were inconsistent across
cartoonists. However, the cost of the intervention was essentially zero, and
the effect was to alter stimuli contacted by hundreds of thousands of people;
therefore, the cost/benefit ratio may have been high. The effects of writing
letters to cartoonists must be largely due to RGB. Several of the cartoonists
confirmed something of this sort when they wrote back to the researchers,
stating that they felt good about “doing the right thing.”

Higher Education
 A few studies have examined how rules may participate in policies
designed to have large-scale effects on student outcomes. Robertson and
Peláez (2016) studied the effects of a graduation success initiative (GSI) on
rate of retention and on-time graduation in undergraduate college students.
The GSI was a university-wide policy that required that students declare a
major upon entering school. It was hypothesized that doing so would set the
occasion for students to develop rules regarding which classes they needed to
take to graduate on time and behave accordingly. Overall, the GSI produced a
sixteen-point increase in on-time graduation.

Rules as Supplements to Behavior Intervention

Plans
It seems probable that adding rules to effective behavior intervention
plans could make those intervention plans work more rapidly, because less
direct contact with contingencies may then be needed. However, very little
research has systematically evaluated the added benefit of including rules in
behavior intervention plans. Watts, Wilder, Gregory, Leon, and Ditzian
(2013) compared differential reinforcement of other behavior (DRO) with
rules to DRO without rules for decreasing excessive engagement with high-
preferred toys in four children with autism. During DRO without rules, the
therapist delivered a reinforcer if the child did not engage with a preferred
toy for twenty seconds. In the DRO with rules condition, the therapist
informed the child that they would receive a reinforcer if they did not touch
the preferred toy for twenty seconds. The DRO with rules condition was
highly effective for all participants and decreased the behavior more rapidly
than the DRO without rules for three of the four participants. For the fourth
participant, the DRO without rules condition was not effective at all. It is
likely that many other applied behavior analysis studies have also included
descriptions of contingencies but did not include that information in the
methods section of the manuscripts and/or did not systematically evaluate
whether the inclusion of instructions was an effective component of the
intervention. Much more research is needed on the effects of rules on larger
behavior intervention plans.
Although adding rules to interventions may enhance effectiveness, it is
possible that it could have unwanted effects as well. Because rules can
decrease the sensitivity of behavior to contingencies, some researchers have
suggested that the use of rules as part of interventions should be considered
carefully (Hayes, Brownstein, Haas, & Greenway, 1986). Particularly when
the goal of an intervention is to strengthen a repertoire that should then be
sensitive to natural contingencies, it may not be optimal to use rules to
establish particular behaviors early on. For example, when teaching social
skills, one could use rules to establish particular topographies quickly (e.g.,
“Introduce yourself,” “Introduce a topic of conversation”), but those
behaviors may not then be sufficiently sensitive to actual social
contingencies. As a person matures and navigates different social contexts
and groups, what works and what doesn’t work will change from time to time
and context to context. If there are methods for establishing social skills that
primarily rely on contingencies, these might be preferred over reliance on
rules.

Establishing Repertoires of Rule Following

Most or all of the basic and applied research described above has
evaluated RGB in people who already demonstrate a repertoire of rule
following. Very little research has established repertoires of RGB in people
who did not already demonstrate it. Tarbox, Zuckerman, Bishop, Olive, and
O’Hora (2011) used an RFT analysis to design an intervention for
establishing the simplest repertoire that might reasonably be conceptualized
as rule following: accurately following novel rules that describe antecedents
and behaviors. The experimenters used multiple exemplar instruction to teach
children with autism, ages three to seven, to follow simple rules that
described mastered antecedent stimuli and mastered motor responses. For
example, therapists would hold up picture cards and present instructions such
as “Clap if this is a carrot” or “If this is a fire truck, then touch your head.”
During half of the trials, the described antecedent was present, and during the
other half it was not (randomly determined). Although the picture cards were
known (i.e., participants could respond accurately to them as listeners, in
other words, receptively) and the motor responses were known (i.e., the
participants could respond accurately to the motor response instructions), the
participants could not respond accurately to the if/then conditional statements
contained in the instructions. During baseline, participants almost exclusively
displayed the motor response described in the rule, regardless of whether the
antecedent described in the rule was present. Nontechnically speaking, the
participants did not “understand if/then conditionality” or they lacked “cause
and effect reasoning.” The experimenters taught participants multiple
exemplars of rules describing antecedents and behaviors until generalization
was observed across untaught rules. Results were variable, but all six
participants acquired the skill of responding to untaught rules describing
antecedents and behaviors by the end of the study. Although the study was
not designed to address pliance directly, the repertoire established in the
study might reasonably be analyzed as pliance because the reinforcers
included were never related directly to the behaviors described in the rules,
nor were the reinforcers described in the rules at all. Instead, the same
generalized reinforcers (e.g., food, praise, tokens) were used across all rules
that were presented.
In a follow-up study, Wymer, Tarbox, Beavers, and Tullis (2016)
extended multiple exemplar instruction to rules that described behaviors and
consequences. In the multiple exemplar teaching, a therapist presented rules
such as “If you do [behavior], then you get [consequence].” On half of the
trials, the described consequence was preferred (e.g., “If you clap hands, you
get a ball”), and on the other half, the described consequence was
nonpreferred (e.g., “If you stomp your feet, you get broccoli”). Preference
was determined at the outset of each session by a mini preference assessment.
All three participants rapidly acquired the generalized repertoire of following
untaught rules that described preferred and nonpreferred consequences.
Although this study was not designed to address tracking directly, the rule
following that the study established might be considered tracking because the
rules described behaviors and the consequences those behaviors would
produce if executed. Put simply, the children with autism were not given
reinforcers merely for doing what the experimenter told them to do, but
rather, were given whatever reinforcers they chose, based on whether they
executed the described behaviors or not. Taken together, the two studies
described above provide evidence that a basic repertoire comprising the
ability to follow antecedent-behavior and behavior-consequence rules is
teachable to children with autism via instruction with multiple exemplars.

When Rule-Governed Behavior Is Maladaptive

Despite the many advantages of RGB, there are potential hazards
associated with it as well. Researchers have pointed out that the rigidity that
rules can engender can be maladaptive (Hayes, Brownstein, Zettle,
Rosenfarb, & Korn, 1986). In other words, even when a rule occasions an
adaptive behavior (e.g., “I have to study to do well in school”), if it is
followed too rigidly, it may prevent the person from contacting the actual
contingencies, for example, overstudying may not actually produce a higher
grade and may decrease contact with important social or family relationships.
A small number of studies have attempted to directly address the relation
between overly rigid RGB and maladaptive behavioral outcomes.
Wulfert, Greenway, Farkas, Hayes, and Dougher (1994) interviewed
undergraduate college students using the Scale for Personality Rigidity and
divided participants into high-rigidity and low-rigidity groups based on their
scores on the scale. They then gave participants a lever-pressing task
consisting of a multiple DRL and fixed ratio (FR) schedule. After the
participants demonstrated stable responding, the examiners switched the
reinforcement schedule to extinction. Half of the participants in each group
received accurate instructions about extinction, whereas the other half
continued to receive instructions telling them to respond on an FR schedule.
Participants in the high-rigidity group continued to respond as though they
were on an FR schedule (they continued to follow the FR instructions),
whereas participants in the low-rigidity group demonstrated decreased rates
of lever pressing (their behavior was sensitive to extinction, as opposed to
continuing to follow an inaccurate rule). The implication of these results is
that individuals who are reported to “have higher rigidity” are likely to
engage in behavior that is more sensitive to control by rules and less sensitive
to the contingencies for their behavior. In other words, these individuals may
be more likely to follow old rules, even when those rules do not accurately
describe real contingencies and even when following those rules may not be
in their best interest.
McAuliffe, Hughes, and Barnes-Holmes (2014) compared rigidity of
rule following in participants with and without reported depressive
symptoms. In the first experiment, the researchers initially gave participants
accurate rules that described how to respond on high- or low-rate
reinforcement schedules, and participants in both groups responded
appropriately. However, when the researchers provided inaccurate rules later,
participants who reported higher depressive symptoms rigidly followed old
inaccurate rules, despite a lack of reinforcement. The behavior of participants
who reported lower depressive symptoms changed when the provided rules
were not effective, thereby showing greater sensitivity to contingencies.
 In the second experiment, McAuliffe et al. (2014) compared rules that
modeled pliance versus tracking. In the pliance condition, experimenters
asked participants to read the experimental instructions out loud, and then the
experimenters told the participants that they would be watching them
carefully and that they would check to see how many points they earned after
the experiment. In the tracking condition, experimenters asked the
participants to read the instructions silently, and the experimenters could not
see the instructions. In addition, experimenters did not indicate that they
would watch the participants; they merely told participants to earn as many
points as they could. In the pliance condition, participants with higher
depressive symptoms were significantly less sensitive to reinforcement
contingencies than were participants with less depressive symptoms.
However, in the tracking condition, participants with higher depressive
symptoms were indistinguishable from participants with lower depressive
symptoms. The results suggest that, even when a person has a history of
maladaptive, rigid rule adherence, their RGB may be amenable to disruption
through strengthening tracking behavior. The larger applied implication of
this study is that strengthening repertoires of tracking, as opposed to pliance,
may help establish more flexible, sensitive repertoires of behavior. These
repertoires could be beneficial to people facing new challenges for which old
rules have not prepared them, such as a parent with a newly diagnosed child
with autism, or a child with autism encountering new social contingencies in
high school or college.

Research on Say-Do Correspondence

Several decades of research have focused on making positive
reinforcement contingent on correspondence between what people say and
what they do. Space does not permit a comprehensive review of the
literature, but readers are encouraged to read Bevill-Davis, Clees, and Gast
(2004) for an in-depth review. In an early study, Risley and Hart (1968)
implemented a “do-say” teaching procedure in which they first gave
preschool children an opportunity to engage in a specific low-frequency play
behavior (e.g., playing with blocks and paint), referred to as the “do”
component, followed by an opportunity to verbally report which play
behaviors they engaged in (the “say” component). The researchers delivered
reinforcement contingent upon correspondence between doing and saying,
resulting in an increase in those low-frequency play behaviors. Furthermore,
these results were maintained, and correspondence was maintained even after
it was no longer required for reinforcement, suggesting the strengthening of a
generalized operant of accurately describing what one did (e.g., say-do
correspondence).
Similar results have been replicated across more socially significant
behaviors (Rogers-Warren & Baer, 1976), subsequent studies have replicated
the finding of generalization (Ward & Stare, 1990), and a small amount of
research has been done on investigating say-do correspondence as
generalized repertoires of RGB (Luciano, Heruzo, & Barnes-Holmes, 2001).
Overall, the literature on say-do correspondence suggests that relatively
young children (i.e., preschool age) can be taught to talk to themselves and
others about what they are going to do and that, when reinforcement is
provided for accurately predicting their behavior (or describing it after the
fact), increases in targeted nonverbal behavior can be produced. Furthermore,
the generalization observed in multiple studies seems to suggest that such
procedures may establish a generalized repertoire of say-do correspondence.
It seems plausible that such behavior may be closely related to, perhaps even
a subtype of, RGB, and future research should more explicitly investigate this
connection.

Practice Recommendations
The literature on RGB is still in its infancy, particularly in the applied
realm. We look forward to a more productive future of research in this area
and offer the following preliminary recommendations for practitioners.

Establishing Repertoires of Rule-Governed

Behavior
The studies by Tarbox and colleagues (2011) and Wymer and colleagues
(2016) provide a relatively straightforward framework for how the repertoire
consisting of the ability to follow novel rules can be taught to someone who
does not already display it: treat that ability like a generalized operant and
teach it through multiple exemplar instruction until generalization to untaught
exemplars is demonstrated. The following are practical tips for establishing
initial repertoires of rule following:
 1. Assess prerequisite skills. Learners should already accurately
follow a large variety of one-step instructions, should demonstrate a
large variety of auditory-visual conditional discriminations (i.e.,
“receptive labels” or listener behavior), and should likely already
display generalized derived symmetry (in RFT jargon, “generalized
derived mutually entailed relations of coordination”).
2. Start small. First teach rules that describe only antecedent-rule
relations or only behavior-consequence relations. After learners
demonstrate generalization to novel rules, teach the other repertoire.
After they demonstrate generalization to novel rules in the second
repertoire, consider teaching rules that describe relations between
antecedents, behaviors, and consequences.
3. Expand complexity. After learners can reliably follow novel rules
that describe antecedents, behaviors, and consequences, consider
adding in other relational operants. Consider gradually fading in less
probable consequences (i.e., not always fixed ratio one). Consider
gradually fading in delays between the presentation of the rule and
the antecedent described in the rule (i.e., opportunity to follow the
rule), and consider gradually fading in delays between the behavior
described in the rule and the consequence described in the rule.
4. Be flexible. Adjust how instructions are presented, what type of
reinforcers are used, how many exemplars are taught at once, and
any other relevant procedural details, in order to maximize the
learner’s response to teaching.

Consider Dimensional Characteristics of Rules

When intervening on a client’s RGB and/or when using rules as a
component of larger intervention packages, consider the dimensional
characteristics of the rules you are providing. As discussed earlier, Peláez
(2013) suggests that the following characteristics are likely to impact how
effective rules are: (1) explicitness, (2) accuracy, (3) complexity, (4) source
(who the speaker is), and (5) time (immediate, delayed, remote). For
example, when delivering rules that describe the contingencies of a
classroom-wide behavior management program in a school setting, to
maximize effectiveness, it might be wise to consider delivering those rules
clearly, honestly, simply, from a trusted source (e.g., teacher or principal who
has a history of delivering honest rules), frequently (not just at the beginning
of the semester), and—for rules that describe consequences—with a
relatively short delay (e.g., as opposed to consequences that occur only at the
end of a month or semester).
However, as learners are successful with rules that optimize the
dimensions described above, consider gradually fading them out. The real-
life settings in which clients live will rarely continue to deliver optimal rules,
and clients would be well served to learn to become sensitive to rules with
less optimal dimensions. As Peláez (2013) notes, little research has
systematically evaluated these characteristics of rules, but these dimensions
are likely a useful resource for designing and problem solving interventions
that contain rules.

Teaching Rules for Flexible Rule-Governed

Behavior
Little or no research has been done on how to establish rules that
optimize a client’s RGB repertoire, but some general practice
recommendations are worth discussing and should serve as directions for
future research. It seems likely that, all other things being equal, flexible rule
following and deriving is probably more adaptive than rigid rule following
and deriving. For example, in most challenging situations, a rule like “I need
to try different approaches and see what works” is probably going to be more
adaptive than “I have to get it right” or “I have to just do what I am supposed
to do.” Therefore, it might be worthwhile both to model these types of rules
during regular everyday interactions with clients and also to explicitly teach
them. It is possible that even varying the topography of rules from one
moment to the next might help foster rule flexibility. For example, rather than
always saying, “Try to finish your work quickly,” one might present a variety
of similar rules, such as “Work fast,” “Go like a racecar,” “Work like
lightening,” and so on. In addition, there is some evidence to suggest that
rules that are more general may engender behavior that is more flexible and
sensitive to contingencies, relative to rules that are more specific and
prescriptive (Hayes, Brownstein, Haas, & Greenway, 1986). For example,
when teaching conversational skills, a rule like “Try noticing what your
friends are into” might engender more flexible responding than “Ask your
friend what they want to talk about.”
Implications and Future Directions
Rule-governed behavior is an area of behavior analysis that is widely
acknowledged as important and yet, at the same time, largely ignored in
research and practice. This circumstance could be partially due to the
previous lack of a behavioral conceptual analysis of rules and RGB that
suggested a clear road ahead for studying RGB. Such a practical analysis is
now available in RFT, and research has started to very slowly progress. A
separate, highly productive line of research is that of acceptance and
commitment therapy (ACT). ACT is a behavior analytic treatment
approach designed to loosen the rigidity with which rules control behavior,
particularly rules that evoke responding toward shorter-term, smaller
negative reinforcement (e.g., avoidance of social anxiety in the moment) at
the cost of longer-term, larger positive reinforcement (e.g., meaningful
social relationships). To a very large extent, ACT works by disrupting
control of overt behaviors by maladaptive and rigidly held rules and
increases control both by less-verbal, direct-acting, short-term
contingencies as well as by verbally constructed, long-term positive
reinforcement contingencies referred to as “values.” Space does not permit
a full treatment of ACT here (see chapter 15 for more on ACT), but it is
worth noting that the ACT literature has largely developed separately from
the applied behavior analysis literature and so the potential of ACT work,
which is designed to make repertoires of RGB more flexible and adaptive,
has hardly been tapped in terms of its contributions to the mainstream ABA
literature. Future researchers should bridge this gap by dismantling package
interventions such as ACT and identifying the specific ways in which they
affect RGB, weakening maladaptive RGB and strengthening flexible and
adaptive repertoires of RGB.
Little or no previous research has investigated the role of RGB in
rigidity in autism, but a connection seems plausible. Rigidity might be
conceptualized as behavior that is relatively insensitive to environmental
contingencies, which is also a well-documented feature of RGB. Future
research might investigate how RGB correlates with rigidity in autism and, in
particular, whether the rigidity of RGB, itself, is related to overall rigidity in
people with autism. More important than mere correlations would be research
that attempts to loosen the rigidity of behavioral repertoires displayed by
people with autism. One fruitful direction for such research might be to
attempt to increase the sensitivity of behavior to contingencies, perhaps by
providing strong differential reinforcement for behaviors that contact
environmental contingencies versus those that follow rules. For example,
perhaps a player who follows the rules of a game might earn a small
preferred consequence (e.g., brief praise), whereas a player who invents new
rules to the game might earn a much larger reinforcer (e.g., more praise and
extended play time). A second and perhaps complimentary direction for
research in decreasing the rigidity of RGB in individuals with autism might
be to use rules to enhance the flexibility of the individual’s RGB itself. For
example, one might present rules such as “It’s fun to do things differently,”
“Let’s try a new way,” and “Let’s see what happens if…” while providing
strong positive reinforcement for stating, following, and ultimately deriving
novel rules of this sort. Much more research in this area is needed.

Conclusion
We hope that the conceptual analyses and practice recommendations
presented in this chapter will spur continued research into RGB, as well as
program development by practitioners, thereby contributing to a fuller
scientific understanding of human behavior. Such an understanding is likely
to be relevant to most repertoires of socially significant, verbally influenced
behavior, and would be especially relevant to the treatment of autism
spectrum disorders. Overly rigid rule following seems to be at the core of
many difficulties displayed by more verbal individuals with autism and
perhaps at the core of many of the challenges that parents of individuals
with autism face as well (Gould, Tarbox, & Coyne, 2017).

Study Questions
1. Rules that describe contingencies can influence behavior even when
the individual has never contacted the contingency. Explain this
with an example.
2. Comment on the benefits of RGB and explain how humanity would
have to “reinvent the wheel” every generation in the absence of
RGB.
3. How can RGB also contribute to maladaptive behavior?
 4. Provide an example that highlights the difference between rule-
governed and contingency-shaped behavior.
5. Why is it difficult to conceptualize a rule as a discriminative
stimulus?
6. How does equivalence research contribute to our understanding of
rules?
7. Provide an example that includes each of the five components
proposed by Peláez (2013).
8. How are the concepts of pliance, tracking, and augmenting useful in
the analysis of RGB?
9. Explain and comment on the results of the study by Shimoff,
Catania, and Matthews (1981). What is the primary implication of
these results?
10. Describe two of the studies in the section titled “Applied Research
on Rule-Governed Behavior.”
11. What are some areas for future research in the area of RGB,
behavioral rigidity, and autism?
12. What are the four recommendations for practice related to
establishing RGB?
13. What do the authors recommend in the section titled “Teaching
Rules for Flexible Rule-Governed Behavior”?
 Chapter 14:

Problem Solving

Thomas G. Szabo
Florida Institute of Technology

Link and Preview

The present chapter builds upon the previous chapter on rule-
governed behavior by focusing on a repertoire that is fundamental
to the daily lives of humans—solving problems. The chapter also
expands upon prior chapters on relational frame theory and
contrasts a behavior analytic approach with cognitive work in the
area. The author provides an overview of recent research on
problem solving and describes implications for practice.

Skinner’s operant analysis of problem solving provides a rich scaffold from

which to help learners develop skills for generating novel solutions in
situations where they lack an effective response. Despite Skinner’s
conceptual advances, many efforts to teach problem solving within a
behavior analytic framework do not veer far from nonbehavioral
approaches. Nevertheless, in recent years, some authors have begun
investigating ways to utilize the more intricate aspects of Skinner’s analysis
in applied studies with children. Relational frame theory offers additional
tools that can be used to teach problem-solving skills.
Problem solving is likely critical to everyday human functioning, given
that everyday situations rarely provide the perfect environment to support
behavioral repertoires exactly as they are first learned. For example, children
may need to derive new ways of solving simple problems during play,
adolescents may need to come up with novel solutions to complex social
conflicts, and adults must find creative solutions to family and work
challenges on a regular basis. Many of these situations seem to require a
person to execute a new behavior, which they were never directly taught to
do, often involving a novel combination of two or more previously learned
behaviors in composites that were never directly learned. Despite the
importance of problem solving, like many constructs, it is not clearly
understood and has not been addressed adequately in behavior analytic
conceptual or empirical work. Empirical evaluations of this process are rare
in the behavioral literature. Nevertheless, the rudiments of a conceptual
analysis were proposed by B. F. Skinner many decades ago, and a small
amount of behavioral research has been done on the topic.
This chapter reviews behavioral and nonbehavioral research on teaching
problem solving and related skills, including research using common-sense
approaches, and discusses the usefulness of extending Skinner’s examination
of problem solving with pragmatic verbal analysis, an approach derived from
the experimental analysis of human behavior. Additionally, the chapter
discusses pragmatic approaches informed by relational frame theory (RFT;
Hayes, Barnes-Holmes, & Roche, 2001) and concludes with
recommendations for practitioners.

Conceptual Approaches to Problem Solving

This section of the chapter reviews conceptual foundations for the study of
problem solving, including Skinner’s analysis and common cognitive and
constructivist approaches.

Skinner’s Analysis of Problem Solving

Skinner (1953) proposed that a problem is a situation in which an
organism in a state of deprivation or aversive stimulation has no immediately
available response that will reduce the deprivation or terminate the aversive
condition. The solution to the problem, according to Skinner, is the response
that reduces or terminates this nonoptimal condition. In turn, solving the
problem involves manipulating the variables of which behavior is a function
such that the solution can be emitted. Skinner’s interpretation of problem
solving involves two parsimonious steps: (1) an establishing operation is in
effect, and (2) variables are manipulated until a response that terminates the
establishing condition is produced. However, Skinner suggests that in many
cases, a third step will be needed in which supplementary stimuli in the form
of probes and prompts will be required to occasion the problem-solving
response. For example, one might talk to oneself about various potential
solutions, draw a diagram, write a list, and so on, all of which produce
supplementary stimuli, to which one might then respond with a successful
solution.

Cognitive and Constructivist Approaches to

Problem Solving
In contrast to Skinner’s focus upon manipulatable environmental
variables, constructivist accounts focus on the architecture of the mind
(Piaget, 1970), and cognitive accounts focus on information-processing
models of problem inputs that lead through various mental processes to a
solution output (Siegler, 1983). Reese (1994) reviewed fifty-seven applied
psychology or education papers dealing with problem solving from these
nonbehavioral traditions and identified seven shared steps described in
common sense terms by researchers. The sequence of steps Reese
documented throughout the constructivist and cognitive problem-solving
literature is as follows: Difficulty is Felt, Define the Problem, Gather
Information, Identify Possible Solutions, Select a Plan, Carry Out the Plan,
Test the Outcome, and Change the Plan.
Skinner’s two-step approach underscores the parsimony of easily
controlled antecedent and consequent variables as opposed to hypothesized
developmental levels at which the steps outlined above can be used or
cognitive maps inferred with which to make use of these steps. Yet despite
the elegance of Skinner’s interpretation, many applied behavior analytic
treatments of problem solving examined herein make use of a sequence of
steps that are conspicuously similar to those emerging from the cognitive and
constructivist perspectives (e.g., Agran, Blanchard, Wehmeyer, & Hughes,
2002; Bernard-Opitz, Sriram, & Nakhoda-Sapuan, 2001; Briscoe, Hoffman,
& Bailey, 1975; Cote et al., 2014; Hughes, 1992). Perhaps this is
unsurprising: there may be pragmatic utility to breaking down Skinner’s
composites into common-sense component repertoires. However, the applied
science of behavior analysis aims to be conceptually systematic (Baer, Wolf,
& Risley, 1968) for the purpose of identifying redundancies and
inconsequential variables. Thus, restating these steps in technical terms
would be a useful procedure for analysts interested in paring the steps down
to those that are necessary and sufficient. For example, the steps discussed
above could be recast as follows: Establishing Operation in Effect, Tact the
Problem, Arrange and Sample Multiple Discriminative Stimuli-
Reinforcement Probabilities, Environment Selects the Most Probabilistically
Effective Solution, Emit Behavior, Contact Reinforcers, and Environment
Refines the Selection of an Optimal Solution.
In some cases, tacting the problem may be unnecessary. In others,
arranging multiple three-term contingency matrices may be redundant
because only the putative establishing operation, discriminative stimulus, and
reinforcer are needed to evoke effective action. To date, no investigators have
conducted component analyses to determine the conditions under which any
of these steps are necessary or sufficient in and of themselves.

Behavioral Studies with Common-Sense Approaches

This section reviews behavioral research that follows a more traditional
approach to studying problem solving. The studies are grouped into four
subsections: a) teaching problem solving in unstructured community
settings, b) the self-determined learning model of instruction, c) using video
models to teach social problem solving, and d) teaching problem solving
strategies for complex categorization tasks.

Teaching Problem Solving in Unstructured

Community Settings
In an early community psychology study involving lower-income policy
board members, Briscoe et al. (1975) taught participants a three-step
procedure that included stating the problem, finding solutions to the problem,
and implementing the action to the solution. The experimenters employed
supplementary prompts in the form of cue cards and faded in the instructional
sequence. In accordance with Skinner’s suggestion that successful problem
solvers generate verbal stimuli to supplement other behavior in their
repertoire, they taught participants to vocalize several possible options and
the response to be made that would solve the problem at hand in upcoming
board meetings.
Problem and solution tacts as well as action behaviors increased during
teaching and during follow-ups. However, some limitations were noted. First,
participants did not maintain these skills at high levels. Second, follow-up
sessions were conducted with the experimenters present; thus, it is not clear
whether the problem-solving behaviors would have been maintained in the
absence of an outside observer. Third, the authors reported that although the
participants learned the skills they were taught, no data were collected as to
whether the skills used by participants led the board to successful
accomplishment of its objectives. Fourth, it is important to note that the
problem-solving repertoire Briscoe and colleagues taught was really a rule-
following repertoire. That is, participants followed the rule to generate
options, but there were no contingencies that brought these options under
stimulus control, and no attempt was made to assess the degree to which
problem solving that followed this pattern was maintained or increased in
frequency over time. Finally, although the procedure involved an establishing
operation and manipulating variables that facilitate the successful emission of
a solution response, the manipulated variables were preselected by the
investigators. Since it is unknown whether the solutions helped the board
accomplish its aims, it is not entirely certain whether the procedure led to
behavior under the control of relevant features of the environment. Taken as a
whole, this was an important example of community behavioral psychology
with specific, measurable behavioral outcomes, in contrast to frequently
reported mental health approaches with affective gains (e.g., Perkins &
Zimmerman, 1995). Nevertheless, the study offered only preliminary
evidence of the utility of a behavior analytic approach to problem solving, as
is likely the norm in a study that is perhaps the first in a particular area of
research.

The Self-Determined Learning Model of

Instruction
A number of investigators have employed multiple variants of the self-
determined learning model of instruction (SDLMI; Agran et al., 2002;
Cote et al., 2014; Hughes, 1992; Wehmeyer, Palmer, Agran, Mithaug, &
Martin, 2000). Common throughout these and other studies from this group
of researchers is the use of scripted problem-solving instruction delivered by
a teacher that follows a sequence predetermined by the investigator. Hughes
(1992) taught individuals with severe intellectual disabilities to verbalize five
statements: the problem, a generic class of solution response, a specific
solution behavior, the outcome, and a self-reinforcing statement. Wehmeyer
et al. (2000) and Agran et al. (2002) taught students with mild disabilities to
vocalize answers to the questions “What is my goal?” and “What can I do to
make this happen?” and, after emitting a response, to vocalize answers to
additional questions, “Did I meet my goal?” and “Do I need to adjust my goal
or plan?” Cote et al. (2014) taught elementary students with autism a
variation that included identifying problems, determining possible solutions,
choosing best solutions, emitting a response, and self-evaluating.
Each of these studies provided evidence that vocal rehearsal of a
problem-solving strategy in the presence of an establishing operation could
lead to successful utilization of the tactics involved in the strategy. In each of
these approaches, too, manipulation of variables with the supplemental
support of formal prompts and verbal probe stimuli proved useful. However,
as was the case in Briscoe et al. (1975), solving the problem did not result
from participants’ manipulating variables until the environment selected an
effective response. Rather, the experimenters taught a predetermined package
of steps, and only after the preplanned response was emitted did participants
evaluate the outcomes. Perhaps rather than problem solving in the Skinnerian
sense, the participants learned to follow rules describing particular scripted
sequences of behaviors. In these studies, problems were solved by the
investigators, who constructed external stimuli used to assist others. Skinner
described such approaches as “public products of problem solving” (Skinner,
1969, p. 139) that are useful in the accumulation and transmission of social
wisdom. As in the constructivist and cognitive models, these top-down
efforts solve the problem for others but do little to assess whether learners
have developed a problem-solving repertoire that they employ independently
in novel situations.
An additional limitation of this body of research is that the majority of
studies included problems that were perhaps not really problems in the
Skinnerian sense. In particular, most of the problems consisted of the
participants’ engaging in behaviors that were less problematic for the learners
themselves than for others (e.g., engaging in stereotypy in the classroom).
Skinner’s analysis of a problem requires that the solution to the problem be
reinforcing to the person solving the problem. The participants’ solving the
problem in the studies described above probably amounted more to
reinforcement for caregivers than for the participants themselves.

Using Video Models to Teach Social Problem

Solving
A slightly closer approximation to teaching what Skinner
conceptualized as problem solving is found in Bernard-Opitz et al. (2001). In
this study, investigators showed children with autism animated social
problems, each with a choice of possible solutions and an additional option to
produce novel, unprompted solutions. The experimenter showed and
discussed pictures presenting both appropriate and inappropriate responses,
prompted innovative solutions by a light and the question, “Do you have any
good ideas?” and reinforced novel answers. Probes in which prompts and
reinforcers were withheld showed gains, but the study did not assess
generalization to novel contexts outside of the laboratory. Additionally,
although the rate of appropriate solution selections increased, the authors did
not supply data on the rate of alternative solutions generated by learners
during probes. Thus, as in the previous studies, children were successfully
taught to select previously modeled appropriate answers to social conflict
questions, but they were not taught to manipulate variables that would make
it possible for them to emit an optimal response of their own.

Teaching Problem-Solving Strategies for

Complex Categorization Tasks
Skinner (1957) defined the intraverbal as a verbal operant that bears no
formal similarity to or point-to-point correspondence with the antecedent
verbal stimulus that evokes the response and is maintained by nonspecific
social reinforcement. Learning to categorize stimuli when asked to do so can
be conceptualized as intraverbal responding that can be hastened by using a
verbal problem-solving strategy. Sautter, LeBlanc, Jay, Goldsmith, and Carr
(2011) taught typically developing preschoolers a problem-solving strategy
for categorization that involved self-prompting with intraverbal chains. The
authors found that multiple tact instruction, intraverbal instruction, and
mediating-response teaching were necessary but insufficient for producing a
problem-solving repertoire. When learners were offered mediating-response
prompting (instructional rule-use prompts and pictorial-tact prompts),
intraverbal categorization increased and prompt fading was successful. An
important finding was that participant use of audible self-prompts, evident
initially, declined with use of the problem-solving strategy.
In a follow-up study, Kisamore, Carr, and LeBlanc (2011) examined one
of Palmer’s (1991) predictions that visual imagining could be taught as a
problem-solving strategy to facilitate intraverbal categorization. The authors
taught typically developing preschoolers the same prerequisite skills as
before (multiple tact instruction, intraverbal instruction) and followed this
with visual imagining instruction, visual imagining prompting, and visual
imagining prompting plus a rule during probes. In the final phase,
experimenters prompted participants with a rule statement prior to beginning
the phase. Rules such as “You can imagine [italics added] all the places
where animals go and tell me what you see” (Kisamore et al., 2011; p. 263)
decreased the number of problem-solving prompts needed and appeared to
augment visual imagining.
As in Sautter et al. (2011), teaching was necessary but insufficient.
Modeling and prompting generated learners’ self-prompting and self-probing
for supplemental stimuli that occasioned an imaginative response to solve the
problem. Skinner’s analysis of operant seeing, or “seeing in the absence of
any identifiable external support” (Skinner, 1953, p. 271), involves the
learner’s performing speaker and listener roles simultaneously and the
presence of a self-rule-following repertoire. The combination of these skills
might have led to automatic reinforcement, curtailing the need to acquire
social reinforcement by solving problems vocally aloud. Nevertheless, also as
in Sautter et al. (2011), it is unclear whether participants actually used the
strategy of visual imagining, since this is a covert and currently
unmeasurable behavioral repertoire.
To summarize the work reviewed thus far, applied behavioral literature
that explicitly harnesses Skinner’s account of verbal problem solving is
limited. Many studies appear on the surface to use Skinner’s analysis, but
upon close examination, they do not meet the inductive criteria that Skinner
established. Said differently, many studies provide top-down rules by which
instructor-defined solutions are generated in a lock-step manner when
learners follow a particular rubric. Although these approaches are grounded
in environmental determinants of behavior and are laudable because they do
not require an appeal to cognitive structures or innate information-processing
formulas, the extent to which they establish generative repertoires of problem
solving is unclear. The two studies attempting to teach problem-solving
strategies for complex categorization tasks are exceptions in that participants
in both experiments used problem-solving strategies to generate their own,
novel supplemental stimuli that evoked effective solution responses. It is
upon the shoulders of these studies that we now turn to RFT extensions of
Skinner’s analysis of verbal behavior and applied RFT problem-solving
studies.

Pragmatic Verbal Analysis

From an RFT lens, Skinner’s definition of problem solving, which “makes
the appearance of a solution more probable” (Skinner, 1953, p. 247), limits
the term to contexts in which there are obstacles to be avoided or removed.
Taken at face value, a rat that pushes a box until it is flush with a cabinet,
stands on the box, and pulls a chain to shift a bag of nuts on a counter until
it falls to the floor is engaging in problem solving. Although this could be
considered one aspect of problem solving, it is a noncomplex form that
involves no verbal behavior. Yet in Skinner’s conceptualization, it is the
intersection of contingency-shaped and rule-governed behaviors that will
lead to the most promising lines of applied research (Skinner, 1969), so this
same line of reasoning applied to human problem solving would likely be
fruitful.
Accounts of problem solving based on RFT analyze problem solving as
a verbal repertoire consisting of framing events relationally under the
antecedent and consequent control of an absence of effective action (Hayes,
Gifford, Townsend, & Barnes-Holmes, 2001). This repertoire involves
verbally relating stimuli in ways such that their functions are altered. Verbal
problem solving defined as such involves pragmatic verbal analysis. Hayes,
Gifford, et al. (2001, p. 90) define pragmatic verbal analysis as “framing
relationally under the control of abstracted features of the nonarbitrary
environment that are themselves framed relationally.” In other words, tacting
features, functions, and classes of stimulus events leads to tacting features,
functions, and classes of the tacts themselves. This categorization of one’s
verbal behavior is the procedure by which supplemental stimuli are generated
such that discriminated, generalized, and abstracted relations among arbitrary
and nonarbitrary stimuli can be brought to bear upon action. Effective verbal
problem solving, then, involves such tacting of one’s own tacts and conjoint
actions, which is perhaps what is meant by the common sense terms “self-
reflection” and “self-evaluation.” For example, given an establishing
operation of being unable to read without glasses and no eyewear at hand, an
individual might generate formal probes such as “Where have I been in the
last hour that I might find my glasses? I’ll check the closet, pantry, and
garage.” Tacting the locations where the individual has been in the past hour
is the first aspect of pragmatic verbal analysis. The second step is to tact
features, functions, or classes of these and other tacts that have been
successful in the past. To do this, the individual might generate additional
supplementary prompts, such as “These are all dark places (feature) where I
didn’t need glasses, so what are some places I’ve been recently where I might
have stopped to read something (function)?”
When verbal problem solving generalizes to novel contexts, it may be
because verbally abstracted physical dimensions of nonarbitrary stimulus
events come to serve as relational cues for the frame of coordination between
them. In other words, tacting a relation between two stimulus events cues
action with respect to that relation. A professor who tells himself, When I
read too many source materials, it’s like I’m flooring the gas pedal in my car;
I need to find ways to prepare lectures more efficiently is solving an
intellectual problem by comparing it to a physical one. Pragmatic verbal
analysis thus produces the context in which acting in accordance with
multiple relational frames (i.e., coordination, distinction, opposition,
hierarchy) between verbal and physical events become repertoires that
promote problem solving.

Strategic Versus Valuative Problem-Solving

From an RFT lens, it can be useful to differentiate strategic from
valuative problem solving. In strategic problem solving, novel solutions are
generated based upon previously successful ones and well-established
networks of verbal relations. The result is the formation of linear, step-like
problem-solving heuristics, such as those discussed earlier. For example, to
solve a problem, a learner might ask themself, What is the problem, what are
two solutions, and what are the pros and cons of each? Valuative problem
solving involves the use of relational framing in contexts where no previous
verbal history can be brought to bear. Therefore, valuative problem solving
results in a number of possible outcomes that can be contacted briefly, either
by metaphorical extension or by direct contingency sampling, until the
contingencies of reinforcement reveal and select the optimal solution. For
example, a writer composing verse about the swells of emotion may insert
lines using alliteration and rhyme into a poem and sample each one
repeatedly for its effect at bringing the listener’s attention to the equivalence
between cadences of ocean and emotion.
The important contribution of RFT to these two types of problem
solving is the categorization of framing repertoires (i.e., equivalence,
distinction, opposition, conditionality, temporality, locality, hierarchy, and
deixis). These repertoires have been well documented in at least 288
empirical studies to date (O’Connor, Farrell, Munnelly, & McHugh, 2017)
and are discussed at greater length in chapters 11 and 12. They are skill sets
that can be taught with multiple exemplars to improve the effectiveness and
efficiency of strategic problem-solving repertoires. In other words, relational
operants permit for an analysis of the broad and temporally drawn-out
behaviors described by Skinner into smaller, more easily analyzed and
potentially more easily taught units.

Reducing Delay Discounting

The ability to delay gratification is widely acknowledged to be an
important predictor to many socially relevant outcomes (e.g., Dixon, Buono,
& Belisle, 2016), and situations that demand self-controlled responses might
reasonably be conceptualized as situations that require problem-solving
behavior that makes impulsive responding less likely. A prediction from the
RFT literature is that the ability to delay responses is, in itself, a problem-
solving skill that can be acquired by multiple exposures to tasks involving
conditional framing. Given Skinner’s analysis, the problem is that the delay
to reinforcement is aversive, and the only available response that reduces the
aversive condition is an impulsive choice. An ideal solution would be a
response that both reduces the aversive stimulation and decreases reliance
upon the nonoptimal impulsive choice. Dixon and Holton (2009) evaluated
the RFT prediction that such conditions can be met with pragmatic verbal
analysis instruction. The authors asked five individuals with gambling
disorder to complete a relational responding task involving learning
conditional discriminations aimed at transforming the functions of irrelevant
stimuli to alter subsequent delay discounting. Arbitrary color stimuli were
related to quantitative gambling stimuli in conditional discrimination
instruction involving worse than/better than relations. Following teaching and
testing, discounting occurred less often; that is, the participants responded to
arbitrary color stimuli correlated with larger, later payoffs more than to those
correlated with smaller, sooner rewards. This study provides very preliminary
support to the possibility that a pragmatic verbal analysis approach to
instruction can reduce impulsive responding to the present situation due to
the intrusion of stimulus functions transformed via language and due to the
increased salience of larger but temporally remote reinforcers after relational
instruction. Put more simply, gamblers’ impulsive behavior in this study
seemed to lessen as they were taught to “reflect back” on poorer and better
choices made in the past.

Teaching Metaphorical Reasoning

Metaphorical language is a particularly strange form of verbal behavior
because it necessarily involves calling a thing something other than what it
literally is. When considered from a behavioral perspective, a speaker might
tact a nonverbal stimulus metaphorically by labeling it with a label that is
conventionally reserved for tacting a completely different stimulus. For
example, a speaker might tact a particularly strong person as a “horse.”
Depending on the history of the listener with respect to metaphorical verbal
behavior, it could be quite challenging to produce a socially successful
listener response. An RFT analysis might prove useful for addressing the
behavior of the listener. Specifically, effective metaphorical “reasoning” by
the listener in such cases may be a skill that will reduce an establishing
operation present when an individual is unable to comprehend metaphor used
in typical conversation. In order to successfully “think through” a novel
metaphor uttered by a speaker, a listener may manipulate verbal variables in
accordance with coordination, distinction, and hierarchy frames to solve the
problem of how to respond intraverbally to speaker metaphors.
Based on the analysis described above, Persicke, Tarbox, Ranick, and
St. Clair (2012) taught children with autism to attend to relevant contextual
cues surrounding a metaphor and to engage in reciprocal relational
responding to respond correctly to metaphorical questions. The metaphorical
questions involved asking the participants to explain the meaning of
metaphors that combined two seemingly unrelated stimuli by identifying the
shared feature of the two stimuli. To do this, the experimenters prompted the
participants to talk to themselves about the features of one stimulus (i.e.,
hierarchical relating) and then the features of the other stimulus (hierarchical
relating), and then identify which feature was the same between the two
stimuli (coordinative relating), thereby providing the answer to the
“meaning” of the metaphor. For example, participants might respond to the
question “If I told you that the snow was a blanket, what would I mean by
that?” with something like “Snow is cold, white, and it covers things, and
blankets are warm, they cover you, and they belong on your bed—you must
mean that the snow was covering things!” The experimenters instructed the
participants on multiple exemplars of metaphors until they demonstrated
accurate responding on the first trial of new metaphors that had never before
been taught. Two children demonstrated generalized accurate responding
after multiple exemplar instruction plus visual prompts (which were faded
out), and a third reached mastery criteria after multiple exemplar instruction
alone. That the findings generalized to untaught metaphors demonstrates that
metaphorical reasoning may be taught to children with language deficits and
that once learned, it may be a flexible, generally applicable skill that solves
the problem inherent in an aversive condition when a speaker is using an
unfamiliar turn of phrase. This strategy for teaching metaphorical reasoning
is a kind of pragmatic verbal analysis that facilitates a generalized problem-
solving repertoire. That is, participants produced supplemental stimuli to
categorize arbitrary and nonarbitrary events in accordance with relational
frames of hierarchy, coordination, and distinction and thus solved the
problem of how to respond to metaphor.
In summary, research informed by an RFT perspective on problem
solving as pragmatic verbal analysis is still in its infancy, but the initial
results are promising and consistent across studies. The general findings
would appear to be that one’s own behavior, one’s current circumstances, and
one’s verbal construction of near future circumstances are all environmental
variables to which one can be taught to respond verbally. Put another way,
the ability to talk to oneself about one’s own behavior, one’s current
circumstances, and which behaviors are likely to produce preferred future
circumstances seems to be a complex repertoire of relational verbal behavior
that is teachable via multiple exemplar instruction.

Relevance of Rule Deriving

Chapter 13 addressed rule-governed behavior (RGB), and rule following
in particular, to a thorough degree, and that will not be repeated here.
However, rule deriving may be particularly relevant to problem solving and
warrants discussion. Rule deriving is a repertoire of behavior consisting of
creating rules that describe potential future behaviors and future
environmental circumstances and then responding with respect to the rules
one just derived. It differs from rule following in that rule following
generally entails the delivery of rules by someone else (often an authority
figure), whereas rule deriving entails articulating novel rules oneself. Rule
deriving fits well in Skinner’s analysis of problem solving in that it likely
comprises some of the behavior that Skinner posited to occur after the
appearance of a problem and before the appearance of the solution. That is,
some of what one likely does to modify one’s own environment such that a
solution is available is talk to oneself about how one might modify one’s
environment and what the outcomes might be. For example, if a child’s
favorite toy is not working (i.e., problem) and the child then says to
themself, “Maybe I could check the batteries, or maybe I could ask an adult,
or maybe it’s just broken—I’ll check the batteries!” (problem solving
behavior), and then checks the batteries and replaces them (solution) and
the toy then works (reinforcer), then it would appear that rule deriving may
have played an important role in the chain of problem-solving behaviors.
It is also worth noting that rule-deriving behavior of this sort might
provide a behavior analytic way to analyze and teach some of the stages of
problem solving proposed by cognitive researchers, including “identify
possible solutions and select a plan,” as socially meaningful overt behaviors,
as opposed to hypothetical constructs. In addition, the RFT analysis of
pragmatic verbal analysis described earlier also fits well with an analysis in
terms of rule deriving because rule deriving is likely an important part of the
self-directed verbal relating that characterizes that repertoire. For all of these
reasons, it would appear that rule deriving could be a useful area for further
research and conceptual analyses into problem solving.
Unfortunately, very little research has been done on rule deriving, so it
is not possible to say with any degree of certainty how the repertoire is
acquired or how relevant it is to problem solving. In a recent study on rule
deriving, Monclus, Uribio, and Szabo (2016) taught three children with
autism (five, seven, and fourteen years of age) to derive their own rules for
solving social behavior problems after viewing multiple video models of
related situations. To occasion rule formation, the models turned away from
the problem, pressed a red button, and looked into the camera as they
vocalized options and selected a response. They then pressed the button again
and emitted the solution in context. The experimenters then exposed the
participants to situations that resembled the problems modeled in videos and
instructed them to press a red button, turn to face the experimenter, and
derive an if/then statement about the contingencies. All three participants
responded effectively to 80% or more of the situations they were exposed to,
although two needed a supplemental prompt before responding in at least one
of the probes. To assess whether nonverbal behavior (such as handing a toy
to a sibling) was under the control of covert rules and whether verbal reports
were functionally equivalent to the covert rules, the authors conducted three
control procedures known as the silent dog method (Arntzen, Halstradtro, &
Halstradtro, 2009; Taylor & O’Reilly, 1997). In the first control, duration
from instruction to completion was measured in three novel problems. The
second control involved measuring response duration when self-talk was
interrupted by another speaker. In the third control, participants taught a
confederate to solve the problem using the same rules that they had derived
when faced with a similar situation. One participant, a seven-year-old girl,
successfully solved all problems and, based on results obtained during the
silent dog control procedure, appeared to have done so using self-generated
rules.
The types of problems solved by participants in this experiment have
applied significance. In addition, pragmatic verbal analysis procedures used
to teach problem solving resulted, in some cases, in a generalized repertoire
of rule deriving. This is one of the few studies that have attempted to analyze
the relationship between covert and overt problem solving and the aspects of
pragmatic verbal analysis necessary for establishing this repertoire. More
work in this area is needed, yet there seems to be great potential for
generating a range of procedures for teaching self-directed verbal relating
and precise protocols for teaching rule deriving in context.

Implications and Future Directions

The literature reviewed in this chapter is a promising start to the behavior
analytic study of problem solving, but it is probably fair to say that the field
of applied behavior analysis is far from conclusively answering how
problem-solving skills can be taught. Accordingly, there are a number of
implications that can be drawn with regard to both research and practice.

Research Directions
The topic of problem solving presents a number of opportunities for
future research and practice. Few studies have examined problem solving
from the perspective of pragmatic verbal analysis as described in this chapter.
Given this, it seems that entire lines of research are needed to explore this
area. In addition, more research is needed in the area of rule deriving,
including further conceptual analysis. This work not only is conceptually
interesting but has large social implications given the importance of problem
solving in day-to-day life.

Practice Guidelines
Despite the limited empirical data on the most effective methods to
teach problem solving from an applied behavior analytic perspective, clients
receiving applied behavior analytic services, across a wide variety of
populations and diagnoses, would benefit from enhanced problem-solving
skills. Therefore, we believe it is prudent for practitioners to address such
deficits when needed (also see Najdowski, 2017, for a practitioner’s manual
that includes guidance on teaching problem solving). Based on conceptual
analyses from Skinner (1953, 1969), Hayes, Gifford, et al. (2001), and the
empirical literature reviewed herein, I offer the following recommendations
to guide ABA practice in developing problem-solving repertoires:
1. Identify relevant establishing operations and controlling variables of
the desired problem-solving behavior.
2. Provide supplemental stimuli (prompts and probes, both formal and
thematic) to aid discrimination. Fade these stimuli when learners
can respond independently to desired controlling variables.
3. Teach learners to vocally generate their own prompts and probes.
When learners vocalize their problem solving, teachers can shape
both verbal and nonverbal responses. The aim is to improve the
probability that learners will learn to manipulate only those
variables most relevant to their immediate context.
4. After problem-solving skills are acquired, teach learners to reduce
the magnitude of their self-prompts and probes. This is important for
at least two reasons. First, many learning environments do not
tolerate vocal self-talk. Second, when problem solving is
competitive, inscrutability is advantageous.
5. Scripted problem-solving instructions serve as strategic
supplements. These stimuli can be introduced in various forms so
that behavior develops contextual sensitivity to the necessary and
sufficient script components for each new problem scenario. The
aim of varying these supplemental stimuli is to induce behavioral
flexibility (response generalization). Therefore, fade out scripts as
quickly as possible and consider using variable multiple exemplars
of scripts, as opposed to only a single script format.
6. Specific elements of traditional problem-solving scripts may or may
not be relevant in a given context. For example, emitting a self-
reinforcing statement (Hughes, 1992) could be beneficial in
situations where solutions do not contact reinforcement
immediately, but not in other circumstances. Another aspect of
scripts is what has been termed “strategic problem solving” (Hayes,
Gifford, et al., 2001). One type of strategic script is systematic
evaluation of the outcomes of problem-solving strategies, which
seems to be crucial for future refinement of and generation of new
problem-solving strategies.
7. Program for and reinforce stimulus generalization. That is, reinforce
the use of problem-solving strategies across novel environments to
solidify and contextualize the problem-solving repertoire.
8. Teach learners to respond in accordance with coordination,
distinction, opposition, hierarchy, temporal, location, and numerical
relational frames. Pragmatic verbal analysis may add appetitive
functions to otherwise aversive problem situations. Practice of this
kind has also been shown to increase the salience of larger,
temporally remote reinforcers.
9. To improve existing problem-solving repertoires and improve the
control exerted by the environment, teach learners to self-prompt in
accordance with coordination, distinction, opposition, hierarchy,
temporal, location, and numerical relational frames.
10. Foster the development of a pragmatic verbal analysis repertoire by
asking learners to tact verbally abstracted physical dimensions of
nonarbitrary stimuli and to tact their tacts in relational classes. For
example, teach children to experiment to see whether baking soda or
sugar produces a livelier volcano; then ask them to experiment to
see whether smiling or frowning produces livelier conversations.
Over multiple exemplars, tacts regarding the relationship between
abstracted properties of stimuli will serve as relational cues for
behavior in accordance with frames applicable to the context. In the
example above, a successful intervention might result in a child’s
telling you later that he is experimenting to see whether looking in
peoples’ eyes or looking away results in more people offering to
play with him.
11. Where possible, create problem-solving contexts in which learners
contact proprioceptive feedback. Problem solving can be taught in
physicalized contexts and later prompted in more abstract, verbally
demanding situations. As in other learning scenarios, establish
contingencies for learners to vocalize physical problem-solving
procedures during learning activities and diminish the magnitude of
their vocalizations during performances. Fostering the dynamic
interplay of rule governance and contingency shaping in this manner
can lead to flexible, generally applicable problem-solving behavior.
12. Problem solving, as analyzed here, involves self-directed verbal
behavior and evaluation. When acquiring these skills, some learners
might emit overly critical or negative self-directed verbal behavior.
Avoid establishing excessive or overly harsh evaluation of one’s
own behavior (e.g., “I am stupid,” “I always mess up”). If you hear
responses such as these, express empathy. Let the learner know that
solving problems is hard and that you make mistakes too. Encourage
learners to be “kind” to themselves by saying things such as, “I
caught my mistakes and now I have a solution that works!”
13. As with everything we do in applied behavior analysis, use powerful
positive reinforcement and make learning fun. In early stages of
instruction, use prompts that are effective enough to prevent
repeated errors, thereby preventing teaching from becoming
aversive.
Conclusion
The present chapter provided an overview of the important topic of problem
solving. Traditional behavior analytic approaches were considered relative
to those informed by cognitive and constructivist foundations. After
reviewing behavioral studies derived from more traditional foundations, an
analysis of problem solving based on relational frame theory—pragmatic
verbal analysis—was reviewed. A summary of select research related to this
conceptualization was provided, as was an analysis of rule deriving in the
context of problem solving. The chapter concluded with implications for
further research and guidelines for practice.

Study Questions
1. Why is problem solving a socially important behavioral repertoire?
2. How does Skinner define a problem?
3. Provide an example of a problem using Skinner’s conceptualization.
4. What are the two steps involved in Skinner’s interpretation of
problem solving?
5. List the cognitive steps to problem solving as well as their behavior
analytic translation.
6. What was the primary shortcoming of the study by Briscoe,
Hoffman, and Bailey (1975)?
7. Comment on the research described in the section “Teaching
Problem-Solving Strategies for Complex Categorization Tasks”; be
sure to include a strength and weakness of this line of research.
8. Distinguish between the RFT emphases on strategic and valuative
problem solving.
9. Describe one of the studies in the section on reducing delay
discounting or teaching metaphorical reasoning.
10. What is rule deriving and how does it pertain to the understanding
of problem solving?
11. Comment on two of the recommendations for practice. Explain why
these two recommendations stand out to you as especially
interesting and important.
 Chapter 15:

Second- and Third-Wave Behavior Therapy

Emily K. Sandoz
Department of Psychology, University of Louisiana at Lafayette
Gina Q. Boullion
Department of Psychology, University of Mississippi
D. Owen Rachal
Department of Psychology, University of Louisiana at Lafayette

Link and Preview

This chapter extends the analyses of verbal and nonverbal
behavior that compose the body of this book into a greater degree
of complexity and a wider scope—specifically, those involved in
the second- and third-wave behavior therapies. The content of this
chapter consists of behavior analytic treatment approaches that
are generally applied outside of the scope of practice of
mainstream behavior analysts. With a focus on the variety of
complexity of the typically developing human behavioral
repertoires, this chapter presents the possibility of extending
mainstream behavior analysis both scientifically and practically. In
particular, the chapter describes acceptance and commitment
therapy, a contemporary behavior analytic third-wave therapy, and
addresses the implications for the science and practice of applied
behavior analysis.
Behavior analysis has long been challenged with two related issues: (1) the
role of private events in the analysis of human behavior, and (2) the
application of behavior analysis for the adult outpatient population (e.g.,
Anderson, Hawkins, & Scotti, 1998; Breger & McGaugh, 1965; Golden,
2009; Wolpe, 1978; Wulfert, 2002; Wynne, 2004). These challenges are
particularly salient when considering behavior therapy from a behavior
analytic perspective, as individuals seek behavior therapy due to concerns
about private events and their impact on behavior, concerns that are relevant
for many of the clinical populations that behavior analysts serve.

Behavior Analysis and Private Events

Despite the common impression that behaviorists somehow deny the
occurrence of behaviors like thinking and feeling (e.g., Shevrin & Dickman,
1980), the issue of private events in behavior analysis has received much
direct and consistent attention in the literature. In 1945, Skinner proposed
private events as part of the stimulating conditions that might occasion
operant behavior. By 1969, he had clarified the role of private events in his
behavioral analysis, including them as part of the behavior being analyzed
(Skinner, 1969; pp. 221–268). In fact, for Skinner, the inclusion of private
events was an important distinction between methodological and radical
behaviorism (Skinner, 1953). Over the years, many behavior analysts
clarified (e.g., Branch & Malagodi, 1980; Creel, 1980; Moore, 1980) or
extended (e.g., Zuriff, 1979) Skinner’s work on private events, while others
offered distinct alternative positions on private events from interbehavioral
(e.g., Parrott, 1983, 1986) and molar (e.g., Rachlin, 1984) perspectives.
Behavior analysts’ positions on the role of private events in radical
behaviorism still vary widely with respect to (1) the relevance of private
events to a behavioral analysis and (2) the causal status of private events.
Many accept Skinner’s position, discussing private events not as causes but
as still relevant as private behavior, private physiological stimuli, or
descriptors of a likelihood of engagement in a particular behavior (see
Moore, 2009). Some (e.g., Baum, 2011a, 2011b; Rachlin, 2011) discard
private events as irrelevant to the scientific analysis of behavior. Some (e.g.,
Palmer, 2009, 2011) emphasize the role of private events in scientific
interpretation, distinct from analysis but necessary for maintaining a
comprehensive science. Others (e.g., Fryling & Hayes, 2015) dispute the
distinction of some events as “private,” instead positing all psychological
events as wholly unique and in-kind observable. And still others (e.g.,
Dougher, 2013) point to the workability of the inclusion of private events in
behavior analysis as it is clinically applied in behavior therapy.

Clinical Behavior Analysis and the Rise of Behavior

Therapy
Less consistent attention has been given to clinical behavior analysis—the
application of behavior analysis to serve adults seeking to address
psychological difficulties in outpatient treatment (see Kohlenberg, Tsai, &
Dougher, 1993; c.f., Kohlenberg, Bolling, Kanter, & Parker, 2002).
Behavior analysts have long been interested in the kinds of difficulties that
typically bring adults into outpatient treatment (e.g., anxiety, depression, or
other struggles categorized as “psychopathology”). This is evident in early
conceptual behavioral accounts of psychopathology. For example, Skinner
(1953) rejected the disease model of psychopathology, describing
disordered behaviors in terms of the reinforcing context. From the Hullian
tradition, Dollard and Miller (1950) provided a book-length account of
psychoanalytic principles in terms of drives. This interest is also evident in
experimental studies supporting early animal models of psychopathology.
For example, Estes and Skinner (1941) attempted to model anxiety along
with its disruption of behavior under appetitive control in rats, a study that
created the foundation for research on conditioned suppression.
From early on, these behavioral accounts of psychopathology came with
behavioral accounts of treatment and behavior change in psychotherapy. Both
Skinner’s (1953) and Dollard and Miller’s (1950) texts, for example,
included not only descriptions of psychopathology, but also descriptions of
psychotherapy processes in behavioral terms. Neither proposed new
approaches or formats to psychotherapy. Instead, they attempted to
understand the phenomenon of psychotherapy as it existed at that time in
terms of behavioral principles. These conceptual accounts also translated into
empirical investigations of psychotherapy from a behavioral perspective in
two veins—understanding how these traditional psychotherapies worked, and
applying behavioral principles in ways that might work better.
 Several lines of inquiry sought to empirically account for psychotherapy
processes and outcomes in behavioral terms. This work mostly involved
exploring the verbal contingencies at play in the therapy room. Krasner and
colleagues conceptualized and modeled experimentally psychotherapy
processes using a series of verbal conditioning studies (see Krasner, 1963).
Kanfer and colleagues conducted a series of clinical interview and analogue
studies manipulating interviewer behavior to directly shape interviewee
verbal behaviors (e.g., Kanfer, 1959; see Kanfer, 1961; Kanfer & McBrearty,
1962; Kanfer, Phillips, Matarazzo, & Saslow, 1960; Phillips, Matarazzo,
Matarazzo, Saslow, & Kanfer, 1961). As behavior analysts explored the
clinical relevance of verbal conditioning processes, however, many came to
challenge the central tenet of psychotherapy—that changing verbal behavior
in the therapy room could result in broad, generalizable behavior change
outside of the therapy room (e.g., Zax & Klein, 1960). In fact, a number of
behavior analysts responded to this challenge by rejecting not just the
theories behind psychotherapy, but the focus on manipulating verbal
contingencies to change verbal behavior at all. And behavior therapy was
born.

First Wave of Behavior Therapy

In its first wave, behavior therapy all but abandoned the focus on
manipulating verbal behavior, focusing instead on changing the direct
contingencies controlling the behavioral excesses and deficits classified as
psychopathology. Offered as a distinct alternative to psychodynamic
psychotherapy, the original behavior therapy movement united behaviorists
from disparate traditions around a focus on changing direct contingencies in
accordance with specific scientific values. In short, behavior therapists held
that behavior change interventions should be
1. supported by coherent theories derived from empirically established
basic principles, and
2. comprised of well-established, empirically supported procedures
(Hayes, 2004).

Thus, the well-established basic behavioral principles of operant and

respondent conditioning were applied in techniques such as
counterconditioning, aversion therapy, negative practice, reactive inhibition,
and systematic desensitization (see Bandura, 1961; Eysenck, 1960, 1964;
Metzner, 1963; Wolpe, 1958, 1961). The evidence supported the
effectiveness of these interventions in terms of reducing subjective ratings of
distress and maladaptive behavior (e.g., see Grossberg, 1964). Further,
research efforts sought to clarify the mechanisms of change in behavioral
terms (e.g., see Rachman, 1967, for an early review of systematic
desensitization), supplementing data from effectiveness studies with data
from animal (e.g., Wilson & Davison, 1971) and analog (e.g., Ross &
Proctor, 1973) studies.
With these early successes, behavior therapy grew in prominence. It did
not, however, take clinical behavior analysis with it. In fact, interest in the
behavioral mechanisms of talk therapy faltered, lying dormant for almost two
decades (Kohlenberg et al., 2002). This can be attributed to two issues. First,
doubt grew regarding the clinical significance of changing verbal behavior in
the therapy session (e.g., see Grossberg, 1964), which characterizes much of
the behavior the therapist directly observes and manipulates in adult
outpatient treatment. Second, clinical behavior analysis did not rise to
behavior therapists’ original challenge that psychotherapy be based on
evidence supporting not only effectiveness but the basic underlying theory.
Accounts of verbal conditioning, and the data supporting them, were of
limited relevance to (1) most of the problems for which adults sought
outpatient treatment, and (2) the fact that discussing these problems in a
therapy session could result in improved functioning outside of a therapy
session (see Kohlenberg et al., 1993; Kohlenberg et al., 2002). Until these
accounts progressed, clinical behavior analysis would have little to offer,
leaving behaviorists to either avoid adult outpatient treatment altogether,
employ conditioning concepts with imprecision, or incorporate
nonbehavioral analyses to support their work.

Second Wave of Behavior Therapy

Meanwhile, behavior therapy continued to evolve independent of clinical
behavior analysis. And as the connections between behaviorism and
behavior therapy grew strained, behavior therapists became increasingly
accepting of accounts of psychopathology and behavior change that
incorporated models of cognitive mediation (e.g., Bandura, 1969).
Accordingly, the second wave of behavior therapy (e.g., Bandura, 1974;
Beck, 1976; Mahoney, 1974; Meichenbaum, 1977) was dominated by
approaches that conceptualized specific psychopathologies as attributable to
specific cognitive errors. Changing these behaviors, then, involved using
language to correct these errors. Technologies from the first wave of
behavior therapy were largely retained but were deemed supplementary to
new technologies designed to alter dysfunctional beliefs and problematic
cognitive processing.
The commitment to the development of empirically supported theories
and clinical procedures was also retained in this era of behavior therapy (see
Beck, 2005). For example, behavior therapy researchers sought empirical
support for the general cognitive theory of depression (e.g., see Haaga, Dyck,
& Ernst, 1991), the content specificity hypothesis (e.g., Greenberg & Beck,
1989; Salkovskis et al., 2000), and the efficacy (e.g., Blackburn, Bishop,
Glen, Whalley, & Christie, 1981) and effectiveness (e.g., Cohen, Sullivan,
Minde, Novak, & Helwig, 1981) of cognitive behavior therapies. And these
efforts were largely successful, fostering the dominance of the second wave
of behavior therapy over other forms of intervention for some time.
All but absent from this program of research, however, was the
development of basic principles of cognitive change (behavioral or
otherwise) and their relationship with overt behavior change. As descriptive
data grew matching diagnostic categories with cognitive functioning, the
burgeoning body of basic data on cognitive processes was rarely consulted
(see Beidel & Turner, 1986). As efficacy data piled up, research on cognitive
mediation was conspicuously lacking (see Stevens & Pfost, 1983). In short,
the basic idea that behavior changed because cognitive errors changed was
simply not considered.
In time, behavior therapy research efforts did take a closer look at
cognitive mediation. And among the data that emerged was a series of
anomalies that seriously challenged this core idea that direct cognitive
changes were necessary for behavior change (see Hayes, 2004). Data on
changes over the course of treatment suggested that early positive responses
to behavior therapy occurred before clients were exposed to key components
(e.g., Ilardi & Craighead, 1994). Results from a large component analysis
suggested that cognitive components do not incrementally improve the
effectiveness of behavior therapy (Gortner, Gollan, Dobson, & Jacobson,
1998; Jacobson et al., 1996). And data that directly examined purported
mechanisms of change were inconsistent at best. For example, several data
sets suggested that cognitions covary with symptoms, but without evidence
of causal links (e.g., Burns & Spangler, 2001). And as these data threatened
the cognitive model that supported behavior therapy’s second wave, several
different bodies of research converged on an alternative.

Third Wave of Behavior Therapy

In applied research on behavior therapy, novel approaches that
deemphasized the role of cognitive mediation began to proliferate. For
example, attempts to adapt behavior therapy to chronically suicidal patients
resulted in the development of dialectical behavior therapy (DBT;
Linehan, 1987), which integrates the explicit focus on behavior change that
characterized first-wave behavior therapies with the shaping of awareness
and acceptance of one’s immediate experience. Others (e.g., Wells &
Matthews, 1994) focused on the increased attention paid to worry,
rumination, and other forms of nonproductive cognition. This resulted in
the development of the metacognitive therapy (Wells, 2000), which replaces
direct strategies to modify cognitive contents with direct strategies to
modify counterproductive metacognitive beliefs (e.g., attributing power to
unwanted thoughts). In a more traditional wing of behavior therapy, Barlow
and colleagues (e.g., Barlow, 1988; Barlow, Craske, Cerny, & Klosko,
1989) incorporated private events into treatment by extending exposure,
based on respondent extinction, to interoceptive cues (e.g., physical
sensations associated with anxiety).
Meanwhile, in behavior analysis, two streams of basic research provided
the foundation for a coherent and productive theory of language and
cognition—relational frame theory (RFT; Hayes, Barnes-Holmes, &
Roche, 2001). First, the data on rule-governed behavior (RGB; see Hayes,
1989) suggested that humans will respond to contingencies contacted via
instruction (i.e., via a “rule”) even when they conflict with directly
experienced contingencies (e.g., Galizio, 1979). In light of the rigidity and
insensitivity that characterizes psychopathology, RGB was offered as a basic
theoretical account that could support continued conceptual and technical
development of behavior therapy (e.g., Zettle & Hayes, 1982). Second, the
data on stimulus equivalence in verbally able humans (Sidman, 1971)
suggested that when the discriminative functions of one stimulus are
conditional upon the presence of another stimulus, the two stimuli come to
share functions in a bidirectional manner. Further, these bidirectional
relationships among stimuli appeared to emerge, untrained, through
combinations of directly trained relationships, and are not limited to
equivalence. And in this phenomenon that came to be called derived
relational responding (DRR; see chapter 4 in this volume) was an account
of verbal conditioning that both explained rule-governed behavior and
extended beyond.
The development of this program of research on DRR as a
conceptualization of verbal behavior also opened the doors for a resurgence
in clinical behavior analysis (Kohlenberg et al., 2002). Many of the kinds of
problems that brought adults to outpatient behavior therapy were analyzable
when viewed according to the basic theory of DRR and transformation of
stimulus functions. The verbal stimuli that those humans had learned to
effectively discriminate in their world came to share functions with the
events themselves. With this, new responses could be learned without direct
experience. Humans could learn to fear and avoid events without ever having
been exposed to them. Likewise, it was clear that the talk that occurred in
therapy could evoke responses that were functionally equivalent to the
problematic behaviors in clients’ lives and shape new, more effective
repertoires. And thus, the third wave of behavior therapy was born.
The third wave of behavior therapy has been described by Hayes (2004)
as including several elements:
1. an emphasis on empirical support, coherent at multiple levels of
analysis;
2. a focus on generalizable principles of behavior change;
3. a sensitivity to context and function of behavior (both public and
private) over form;
4. the aim to expand repertoires to produce breadth and flexibility; and
5. a consistency between analysis of client behavior and that of
clinician behavior.

In this way, third-wave behavior therapies returned to the scientific

strategy and principles that defined the first wave, but with an explicit
account of the role of private events. In short, third-wave behavior therapies
tend to focus on cognition and emotion indirectly, changing the context to
alter the relationships among private events and overt behaviors, such that the
functions of those private events are expanded.
The trends that typify the third wave are evident in behavior therapies
from varying theoretical orientations. However, two of the third-wave
behavior therapies retained explicit philosophical and theoretical foundations
within clinical behavior analysis: functional analytic psychotherapy and
acceptance and commitment therapy. Functional analytic psychotherapy
(FAP; Kohlenberg & Tsai, 1991) focuses on using the therapeutic
relationship to analyze, evoke, and shape clinically relevant behaviors.
Acceptance and commitment therapy (ACT; pronounced “act”; Hayes,
Strosahl, & Wilson, 1999, 2012) focuses on undermining problematic
impacts of DRR and increasing personally meaningful actions in the presence
of unwanted private events. The remainder of this chapter will focus
exclusively on ACT.

Acceptance and Commitment Therapy

Acceptance and commitment therapy (Hayes, Strosahl, & Wilson, 1999,
2012) applies RFT and the basic findings on DRR to help people engage in
their lives in more personally meaningful ways. ACT is based on RFT’s
idea that most of the behavioral tendencies typically categorized as
psychopathologies are acquired and maintained by complex relational
networks (Dymond & Roche, 2009). Specifically, repertoires become
dominated by behaviors under rule-governed aversive control (Levin,
Hayes, & Vilardaga, 2012). Thus, from the ACT perspective,
psychopathology is characterized functionally by psychological
inflexibility—a response class that involves persistent avoidance, excessive
rule governance, and subsequent difficulty taking meaningful action
(Blackledge & Drake, 2013; Hayes & Wilson, 1994). Psychological well-
being, on the other hand, is characterized functionally by psychological
flexibility—a response class that reflects sensitivity to both immediate
contingencies and verbally constructed values (Blackledge & Drake, 2013).

Scientific Strategy and Support

ACT is grounded in a particular extension of behavior analysis called
contextual behavioral science (CBS; Hayes, Barnes-Holmes, & Wilson,
2012), which is distinguished by its explicit grounding in functional
contextualistic philosophy. Functional contextualism (FC) emphasizes a
pragmatic approach to scientific pursuit, with the aim being to reliably
predict the conditions under which behaviors of social significance occur and
change those conditions to exert experimental control or influence. FC
further holds that ideally, the behavioral analyses that allow for prediction
and control do so with precision (a high degree of specificity), scope (broad
applicability), and depth (cohering with other levels of analysis). CBS also
explicates a reticulated research approach where theoretical, practical, and
multifaceted (i.e., basic, analogue, and applied) empirical efforts are mutually
influential in fostering CBS’s overall progress. The CBS scientific approach
has resulted in the rapid proliferation of studies demonstrating support for the
efficacy of ACT in fostering behavior change in a number of domains (see
Hayes, Luoma, Bond, Masuda, & Lillis, 2006; Powers, Vörding, &
Emmelkamp, 2009; Ruiz, 2010) and for psychological flexibility as the
mechanism for improvement (Ruiz, 2012).

Processes
ACT is a process-based therapy, defined not by a specific set of
techniques but by a focus on increasing the probability of psychologically
flexible behaviors. This is quite distinct from the common focus in
psychotherapy on reducing psychological distress or changing problematic
thoughts. Instead of aiming to reduce or change the aversive private events
associated with various diagnostic categories, ACT practitioners elicit these
events and broaden their functions by shaping new, personally meaningful
behaviors in their presence. This might involve the integration of any number
of techniques from first- or second-wave behavior therapies, from other third-
wave behavior therapies, or from outside of behavior therapy altogether.
Fidelity to the ACT model simply involves assessing and intervening on the
repertoire in terms of psychological flexibility.
In order to assist practitioners in the implementation of ACT,
psychological flexibility is often conceptualized as involving six
interdependent facets (Hayes, Strosahl, Bunting, Twohig, & Wilson, 2004).
Each facet describes a functional dimension of behavior that together
represent a unified model of behavior change (Hayes, Pistorello, & Levin,
2012). These facets are described as mid-level terms. Like psychological
flexibility (and inflexibility), they are not proposed as technical terms, but
rather serve to orient practitioners to assessing and intervening on functional
dimensions of behavior consistent with the underlying technical account
described in RFT (Levin et al., 2012). These six mid-level terms are valued
living, present moment awareness, cognitive defusion, experiential
acceptance, self-as-context, and committed action.

Valued living. Valued living has been described as the primary purpose of
ACT (Hayes et al., 1999), such that psychological flexibility and each of the
other core processes are evaluated in terms of the extent to which they
foster valued living (Hayes, Strosahl, Bunting, et al., 2004). Valued living
refers to any behaviors that are acquired and maintained due to their
correspondence with a person’s verbally established values (Wilson,
Sandoz, Kitchens, & Roberts, 2010). Thus, in ACT, clients verbally clarify
their chosen purpose, or the abstract qualities they desire their ongoing
behavior to have, along with the specific topographies or goals they can
track as potential values-consistent behaviors (see Sandoz & Anderson,
2015). These verbal constructions then serve as formative and motivative
augmentals, transforming the functions of the inherent stimulus products
and natural consequences of values-consistent behaviors so that they serve
as intrinsic reinforcers. In this way, values can foster behavior under rule-
governed appetitive control in any conditions (Wilson et al., 2010). Values
avoid the rigidity and insensitivity typically associated with verbal control
by emphasizing abstract qualities of behavior over specific topographies or
outcomes (Levin et al., 2012). For example, the value of “being an
extraordinary mother” might involve the acute goal of ignoring tantrums.
Once established as such, the entire experience of ignoring, including the
sound of the child’s cries, could be transformed to reinforce the entire class
of values-consistent behaviors. Shaping a more flexible repertoire, then,
involves shaping verbalization of both purposes across a range of contexts.
In practice, intervening on valued living often involves evoking client
descriptions of (1) a specific meaningful moment from memory (e.g., a
sweet moment; Wilson & Sandoz, 2008) or imagination (e.g., one’s own
funeral; see Hayes, Strosahl, & Wilson, 1999), (2) some quality of the
client’s behavior that makes or has made that moment meaningful, and (3)
examples of behaviors that would also be characterized by that meaningful
quality.

Present moment awareness. ACT practitioners intervene to build present

moment awareness, or flexible, focused awareness of public and private
events as they occur (Hayes, Strosahl, Bunting, et al., 2004; Hayes,
Pistorello, & Levin, 2012; Levin et al., 2012). Psychological flexibility is
characterized by sensitivity to a broad range of stimuli in the changing
context, whereas psychological inflexibility is sensitive to a narrow range of
aversive events. From the ACT perspective, discrimination of events, both
public and private, as they occur, will broaden the aspects of context likely
to influence behavior. ACT emphasizes both flexibility and focus, as
flexibility of awareness without focus is essentially distractibility, and
focused awareness without flexibility is essentially hypervigilance. Instead,
present moment awareness as trained in ACT allows for focus on particular
stimuli that are significant, and for the shifting of attention when necessary
for effective values-consistent behavior. In practice, intervening on present
moment awareness often involves guiding clients’ attention to aspects of
the immediate experience. This could be carried out in an eyes-closed
meditation (e.g., a breathing meditation) or simply in conversation (e.g.,
“What do you notice happening in your body as we speak about this?”).

Experiential acceptance. Experiential acceptance involves approaching

events that have previously functioned as aversive without attempts to
manage or control them (Blackledge & Drake, 2013). Psychological
inflexibility is characterized by persistent experiential avoidance, when a
small number of difficult, unwanted private events exert extensive aversive
control over behavior. This emerges through DRR, where individuals learn
to tact certain co-occurring private events (i.e., thoughts, bodily sensations,
memories) as emotions (e.g., “anxiety”) that should be avoided (Hayes,
Strosahl, Wilson, et al., 2004). Building flexibility involves changing the
functional relationship between behavior and the difficult, unwanted private
events that have previously functioned as aversive (Levin et al., 2012). In
ACT, we shape valued living in the context of unwanted private events, as it
is often values-relevant contexts that foster vulnerability by eliciting painful
private events. In practice, intervening on experiential acceptance often
involves use of metaphor to practice approach of unwanted experiences. For
example, the therapist might instruct a client to interact with an object that
symbolizes their pain as they typically do (e.g., by swatting or pushing it
away), then to interact with it as they would a butterfly (e.g., gently letting
it land in cupped hands and holding still to observe it; Boone & Cannici,
2013).
Cognitive defusion. Cognitive defusion (originally called
“deliteralization”; Hayes et al., 1999) involves awareness of thinking as
ongoing behavior, such that the functions of cognitive events broaden and
no one verbal function dominates. Psychological inflexibility is
characterized by cognitive fusion, where the context supports literal
functions (i.e., the content of thoughts) as the dominant source of control
over the repertoire (Blackledge, 2007). In this way, cognitive fusion refers
to the same phenomenon described in the rule-governed behavior literature
and produces the same insensitivity to other sources of stimulation
(Blackledge & Drake, 2013). Building flexibility involves disrupting the
literal functions such that other functions (e.g., sensory qualities, memories
of similar thoughts, other relevant perspectives) can come to influence
behavior along with other events in the context. In practice, intervening on
cognitive defusion often involves drawing attention to the process of
thinking. This could involve reflecting the client’s description of thoughts
with explicit labeling of them as thoughts (e.g., “So you feel compelled by
this thought that you are unlovable”) or asking the client to describe
thoughts (e.g., “It looked like something shifted for you just then. What’s
coming up for you as we discuss this?”).

Self-as-context. Self-as-context involves awareness of one’s own behavior

as distinct from other internal or external events, including self-relevant
private events (Hayes, Pistorello, & Levin, 2012; Levin et al., 2012). The
self emerges through discrimination of one’s own behavior as distinct from
other events (Skinner, 1974b). Once a client verbally discriminates the self
(i.e., as a concept), however, they can easily relate to it, and it can take on
functions of other events through DRR. Psychological inflexibility involves
private events (e.g., evocative emotions or self-relevant cognitions) coming
to dominate functions of the self (Levin et al., 2012). Thus, like a highly
generalized rule, these functions can impact behavior in any situation, as the
self is an event that is always relevant (i.e., “If me, then…”). Psychological
flexibility, however, involves awareness of the self as an event that is
present for and relatable to all other events, but is uniquely discriminable
and uniquely invariant in terms of its location in time, place, and person
(Blackledge & Drake, 2013). As such, building flexibility in ACT involves
shaping awareness of self in the presence of private events that vary across
time (e.g., memories of the past and concerns for the future), place (e.g.,
different situations, relationships, roles), and person (e.g., different
perspectives on the same event). In practice, intervening on self-as-context
might involve having the client list different roles, concepts, labels, or
characteristics they associate with themselves (e.g., by completing a list of
“I am” sentence stems). Next the therapist might ask the client to notice
conflicts among the different descriptions, notice which others would know
and which they wouldn’t, imagine contexts in which these different
descriptions dominate their experience of themselves, and imagine that
certain descriptions would not apply in the future.

Committed action. Committed action describes a large response class of

values-consistent behaviors for which the stimulus control over different
topographies is well discriminated between contexts. The dominance of
aversive control that characterizes psychological inflexibility often prevents
contact with appetitive events that could serve to discriminate and
consequate effective, values-consistent behavior. Building flexibility, then,
involves shaping (1) the discrimination aspects of the immediate context
related to values and (2) the engagement in values-consistent behaviors as
appropriate to that context. In practice, intervening on committed action
often involves having a client describe the function of their current
behavior, describe immediately available behaviors that would be more
values consistent (and likely evoke some discomfort), and set a goal to
engage in a chosen behavior more often outside of session (despite the
discomfort).

Implications and Future Directions

The material covered in this chapter thus far—and in particular the
discussion of ACT—is decidedly relevant to practitioners and to the field of
applied behavior analysis as a whole. We explore both of these areas below.

Implications for Practitioners

For behavior analytic practitioners, the implications of the evolution of
behavior therapy and ACT can be quite significant. Applied behavior
analysis has always been committed to addressing problems of social
significance (Baer, Wolf, & Risley, 1968). Behavioral difficulties categorized
as psychopathology account for 30 percent of the nonfatal disease burden and
10 percent of the overall disease burden worldwide (World Health
Organization, 2016). This is despite the proliferation of psychological
treatments and the ever-increasing availability of psychotropic medications.
Further, this excludes individuals whose suffering is not categorized by the
socioverbal community as pathological. The basic research underlying ACT
would suggest that psychological suffering is, in fact, universal, as it is a
product of normal learning processes. Despite its universality, however,
psychological suffering is of tremendous social significance, and behavior
analysis clearly has something to offer. We encourage practitioners, whether
trained as behavior therapists or not, to rise to the challenge of the human
condition in at least three ways.
First, let us be open to the inclusion of private events as relevant in
behavioral analyses. Mentalism, or the appeal to private events as causes of
overt behavior, is rejected by behavior analysis. However, private events and
their relationships with overt behavior can be analyzed and intervened upon
from a behavior analytic perspective. Just as with any other behavior outside
of the focus of the intervention, considering private events is often
unnecessary to gain prediction and influence over the behavior of interest.
But when private events are relevant, referring clients to practitioners
working from perspectives we reject as mentalist (e.g., cognitive therapists)
seems a bizarre course of action. Where our direct analyses of overt behavior
alone fall short, let us be open to considering private events in relation to the
behaviors of interest, with a focus on shifting the contexts controlling those
behavior-behavior relations.
Second, let us consider verbal behavior in terms of DRR and the
transformation of stimulus functions. Behavior analysts, like members of any
profession, use language to intervene on their clients’ behavior as well as that
of other stakeholders (e.g., employees, colleagues, students, and parents).
The difference, however, is that we have at our disposal, in RFT, an analysis
of verbal behavior that accounts for the conditions under which that language
might be most effective. We can change behavior by using language to help
transform the functions of the contexts in which individuals live and work.
Whether training a parent, teaching a class, giving a presentation, or
marketing an intervention, let us harness the incredible power of language to
increase our effectiveness.
Third, let us consider psychological flexibility as a dimension of the
behavioral repertoire, even outside of the behavior therapy setting. Often the
rigidity and insensitivity of client behavior is what makes it problematic. For
example, a client might talk about trains or dinosaurs regardless of the
context or their partner’s interest, and at the exclusion of other topics. The
psychological flexibility model proposes conditions under which rigidity like
this might emerge (i.e., in the context of aversive private events where this
pattern allows for temporary escape and avoidance). ACT proposes how we
might intervene to improve flexibility without having to avoid contexts that
elicit aversive private events. Psychological flexibility, by definition,
involves increased sensitivity to direct contingencies and verbally established
values to allow for an increase in personally meaningful behavior. This
quality of behavior is associated not only with recovery from
psychopathology but with thriving in a number of different domains of
living. Certainly, this level of engagement in life is relevant to behavior
analysts intervening in all sorts of settings with all sorts of clients.

Implications for Our Field

If practitioners are going to easily incorporate these implications into
their work, there are developments that behavior analysis as a field must
make. First, we need to expand our approaches to assessment. In particular,
to the extent that we are dissatisfied with the tools commonly used to assess
private events and relationships between private events and overt behavior
(e.g., questionnaires), we are challenged to develop new approaches. Second,
we need to integrate theory, research, and practical training in complex
human behavior into typical behavior analysis curricula in order for the field
of behavior analysis to progress and evolve. This text represents an initial
effort, yet it is not enough, nor is the work presented in this text complete.
Third, we need pioneer behavior analysts to adapt third-wave behavior
therapy technologies to the practice of behavior analysis in the various
settings in which behavior analysts work. ACT is topographically different
when implemented by medical doctors, managers, or coaches, and likewise,
most behavior analysts will not seek the training and licensure necessary to
incorporate talk therapy into their practice. Behavior analysts, however, do
not need to be trained in talk therapy to adapt ACT where it would serve their
work. For example, Gould, Tarbox, and Coyne (2017) demonstrated
increased overt values-directed behaviors following a brief ACT training
intervention implemented by behavior analysts.
Despite challenges, behavior analysis has made substantial progress in
how to apply behavior analytic approaches to complex human suffering. It
may now be time to integrate these advances into the mainstream of behavior
analytic research, training, and practice.

Conclusion
The present chapter considered the historical development of contemporary
behavior therapy. Beginning with the consideration of private events in
behavior analysis, this chapter provided an overview of the development of
clinical behavior analysis, and in particular its role in analyzing the
complex problems faced by adults in outpatient psychotherapy. Behavior
therapy can be seen as having developed in three phases, or “waves”—each
of which was described in this chapter, with particular emphasis given to
acceptance and commitment therapy. A number of implications were
described, including those focused on practice and the field of behavior
analysis as a whole.

Study Questions
1. Define private events. Are they the same as or different from mental
events?
2. Can private events be included in the science of behavior analysis?
Why or why not?
3. What are some of the defining features of the first-wave behavior
therapies?
4. What are some of the defining features of the second-wave behavior
therapies? How do these depart from the first wave?
5. What are the some of the defining features of the third-wave
behavior therapies? How do these depart from the second wave?
6. Do the third-wave behavior therapies move toward, away from, or
make no movement with respect to the foundational philosophy and
principles of behavior analysis, compared to the second wave?
7. What is acceptance and commitment therapy?
8. What are some of the defining processes of ACT?
 9. Compare and contrast the defining features of ACT with mainstream
applied behavior analysis, in terms of being conceptually systematic
with behavioral principles, focusing on behavior, focusing on
generalization, and insistence on being effective in producing
substantial changes in socially significant behavior.
10. What are the implications of ACT for your current practice in
applied behavior analysis and for what you see as the future
evolution of mainstream applied behavior analysis?
 Chapter 16:

Implicit Cognition and Social Behavior

Dermot Barnes-Holmes
Colin Harte17
Ghent University, Belgium
Ciara McEnteggart
Ghent University, Belgium

Link and Preview

Chapter 16 builds upon previous chapters on derived stimulus
relations, and relational frame theory in particular. The focus of this
chapter—implicit cognition and social behavior—is of great
importance as it relates to many of the most pervasive social
issues around the world. The authors review research on a
number of consequential topics and describe ongoing conceptual
development.

The human species is composed of relatively weak and slow-moving

primates, and yet we have, in one sense, come to dominate the planet in a
very short period of time. Arguably, the propelling force behind our rise in
dominance is our ability to learn how to relate stimuli and events in
increasingly abstract, or arbitrarily applicable, ways. Irrespective of how
this ability evolved so strongly in our species (see Hayes & Sanford, 2014),
it appears to lie at the core of human language and cognition (Hayes,
Barnes-Holmes, & Roche, 2001). On the one hand, the human ability to
engage in abstract relational responding has been the key to our success in
developing modern civilization with increasingly sophisticated sciences and
technology. On the other hand, abstract relational responding allows
humans to categorize other members of their own species into in- and out-
groups and to react to members of an out-group in a highly negative manner
without having a direct negative experience with any member of that group.
Such prejudicial behaviors appear to be widespread and common among the
human species. Indeed, at the time of writing we had just witnessed populist
appeal for literally building a wall along the border between the United
States of America and Mexico and the withdrawal of the United Kingdom
from the European Union to allow for greater control over its borders. And
in the last 100 years of human history, we have witnessed two world wars in
which millions were slaughtered, and extremist attempts at mass genocide,
based, at least in part, on the human ability to categorize out-groups as
sufficiently threatening to warrant warfare.
The ability to categorize other members of our species into in- and out-
groups may have served us well as small groups of hunter-gatherers, because
it supported necessary cooperation within a group against another in
competing for potentially limited resources. However, when human groups
become increasingly abstract and increase in size (e.g., to nation states) and
the level of science and technology allows for the annihilation of millions at
the press of a button, the evolutionary value of prejudicial behavior, or social
categorization, seems completely lost. Indeed, one could argue that in the
modern context of a genuinely globalized world there is only one in-group—
the entire human species and even more broadly life itself on this planet.
Clearly, the human ability to engage in social categorization is a
potentially lethal behavior that we need to study and to understand if we are
to predict and influence it in a manner that will serve to protect life on this
planet. The science of psychology, and social psychology in particular, has
devoted considerable time and effort in tackling this issue (see Tajfel, 1981,
for seminal work). Behavior analysis has been less engaged, empirically, in
this area, but its main progenitor, B. F. Skinner, became a household name for
publishing books that described how behavioral principles could be used to
engineer human cultures in a positive and peaceful direction. Empirical
efforts at understanding social categorization within behavior analysis have
not been entirely absent, however. As we shall see, relatively early in the
study of stimulus equivalence and arbitrarily applicable relational
responding, behavior analysts made attempts to create experimental models
of human prejudicial behavior. It is fair to say that those conducting this early
work had little appreciation of how rapidly social categorization effects can
occur in human behavior. Only relatively recently has the “split-second”
nature of such responding come to light. As such, the pervasive and
uncontrolled nature of human prejudicial behavior is rendered even more
threatening because it seems to occur “involuntarily” (because it is so rapid).
The area of research that has focused on this phenomenon has been labeled
implicit social cognition, and the current chapter will provide an overview
of this work.
Specifically, this chapter will focus on the research investigating the role
of derived relations in the development of social stereotypes, prejudices, and
beliefs. We will briefly review the large body of literature that has employed
a method derived from relational frame theory (RFT; Hayes, Barnes-
Holmes, & Roche, 2001), known as the implicit relational assessment
procedure (IRAP). The social domains upon which we will focus comprise
national identity, religion, race, gender, sexuality and sexual preferences, age,
body image, and smoking as a stigmatized behavior. We will then discuss the
existing conceptual attempts to understand these findings in the context of
RFT, including a recently proposed model for analyzing IRAP effects—the
differential arbitrarily applicable relational responding effects
(DAARRE) model.

Implicit Social Cognition

Work on implicit social cognition, from an RFT perspective, actually began
in the early 1990s. Specifically, researchers sought to examine social
categorization in Northern Ireland, where family names and sectarian
symbols are often associated exclusively with either Catholic or Protestant
communities (Watt, Keenan, Barnes, & Cairns, 1991). The study involved
training participants in a series of match-to-sample tasks that were designed
to generate derived equivalence relations between Catholic names and
Protestant symbols that would be inconsistent with the verbal/social
histories of participants who resided in Northern Ireland. The results
showed that some Northern Irish residents did indeed demonstrate difficulty
in forming these equivalence relations, whereas individuals from outside
Northern Ireland did not. Numerous studies since then have reported
broadly similar outcomes in which participants with specific
preexperimental histories appear to show difficulty forming derived
relations that are inconsistent with those histories (e.g., Barnes, Lawlor,
Smeets, & Roche, 1996; Dixon, Rehfeldt, Zlomke, & Robinson, 2006;
Leslie et al., 1993; Merwin & Wilson, 2005). Interestingly, recent research
in this area has also shown that it is possible to undermine these types of
effects with appropriate match-to-sample training designed to counter
racially biased responding in white children (Mizael, de Almeida, Silveira,
& de Rose, 2016).
The general strategy of comparing patterns of responding that are
consistent versus inconsistent with participants’ preexperimental histories led
to more recent efforts to develop behavior analytic procedures that may be
used to assess verbal relations. Currently, the most widely used method in
this regard is the IRAP (Barnes-Holmes, Murphy, Barnes-Holmes, &
Stewart, 2010). In the IRAP procedure, experimenters present pairs of stimuli
(e.g., words, pictures, statements) on each trial, require participants to
confirm or disconfirm the relation between these pairs within a short
response latency window, and provide corrective feedback after each
response. In general, the feedback is designed to be consistent with
participants’ preexperimental verbal histories on half of the blocks of trials,
and inconsistent on the other half. For example, an IRAP might require
responding “True” to a picture of a flower and the word “pleasant” (history
consistent) on one block, and “False” (history inconsistent) on another block.
The basic logic of the IRAP is that, all things being equal, participants should
show a tendency to respond more quickly on history-consistent, relative to
history-inconsistent, blocks. This difference in latencies across the two types
of blocks is often referred to as the IRAP effect or a positive or negative
response bias, depending on whether the effect is above or below zero. It is
important to understand that the terms “IRAP effect” and “response bias”
should not be interpreted as a proxy for a mental construct or implicit attitude
in a cognitive or social psychological sense. Instead, these terms simply
denote a tendency to respond in one particular direction over another on the
IRAP.
There are several response-time measures for assessing implicit
cognition, such as the Implicit Association Test (IAT), evaluative priming,
and the Extrinsic Affective Simon Task (EAST). However, all of these other
measures emerged from the “mainstream” cognitive tradition and therefore
cannot be attributed to RFT, or behavior analysis more generally. There is
one other task, however, that has emerged within behavior analysis known as
the Function Acquisition Speed Test (FAST; O’Reilly, Roche, Ruiz, Tyndall,
& Gavin, 2012). The FAST has been used to assess implicit social cognition,
but at the time of writing only one published study had used the measure in
this domain.
One advantage of the IRAP over many of the other measures is that it
allows for a more detailed analysis of the relations being measured. This is
made possible through the separation of the measured relations into four
individual trial-types. In an IRAP designed to measure racial response biases,
for example, the following four trial-types might be presented: White People-
Positive-True/False; White People-Negative-True/False; Black People-
Positive-True/False; and Black People-Negative-True/False. Therefore, the
IRAP permits a functional distinction between, for example, black people as
both positive and negative, which may be conceptually important in socially
sensitive domains.

Social Research
This section will address research within the various social domains listed
earlier. We will focus largely on IRAP research because it is clearly
contained within the behavior analytic tradition.

National Identity
One of the first published IRAP studies assessed biases toward different
nationalities (i.e., Irish, Scottish, American, and African) by native Irish and
Irish-American participants (Power, Barnes-Holmes, Barnes-Holmes, &
Stewart, 2009). Across two experiments, researchers gave participants an
IRAP in which the label stimuli “more likeable” and “less likeable” appeared
with a target consisting of two nationalities to be compared in terms of
likeability on each trial. In Experiment 1, the pairs were “Irish-Scottish,”
“Scottish-American,” “American-African,” “Scottish-Irish,” “American-
Scottish,” and “African-American,” and participants were required to choose
either “True” or “False” as response options on each trial. As an example, the
first trial listed above could be read as “Irish are more likeable than Scottish;
true or false?” Results indicated a strong “in-group response bias” for Irish
over Scottish, and American over African nationalities. That is, participants
tended to respond more quickly when they were required to respond “True”
rather than “False” when these two pairs were presented with “More
Likeable.” In Experiment 2, the target pairs were “American-Irish,” “Irish-
Scottish,” “Scottish-African,” “Irish-American,” “Scottish-Irish,” and
“African-Scottish,” and the participants were Irish-Americans (i.e., American
citizens who claimed Irish heritage). Results indicated an in-group response
bias for American over Irish, Irish over Scottish, and Scottish over African.
In both experiments, participants’ self-report measures rating preferences of
these nationalities diverged from the IRAP biases. For example, in the self-
report measure, the Irish Americans rated Irish as more likeable than
Americans (or any other group), but on the IRAP they responded “True”
more quickly than “False” when responding to the “More Likable; American-
Irish” trial-type. This study was one of the first to highlight the potential
utility of the IRAP in the assessment of potentially sensitive social issues.
That is, the IRAP seemed to provide a method for assessing the natural
verbal relations at play over and above self-report measures.

Religion
Hughes, Barnes-Holmes, and Smyth (2017) employed the IRAP to
assess the biases of Catholics and Protestants in Northern Ireland, a post-
conflict context in which sectarian tensions had been the source of violence
and discrimination across three decades. Specifically, in a known-groups
design, the researchers presented Catholic and Protestant participants in
Northern Ireland with one of two label stimuli (i.e., “Catholics” or
“Protestants”) on each trial with one of six positive (e.g., friendly, safe) or
negative (e.g., bad, dishonest) words as target stimuli. The IRAP required
responding in a pro-Catholic and anti-Protestant pattern on some blocks of
trials (e.g., pressing a key for “True” when “Safe” appeared with the word
“Catholic”) and a pro-Protestant and anti-Catholic pattern on the remaining
blocks of trials (e.g., pressing a key for “False” when “Safe” appeared with
the word “Catholic”). The IRAP revealed pro-Catholic and pro-Protestant
biases across both groups; however, in-group biases were stronger. For
example, the size of the difference in reaction times was larger when Catholic
participants were required to confirm rather than deny that Catholics were
positive compared to when they were required to confirm rather than deny
that Protestants were positive; the opposite was the case for the Protestant
participants.
 Similar research was also conducted by Drake et al. (2010) on attitudes
toward Christians and Muslims in an undergraduate sample in the United
States. Results revealed a distinct positive bias for Christians on both the
Christian-Positive and Christian-Negative trial-types, and a negative bias for
Muslims on Muslim-Negative, but no significant effect was observed on
Muslim-Positive. Moreover, Scheel, Roscoe, Scaewe, and Yarbrough (2014)
found a broadly similar pro-in-group Westerner bias when they assessed
attitudes toward Muslims and Westerners in a US-based undergraduate
sample.

Race
One of the earliest IRAP studies focused on racial bias and examined
the response patterns of white participants toward pictures of black and white
individuals (Barnes-Holmes, Murphy, et al., 2010). The researchers presented
participants with one of two label stimuli (i.e., “Safe” and “Dangerous”) on
each trial with a picture of a white or black man holding a gun as a target
stimulus. The IRAP revealed pro-white and anti-black biases, although the
anti-black effect was restricted to one trial-type (i.e., the Dangerous-Black
trial-type). Indeed, four other studies have also examined racial bias using the
IRAP, and broadly similar positive in-group biases were found for white
participants (Drake et al., 2010; Drake et al., 2015; Power, Harte, Barnes-
Holmes, & Barnes-Holmes, 2017a). The most recent of these combined the
IRAP for examining racial biases with the measurement of
electroencephalograms (EEGs; Power, Harte, Barnes-Holmes, & Barnes-
Holmes, 2017b). The EEG recordings revealed that the event-related
potentials (ERPs; i.e., brain responses) were more positive for the pro-black
trials than for the pro-white trials across six of the frontal sites. Indeed, it is
interesting that the differential ERP patterns observed in this study were
restricted to the frontal sites and that greater positivity was recorded for the
IRAP performances that required responding in a manner that was
inconsistent with a white in-group racial bias. Specifically, as the authors
point out, the findings are broadly consistent with research in the
neurocognitive literature indicating that the prefrontal areas of the cortex
(i.e., more primitive areas of the brain) may be involved in suppressing
emotional reactions that are deemed to be undesirable in some way—in this
case a pro-white/anti-black response.
 Gender
The IRAP has also been used across a number of studies in the
assessment of gender biases. For example, a study by Cartwright, Hussey,
Roche, Dunne, and Murphy (2016) evaluated the potential utility of the IRAP
for assessing gender binary beliefs. Specifically, participants completed two
IRAPs, in which one of two label stimuli were presented (i.e., “Men” and
“Women”) on each trial. One IRAP presented positive masculine (e.g.,
“competitive”) or feminine (e.g., “nurturing”) traits as target stimuli; the
other IRAP presented negative traits (e.g., masculine “aggressive”; feminine
“bossy”). Participants also completed a number of self-report measures on
sexism and heteronormativity and a hiring task to assess hiring preference.
The IRAPs revealed that participants readily paired masculine traits to men
(e.g., saying men-competitive-true, more easily than false) and feminine
traits to women (e.g., saying women-nurturing-true, more easily than false),
and rejected pairing feminine traits with men (e.g., saying men-nurturing-
false, more easily than true) and masculine traits with women (e.g., saying
women-aggressive-false, more easily than true). Interestingly, 83% of
participants rated male traits as making someone more hirable. Similar results
were found in another study using a FAST (Cartwright, Roche, Gogarty,
O’Reilly, & Stewart, 2016).
Indeed, similar gender stereotyping biases have also been found using
the IRAP concerning gendered house chores (e.g., chopping wood is for men
and cooking is for women; Drake et al., 2010), gendered toys in young boys
and girls (e.g., dolls are for girls and toy cars are for boys; Rabelo, Bortoloti,
& Souza, 2014), and gendered university disciplines (i.e., men are related to
science and arts but women to arts only; Farrell, Cochrane, & McHugh,
2015).

Sexuality and Sexual Preferences

The IRAP has also been used in the assessment of attitudes toward
sexuality and sexual preferences. For example, in a study of sexual
orientation, Cullen and Barnes-Holmes (2008) assessed biases among a
group of heterosexual and homosexual male participants. The IRAP
presented one of two label stimuli (i.e., “Straight” and “Gay”) along with a
positive stereotypical target for straight people (e.g., “normal,” “safe”) or a
negative stereotypical target for gay people (e.g., “abnormal,” “dangerous”).
The IRAP revealed pro-gay and pro-straight biases for both groups; however,
an anti-gay bias was observed on the gay-negative trial-type among
heterosexuals only. In three further studies, researchers successfully used the
IRAP to predict sexual orientation by asking participants to relate the labels
“straight” or “gay” with attractiveness and unattractiveness (Ronspies et al.,
2015; Timmins, Barnes-Holmes, & Cullen, 2016).
In a study on sexual stereotyping, Scheel, Fischer, McMahon, and Wolf
(2011) reported IRAP findings in which women confirmed more quickly than
they denied that traits stereotypical of gay men (e.g., “artistic,” “feminine”)
corresponded with this group; a similar bias was obtained for associating
straight men with stereotypical traits (e.g., “assertive,” “masculine”). And in
a study on BDSM (bondage and discipline, sadism and masochism; or acts of
domination and submission) practice, students and clinicians who reported no
BDSM tendencies produced greater anti-BDSM/pro-normal IRAP effects
than those who reported BDSM tendencies (Stockwell, Hopkins, & Walker,
2017; Stockwell, Walker, & Eshleman, 2010).

Age
One of the earlier IRAP studies examined ageist attitudes among young
people and revealed pro-young and anti-old biases in the first experiment
(Cullen, Barnes-Holmes, Barnes-Holmes, & Stewart, 2009). A second
experiment attempted to assess the malleability of the IRAP effects.
Researchers randomly assigned participants to either a pro-young or pro-old
condition in which they were exposed to pictures of admired old and disliked
young people (pro-old condition) or to pictures of admired young and
disliked old people (pro-young condition). Participants completed one IRAP
immediately after exposure to the pictures and then a second identical IRAP
twenty-four hours later. The pro-young condition produced strong pro-young
and anti-old effects on both days, whereas the pro-old condition produced
relatively weak pro-young effects and relatively strong pro-old effects on
both days.

Body Image
A substantial body of IRAP research has also emerged in the domain of
body image in recent years. One of the first studies was conducted by Roddy,
Stewart, and Barnes-Holmes (2010), and results revealed a pro-slim, but not
an anti-fat, bias. This effect was replicated in another study by Roddy,
Stewart, and Barnes-Holmes (2011), which also employed facial
electromyography (EMG) to gauge emotional responses of participants while
they conducted the IRAP, and both measures found a pro-slim bias.
Interestingly, in a similar study by Nolan, Murphy, and Barnes-Holmes
(2013), this effect was found only among males. Similarly, in a study that
employed only female participants and female stimuli, Exposito, Lopez, and
Valverde (2015) found no pro-slim or anti-fat bias, potentially suggesting that
women demonstrate less weight-related prejudice than men.
Body-image IRAP effects have also been shown to predict body
dissatisfaction, body weight, and disordered eating, beyond that of self-report
measures (Juarascio, Forman, Timko, & Herbert, 2011). Furthermore, Heider,
Spruyt and De Houwer (2015) employed two IRAPs, one assessing actual
body image (i.e., I am thin) and one assessing ideal body image (i.e., I want
to be thin). Results indicated actual self-thin bias was lower in those with
higher levels of body dissatisfaction, whereas ideal self-thin bias was higher
in those with higher levels of body dissatisfaction. Similar research on
attractiveness, rather than body image, found broadly consistent findings in
which pro-attractive biases were observed (Murphy, Hussey, Barnes-Holmes,
& Kelly, 2015; Murphy, MacCarthaigh, & Barnes-Holmes, 2014).

Smoker Status
Two published studies have used the IRAP to assess attitudes toward
smokers and non-smokers (Cagney, Harte, Barnes-Holmes, Barnes-Holmes,
& McEnteggart, 2017; Vahey, Boles, & Barnes-Holmes, 2010). In the first of
these, Vahey et al. (2010) presented to adolescents an IRAP in which the
labels “smoker” and “non-smoker” appeared with either social acceptance or
social rejection words (identified from tobacco marketing campaigns), along
with “similar” or “opposite” as response options. Results showed that
smokers responded more quickly when confirming that smokers were similar
to social acceptance words, but no difference emerged between acceptance or
rejection words among the non-smokers.
In the second study, Cagney et al. (2017) attempted to conduct a more
detailed analysis of attitudes toward smokers using the IRAP in both adult
and adolescent smokers and non-smokers, and they also explored the impact
of parental smoking status. The results showed a pro-smoker bias for both
adults and adolescents smokers, with a neutral bias for non-smokers in both
groups. However, adult smokers and non-smokers were more clearly
differentiated than the two adolescent groups on the IRAP, whereas the
opposite was true in terms of the self-report measures. Post-hoc analyses
indicated that non-smokers showed more positive bias scores toward smokers
on the IRAP if their parents were smokers (relative to non-smoking parents).
Non-smokers also showed more positive bias scores toward non-smokers if
their parents did not smoke (relative to smoking parents). Overall, therefore,
smoking status and parental smoking status appeared to influence social
attitudes toward a socially stigmatized group (i.e., smokers).

Understanding IRAP Effects from a Relational Frame

Theory Perspective
As noted at the beginning of this chapter, the behavior analytic basis of the
IRAP involves comparing two opposing patterns of relational responding,
one of which is deemed to be generally consistent with the preexperimental
history of the participant, the opposing pattern, by definition, which is not.
The so-called IRAP effect is therefore deemed to reflect this difference in
preexperimental relational responding. This is, in essence, the basic
assumption underlying the IRAP. Over the years, however, increasingly
sophisticated behavior analytic accounts of the IRAP have emerged, and the
remaining half of the current chapter will consider these, with a particular
focus on social implicit cognition.

The Relational Elaboration and Coherence

(REC) Model
The types of effects that have been observed with the IRAP have been
referred to as brief and immediate relational responses (BIRRs), in that
they are emitted within a short response window of time after the onset of
each trial. In contrast, extended and elaborated relational responses
(EERRs) are more complex and emitted more slowly and as such occur over
a longer period of time (Barnes-Holmes, Barnes-Holmes, Stewart, & Boles,
2010; Hughes, Barnes-Holmes, & Vahey, 2012). The distinction between
BIRRs and EERRs was conceptualized within the context of the relational
elaboration and coherence (REC) model, an initial RFT approach to
implicit cognition (Barnes-Holmes, Barnes-Holmes et al., 2010; Hughes et
al., 2012). The basic idea behind the model is that the IRAP (and indeed
other implicit measures) produces the types of effects it does because the task
forces participants to emit BIRRs rather than EERRs. The latter are generally
assumed to be evoked by traditional self-report measures, when participants
are not under time pressure to respond to each item in a questionnaire. The
strength or probability of the BIRRs emitted toward experimental stimuli is
deemed to be functionally similar to that of responses toward these stimuli in
the participant’s preexperimental history, particularly in situations in which
individuals have to respond relatively quickly or have little motivation to
reflect on how they are responding in that particular moment in time.
Imagine, for example, a white individual who has resided exclusively in
white neighborhoods, has no non-white friends or family members, and has
been exposed to many media images of black people as violent drug dealers
and inner-city gang members. When presented with an IRAP that displayed
pictures of black males carrying guns, according to the REC model, BIRRs
confirming that black men are “dangerous” and “criminals” are likely to be
more probable for this individual than those denying such relations, thus
revealing an anti-black racial bias (see Barnes-Holmes, Murphy, et al., 2010).
Indeed, this bias might not be observed if the same individual were asked to
rate the same pictures of the black men in the absence of time pressure. In the
latter context, there is sufficient time to respond in accordance with an
extended and elaborated relationally coherent network (i.e., an EERR), which
by definition extends beyond the initial BIRR. In this case, the REC model
suggests that the individual may fail to report the initial BIRR based on the
additional relational responding, such as “It is wrong to discriminate on the
basis of race” and “I am not a racist.” Hence, the IRAP reveals the BIRR, and
a questionnaire reveals the EERR.
It’s worth noting some limitations to the REC model. In concluding that
the IRAP reveals BIRRs rather than EERRs, the REC model assumes that
participants respond to each of the four IRAP trial-types in more or less the
same manner. This basic assumption has not been upheld empirically,
however, and thus behavior analysts have recently attempted to develop a
more sophisticated RFT-based understanding of the behavioral effects
observed with the IRAP. In presenting this more recent account, we will
focus first on an IRAP that is not relevant to social cognition, but thereafter
consider the implications for the social-psychological domain.
Imagine an IRAP that aimed to assess the response probabilities of four
well-established verbal relations pertaining to non-socially valenced stimuli,
such as shapes and colors. Across trials, the two label stimuli, “Color” and
“Shape,” could be presented with target words consisting of specific colors
(“Red,” Green,” and “Blue”) and shapes (“Square,” “Circle,” and
“Triangle”). As such, the IRAP would involve presenting four different trial-
types that could be designated as (1) Color-Color, (2) Color-Shape, (3)
Shape-Color, and (4) Shape-Shape. During such a Shapes-and-Colors IRAP,
participants would be required to respond in a manner that was consistent
with their preexperimental histories during some blocks of trials: (1) Color-
Color-Yes, (2) Color-Shape-No, (3) Shape-Color-No, and (4) Shape-Shape-
Yes. On other blocks of trials, the participants would have to respond in a
manner that was inconsistent with those histories: (2) Color-Color-No, (2)
Color-Shape-Yes, (3) Shape-Color-Yes, and (4) Shape-Shape-No. Thus, when
the four trial-type effects are calculated, by subtracting response latencies for
history-consistent from history-inconsistent blocks of trials, one might expect
to see four roughly equal trial-type effects. In other words, the difference
scores for each of the four trial-types should be broadly similar. Critically,
however, the pattern of trial-type difference scores obtained with the IRAP
frequently differs across the four trial-types (e.g., Finn, Barnes-Holmes,
Hussey, & Graddy, 2016).
The REC model always allowed for the potential impact of the functions
of the response options on IRAP performances, in which there may be a bias
toward responding “Yes” over “No,” for example, and that this interacted
with the stimulus relations presented in the IRAP (Barnes-Holmes, Murphy,
et al., 2010). As such, one might expect to observe larger differences in
response latencies for trial-types that required a “Yes” rather than a “No”
response during history-consistent blocks of trials. In the case of the Shapes-
and-Colors IRAP described above, therefore, larger IRAP effects for the
Color-Color and Shape-Shape trial-types might be observed relative to the
remaining two trial-types (i.e., Color-Shape and Shape-Color). The REC
model does not predict, however, that the IRAP effects for the Color-Color
and Shape-Shape trial-types will differ (because they both require choosing
the same response option within blocks of trials), but in fact our research has
shown that they do (e.g., Finn et al., 2016, Experiment 3). Specifically, we
have found what we call a single-trial-type-dominance-effect for the Color-
Color trial-type. That is, the size of the difference score for this trial-type is
often significantly larger than for the Shape-Shape trial-type. This finding has
led us to propose an updated model of the relational responding that we
typically observe on the IRAP, which we will briefly outline subsequently. A
complete description of the model and its implications for research using the
IRAP is beyond the scope of the current chapter (but see Finn, Barnes-
Holmes, & McEnteggart, 2018). However, it is important to consider the
model here simply to highlight how an ongoing focus on relational
responding is continuing to contribute toward a behavior analytic approach to
human social implicit cognition.
In attempting to explain the single-trial-type-dominance-effect for the
Shapes-and-Colors IRAP, it is important to note that the color words we used
in our research tend to occur with higher frequencies than the shape words in
natural language (Keuleers, Diependaele, & Brysbaert, 2010). We therefore
assume that the color words evoke relatively strong orienting responses
relative to the shape words. Or more informally, participants may experience
a type of confirmatory response to the color stimuli that is stronger than for
the shape stimuli. Such responses may be interpreted as arising from the
Cfunc rather than the Crel properties of the stimuli.18 Critically, a
functionally similar confirmatory response may be likely for the “Yes”
relative to the “No” response option (because “Yes” frequently functions as a
confirmatory response in natural language). A high level of functional
overlap, or what we define as coherence, thus emerges on the Color-Color
trial-type among the orienting functions of the label and target stimuli and the
“Yes” response option. During history-consistent blocks of trials on the
IRAP, this coherence also coordinates with the relational responding that is
required between the label and target stimuli (e.g., Color-Red-Yes). During
history-consistent blocks, therefore, the trial-type could be defined as
involving a maximum level of coherence because all of the responses to the
stimuli, both orienting and relational, are confirmatory. During inconsistent
blocks, however, participants are required to choose the “No” response
option, which does not cohere with any of the other orienting or relational
responses on that trial-type, and this difference in coherence across blocks of
trials yields relatively large difference scores. The model we have developed
that aims to explain the single-trial-type-dominance-effect, and a range of
other effects we have observed with the IRAP, is named the differential
arbitrarily applicable relational responding effects (DAARRE) model
(pronounced “dare”). In the following sections we will outline the DAARRE
model for the Shapes-and-Colors IRAP and then consider an example of how
the model may apply to the results obtained from an IRAP that targeted
social cognition (i.e., race).

The DAARRE Model

A core assumption of the DAARRE model is that differential trial-type
effects may be explained by the extent to which the Cfunc and Crel
properties of the stimuli contained within an IRAP cohere with specific
properties of the response options across blocks of trials. Response options,
such as “Yes” and “No,” are referred to as relational coherence indicators
(RCIs) because they are often used to indicate the coherence or incoherence
between the label and target stimuli that are presented within an IRAP (see
Maloney & Barnes-Holmes, 2016, for a detailed treatment of RCIs).
The basic DAARRE model as it applies to the Shapes-and-Colors IRAP
is presented in figure 16.1. The model identifies three key sources of
behavioral influence: (1) the relationship between the label and target stimuli
(labeled as Crels), (2) the orienting functions of the label and target stimuli
(labeled as Cfuncs), and (3) the coherence functions of the two RCIs (e.g.,
“Yes” and “No”). Consistent with the earlier suggestion that color-related
stimuli likely possess stronger orienting functions than shape-related stimuli
(based on differential frequencies in natural language), the Cfunc property for
Colors is labeled as positive and the Cfunc property for Shapes is labeled as
negative. The negative labeling for shapes should not be taken to indicate a
negative orienting function; rather, it simply indicates an orienting function
that is weaker than that of colors. The labeling of the relations between the
label and target stimuli indicates the extent to which they cohere or do not
cohere based on the participants’ relevant history. Thus, a color-color relation
is labeled with a plus sign (i.e., coherence), whereas a color-shape relation is
labeled with a minus sign (i.e., incoherence). Finally, the two response
options are each labeled with a plus or minus sign to indicate their functions
as either coherence or incoherence indicators. In the current example, “Yes”
(+) would typically be used in natural language to indicate coherence, and
“No” (-) to indicate incoherence. Note, however, that these and all of the
other functions labeled in figure 16.1 are behaviorally determined, by the past
and current contextual history of the participant, and should not be seen as
absolute or inherent in the stimuli themselves.
 Figure 16.1 The DAARRE model as it applies to the Shapes-and-
Colors stimulus set. The positive and negative labels refer to the
relative positivity of the Cfuncs for each label and target, the relative
positivity of the Crels, and the relative positivity of the RCIs in the
context of the other Cfuncs, Crels, and RCIs in that stimulus set.19

As can be seen in figure 16.1, each trial-type differs in its pattern of

Cfuncs and Crels, in terms of plus and minus properties, that define the trial-
type for the Shapes-and-Colors IRAP. The single-trial-type-dominance-effect
for the Color-Color trial-type may be explained, as noted above, by the
DAARRE model based on the extent to which the Cfunc and Crel properties
cohere with the RCI properties of the response options across blocks of trials.
To appreciate this explanation, note that the Cfunc and Crel properties for the
Color-Color trial-type are all labeled with plus signs; in addition, the RCI
that is deemed correct for history-consistent trials is also labeled with a plus
sign (the only instance of four plus signs in the diagram). In this case,
therefore, according to the model, this trial-type may be considered as
maximally coherent during history-consistent trials. In contrast, during
history-inconsistent trials there is no coherence between the required RCI
(minus sign) and the properties of the Cfuncs and Crel (all plus signs).
According to the DAARRE model, this stark contrast in levels of coherence
across blocks of trials serves to produce a relatively large IRAP effect.
Now consider the Shape-Shape trial-type, which requires that
participants choose the same RCI as the Color-Color trial-type during
history-consistent trials, but here the property of the RCI (plus signs) does
not cohere with the Cfunc properties of the label and target stimuli (both
minus signs). During history-inconsistent trials, the RCI does cohere with the
Cfunc properties (minus signs) but not with the Crel property (plus sign).
Thus, the difference in coherence between history-consistent and history-
inconsistent trials across these two trial-types is not equal (i.e., the difference
is greater for the Color-Color trial-type) and thus favors the single-trial-type-
dominance-effect (for Color-Color). Finally, as becomes apparent from
inspecting figure 16.1 for the remaining two trial-types (Color-Shape and
Shape-Color), the differences in coherence across history-consistent and
history-inconsistent blocks is reduced relative to the Color-Color trial-type
(two plus signs relative to four), thus again supporting the single-trial-type-
dominance-effect.
At this point, it seems important to consider how the DAARRE model
might be used to interpret a single-trial-type-dominance-effect that was
obtained in an early IRAP study that focused on racial bias. Specifically,
Barnes-Holmes, Murphy, et al. (2010) reported four trial-type effects for a
study that presented pictures of white and black males carrying guns with
words related to safety and danger. The two critical trial-types in this context
were Safe-White and Dangerous-Black because participants were required to
press “True” during pro-white and “False” during pro-black blocks of trials.
The participants (who were all indigenous white Irish individuals) tended to
respond “True” more quickly than “False” on these two trial-types. However,
the size of the difference for the Safe-White trial-type was approximately
twice the size of the difference for the Dangerous-Black trial-type. In effect,
a single-trial-type-dominance-effect was observed. One could interpret this
result as indicating that participants were simply more certain about the
safety of white men carrying guns than they were about the danger of black
men carrying guns. Our recent work with the DAARRE model, however,
suggests that we should be more cautious in drawing such a conclusion. For
illustrative purposes, consider the following (speculative) interpretation.
Let us assume that the pictures of the white men and the safety words
possessed relatively positive evaluative functions, whereas the pictures of
black men and danger words possessed relatively negative evaluative
functions. If we translate these assumptions into a figure similar to the one
we used above for the Shapes-and-Colors IRAP, it quickly becomes apparent
that the single-trial-type-dominance-effect may have arisen, in part, from
differences in coherence across the two trial-types, rather than purely from
racially biased responses (see figure 16.2). Specifically, note that the Cfunc
and Crel properties for the Safe-White trial-type are all labeled with plus
signs; in addition, the RCI that is deemed correct for pro-white trials is also
labeled with a plus sign (the only instance of four plus signs in the diagram).
In this case, therefore, according to the model, this trial-type may be
considered as maximally coherent during pro-white blocks. In contrast, the
Dangerous-Black trial-type involves a “mixture” of Cfunc and Crel
properties; the pictures and words possess negative Cfunc properties but a
positive Crel property between them, and participants are required to choose
the positive RCI on pro-white blocks. Similar to the Shapes-and-Colors
IRAP, therefore, the difference between the size of the IRAP effects may be
attributed in part to a relative difference in the coherence among the Cfunc,
Crel, and RCI properties of the stimuli presented within the IRAP.
In making this argument it is important to understand that we are not
suggesting that IRAP effects are therefore irrelevant procedural artifacts that
do not reflect potentially important behavioral histories with regard to racial
differences and other social psychological phenomena. Indeed, the relative
difference in coherence that we have just suggested using the DAARRE
model requires that participants evaluate the pictures of the white men
positively and the black men negatively. What the DAARRE model provides,
therefore, is the potential for a more precise analysis of the functional
relations that are in play when participants are required to complete an IRAP.

Figure 16.2 The DAARRE model as it applies to the race stimulus

set.
Implications and Future Directions
On balance, there has been very little direct study of the variables that may
be used to change or manipulate the types of relational responding that
appear to be involved in implicit social cognition. One notable exception
was the study reported by Cullen et al. (2009) in which exposure to positive
and negative exemplars (of old and young individuals) was shown to impact
relevant IRAP performances. It seems important, therefore, that future
research in this area begin to focus on methods, techniques, and general
strategies for changing implicit social cognition that may be deemed
problematic in the natural environment. Indeed, it should be noted that
some success has been reported in manipulating IRAP effects, or their
relationship to other behavioral measures, in the clinical domain (e.g., Bast
& Barnes-Holmes, 2015; Bast, Linares, Gomes, Kovac, & Barnes-Holmes,
2016; Hussey & Barnes-Holmes, 2012; Leech, Barnes-Holmes, &
McEnteggart, 2017; Ritzert, Forsyth, Berghoff, Barnes-Holmes, &
Nicholson, 2015). Future behavior analytic research in implicit social
cognition is therefore well placed to explore and develop “interventions”
for changing the types of relational responding that seem to be involved in
implicit cognition (e.g., Mizael et al., 2016).

Conclusion
The current chapter aimed to provide an overview of the main body of work
on what may be described as implicit social cognition, but from a behavior
analytic perspective. Much of the empirical research has emerged from RFT
and a procedure (the IRAP) arising from the theory. Conceptual analyses,
such as those provided by the REC and DAARRE models, have also
emerged. The research, both empirical and conceptual, has clearly become
increasingly sophisticated since the initial studies reported in the early
1990s, and it appears our understanding of the relational responding
involved is gradually becoming clearer. As noted, however, further research
aimed at changing implicit social cognition is clearly warranted. Indeed, at
a time when there is increasing populist appeal for building walls and
withdrawing from a forty-year European Union, there seems to be even
greater urgency to tackle the human capacity for categorizing each other in
potentially negative and ultimately lethal and toxic ways.
Study Questions
1. Construct a list of the studies on implicit cognition and social
behavior that have been published within the behavior analytic
literature.
2. Name the methodology that has been used most extensively in the
studies you listed in response to question 1.
3. Provide a brief description of one implicit relational assessment
procedure (IRAP) that has been used to study implicit cognition and
social behavior.
4. Briefly describe the findings from three studies that used the IRAP
to examine implicit cognition and social behavior.
5. Briefly describe the relational elaboration and coherence (REC)
model.
6. Explain how the REC model proved to be limited in accounting for
IRAP findings.
7. Using an appropriate diagram, describe the differential arbitrarily
applicable relational responding effects (DAARRE) model.
8. Explain how the DAARRE model accounts for IRAP effects that the
REC model does not.
9. Provide an example of a DAARRE model that explains the findings
from a study on implicit cognition and social behavior.
10. Provide a brief outline of a study in implicit cognition and social
behavior that could be useful in achieving the goals of prediction
and influence.
 Chapter 17:

Perspective Taking, Empathy, and

Compassion

Carmen Luciano
University of Almería, Spain; Madrid Institute of Contextual Psychology
Bárbara Gil-Luciano
University of Nebrija, Madrid; Madrid Institute of Contextual Psychology
Adrián Barbero
Madrid Institute of Contextual Psychology; University of ICADE Comillas,
Spain
Francisco Molina-Cobos
University of Almería, Spain

Link and Preview

This chapter concludes the book and addresses three complex but
critically important repertoires of social behavior—perspective
taking, empathy, and compassion—from the perspective of
traditional psychology, radical behaviorism, and functional
contextualism. The authors’ analyses suggest a number of
directions for practical action and provide fruitful insights for
expanding behavior analytic research and practice into an area
that has traditionally been dominated by other branches of
psychology.
Perspective taking (PT), empathy, and compassion are three ordinary terms
that refer to concepts that have existed for centuries, but researchers only
began to study them in the last years of the twentieth century. In the last
decade, the number of published studies has shown an accelerated curve,
and most of them point to the positive impact of the behaviors defining
these terms in several domains, including developmental, social, health,
clinical, political, and economic arenas. The interplay of one’s behavior
with others’ human behavior involved in PT, empathy, and compassion is at
the core of complex inter- and intrapersonal abilities for problem solving,
effective living, and evolution. Although the massive trend of papers adds
relevant information, it also presents a lack of consensus either in defining
these abilities or in identifying the conditions under which they are
developed. It is our contention that the behavior of PT is the core ability in
empathy and compassion. The examples below illustrate this point.
Pedro is nine years old. One of his favorite games to play with his
friends is to portray their favorite fictional characters or their
heroes. He portrays his sister, his teachers, or people from several
different professions. When he represents them, he thinks and acts
just like they would, while discriminating his own perspective from
those of the character he is representing. We might say that Pedro
can place himself in the perspective of those characters. On one
occasion, Pedro arrives at school and sees that his friend Luis has
just wounded his knee. He is bleeding. Pedro goes to him and, even
though he does not feel pain in his own knee, he feels like he
understands Luis when watching him complain about the pain—
Pedro has gone through a similar situation before. He acts as
quickly as possible to make his friend’s pain go away. He is acting
the way others acted when he was injured, and in so doing, he is
compassionate toward Luis by assisting him.
Alfonso recently lost his sister in a terrible train accident. Among
other things, he feels disconnected and isolated from others, even
though everybody worries about him and tries to make him feel
better. Suddenly, he hears about his good friend Patricia, with whom
he has always had a special relationship. Patricia’s father has been
hospitalized for several months, struggling between life and death.
When they meet, Patricia tells Alfonso what is happening to her. “I
don’t seem to connect to practically anyone since I have been going
through my father’s situation.” Suddenly, Alfonso connects with
 Patricia. He feels his own pain, recognizes the similarity to what she
is feeling right now, and feels closer to her. He tells her this and
discovers that Patricia also feels closer to him.
Maria is an architect who builds houses with a purpose in mind, a
purpose derived from enjoying step-by-step creative building so that
people will enjoy living in warm houses with adequate lighting,
ventilation, and green spaces. She loves making buildings that can
offer other people not only beauty but usefulness. It is a
compassionate act, an act with personal meaning.
It is quite possible that, when reading these examples, most of us will
realize that they have common elements and that PT (I-You) is fundamental
to all of them. In Pedro’s case, PT is present for a specific purpose, but it was
also present when empathy emerged, in which case, his own emotion was
coordinated with the emotional behavior of another person. In another case,
PT is present when compassion is also present, like when the motivation of
helping others is driven by feeling connected to them. The key question is
whether some fundamental behavioral repertoire is critical to all of these
examples and, if so, what are the necessary behavior-environment functional
relations for these behaviors to take place?
These critical social repertoires described above seem to concern the
development of the self and self-knowledge, invoking a functional
perspective that Skinner specified long ago. Skinner (1974a, p. 35) wrote, “A
person who has been made aware of himself by the questions he has been
asked is in a better position to predict and control his own behavior.” One
relevant derivation would be that, without the development of these complex
abilities, our species would have been extinguished. That is, the ability to
discriminate one’s own emotions and the emotions of others, as well as the
chosen action and the responsibility that comes with it, is developed through
interactions with others. These interactions promote the building of flexible
patterns of interactions with one’s own emotions and thoughts so that
effective or meaningful actions take place. And perhaps in the absence of
these repertoires developing, the development of behavioral patterns defined
by suffering or psychological inflexibility may be more probable.
In this chapter, we address PT, empathy, and compassion from the
standpoints of traditional psychology, radical behaviorism, and functional
contextualism. In doing so, we track the verbal processes involved in all three
of these repertoires, along with their implications in applied contexts.

Perspective Taking
Perspective taking refers to the ability to respond to another’s perspective
without losing one’s own perspective. It is known as the ability “to be in the
other person’s shoes” or as “reading minds.” For instance, if Pedro, in the
above example, were asked, “How would you feel if you were your sister
when she wakes up in the middle of the night with a nightmare?” he would
feel somehow similar to his sister and might respond to his feeling by
helping his sister in some way.
PT is a key ability that is likely critical for effective daily living
interactions. PT has been found to be a core element in many areas that
involve relating to others as well as relating to one’s own emotions and
feelings. PT has been associated with effective social interactions, resolving
health problems, effective clinical interactions, improving economic
protocols, and spirituality and compassion (Barnes-Holmes, Barnes-Holmes,
Roche, & Smeets, 2001b; Biglan, 2015; Hayes, 1984; Luciano, Valdivia, &
Ruiz, 2012; Törneke, 2010; Strosahl, Robinson & Gustavsson, 2015; Wilson,
Bordieri, & Whiteman, 2012; Villatte, Villatte, & Hayes, 2016; M. Villatte,
Vilardaga, & Monestes, 2012). However, in spite of the relevance of PT
abilities and the large amount of published research that addresses it, only in
the last few decades has PT begun to be approached in a way that focuses on
the conditions under which it can be promoted.

The Traditional Approach to Perspective Taking

Early attempts to understand PT were dominated by research coming
out of the theory of mind (ToM) tradition, which refers to the capacity of
inferring the beliefs, intentions, and emotions of others in order to explain
and predict their behavior (Premack & Woodruff, 1978). In the classical
Sally-Anne ToM task (Baron-Cohen, Leslie, & Frith, 1985), children witness
a doll named Sally putting a marble in a box and then leaving the room.
Then, another doll named Anne moves the marble to a drawer. Children are
then asked where Sally will look for the marble when she returns. Children
who have not yet developed ToM abilities cannot distinguish what Sally
knows from what they know, and thus respond incorrectly that Sally will look
for the marble in the drawer. With some exceptions, most of the research on
ToM has been conducted under the cognitive conceptualization of levels,
identified as simple and complex visual PT abilities, as the basis of
knowledge and true and false beliefs (Howlin, Baron-Cohen, & Hadwin,
1999). ToM has commonly been used to evaluate and train the development
of PT abilities, and ToM deficits have been shown in children with autism
spectrum disorder (Baron-Cohen et al., 1985) and in persons diagnosed with
schizophrenia, schizotypy, and social anhedonia (Corcoran, Mercer, & Frith,
1995; Villatte, Monestès, McHugh, Freixa i Baqué, & Loas, 2008). A large
amount of ToM research has been published on PT, but all of it has been
dedicated to documenting deficits in PT, as opposed to remediating them.
The ToM model has not been fruitful in leading to effective intervention
research, arguably because it did not provide the key learning processes
involved in ToM that affect PT (see Barnes-Holmes et al., 2001b; McHugh,
Stewart, & Hooper, 2012; and Montoya, Molina-Cobos, & McHugh, 2017,
for reviews). However, the ToM literature contributed substantially to
drawing research attention to the topic of PT, and as well as describing the
various overt behaviors involved in PT. This helped lay a foundation for a
functional approach to the study of PT.

A Functional Approach to Perspective Taking

Functionally speaking, PT is the ability to notice one’s mind (i.e.,
private behaviors and stimuli) while inferring others’ minds, without losing
perspective of oneself, the one who is experiencing all the thoughts and
emotional content across time and places as a unique experiential locus or
self (Barnes-Holmes et al., 2001b; Hayes, 1984; Kohlenberg & Tsai, 1991;
Luciano, Valdivia-Salas, Cabello, & Hernández, 2009; Skinner, 1974a; J. L.
Villatte, Villatte, & Hayes, 2012).
Early on in the behavioral tradition, Skinner (1974a) pointed to the
relevant role of the self and self-knowledge. He wrote that the self is a locus
where a diversity of genetic and environmental conditions converges to
organize a unique and characteristic repertory. Earlier, Skinner (1945)
provided some of the conditions that give rise to private events as part of the
self. And moreover, he emphasized the role of self-awareness as central to
the human condition. He wrote that the person who is conscious of himself or
herself is in a better position to predict and control his or her own behavior.
These ideas crystallize in the proposal of the self of Kohlenberg and Tsai
(1991), who specify the problems of the self and the therapy created to
overcome such problems, functional analytic psychotherapy (FAP).
However, there was no specific research on this complex behavior, leaving
the interactions to prevent difficulties in the self’s behaviors and the client-
therapist interactions in an unknown scenario. That is, the verbal processes
involved between the therapist’s and the client’s behavior to build new
repertoires were unknown.
The main problem in the difficulty researching self-behaviors probably
had to do with the consideration that discriminating one’s own behavior was
a simple discrimination, rather than one that involved verbal processes. But
for that behavior to take place, perspective taking is needed between the one
performing the discrimination and the thing being discriminated so that the
functions of these events are transformed (Barnes-Holmes, et al., 2001b). Not
an easy problem in those times, when there was no effective
conceptualization for this complex behavior. The early behavioral view of the
self was consistent with the way these behaviors seem to be produced.
However, research in this area was almost absent until behavior analysts took
a contextually oriented approach to PT and conducted research on the
emergence or derived production of functions and behavior. This set the stage
for the development of relational frame theory (RFT) to give an account of
language as relational behavior. And it was then that one of the relational
repertoires of behavior, deictic framing, was developed and became
available to provide a behavior analytic account of PT—one that could be
applied to different human behaviors including the problems of the self and
the therapeutic relationship.

Perspective Taking as Deictic Framing

The RFT perspective states that language is framing, which means
responding to one stimulus in terms of the contextual relation established
with another stimulus. As described in earlier chapters, framing is a relational
operant established through multiple exemplars to abstract a number of
relational cues that, when learned, can be arbitrarily applied to any type of
stimulus. New relations derive from the effects of this application and
indirectly provide or change stimulus functions. Consider, for instance, a
child who has learned the relational cues “more than,” “less than,” and “same
as” and is told, “These two toy cars are the same color, red, but the size is
different. The big car is as fast as the one you play with at home, but the
small one is faster than the big one. Which of them would you choose if you
want the fastest?” A child with an intact framing repertoire can respond by
picking up the small car the first time without any previous interactions with
that car or that particular question. He will also probably be able to learn the
rule “These cars are both red, but the small one goes faster than the big one.”
That is, the small car has acquired a derived appetitive and discriminative
function on the basis of the comparative relation applied to the two cars and
the sameness relation with directly established functions (the one he plays
with at home).
Fully formed human verbal repertoires involve responding to a number
of relational cues: coordination (e.g., same as), distinction (e.g., different),
opposition (e.g., opposite), comparison (e.g., more than, less than),
hierarchical (parts of something, or inclusion), temporal (now, then/before,
later), causality (if…, then), and deictic (I-You, Here-There, Now-Then).
Advanced framing ability goes beyond a particular relation to combinations
of relations, as is the case when we make or understand analogies or
metaphors, or we form problem-solving rules and follow them. The relational
behavior that is particularly relevant to PT is what is known as deictic
framing. Deictic framing specifies a relation in terms of the perspective of
the person who is behaving (Barnes-Holmes et al., 2001b), for instance, the
perspective I versus Other; the perspective I before and I later, or You before
and You now; or the perspective I here and My stomachache, there. Deictic
framing is at the core of most complex human behaviors that involve
responding to the relation between oneself and others.
Three deictic frames were identified almost twenty years ago (Barnes-
Holmes et al., 2001b): I versus YOU (Other), NOW versus THEN, and
HERE versus THERE. However, the most relevant point is that they are not
separated, but rather they promote the unique perspective of the I as invariant
(even when noticing the Other perspective) and always occurring in the
NOW (even when noticing a THEN, in the past or future), and always
occurring HERE, NOW (even when noticing I-THERE in other places;
Barnes-Holmes et al., 2001b; Luciano et al., 2009). This occurs through the
many opportunities that other people in the community provide for the
combination of these perspectives. Thus, the perspective of I-Here-Now
emerges as invariant through any other perspective that might be involved,
for instance, I-Here-Now and You-Here-Now (e.g., I am here, now, in the
kitchen and I notice that you are in the bathroom), or I-Here-Now and I-
Here-Then (e.g., I am making the pizza now and I will clean the kitchen
later). The establishment of deictic relational cues is a social learning
process. In other words, across multiple exemplar training provided by the
people around us who ask questions, model responding, and give feedback,
we learn to abstract the deictic relational cues I-You, Here-There, and Now-
Then in multiple combinations. The questions that Skinner identified to
establish self-knowledge are the key verbal interactions that establish deictic
framing in young children, for instance, What I am doing now? What are
YOU doing now? and What is HE doing now? What am I doing Here? and
What did I do Yesterday?
The implications of deictic framing for understanding human behavior
are enormous because deictic cues are applied in thousands of ways in daily
life. That is, when this repertoire is fluently in place, we can apply the deictic
cues when we understand or formulate rules, when we learn to differentiate
our own emotions from those of others, when we look back or forward in our
life, when we know what the other person knows or sees, when we feel we
connect to other people, and so on. We apply deictic cues when we feel what
it is being said by another person or we formulate questions and statements
such as the following:
“What are you going to do later?”
“You and I cried before, and now I am still crying and you are
laughing.”
“I was better before than I am now.”
“If I were where you are now, I would be able to see what you see,
but I cannot from here.”
“How do you think your brother is feeling when you pick up his car
without asking permission?”
“I cannot feel what you are feeling now. But, you know, if I were in
your shoes and had the same history you have had, I would be
feeling the same.”
Finally, beyond the core role of deictic framing to promote PT and to
understand the formation of beliefs about oneself or others is the most
relevant role that deictic framing promotes: it is the formation of the self as a
unique and integrative context for any thought or emotion that develops. That
is, deictic framing allows the formation of a hierarchical or inclusion relation
between oneself and one’s own behavior, which is essential for the human
being to be conscious, to be mindful, and to have a sense of spirituality. And
this repertoire is what is likely needed to respond to one’s own behavior
without responding as though I AM it. In other words, the ability to respond
to oneself as a context in which behavior occurs is broader and more flexible
than responding to oneself by labeling oneself as one’s behavior. To respond
to a momentary emotion as though there is some space between oneself and
that emotion likely allows the opportunity for a flexible response to that
emotion. For example, when one experiences rage, one might notice that
emotion as something one is experiencing, which could leave room for a
variety of responses, rather than defining oneself by the emotion (e.g., “I am
just the kind of person who can’t control their anger,” which may be more
likely to evoke a rigid, maladaptive response (e.g., aggression). As a whole,
flexible deictic relational framing likely facilitates problem solving and
interpersonal relations skills such as empathy and compassion.

Empirical Research on Deictic Framing and

Implications
Research on deictic framing has increased in recent years and has been
extended to different areas. The developmental domain was the first to be
addressed by the analysis of protocols of deictic combinations in order to
evaluate and train PT in normal developing children and in children with
autism spectrum disorder (ASD). The protocol was designed to target the
three deictic relations, I-You, Here-There, and Now-Then, through three
levels of complexity (simple, reversed, and double-reversed). A simple trial
of I-You would be If I have a ball and YOU have a cookie, then what do I
have and what do YOU have? Similar simple trials are designed to respond to
Here-There (e.g., I am standing HERE at the door and you are standing
THERE at the table) and to Now-Then as relating I-Now and I-Then (e.g.,
Now I am playing with the ball and yesterday I was reading).
Reversed trials require more complexity. For example, a reversed I-You
trial would be I have a ball and YOU have a cookie, however if YOU were me
and if I were YOU, then what would I have? A reversed Here-There example
would be I am HERE at the door and you are THERE at the window, but if
THERE were HERE and HERE were THERE, then where would I be, and
where would you be? The double-reversed trial, in which two deictic
relations are reversed, requires still more complexity. Following the above
example, the trial would be I am HERE at the door and you are THERE at
the window, but if I am YOU and YOU are me and if HERE is THERE and
THERE is HERE, then where would I be and where would YOU be?
Research has supported a general developmental sequence in the
learning of these levels of complexity, with older children and adults showing
fewer errors in spatial and temporal reversed and double-reserved relations
(McHugh, Barnes-Holmes, & Barnes-Holmes, 2004). Children with ASD
have shown difficulties with reversed deictic relations, especially the Now-
Then relations, which have been correlated with poor social skills (Rehfeldt,
Dillen, Ziomek, & Kowalchuk, 2007). The deictic protocols have also been
adapted to evaluate false-belief and deception repertories, showing a clear
developmental trend in both areas (Barnes-Holmes et al., 2001b; Heagle &
Rehfeldt, 2006; McHugh et al., 2004; Weil, Hayes, & Capurro, 2011). As a
whole, the available research presents a fruitful panorama in remediating
deficits in PT. However, at the same time, more natural protocols are needed
to increase generalization to the natural setting beyond the specific deictic
tasks designed for the protocols (see reviews by McHugh et al., 2012;
Montoya et al., 2017; Mori & Cigala, 2016). In this line, researchers use a
precise analysis of the deictic relations to detect flexibility in PT using
different tasks, such as, for instance, the implicit relational assessment
procedure (IRAP; Barbero-Rubio, López-Lopez, Luciano, & Eisenbeck,
2016).
Finally, M. Villatte et al. (2012) used the deictic protocol designed by
McHugh et al. (2004) to evaluate PT in persons diagnosed with
schizophrenia in comparison to persons without such a diagnosis. The
performance of the former group was poorer on all reversed and double-
reversed deictic relations, and they had more difficulties in taking the
perspective of another and correctly attributing the intention of others (M.
Villatte et al., 2012). To summarize the above, deictic protocols seem to be
sensitive to measuring and treating PT abilities and appear to provide a
behavioral foundation underlying classical ToM tasks and other attribution
tasks, where no specific processes had previously been identified.
Beyond the relevant impact of the deictic protocol for evaluating and
teaching PT or ToM, deictic framing has been identified as being at the core
of the process responsible for the formation of thoughts and emotions and,
more importantly, of the self as an integrated set of behavioral repertoires.
More specifically, deictic framing has been found to be the key ability for the
formation of self-rules (thoughts and emotions about oneself and the world
and others around one). Developing content about one’s own behavior and
life require the deictic-I in order to experience and notice that it is a behavior
about oneself. To feel in one’s own skin and say sentences to oneself such as
I am very good at this and they are not, I am not good enough, Nobody loves
me, I will never have a successful life, I am a person who loves to help others
is not the same as repeating sentences we all know. It means that emotions or
feelings become present according to one’s history and one can feel who is
feeling them, in contrast to other agents, moments, and places. The formation
of thoughts about oneself is only possible through multiple interactions with
others where the I-You, Here-There, and Now-Then are interchanged.
Consequently, the content about oneself and its function depends on one’s
personal history.
The final point is that deictic interplay is the key to establishing the
content and functions of self-rules, and it is the key to establishing a
hierarchical perspective between the self-rules and the very person who is
noticing the self-rule (the deictic-I). The advantages of self-knowledge and
the self have already been mentioned, although the processes were not
established. The main point is that having a fluent hierarchical deictic-I
repertoire establishes a context where one can be less attached to or trapped
by emerging self-rules. That is, such a repertoire allows us to overcome these
momentary events and to connect with more relevant and meaningful
functions. In other words, with this ability intact, we would react by noticing
and integrating the self-evaluations that emerge and we would respond under
the influence of those other functions.
Based on this analysis, researchers have designed deictic framing
protocols to promote the deictic/hierarchical framing of one’s own behavior.
In the first of these studies, Luciano et al. (2011) proposed a relational-based
account of defusion interactions, typically used in acceptance and
commitment therapy (ACT; Hayes, Strosahl, & Wilson, 1999). This study
compared a deictic-I protocol with a deictic-I protocol plus hierarchical cues
in adolescents with problematic behaviors as participants (Luciano et al.,
2011). Results showed that the latter protocol had a greater effect than the
former on reducing the frequency of problematic behaviors and promoting
psychological flexibility at a four-month follow-up. These results have been
replicated, improved, and extended using specific experimental tasks (such as
analogues of daily situations). For instance, Gil-Luciano, Ruiz, Valdivia-
Salas, and Suarez-Falcon (2016) analyzed the effect of these deictic protocols
on a behavioral pain tolerance task; López and Luciano (2017) analyzed their
effect on attention tasks; and Floody, Barnes-Holmes, Barnes-Holmes, and
Luciano (2013) tested the effect on the functions of negative self-evaluative
rules. So far, these authors’ findings suggest the significant role of
hierarchical and deictic-I framing in changing functions to self-content in
establishing the ability of noticing and behaving according to what matters.
Deictic framing seems to be key in many other psychological phenomena
such as empathy, compassion, psychological flexibility, spirituality,
transcendence, and mindfulness (Hayes, 1984; Törneke, Luciano, Barnes-
Holmes, & Bond, 2016; J. L. Villatte et al., 2012; Wilson et al., 2012).
For instance, psychological flexibility has been defined as behaving on
the basis of hierarchical framing with the deictic-I (Törneke et al., 2016).
This can be explained as the ability of noticing my own ongoing emotion and
thoughts from the deictic I-Here-Now and my behavior, I-There, at the same
time that noticing the integration or inclusion of such behavior as part of ME
as the context for the higher-order appetitive functions (valued or meaningful
functions) be present and to act accordingly. At this point, it is worth pointing
out that several experimental-clinical studies have already been designed
according to this view with very positive results (e.g., Ruiz, Riaóo, Suárez, &
Luciano, 2016). Moreover, the relevant role of deictic and hierarchical
framing has recently been used in a very specific way to approach the
therapeutic relationship and clinical problems (Barnes-Holmes et al., 2018)
that are central in client-therapist interactions promoted in FAP. In a similar
but less specific manner, this role is also present in the relational approach to
clinical interactions of J. L. Villatte et al. (2016) and of Törneke et al. (2016).
Putting all this together, deictic framing is involved in the formation of
rules, whatever their content, and general beliefs about oneself as a
hierarchical experience of ME with the functions coming from one’s personal
history. And, at the same time, deictic framing is needed to establish the
experience of ME noticing whatever self-content is present at the moment
(e.g., Barnes-Holmes et al., 2001b; Luciano et al., 2009; Wilson et al., 2012).
As mentioned, becoming fluent in framing one’s thoughts or emotions in the
deictic I-Here-Now and one’s thoughts and feelings There-Now establishes a
hierarchical perspective (in other places called self-as-context) that enhances
flexible repertories. This flexible repertoire serves an important function
within a verbal community because we cannot escape from interacting with
our own feelings, thoughts, and functions emerging in the context of other
people’s behaviors and events in the world according to our own personal
history. Put simply, it is very easy to become attached to such functions of
one’s own behavior and, consequently, to act blindly concerning what
matters. However, learning to frame in the hierarchical/deictic-I is central to
generating psychological flexibility because of its relevance to so many other
verbal relational repertoires, as described above.
To sum up, most of the studies on promoting PT have been carried out
under the umbrella of deictic framing, pointing not only to the classical ToM
tasks but also to the design of more naturalistic deictic protocols. In addition,
deictic framing has been identified as being at the core of the emergence and
extension of networks of thoughts and emotions, for better and for worse—
that is, for the emergence of both positive and negative self-content and for
the emergence of a secure or insecure sense of self. Moreover, a good number
of experimental and clinical protocols have shown that the combination of
deictic and hierarchical framing of one’s behavior generates the functional
hierarchical context where one connects with one’s own emotions and
thoughts and experiences the awareness of behaving according to what
matters to oneself.
To conclude, PT as deictic framing seems to be at the core of prosocial
behavior for good, and this is for better but it might be for worse (the former,
as Skinner pointed out years ago, because it places the person in a better
position to behave according to what matters; the latter, because PT might
also be used for purposes other than prosocial ones). We will now turn our
attention toward the deictic processes that might be involved in empathy and
compassion.

Empathy
Empathy is a term that was introduced into the German language around
1858, originating from the Greek term “empatheia,” meaning “feeling-in,”
as in merging with an object or being connected to some kind of feeling or
emotion. The meaning of “empathy” evolved to underlining the
interpersonal connection, and then emphasizing its emotional component;
this latter aspect was formulated by Titchener, who pointed to PT as the
relevant ability (Batson, 2009; Lanzoni, 2015; Vilardaga, 2009). The
interest in studying empathy during the twentieth century reflects its
definition in Reader’s Digest in 1955 as the “ability to appreciate the other
person’s feelings without yourself becoming so emotionally involved that
your judgment is affected” (Lanzoni, 2015).
Today, the large number of studies that have been published on empathy
show a vast panorama that crystallize in the lack of consensus about its
definition and, more importantly, about the conditions under which empathic
behavior develops and could be changed. As Batson (2009) states explicitly,
after decades of studying empathy, the difficulty lies in establishing what is
being considered as empathy by different researchers. For example,
sometimes empathy means labeling or imagining another’s thoughts and
feelings, whereas other times it means adopting another’s position and
feeling like the other person feels.
Empathy appears significant in many areas of human functioning. A
brief overview shows the link between empathy and emotional intelligence
and with functioning well in daily living (e.g., Eisenberg, 2000; Goleman,
1998). It has also been correlated with prosocial behaviors and conflict
resolution (Batson, 2009; Garaigordobil & Maganto, 2011), and with
improving sports and artistic performances (Sevdalis & Raab, 2014).
Conversely, deficits in empathy have been found in children diagnosed with
autism (Baron-Cohen et al., 1985) and in people diagnosed with psychosis
(Vilardaga, Estévez, Levin, & Hayes, 2012; M. Villatte et al., Monestès,
2012). However, the lack of consensus in defining empathy makes analyzing
and studying it a very difficult task.

Traditional Approach and Empirical Context

The numerous different definitions, conceptualizations, measures, and
methodologies used to study empathy prevent a precise analysis and
interpretation of the results obtained across studies. These same problems
hinder the interpretation of studies ­combining neuroscience and empathy. For
example, the review by Zaki and Ochsner (2012) shows that most of the
studies have used measures focused on what is happening in the brain when
an individual watches videos of people talking about highly personal
emotional events. The question here is whether the reaction to such videos is
a precise example of empathy. In other words, what is the conceptualization,
focus, or goal by which empathy is being defined in these studies? The same
problems might be found in those areas oriented to improving empathic skills
in practitioners in medical and clinical contexts, where empathy is defined,
for instance, by active listening, initiating statements, and discussing and
writing about empathy (e.g., Brunero, Lamont, & Coates, 2010; Gerdes,
Segal, & Lietz, 2010).
However, even with different definitions of the term “empathy,” deficits
in empathy have been addressed in children using the ToM protocol (Baron-
Cohen et al., 1985) with no evidence of the positive effects of remediating
empathic responding beyond the specific task that was directly trained. That
is, there was no generalization to novel empathy situations or false-belief
tasks, or any signs of promoting social interactions (e.g., Golan & Baron-
Cohen, 2006; Silver & Oakes, 2001). Behaviorally oriented program
protocols have shown similar results (e.g., Leblanc, Coates, et al., 2003). This
means that targeting emotional labels while training in PT does not help the
learner to be more effective in responding to their emotional interactions. As
Vilardaga et al. (2012) pointed out, the problems with these empathy
protocols is that there is no experimental approach to the psychological
phenomenon of empathy that allows the production of rules and formulas
with increasing levels of precision, scope, and depth. In other words,
empathy has often been addressed as either a hypothetical internal construct
or an oversimplified overt behavior.

A Relational Framing Account of Empathy

Even the most general meaning of empathy as understanding how others
feel, sharing feelings, and responding to those feelings without losing one’s
own perspective (Valdivia-Salas, Luciano, Gutiérrez, & Visdómine, 2009)
reveals dark zones. For example, it is one thing to feel like the other person is
feeling, which might be conceptualized as a coordination between one’s own
feelings and the other’s feelings. But the function that such coordination
promotes in the very person who is behaving this way is a different thing.
That is, feeling like the other person may feel does not have the same
function for everyone, and for the person themself, because it depends on the
person’s history in the present context. For example, let’s imagine that Maria
feels loneliness framed in coordination with Peter’s signals of loneliness to
Maria, but, for Maria, this feeling has the function of remaining in the
situation and doing something for Peter. However, for Carlos, the same
feeling has avoidance functions, and he withdraws from contact with Peter to
reduce the effect of the feeling in coordination between them. And, for
Esther, these same feelings entail actions to manipulate Peter’s behavior.
Perhaps the lack of consensus about what empathy is goes hand in hand
with the need for differentiating two elements. Specifically, we should
differentiate feeling an emotion—in the sense of the function that occurs in
response to the other’s emotional behavior—from the function of this
emotion and the particular action under its control. Differentiating the two
processes should be a fruitful alternative when evaluating empathic behavior
and designing intervention protocols. Following this line, Valdivia-Salas et
al. (2009) proposed a promising protocol to establish empathic behavior by
translating the core features of empathy into behavioral-relational language
with deictic framing as the core ability. In this proposal, deictic framing
occurs throughout the process of experiencing the coordination of emotion
and responding to it without losing one’s own perspective. Three consecutive
strategies are indicated to achieve empathy.
A first step is the establishment and understanding of one’s own
emotions. This process is promoted by agents in the verbal community so
that we learn to verbally discriminate our own private events from others’
behavior. That is, the community reinforces the coordination of internal states
(e.g., pain when I fall) with a label (e.g., I feel pain) with those of the other
(e.g., my brother did not fall and feels no pain). Through multiple exemplars
of these types of discriminations, we not only learn to label a name, but we
also learn to respond in accordance with deictic framing, such as I-Here-Now
and my Pain-There (in the knee), and at other times, I-Here and my Pain-
There (in my head), and with other events, I-Here and my Hungry-There, and
so on. And very importantly, through these multiple exemplars of
discriminating one’s own behavior, we learn to discriminate our own
experience from the unique perspective of I-Here-Now, so that the self is
developed as a hierarchical perspective of many events occurring to me
across time and places. This process occurs along with the discrimination of
the “YOU” and the respective discriminations.
A second step is the establishment and understanding of the emotions of
others as another element for empathic responding. This requires a shift in
one’s perspective to that of another’s perspective but without losing the
former—that is, shifting from I-Here-Now to “as if I were you now” (e.g.,
responding as if now I were my brother), or “if I were you then and then
should be now” (e.g., responding as if I were my brother when he was a
baby), and, importantly, without losing the perspective of I-Here-Now.
Beyond this shift of moving around the deictic reversal I-You, one must learn
to discriminate cues that suggest that some private events are occurring in
others. For example, if Luis’s mother cries, Luis might feel that she is hurt on
the basis of coordination between a particular experience of being hurt with
the label “being hurt” and a subsequent action of crying when he was hurt.
Luis might also have learned the conditional rule “If something hurts, then I
cry.”
A third step is to learn to respond empathically by virtue of
differentiating my own experienced feelings, connected to those of the other
person, while at the same time realizing that the experience I am feeling is
occurring due to my history with similar experiences, by definition not
occurring now, but nonetheless this feeling is occurring, or the functions
stemming from it are present. Expressed in relational terms, the following
deictic framings are in place: “I-Now-Then (before) and I-Now-Here” and “I-
Now-Here and You-Now-Here,” and I can react to my emotions in a useful
way. That is, when deictic framing develops, the functions of one’s behavior
(the feeling I am experiencing) change and alter the feeling I am having Now
as if being There-Then. This increases the likelihood of valued functions
connected to the deictic I-Here-Now emerging and, then, of behaving
accordingly. For instance, in our example, the transfer of the emotional
functions I-You might make Luis reexperience his pain as being similar to his
mother’s pain and, at the same time, know that he has not hurt himself like
his mother has and that he is the one noticing such a pain as I-Here and my
pain as There. As a result, he will be in a good position to help his mother by
asking for help instead of crying, as might be the case if he had been hurt at
that moment or if he had not learned to differentiate himself from his own
pain.
In spite of the conceptualization of empathy from the behavioral-
contextual perspective, no specific and basic experimental designs have yet
been designed to analyze these components so that clear rules for promoting
empathic behavior can be experimentally established. Having said this, we
think that the conceptualization is ready to be used to design protocols for
teaching empathy, which can then be tested empirically.
 Summing up, different researchers have noted the variability in the
definitions, conceptualization, and measures used to study empathy, or in the
applied protocols promoted to improve empathy. However, despite the
confusion, it seems clear that PT skills are at the core of empathy. Expressed
more technically, deictic framing establishes the main conditions for
experiencing one’s own feelings as if sharing emotions in the presence of
another’s emotion. However, empathetically responding to such emotional
feelings is another context, which requires more functions to be brought into
play. The point here is that responding to one’s own emotions as sharing
another’s emotion might have different functions depending on the history of
the person empathizing. Feeling in the sense of sharing emotions might have
avoidance and limiting functions, or it might have avoidance or approximate
functions oriented to, or in the context of, more relevant and meaningful
functions. For some, it might be the case that one’s own history brings
functions to manipulate or cheat others in such contexts.
To conclude, clarification of which behaviors are defined under the
rubric of empathy is needed. Beyond this, it is clear that the development of
empathy should run in tandem with the verbal sophistication connected to PT
in order to develop an effective repertoire of responding to one’s own
emotions in relation to others. The implications of not developing empathic
actions are substantial. In our view, the most important implication is that the
development of a unitary self in responding to others will be highly
compromised. In other words, interpersonal problem solving and, more
generally, prosocial interactions may not be established. The relational
literature on deictic and hierarchical responding might open the door to
designing experiments to isolate the conditions for adequate coordination
between one’s own and others’ behavior, as well as the conditions to frame
one’s own feelings and thoughts in such a way that the relevant functions for
effective action are in place.

Compassion
The word “compassion” has its origin in “com” (with/together) and “pati”
(to suffer), meaning to suffer together with. Most definitions of compassion
share the idea that it is a feeling of understanding someone else’s suffering,
accompanied by the urge to help or the desire to alleviate the other’s pain,
meaning to suffer and love with. Broadly speaking, the identification with
others through compassion seems to lead to increased motivation to do
something in an effort to relieve their suffering. That is, it induces feelings
of kindness and forgiveness, which might allow one to get in touch with
such feelings, integrate them, and shift attention to acting toward valued
goals. Although lacking unanimity about the core concept or agreement on
its nature, the current literature on compassion is massive (e.g., see Goetz,
Keltner, & Simon-Thomas, 2010, for a review). On the one hand,
compassion has been considered a synonym for “empathic distress”
(Ekman, 2003). From this perspective, compassion would be a vicarious
experience of distress in response to another’s suffering. On the other hand,
compassion has been related to emotional responses that normally arise in
the context of witnessing another person’s suffering and embracing them in
a benevolent action toward another in need. These responses have been
gathered under several different names, like “pity,” “sympathy”, “empathy,”
“kindness,” “caring,” “warmth,” and many more (see Wispé, 1986, for a
review).
The relevance of compassion has grown significantly, as will be
indicated later. From a broad social perspective, compassion has been
claimed as one of the greatest virtues of the human condition, and it has
frequently been advocated in all the major religious traditions. It has also
appeared on the scene of political and business organizations, in evolution
studies, and as a primary component in human social interactions. That is, it
has been the core of psychological treatments and is actually one of the
relevant cornerstones of the therapist-client interaction in the contextual
therapies previously discussed in this chapter.

Traditional Approach and Empirical Status

A brief view of the vast research conducted on compassion shows that
most of it is linked to empathy. Beginning with the level of analysis of
neuropsychology, several brain activities have been related to strong feelings
of compassion in the context of observing pain in others (Immordino-Yang,
McColl, Damasio, & Damasio, 2009). Perhaps more informative than the
studies correlating brain activation with the observation of others’ pain are
the studies correlating the reward of neuro-activation when subjects are
performing compassionate acts (Rilling et al., 2002). However, although
these studies reflect an advancement in detecting brain activation while
persons are behaving in a certain way, it is unclear what the participants are
doing. Moreover, the processes by which compassionate acts occur and
activate the brain are unclear.
The evolutionary perspective is another domain where compassion has
been found to play a relevant role. For instance, compassion seems to be
more likely to occur when one is related to another person in need and/or
when the benefits of being compassionate exceed the potential costs
(Henrich, 2004). In political domains, compassion has been identified as a
characteristic of democratic societies, in that they value the recognition of
and identification with others’ suffering, as well as knowledge of the sources
of suffering and a commitment to reduce it. In the opposite context, the
absence of compassion has been found to be a key element of social atrocities
(e.g., genocides) by aggressors whose behavior is under the control of
exclusion rules such that some persons are nonhuman or are people who do
not belong to “good” races (e.g., Hoffman, 1981). This is why the value of
connecting with others is at the core of compassion and self-compassion
(Gilbert, 2010; Ricard, 2015). Consistent with this view, McHugh, Barnes-
Holmes, Barnes-Holmes, Stewart, and Dymond (2006) found that the
identification with another codevelops with the ability to adopt the other’s
perspective.
More concrete studies on compassion spread from medicine to
psychology. For instance, compassion displayed by physicians plays a
mediating role in the effects of sickness and suffering on human behavior. In
the same vein, compassion and empathy between patients and their
caregivers seems to be a primary motivation to relieve the physical, mental,
or emotional suffering of each other (Pedersen, 2009). In spite of the positive
correlations between compassion and different social improvements, a dark
side has been identified and labeled “compassion fatigue.” This has been
found in professionals interacting mostly with suffering (e.g., oncology
patients), who become entangled in the understanding of others’ feelings
(Figley, 1995). However, to overcome this, researchers and medical experts
have promoted training among practitioners to improve the ability to tolerate
one’s own pain (e.g., Kraemer, Luberto, O’Bryan, Mysinger, & Cotton,
2016).

Distinguishing Empathy and Compassion

 Still, it is difficult to differentiate empathy from compassion. As
Sinclair, Hack, McClement, and Raffin-Bouchal (2016) concluded, while the
key feature of empathy is an affective state that is isomorphic with the
affective state of another, the defining feature of compassion is a prosocial
motivation or behavior that is not necessarily linked with such an isomorphic
affect. At the same time, for effective prosocial actions to take place, people
must somehow feel something related to what others are feeling, either by
understanding the sources of suffering or, perhaps more effective, by having
the ability to picture themselves in similar experiences.

The Functional Contextual Approach

The role of compassion is perhaps more explicit now than before, and it is
becoming a key clinical element in psychological therapy, especially in
contextual therapies. In this arena, the term “compassion” opens the way to
self-compassion, and it has emerged strongly with a myriad of interactions
focused on promoting a context of compassion and self-compassion. FAP
(Kohlenberg & Tsai, 1991) and ACT (Hayes, Strosahl, & Wilson, 1999,
2012) involve frequent intimate, empathic, or compassionate interactions
with the client so that the therapist learns the repertoire of being open to
feeling whatever might emerge in a genuine interaction with oneself
(Gilbert, 2010), and so that client and therapist can feel that they are in the
same boat to do the job of building a flexible or effective repertoire.
For this to happen, protocols inherently link openness to empathy and
compassion toward others and oneself (Hayes, Strosahl, & Wilson, 1999,
2012; Strosahl et al., 2015; Villatte et al., 2016; Wilson et al., 2012).
According to this approach, a repertoire of self-compassion is needed, as
defined by the integration of one’s own whole ongoing experience—the
thoughts and emotions that emerge in coordination with the client’s
experience—and responding under the control of higher or valued significant
functions oriented to alleviating the suffering of others. As a collateral effect,
these kinds of interactions with one’s painful and/or appetitive emotions and
thoughts in the context of personal valued actions repeatedly lead to a feeling
of personal satisfaction and fulfillment. Ultimately, compassionate acts are
equivalent to valued actions, and these actions show a great variability. For
instance, Sinclair et al. (2016) point to the variability of actions that are
categorized as compassionate ability, and they conceptualize compassion as
being as unique and particular as the individuals who enter a profession or
engage in research or political activities.
Considering compassion as a pattern or series of different actions under
the control of significant or valued functions, it can be categorized as a
functional class of behavior built across personal historical interactions.
Analyzing behavior in this way emphasizes the function of the actions
instead of their topographies or formal aspects. This is useful because it
establishes a bridge to understand the variability of actions that can be
displayed to connect with another human being, and to understand the key
point in compassion as the different actions under the influence of a
motivating, valued, significant, and personal function. This ability is at the
heart of psychological flexibility, of psychological well-being, and it opens
the door to somehow connecting different terms and behaviors—in this case,
compassion and psychological flexibility—as a functional class of valued or
significant personal actions.

Implications and Future Directions

As we have suggested throughout the chapter, perspective taking, empathy,
and compassion are topics with important implications for understanding
human behavior, future research, and the provision of therapeutic support.
In a nutshell, deictic framing is involved in all the applications where we
need to identify our own and others’ mental and emotional states to act
effectively. This extends to building an empathic, compassionate, and self-
compassionate response, but it also extends to many areas where we detect
problems and do something to resolve them, either personally or otherwise.
This is so because, like any other framing, deictic framing can be used to
relate any kind of events, either between oneself and others or between
oneself and one’s own behavior. That is why it is at the core of
psychological flexibility. Consequently, without these skills, the
development of a unitary and effective self in responding to others becomes
highly compromised. In other words, interpersonal problem solving, and,
more generally, prosocial interactions, may not be established. This
rationale actualizes Skinner’s approach to self-knowledge, as we indicated
at the beginning of this chapter, by elaborating on his analysis and
extending it to relational frames that can be taught and tested. It is our
contention that the research and application of deictic and hierarchical
framing of one’s own behavior might be a never-ending story, not only for
analyzing the behaviors addressed in this chapter but for creative functional
analytic research that might shed some light on the many obscure areas in
the human condition.

Conclusion
“Empathy” and “compassion” are ordinary terms rooted in natural
language. Interest in comprehending these concepts has caused the number
of publications in the last twenty years to increase, as these concepts appear
to be relevant to functioning in almost any human domain. No consensus
has been reached concerning the definitions and meanings of these terms,
nor about their application. However, the behaviors that constitute PT seem
to be involved in all of them.
We have argued that PT is the foundation of empathy and compassion.
Empathy involves PT because it focuses on the occurrence of an equivalent
or coordinated emotion brought into contact with another’s signs of
emotions, as an affective resonance with someone else. Compassion involves
PT because it focuses on connecting with the other’s suffering, and on
responding according to an overreaching significant value of helping others.
In view of the relevant role of PT, we have conducted a conceptual
analysis beyond the classical context of ToM, in an attempt to identify and
analyze the likely behaviors involved, particularly deictic relational framing.
The research conducted on deictic framing has shown how to build the skill
through multiple exemplar training in order to establish the perspective of I-
Here-Now and You-There-Then, and many other combinations that enable us
to switch from one perspective to another—without losing the unique and
transcendent experience of I-Here-Now in which I integrate the ongoing
behavior as a hierarchical perspective.
In conclusion, this chapter provided an overview, conceptual analysis,
and review of research, especially that informed by relational frame theory,
related to the topics of perspective taking, empathy, and compassion. These
topics are central to understanding complex human behavior and many
clinically important issues encountered in the lives of humans.
Study Questions
1. Describe traditional concepts of perspective taking (PT) from
outside behavior analysis.
2. Describe Skinner’s concept of the self and self-knowledge.
3. What are the three key deictic relational frames involved in PT?
4. Distinguish between simple, reversed, and double-reversed deictic
relations involved in PT.
5. Describe traditional concepts of empathy from outside behavior
analysis.
6. Describe the three stages of the development of empathy, according
to an RFT perspective.
7. Interpret how you might extrapolate the three stages of empathy
development into a protocol you could use to strengthen the
behaviors involved in empathy in the clients you work with.
8. Describe traditional concepts of compassion outside of behavior
analysis.
9. Describe the RFT analysis of compassion explained in this chapter.
10. Distinguish between empathy and compassion, as conceptualized
through RFT in this chapter.
 Glossary of Acronyms

abolishing operation (AO): An antecedent with two functions: (1)

decrease the effectiveness of a future consequence as a reinforcer, and (2)
temporarily decrease the probability of responses that have been reinforced
by that reinforcer in the past.
acceptance and commitment therapy (ACT): A contemporary behavior
analytic approach to psychotherapy, based on weakening rigid control by
rules involving short-term escape/avoidance contingencies and
strengthening repertoires of behaving flexibly with respect to long-term
positive reinforcers. When implemented by behavior analysts outside of
psychotherapeutic settings, ACT is often referred to as acceptance and
commitment training.
applied behavior analysis (ABA): The application of behavioral principles
of learning and motivation to solving problems of human behavior that are
of significance to society.
arbitrarily applicable relational responding (AARR): The term
“relational” refers to the fact that AARR is behavior that occurs in response
to the relation between two or more antecedent stimuli, rather than in
response to a single stimulus. The term “arbitrarily applicable” refers to the
fact that AARR does not need to be controlled by the physical
characteristics of stimuli, but rather by the arbitrary relational properties.
For example, a child learns to say that a dime is bigger than a nickel
because it corresponds to a larger amount of money, even though it is
physically smaller.
autism spectrum disorder (ASD): A neurodevelopmental disorder
characterized by challenges with daily functioning that arise from excessive
repetitive behavior and interests, as well as deficits in social communication
skills.
behavioral skills training (BST): A highly researched training procedure
that involves three components: (1) verbal explanation, (2) modeling, and (
3) role-play rehearsal with feedback.
bidirectional naming (BiN): An operant repertoire that involves
responding as both speaker and listener to the same environmental stimuli
and the ability to switch back and forth between listener and speaker
behavior. BiN is characterized by the observation that, when taught one
repertoire (i.e., either speaker or listener), the learner can then exhibit the
other repertoire, despite not having been taught that repertoire. For
example, upon learning to label an apple as “apple,” the learner can then
point to an apple when asked to “point to apple.”
brief and immediate relational responses (BIRRs): Relational responses
that occur under time pressure and therefore tend to be relatively brief and
unelaborated; that is, the person behaving has not had the opportunity to
self-edit the responses. BIRRs are often different from relational responses
that the same person would emit if they had a longer duration of time in
which to self-edit the response.
complex-to-simple (CTS) protocol: A stimulus equivalence training
procedure that involves consecutive training phases followed by testing for
the emergence of derived relations individually, starting with the most
complex relations.
conditioned stimulus (CS): A stimulus that has respondent functions due
to past pairing with another stimulus, for example, a tone that elicits
salivation because that tone was paired with food in the past.
conditioned response (CR): A response that occurs in the presence of a
stimulus due to past pairing of that stimulus with another stimulus, for
example, salivation that occurs in response to a tone because that tone was
paired with food in the past.
contextual behavioral science (CBS): CBS is a scientific tradition that is
based on functional contextualism and therefore adopts pragmatism as its
truth criterion and the ongoing act-in-context as its root metaphor for
scientific analysis. CBS contains a reticulated research approach where
theoretical, practical, and multifaceted (i.e., basic, analogue, and applied)
empirical efforts are mutually influential in fostering CBS’s overall
progress. The larger purpose of CBS is to produce information that can be
applied to successfully reduce human suffering.
derived relational responding (DRR): Behavior that occurs in response to
the relation between two or more stimuli, rather than in response to a single
stimulus, and that was not directly trained and cannot be accounted for by
the physical properties of the stimuli (i.e., is not stimulus generalization).
For example, if one is directly taught that Sam is older than Jill and that Jill
is older than Mario, one can derive that Mario is younger than Sam, even
though one has never been directly taught anything about the relation
between Mario and Sam.
dialectical behavior therapy (DBT): A behavioral approach to
psychotherapy that integrates the explicit focus on behavior change that
characterized first-wave behavior therapies with the shaping of awareness
and acceptance of one’s immediate experience (Linehan, 1987).
differential arbitrarily applicable relational responding effects
(DAARRE): An analytic model that considers how variables affect
responding in the implicit relational assessment procedure. The model
identifies three key sources of behavioral influence: (1) the relationship
between the label and target stimuli (relational cues), (2) the orienting
functions of the label and target stimuli (functional cues), and (3) the
coherence functions of the two relational coherence indicator response
options (e.g., “True” and “False”).
differential observing response (DOR): A procedure in which the learner
is required to indicate in some way they have observed a stimulus before
the procedure continues. For example, a learner may be required to touch a
visual stimulus or repeat an auditory stimulus before the next stimulus is
presented.
differential reinforcement of low rate (DRL): A reinforcement schedule
in which reinforcement is delivered if responding occurs at or below a
particular rate.
differential reinforcement of other behavior (DRO): A reinforcement
schedule in which a reinforcer is delivered contingent on the absence of a
behavior for a specified interval.
discriminative stimulus (SD or S+): An antecedent stimulus that
occasions a response because in the past that response was reinforced in the
presence of that stimulus more often than in the absence of that stimulus.
early intensive behavioral intervention (EIBI): A highly researched
treatment approach for children with autism that contains at least the
following elements: (1) starts before the age of three and a half years old,
(2) contains thirty or more hours per week of one-to-one intervention, (3)
continues for two years or more, (4) is based on the principles of behavior
analysis, and (5) is individualized to clients’ needs. EIBI is sometimes used
with populations other than autism, but the vast majority of published
research has been done with children with autism.
English language learner (ELL): An individual whose first language is
not English.
equivalence-based instruction (EBI): Teaching procedures based on
stimulus equivalence research. Such procedures generally involve directly
training relations between some stimuli, usually utilizing match-to-sample
procedures, and testing for untrained relations.
establishing operations (EOs): An antecedent with two functions: (1)
increase the effectiveness of a future consequence as a reinforcer, and (2)
temporarily increase the probability of responses that have been reinforced
by that reinforcer in the past.
extended and elaborated relational responses (EERRs): Behaviors that
occur under relatively lower time constraints, such that the person
responding had adequate time to self-edit their response. Such responses are
relatively more likely to be influenced by their social consequences; for
example, someone is less likely to emit a response that would be judged as
racist by others when engaging in EERRs.
Extrinsic Affective Simon Task (EAST): A task used in mainstream
psychology to assess implicit cognitions.
fixed ratio (FR): A reinforcement schedule in which reinforcement is
delivered contingent upon a specified number of behaviors occurring.
functional analytic psychotherapy (FAP): A behavioral approach to
psychotherapy, based in contextual behavioral science, that focuses on
improving interpersonal relationships to help individuals lead more
fulfilling lives.
functional contextualism (FC): A philosophy of science that adopts
pragmatism as a truth criterion and the ongoing act-in-context as its root
metaphor for scientific analysis. Functional contextual science strives to
produce increasingly organized statements about nature, with precision,
scope, and depth.
Implicit Association Test (IAT): A test developed in mainstream cognitive
psychology that tests implicit cognitions.
implicit relational assessment procedure (IRAP): An assessment
procedure that requires rapid responding to true or false relations between
stimuli. Evaluating the difference in various response latencies provides a
procedure for assessing implicit relations, that is, relations that the
participant may self-edit if given sufficient opportunity. For example, a
person is highly unlikely to report that they are racist, but their latency to
respond negatively to a statement that a particular ethnic group is bad may
be shorter than their latency to respond saying that same ethnic group is
good.
instructive feedback (IF): A procedure that involves adding additional
nontarget stimuli to the antecedent or consequence events of direct
instructional trials.
inter-stimulus interval (ISI): The time that elapses from the presentation
of one stimulus to the presentation of the next.
linear series: A sequence for training relations in a stimulus equivalence
experiment. For example, one might train that A = B and then B = C.
many-to-one (MTO): A format for training stimulus relations in an
experiment in derived relational responding. One stimulus set is used as the
comparison stimuli, whereas two or more other stimulus sets are used as the
sample stimuli. For example, one might train A = B, C = B, and D = B.
match-to-sample (MTS): A teaching procedure that involves presenting a
single sample stimulus, then an instruction, and then multiple comparison
stimuli. For example, a teacher might give a learner a picture of a car, then
ask the learner to find the one that is the same, and then provide the
opportunity for the learner to scan comparison stimuli consisting of car,
truck, and airplane.
motivating operation (MO): The umbrella term that encompasses both
abolishing operation and establishing operation.
multiple exemplar instruction (MEI): A teaching procedure that involves
teaching more than one example at the same time. MEI can be conducted
across multiple different verbal operants for the same target, for example,
learning to respond as both speaker and listener to blue objects. MEI can
also be conducted within a single operant repertoire but across multiple
exemplars of stimuli, for example, by learning to label the color blue by
being taught to label ten different blue objects as “blue.” Occasionally, MEI
is used to distinguish procedures where multiple exemplars are trained
across operant classes of behavior, as opposed to within.
multiple exemplar training (MET): Another term for MEI (see above
definition). Occasionally, MET is used to refer to procedures where
multiple exemplars are trained within operant classes of behavior, as
opposed to across.
observational learning (OL): A form of learning that involves behavior
change as a function of observing happenings in one’s environment.
one-to-many (OTM): A stimulus equivalence training procedure, wherein
one stimulus set is used as the sample stimuli and relations are trained
between that set as sample stimuli and two or more other sets as comparison
stimuli. For example, one might train A = B, A = C, and A = D.
peer-mediated interventions (PMIs): Interventions in which peers are
actively involved in delivering relevant antecedents, consequences, or both.
Picture Exchange Communication System (PECS): A communication
system commonly used by learners who have difficulty producing vocal
speech. PECS involves selecting and giving pictures to others as an
alternative form of communication. PECS also actively encourages learners
to attempt to produce vocal speech while communicating with pictures.
Promoting the Emergence of Advanced Knowledge (PEAK): A recently
developed assessment and curriculum, based on relational frame theory, that
is intended to be used by practitioners to teach verbal repertoires, including
derived relational responding.
random interval (RI): A reinforcement schedule in which reinforcement is
delivered contingent upon the first response that occurs after a specified
interval of time has elapsed since the last reinforcer was delivered, and
where the duration of that interval varies randomly, within specified
parameters, after each reinforcer is delivered.
random ratio (RR): A reinforcement schedule in which reinforcement is
delivered contingent on the occurrence of a specific number of responses,
and where that number varies randomly, within specified parameters, after
each reinforcer is delivered.
randomized controlled trial (RCT): An experimental design in which
participants are randomly assigned to two or more groups and where those
responsible for measuring the dependent variable are generally blind to
group assignment. This design is considered the strongest experimental
design by most researchers in mainstream psychology and medicine.
relational coherence indicator (RCI): Response options, such as “True”
and “False,” that are often used to indicate the coherence or incoherence
between labels and target stimuli that are presented within an implicit
relational assessment procedure.
relational elaboration and coherence (REC) model: An initial relational
frame theory approach to implicit cognition, based on the idea that the
implicit relational assessment procedure forces participants to emit rapid
relational responses that are less likely to be influenced by social variables
than the slower, more elaborated relational responses emitted on traditional
assessments of bias. The strength of the rapid relational responses emitted
toward experimental stimuli was thought to be functionally similar to that
of responses toward these stimuli in the participant’s preexperimental
history, when one does not have the opportunity to edit one’s own responses
in anticipation of social contingencies.
relational fame theory (RFT): A behavior analytic theory of language and
cognition that addresses verbal behavior that occurs in response to relations
between stimuli, rather than to single stimuli as antecedents. Such behavior
is said to comprise generalized relational operants, learned through a history
of multiple exemplar training. These operants are referred to as “relational
frames” and demonstrate the properties of being arbitrarily applicable (i.e.,
they do not depend on the physical properties of antecedent controlling
stimuli), displaying mutual entailment (i.e., they are bidirectional),
displaying combinatorial entailment (i.e., relations propagate across three or
more stimuli in a network), and involving transformation of stimulus
functions (i.e., operant and respondent functions transform in accordance
with the mutually entailed and combinatorially entailed functions).
rule-governed behavior (RGB): Behavior that occurs due to contact with
an antecedent rule and not due to prior contact with the contingencies the
rule describes.
S-delta (SΔ or S-): A stimulus that does not occasion a response because
the response has not been reinforced in the presence of that stimulus in the
past.
self-determined learning model of instruction (SDLMI): A program for
teaching problem solving that involves scripted problem-solving instruction
delivered by a teacher who follows a sequence predetermined by the
investigator.
simple-to-complex (STC) protocol: A stimulus equivalence training
protocol that intersperses training phases and derived relations probes.
Across the whole program, derived relations are tested individually and
tests are arranged from the simplest type of derived relations (i.e.,
symmetry) to the most difficult (i.e., equivalence).
simultaneous (SIM) protocol: A stimulus equivalence training protocol in
which all training phases are grouped and all testing phases are grouped,
with no overlap between.
social listener reinforcement (SLR): A procedure designed to build
reinforcement for speaking and listening to peers and to increase the
learner’s conversational units with others. Such procedures often involve a
sequence of increasingly complex games, ending with an empathy phase,
that use a yoked contingency game board to establish social listener
reinforcement.
stimulus-pairing observation (SPO): A procedure for teaching
discriminations in which an experimenter briefly presents a stimulus in
isolation to the learner, followed by an inter-stimulus interval (ISI) and then
presentation of a second stimulus. Thus, the procedure involves pairing of
stimuli with no overt response requirement from the participant other than
attending to the stimuli presented by the experimenter.
stimulus-stimulus pairing (SSP): A procedure in which an experimenter
pairs auditory stimuli (e.g., word sounds) with already established
reinforcers in an attempt to condition those word sounds as sources of
automatic reinforcement. This procedure often results in temporary increase
in the probability of the learner’s making those same word sounds.
talk aloud problem solving (TAPS): A problem-solving training
procedure that teaches students to perform both as problem solvers and as
active listeners.
theory of mind (ToM): An evolutionary-developmental theory that
explains perspective-taking behavior in terms of inferred brain mechanisms.
think aloud pair problem solving (TAPPS): A problem-solving training
procedure that teaches students to dialogue with themselves to analytically
talk themselves through solving problems. The procedure is often used to
teach academic skills.
transfer of stimulus control (ToSC): A procedure in which stimulus
control is transferred from one stimulus to another, often through gradually
changing the topographical properties of the first stimulus until it resembles
the second stimulus.
transitive conditioned establishing operations (CMO-T): Previously
neutral conditions whose occurrences alter the effectiveness of another
stimulus and evoke responses that produce or suppress that stimulus.
unconditioned response (UR): A respondent behavior that occurs in
response to unconditioned stimuli and that does not depend on a history of
conditioning to occur.
unconditioned stimulus (US): An eliciting stimulus that does not depend
on a history of conditioning to be effective at eliciting an unconditioned
response.
unidirectional naming (UiN): In the context of rudimentary verbal
behavior, UiN refers to the demonstration of the listener component, which
involves hearing a speech sound and then responding in a nonvocal way to
a nonverbal stimulus (e.g., hearing the word “car” and then pointing to a
car).
verbal behavior developmental theory (VBDT): A theoretical approach
that combines Skinnerian verbal behavior, stimulus equivalence, naming
theory, and relational frame theory to analyze the development of verbal
repertoires over the lifetime of the child, emphasizing the transition from
unidirectional verbal operants to complex bidirectional operants.
verbal conditional discrimination (VCD): Behavior that is under the
control of two antecedent stimuli, where the control that one exerts is
conditional upon the presence of the other. For example, when hearing the
stimuli “vegetable” and “purple,” one might reply “eggplant.”
 References

Adams, B. J., Fields, L., & Verhave, T. (1993). Effects of the test order on intersubject variability
during equivalence class formation. Psychological Record, 43, 133–152.
Adkins, T., & Axelrod, S. (2001). Topography-versus selection-based responding: Comparison of
mand acquisitions in each modality. The Behavior Analyst Today, 2(3), 259–266.
doi:10.1037/h0099941
Agran, M., Blanchard, C., Wehmeyer, M., & Hughes, C. (2002). Increasing the problem-solving
skills of students with developmental disabilities participating in general education.
Remedial and Special Education, 23, 279–288.
Aguilar, J. M., Chan, J. M., White, P. J., & Fragale, C. (2017). Assessment of the language
preferences of five children with autism from Spanish-speaking homes. Journal of
Behavioral Education, published online first, 1–14. doi:10.1007/s10864-017-9280-9
Aguirre, A. A., & Rehfeldt, R. A. (2015). An evaluation of instruction in visual imagining on the
written spelling performance of adolescents with learning disabilities. The Analysis of Verbal
Behavior, 31(1), 118–125. doi:10.1007/s40616-015-0028-0
Aguirre, A. A., Valentino, A. L., & LeBlanc, L. A. (2016). Empirical investigations of the
intraverbal: 2005–2015. The Analysis of Verbal Behavior, 32, 139–153. doi:10.1007/s40616-
016-0064-4
Albright, L., Reeve, K. F., Reeve, S. A., & Kisamore, A. N. (2015). Teaching statistical variability
with equivalence-based instruction. Journal of Applied Behavior Analysis, 48, 883–894.
Albright, L., Schnell, L., Reeve, K. F., & Sidener, T. M. (2016). Using stimulus equivalence-
based instruction to teach graduate students in applied behavior analysis to interpret operant
functions of behavior. Journal of Behavioral Education, 25, 290–309.
Alessi, G. (1987). Generative strategies and teaching for generalization. The Analysis of Verbal
Behavior, 5, 15–27.
Allan, A. C., Vladescu, J. C., Kisamore, A. N., Reeve, S. A., & Sidener, T. M. (2015). Evaluating
the emergence of reverse intraverbals in children with autism. Analysis of Verbal Behavior,
31(1), 59–75. doi:10.1007/s40616-014-0025-8
Anderson, C. M., Hawkins, R. P., & Scotti, J. R. (1998). Private events in behavior analysis:
Conceptual basis and clinical relevance. Behavior Therapy, 28(1), 157–179.
Andronis, P. T., Layng, T. V. J., & Goldiamond, I. (1997). Contingency adduction of “symbolic
aggression” by pigeons. The Analysis of Verbal Behavior, 14, 5–17.
Archer, A., Gleason, M., & Isaacson, S. (2008). REWARDS Writing: Sentence Refinement
[Curriculum Program]. Longmont, CO: Sopris West.
Arntzen, E. (2006). Delayed matching-to-sample: Probability of stimulus equivalence as a
function of delays between sample and comparison stimuli during training. The
Psychological Record, 56, 5–167.
Arntzen, E., & Almås, I. K. (2002). Effects of mand-tact versus tact-only training on the
acquisition of tacts. Journal of Applied Behavior Analysis, 35(4), 419–422.
doi:10.190/jaba.2002.35-419
Arntzen, E., Grondahl, T., & Eilifsen, C. (2010). The effects of different training structures in the
establishment of conditional discriminations and subsequent performance on tests for
stimulus equivalence. The Psychological Record, 60, 437–462.
Arntzen, E., Halstadtro, L.-B., & Halstradtro, M. (2009). The “silent dog” method: Analyzing the
impact of self-generated rules when teaching different computer chains to boys with autism.
The Analysis of Verbal Behavior, 25, 51–65.
Arntzen, E., Halstadtro, L.-B., Bjerke, E., & Halstadtro, M. (2010). Training and testing music
skills in a boy with autism using a matching-to-sample format. Behavioral Interventions, 25,
129–143. doi:10.1002/bin.301
Arntzen, E., & Holth, P. (1997). Probability of stimulus equivalence as a function of training
design. Psychological Record, 47, 309–320.
Arntzen, E., & Holth, P. (2000). Equivalence outcome in single subjects as a function of training
structure. The Psychological Record, 50, 603–628.
Asaro-Saddler, K. (2016). Using evidence-based practices to teach writing to children with autism
spectrum disorders. Preventing School Failure: Alternative Education for Children and
Youth, 60(1), 79–85. doi:10.1080/1045988x.2014.981793
Austin, J. L. (1962). How to do things with words. Oxford, England: Oxford University Press.
(Original work published 1955)
Axe, J. B. (2008). Conditional discrimination in the intraverbal relation: A review and
recommendations for future research. The Analysis of Verbal Behavior, 24(1), 159–174.
doi:10.1007/BF03393064
Ayllon, T., & Azrin, N. H. (1966). Punishment as a discriminative stimulus and conditioned
reinforcer with humans. Journal of the Experimental Analysis of Behavior, 9, 411–419.
doi:10.1901/jeab.1966.9-411
Ayllon, T., & Azrin, N. H. (1968). The token economy: A motivational system for therapy and
rehabilitation. New York, NY: Appleton Century Crofts.
Azrin, N. H., & Foxx, R. M. (1971). A rapid method of toilet training the institutionalized
retarded. Journal of Applied Behavior Analysis, 4, 89–99.
Azrin, N. H., & Holz, W. C. (1966). Punishment. In W. K. Honig (Ed.), Operant behavior: Areas
of research and application (pp. 380–447). New York, NY: Appleton-Century-Crofts.
Baer, D. M., Wolf, M. M., & Risley, T. R. (1968). Some current dimensions of applied behavior
analysis. Journal of Applied Behavior Analysis, 1(1), 91–97. doi:10.1901/jaba.1968.1-91
Baker, K. A. (2014). The effects of social listener reinforcement and video modeling on the
emergence of social verbal operants in preschoolers diagnosed with autism and language
delays (Doctoral dissertation, Order No. 3621578). Available from ProQuest Dissertations &
Theses Global. (1552714267). doi:10.7916/D85T3HNQ
Bandura, A. (1961). Psychotherapy as a learning process. Psychological Bulletin, 58(2), 143–159.
Bandura, A. (1965). Influence of models’ reinforcement contingencies on the acquisition of
imitative responses. Journal of Personality and Social Psychology, 1, 589–595.
doi:10.1037/h0022070
Bandura, A. (1969). Principles of behavior modification. New York, NY: Holt, Rinehart, &
Winston.
Bandura, A. (1974). Behavior theory and the models of man. American Psychologist, 29(12),
859–869.
Bandura, A., Grusec, J. E., & Menlove, F. L. (1966). Observational learning as a function of
symbolization and incentive set. Child Development, 37, 499–506.
Bandura, A., & Jeffrey, R. W. (1973). Role of symbolic coding and rehearsal processes in
observational learning. Journal of Personality and Social Psychology, 26, 122–130.
Barbera, M. L., & Kubina, R. M. (2005). Using transfer procedures to teach tacts to a child with
autism. The Analysis of Verbal Behavior, 21(1), 155–161. doi:10.1007/BF03393017
Barbero-Rubio, A., López-López, J. C., Luciano, C., & Eisenbeck, N. (2016). Perspective-taking
measured by implicit relational assessment procedure (IRAP). The Psychological Record,
66(2), 243–252.
Barlow, D. H. (1988). Anxiety and its disorders: The nature and treatment of anxiety and panic.
New York, NY: Guilford Press.
Barlow, D. H., Craske, M. G., Cerny, J. A., & Klosko, J. S. (1989). Behavioral treatment of panic
disorder. Behavior Therapy, 20, 261–282.
Barlow, K. E., Tiger, J. H., Slocum, S. K., & Miller, S. J. (2013). Comparing acquisition of
exchange-based and signed mands with children with autism. The Analysis of Verbal
Behavior, 29(1), 59–69. doi:10.1007/BF03393124
Barnes, D., Hegarty, N., & Smeets, P. M. (1997). Relating equivalence relations to equivalence
relations: A relational framing model of complex human functioning. The Analysis of Verbal
Behavior, 14, 57–83.
Barnes, D., Lawlor, H., Smeets, P. M., & Roche, B. (1996). Stimulus equivalence and academic
self-concept among mildly mentally handicapped and nonhandicapped children. The
Psychological Record, 46, 87–107.
Barnes-Holmes, D., & Barnes-Holmes, Y. (2000). Explaining complex behavior: Two
perspectives on the concept of generalized operant classes. The Psychological Record, 50,
251–265. doi:10.1007/bf03395355
Barnes-Holmes, D., Barnes-Holmes, Y., & Cullinan, V. (2000). Relational frame theory and
Skinner’s Verbal Behavior: A possible synthesis. The Behavior Analyst, 23(1), 69–84.
doi:10.1007/BF03392000
Barnes-Holmes, D., Barnes-Holmes, Y., Stewart, I., & Boles, S. (2010). A sketch of the Implicit
Relational Assessment Procedure (IRAP) and the Relational Elaboration and Coherence
(REC) model. The Psychological Record, 60, 527–542.
Barnes-Holmes, D., Finn, M., McEnteggart, C., Barnes-Holmes, Y. (2017). Derived stimulus
relations and their role in a behavior-analytic account of human language and cognition.
 Perspectives on Behavior Science, 41, 155–173. doi:10.1007/s40614-017-0124-7
Barnes-Holmes, D., Murphy, A., Barnes-Holmes, Y., & Stewart, I. (2010). The Implicit Relational
Assessment Procedure: Exploring the impact of private versus public contexts and the
response latency criterion on pro-white and anti-black stereotyping among white Irish
individuals. The Psychological Record, 60, 57–66.
Barnes-Holmes, D., O’Hora, D., Roche, B., Hayes, S. C., Bissett, R. T., & Lyddy, F. (2001).
Understanding and verbal regulation. In S. Hayes, D. Barnes-Holmes, & B. Roche (Eds.),
Relational frame theory: A post-Skinnerian account of human language and cognition (pp.
103–117). New York, NY: Springer.
Barnes-Holmes, D., Regan, D., Barnes-Holmes, Y., Commins, S., Walsh, D., Stewart, I., . . .
Dymond, S. (2005). Relating derived relations as a model of analogical reasoning: Reaction
times and event-related potentials. Journal of the Experimental Analysis of Behavior, 84(3),
435–451.
Barnes-Holmes, Y., Barnes-Holmes, D., & McHugh, L. (2004). Teaching derived relational
responding to young children. Journal of Early and Intensive Behavior Intervention, 1, 3–12.
Barnes-Holmes, Y., Barnes-Holmes, D., Roche, B., & Smeets, P. M. (2001a). Exemplar training
and a derived transformation of function in accordance with symmetry: II. The
Psychological Record, 51, 589–603.
Barnes-Holmes, Y., Barnes-Holmes, D., Roche, B., & Smeets, P. M. (2001b) The development of
self and perspective-taking: A relational frame analysis. Behavioral Development Bulletin,
10, 42–45.
Barnes Holmes, Y., Barnes Holmes, D., & Smeets, P. M. (2004). Establishing relational
responding in accordance with opposite as generalized operant behavior in young children.
International Journal of Psychology and Psychological Therapy, 4. 559–586.
Barnes-Holmes, Y., Barnes-Holmes, D., Smeets, P. M., Strand, P., & Friman, P. (2004).
Establishing relational responding in accordance with more-than and less-than as generalized
operant behavior in young children. International Journal of Psychology and Psychological
Therapy, 4. 531–558.
Barnes-Holmes, Y., Boorman, J., Oliver, J. E. Thompson, M., McEnteggart, C., & Coulter, C.
(2018). Using conceptual developments in RFT to direct case formulation and clinical
intervention: Two case summaries. Journal of Contextual Behavioral Science, 7, 89–96.
doi:10.1016/j.jcbs.2017.11.005
Baron, A., & Galizio, M. (1983). Instructional control of human operant behavior. The
Psychological Record, 33, 495–520.
Baron-Cohen, S., Leslie, A. M., & Frith, U. (1985). Does the autistic child have a “theory of
mind”? Cognition, 21(1), 37–46.
Barrera, R. D., Lobato-Barrera, D., & Sulzer-Azaroff, B. (1980). A simultaneous treatment
comparison of three expressive language training programs with a mute autistic child.
Journal of Autism and Developmental Disorders, 10(1), 21–37. doi:10.1007/BF02408430
Barrera, R. D., & Sulzer-Azaroff, B. (1983). An alternating treatment comparison of oral and total
communication training programs with echolalic autistic children. Journal of Applied
Behavior Analysis, 16(4), 379–394. doi:10.1901/jaba.1983.16-379
Barrish, H. H., Saunders, M. S., & Wolf, M. M. (1969). Good behavior game: Effects of
individual contingencies for group consequences on disruptive behavior in a classroom.
Journal of Applied Behavior Analysis, 2, 119–124. doi:10.1901/jaba.1969.2-119
Bast, D. F., & Barnes-Holmes, D. (2015). Priming thoughts of failing versus succeeding and
performance on the Implicit Relational Assessment Procedure (IRAP) as a measure of self-
forgiveness. The Psychological Record, 65(4), 667–678. doi:10.1007/s40732-015-0137-0
Bast, D. F., Linares, I. M. P., Gomes, C., Kovac, R., & Barnes-Holmes, D. (2016). The Implicit
Relational Assessment Procedure (IRAP) as a measure of self-forgiveness: The impact of a
training history in clinical behavior analysis. The Psychological Record, 66(1), 177–190.
doi:10.1007/s40732-016-0162-7
Batson, C. D. (2009). These things called empathy: Eight related but distinct phenomena. In J.
Decety & W. Ickes (Eds), Social neuroscience. The social neuroscience of empathy (pp. 3–
15). Cambridge, MA: MIT Press.
Baum, W. M. (2011a). Behaviorism, private events, and the molar view of behavior. The Behavior
Analyst, 34(2), 185–200.
Baum, W. M. (2011b). No need for private events in a science of behavior: Response to
commentaries. The Behavior Analyst, 34(2), 237–244.
Baum, W. M. (2012). Rethinking reinforcement: allocation, induction, and contingency. Journal
of the Experimental Analysis of Behavior, 97(1), 101–124. doi:10.1901/jeab.2012.97-101
Beaulieu, L., Hanley, G. P., & Santiago, J. L. (2014). Improving the conversational skills of a
college student with peer-mediated behavioral skills training. The Analysis of Verbal
Behavior, 30(1), 48–53. doi:10.1007/s40616-013-0001-8
Beck, A. T. (1976). Cognitive therapy and the emotional disorders. New York, NY: Meridian.
Beck, A. T. (2005). The current state of cognitive therapy: a 40-year retrospective. Archives of
General Psychiatry, 62(9), 953–959.
Beck, I. L., McKeown, M. G., Hamilton, R. L., & Kucan, L. (1997). Questioning the author: An
approach for enhancing student engagement with text. Newark, DE: International Reading
Association.
Beidel, D. C., & Turner, S. M. (1986). A critique of the theoretical bases of cognitive-behavioral
theories and therapy. Clinical Psychology Review, 6(2), 177–197. doi:10.1016/0272-
7358(86)90011-5
Belanich, J., & Fields, L. (1999). Tactual equivalence class formation and tactual-to-visual cross-
modal transfer. The Psychological Record, 49, 75–91.
Belisle, J., Dixon, M. R., Stanley, C. R., Munoz, B., & Daar, J. H. (2016). Teaching foundational
perspective-taking skills to children with autism using the PEAK-T curriculum: single-
reversal “I–You” deictic frames. Journal of Applied Behavior Analysis, 49, 965–969.
Belloso-Díaz, C., & Pérez-González, L. A. (2015). Effect of learning tacts or tacts and
intraverbals on the emergence of intraverbals about verbal categorization. The Psychological
Record, 65(4), 749–760. doi:10.1007/s40732-015-0145-0
Bentall, R. P., Lowe, C. F., & Beasty, A. (1985). The role of verbal behavior in human learning: II
Developmental differences. Journal of the Experimental Analysis of Behavior, 43, 165–181.
doi:10.1901/jeab.1985.43-165
Berens, N. M., & Hayes, S. C. (2007). Arbitrarily applicable comparative relations: Experimental
evidence for a relational operant. Journal of Applied Behavior Analysis, 40, 45–71.
doi:10.1901/jaba.2007.7-06
Bernard-Opitz, V., Sriram, N., & Nakhoda-Sapuan, S. (2001). Enhancing social problem solving
in children with autism and normal children through computer-assisted instruction. Journal
of Autism and Developmental Disorders, 31, 377–384.
Bevill-Davis, A., Clees, T. J., & Gast, D. L. (2004). Correspondence training: A review of the
literature. Journal of Early and Intensive Behavior Intervention, 1, 13–26.
doi:10.1037/h0100276
Biglan, A. (2015) The nurture effect. How the science of behaviour can improve our lives and our
world. Oakland, CA: New Harbinger.
Bijou, S. W. (1983). The initial development of linguistic behavior. In N. W. Smith, P. T.
Mountjoy, & D. H. Ruben (Eds.), Reassessment in psychology: The interbehavioral
alternative (pp. 213–232). Lanham, MD: University Press of America.
Bijou, S. W. (1989). Psychological linguistics: Implications for a theory of initial development
and a method for research. Advances in child behavior and development, 21, 221–241.
Bijou, S. W. (1996). Setting factors in the behavior analysis of child development. In S. W. Bijou
& E. Ribes (Eds.), New directions in behavior development (pp. 147–154). Reno, NV:
Context Press.
Bijou, S. W., & Baer, D. M. (1978). Behavior analysis of child development. Englewood Cliffs,
NJ: Prentice Hall.
Bijou, S. W., Chao, C. C., & Ghezzi, P. M. (1988). Manual of instructions for identifying and
analyzing referential interactions II. The Psychological Record, 38, 401–414.
Bijou, S. W., Umbreit, J., Ghezzi, P. M., & Chao, C. C. (1986). Manual of instructions for
identifying and analyzing referential interactions. The Psychological Record, 36, 491–518.
Binder, C. (1996). Behavioral fluency: Evolution of a new paradigm. The Behavior Analyst, 19,
163–187.
Binder, C., & Bloom, C. (1989, February). Fluent product knowledge: Application in the financial
services industry. Performance and Instruction, 28, 17–21.
Binder, C. & Sweeney, L. (2002). Building fluent performance in a customer call center.
Performance Improvement, 41(2), 29–37.
Binet, A., & Simon, T. (1973). The development of intelligence in children: The Binet-Simon
Scale. New York, NY: Arno Press. (Original work published 1916)
Bishop, C. D., Erezyilmaz, D. F., Flatt, T., Georgiou, C. D., Hadfield, M. G., Heyland, A., . . .
Youson, J. H. (2006). What is metamorphosis? Integrative and Comparative Biology, 46,
655–661. doi:10.1093/icb/icl004
Blackburn, I. M., Bishop, S., Glen, A. I., Whalley, L. J., & Christie, J. E. (1981). The efficacy of
cognitive therapy in depression: A treatment trial using cognitive therapy and
pharmacotherapy, each alone and in combination. The British Journal of Psychiatry, 139(3),
181–189.
Blackledge, J. T. (2007). Disrupting verbal processes: Cognitive defusion in acceptance and
commitment therapy and other mindfulness-based psychotherapies. The Psychological
Record, 57(4), 555–576.
Blackledge, J. T., & Drake, C. E. (2013). Acceptance and commitment therapy: Empirical and
theoretical considerations. In S. Dymond & B. Roche (Eds.), Advances in relational frame
theory: Research and application (pp. 219–252). Oakland, CA: New Harbinger.
Blewitt, P. (1994). Understanding categorical hierarchies: The earliest levels of skill. Child
Development, 65. 1279–1298. doi:10.1111/j.1467-8624.1994.tb00817.x
Bloom, L. (1993). Transcription and coding for child language research: The parts are more than
the whole. In J. A. Edwards & M. D. Lampert (Eds.), Taking data: Transcription and coding
in discourse research (pp. 149–166). Hillsdale, NJ: Lawrence Erlbaum.
Bod, R. (2009). From exemplar to grammar: A probabilistic analogy-based model of language
learning. Cognitive Science, 33, 752–793.
Boë, L-J., Berthommier, F., Legou, T., Captier, G., Kemp, C., Sawallis, T. R., . . . Fagot, J. (2017).
Evidence of a vocalic proto-system in the baboon (Papio papio) suggests pre-hominin speech
precursors. PlosOne, January 11, 2017. doi:10.1371/journal.pone.0169321
Boone, M. S., & Cannici, J. (2013). Acceptance and commitment therapy (ACT) in groups. In J.
Pistorello (Ed.). Acceptance and mindfulness for counseling college students: Theory and
practical applications for intervention, prevention, and outreach. Oakland, CA: New
Harbinger.
Bornstein, M. H., & Mash, C. (2010). Experience-based and on-line categorization of objects in
early infancy. Child Development, 81, 3, 884–897. doi:10.1111/j.1467-8624.2010.01440.x
Bourret, J., Vollmer T., & Rapp, J.T. (2004). Evaluation of a vocal mand assessment and vocal
mand training procedures. Journal of Applied Behavior Analysis, 37(2), 129–144.
doi:10.1901/jaba.2004.37-129
Bouton, M. E., & Trask, S. (2016). Role of the discriminative properties of the reinforcer in
resurgence. Learning & Behavior, 44, 137–150. doi:10.3758/s13420-015-0197-7
Bouton, M. E., Winterbauer, N. E., & Todd, T. P. (2012). Relapse processes after the extinction of
instrumental learning: renewal, resurgence, and reacquisition. Behavioural Processes, 90(1),
130–141. doi:10.1016/j.beproc.2012.03.004
Bower, G.H., & Hilgard, E.R. (1981). Theories of learning. Englewood Cliffs, NJ: Prentice Hall.
Brady, D. O., & Smouse, A. D. (1978). A simultaneous comparison of three methods for language
training with an autistic child: An experimental single case analysis. Journal of Autism and
Developmental Disorders, 8(3), 271–279. doi:10.1007/BF01539630
Branch, M. N., & Malagodi, E. F. (1980). Where have all the behaviorists gone? The Behavior
Analyst, 3(1), 31–38.
Breger, L., & McGaugh, J. L. (1965). Critique and reformulation of “learning-theory” approaches
to psychotherapy and neurosis. Psychological Bulletin, 63(5), 338–358.
doi:10.1037/h0021788
Briscoe, R. V., Hoffman, D. B., & Bailey, J. S. (1975). Behavioral community psychology:
Training a community board to problem solve. Journal of Applied Behavior Analysis, 8,
157–168.
Brodhead, M. T., Higbee, T. S., Gerencser, K. R., & Akers, J. A. (2016). The use of
discrimination-training procedure to teach mand variability to children with autism. Journal
of Applied Behavior Analysis, 49(1), 34–48. doi:10.1002/jaba.280
Brodsky, J., & Fienup, D. M. (2018). Sidman goes to college: A meta-analysis of equivalence-
based instruction. Perspectives on Behavior Science, 41, 95–119. doi:10.1007/s40614-018-
0150-0
Brown, A. L. (1989). Analogical learning and transfer: What develops? In S. Vosniadou & Ortony
(Eds.), Similarity and Analogical Reasoning (pp. 369–412). Cambridge, UK: Cambridge
University Press.
Brunero, S., Lamont, S., & Coates, M. (2010). A review of empathy education in nursing.
Nursing Inquiry, 17(1), 65–74.
Burchard, J. D., & Barrera, F. (1972). An analysis of timeout and response cost in a programmed
environment. Journal of Applied Behavior Analysis, 5, 271–282. doi:10.1901/jaba.1972.5-
271
Burns, D. D., & Spangler, D. L. (2001). Do changes in dysfunctional attitudes mediate changes in
depression and anxiety in cognitive behavioral therapy? Behavior Therapy, 32(2), 337–369.
Byrne, B. L., Rehfeldt, R. A., & Aguirre, A. A. (2014). Evaluating the effectiveness of the
stimulus pairing observation procedure and multiple exemplar instruction on tact and listener
responses in children with autism. The Analysis of Verbal Behavior, 30(2), 160–169.
doi:10.1007/s40616-014-0020-0
Cagney, S., Harte, C., Barnes-Holmes, D., Barnes-Holmes, Y., & McEnteggart, C. (2017).
Response biases on the IRAP for adults and adolescents with respect to smokers and
nonsmokers: The impact of parental smoking status. The Psychological Record, 67(4), 473–
483. doi:10.1007/s40732-017-0249-9
Cahill, C. S., & Greer, R. D. (2014). Actions vs. words: How we can learn both. Acta
Investigacion Psichologia, 4, 1716–1745. doi:10.1016/S2007-4719(14)70976-7
Cameron, G., Roche, B., Schlund, M. W., & Dymond, S. (2016). Learned, instructed and
observed pathways to fear and avoidance. Journal of Behavior Therapy and Experimental
Psychiatry, 50, 106–112. doi:10.1016/j.jbtep.2015.06.003
Cameron, G., Schlund, M. W., & Dymond, S. (2015). Generalization of socially transmitted and
instructed avoidance. Frontiers in Behavioral Neuroscience, 9, 159.
doi:10.3389/fnbeh.2015.00159
Carbone, V. J., Lewis, L., Sweeney-Kerwin, E. J., Dixon, J., Louden, R., & Quinn, S. (2006). A
comparison of two approaches for teaching VB functions: Total communication vs. vocal-
alone. The Journal of Speech and Language Pathology–Applied Behavior Analysis, 1(3),
181–192. doi:10.1037/h0100199
Carbone, V. J., Sweeney-Kerwin, E. J., Attanasio, V., & Kasper, T. (2010). Increasing the vocal
responses of children with autism and developmental disabilities using manual sign mand
training and prompt delay. Journal of Applied Behavior Analysis, 43(4), 705–709.
doi:10.1901/jaba.2010.43–705
Cariveau, T., Kodak, T., & Campbell, V. (2016). The effects of intertrial interval and instructional
format on skill acquisition and maintenance for children with autism spectrum disorders.
Journal of Applied Behavior Analysis, 49(4), 809–825. doi:10.1002/jaba.322
Carpentier, F., Smeets, P. M., & Barnes-Holmes, D. (2002). Matching functionally same relations:
Implications for equivalence-equivalence as a model of analogical reasoning. The
Psychological Record, 52, 351–370.
Carpentier, F., Smeets, P. M., & Barnes-Holmes, D. (2003). Equivalence-equivalence as a model
of analogy: Further analysis. The Psychological Record, 53, 349–371.
Carpentier, F., Smeets, P., Barnes-Holmes, D., & Stewart, I. (2004). Matching derived
functionally-same stimulus relations. Equivalence-equivalence and classical analogies. The
Psychological Record, 54, 255–273.
Carr, J. E., & Miguel, C. F. (2012). The analysis of verbal behavior and its therapeutic
applications. In G. J. Madden (Ed.), APA handbook of behavior analysis (Vol. 2, pp. 329–
352). Washington, DC: American Psychological Association.
Carroll, R. A., & Kodak, T. (2015). Using instructive feedback to increase response variability
during intraverbal training for children with autism spectrum disorder. The Analysis of
Verbal Behavior, 31(2), 183–199. doi:10.1007/s40616-015-0039-x
Carter, M., & Grunsell, J. (2001). The behavior chain interruption strategy: A review of research
and discussion of future directions. Journal of the Association for Persons with Severe
Handicaps, 26(1), 37–49. doi:10.2511/rpsd.26.1.37
Cartwright, A., Hussey, I., Roche, B., Dunne, J., & Murphy, C. (2016). An investigation into the
relationship between gender binary and occupational discrimination using the Implicit
Relational Assessment Procedure. The Psychological Record, 67, 121–130.
doi:10.1007/s40732-016-0212-1
Cartwright, A., Roche, B., Gogarty, M., O’Reilly, A., & Stewart, I. (2016). Using a modified
Function Acquisition Speed Test (FAST) for assessing implicit gender stereotypes. The
Psychological Record, 66(2), 223–233. doi:10.1007/s40732-016-0164-5
Castro, M., & Rehfeldt, R. A. (2016). Comparing the efficacy of peers versus staff models on
observational learning in adults with developmental disorders. Journal of Developmental
and Physical Disabilities, 28, 609–622. doi:10.1007/s10882-016-9498-9
Catania, A. C. (1997). Learning. Upper Saddle River, NJ: Prentice Hall.
Catania, A. C., Shimoff, E., & Matthews, B. A. (1989). An experimental analysis of rule-
governed behavior. In S. Hayes (Ed.), Rule-governed behavior: Cognition, contingencies,
and instructional control (pp. 119–150). Reno, NV: Context Press.
Chambers, M., & Rehfeldt, R. A. (2003). Assessing the acquisition and generalization of two
mand forms with adults with severe developmental disabilities. Research in Developmental
Disabilities, 24(4), 265–280. doi:10.1026/S0891-4222(03)00042-8
Chang, Y. C., & Locke, J. (2016). A systematic review of peer-mediated interventions for children
with autism spectrum disorder. Research in Autism Spectrum Disorders, 27, 1–10.
doi:10.1016/j.rasd.2016.03.010
Charlop, M. H., & Milstein, J. P. (1989). Teaching autistic children conversational speech using
video modeling. Journal of Applied Behavior Analysis, 22(3), 275–285.
doi:10.1901/jaba.1989.22-275
Chase, P. N., & Danforth, J. S. (1991). The role of rules in concept learning. In L. J. Hayes & P.
N. Chase (Eds.), Dialogues on verbal behavior (pp. 205–236). Reno, NV: Context Press.
Chiasson, C. C., & Hayes, L. J. (1993). The effects of subtle differences between listeners and
speakers on the referential speech of college freshman. The Psychological Record, 43, 13–
24.
Chomsky, N. (1959). A review of B. F. Skinner’s Verbal Behavior. Language, 35(1), 26–58.
Chomsky, N. (1966). Topics in the theory of generative grammar. The Hague, Paris: Mouton &
Co. N.V., Publishers.
Chomsky, N., & Place, U. T. (2000). The Chomsky—Place correspondence 1993–1994. The
Analysis of Verbal Behavior, 17, 7–38.
Christie, S., & Gentner, D. (2014). Language helps children success on a classic analogy task.
Cognitive Science, 38, 383–397.
Cihon, T. M., White, R., Zimmerman, V. L., Gesick, J., Stordahl, S., & Eshleman, J., (2017). The
effects of precision teaching with textual or tact relations on intraverbal relations. Behavioral
Development Bulletin, 22(1), 129–146. doi:10.1037/bdb0000056
Clayton, M. C., & Hayes, L. J. (2004). A comparison of match-to-sample and respondent type
training of equivalence classes. The Psychological Record, 54, 579–602.
Clayton, M. C., Hayes, L. J., & Swain, M. A. (2005). The nature and value of scientific system
building: The case of interbehaviorism. The Psychological Record, 55(3), 335–359.
doi:10.1007/BF03395515
Cohen, H. L. (1972). Programming alternatives to punishment: The design of competence
through consequences. In Bijou, S. W. & Ribes-Inesta, E. (Eds.), Behavior modification:
Issues and extensions (pp. 63–84). New York, NY: Academic Press.
Cohen, N. J., Sullivan, J., Minde, K., Novak, C., & Helwig, C. (1981). Evaluation of the relative
effectiveness of methylphenidate and cognitive behavior modification in the treatment of
kindergarten-aged hyperactive children. Journal of Abnormal Child Psychology, 9(1), 43–
54.
Connell, J. E., & Witt, J. C. (2004). Applications of computer-based instruction: Using
specialized software to aid letter-name and letter-sound recognition. Journal of Applied
Behavior Analysis, 37, 67–71.
Contreras, B. P., & Betz, A. M. (2016). Using lag schedules to strengthen the intraverbal
repertoires of children with autism. Journal of Applied Behavior Analysis, 49(1), 3–16.
doi:10.1002/jaba.271
Cook, C., & Adams, H. E. (1966). Modification of verbal behavior in speech deficient children.
Behaviour Research and Therapy, 4, 265–271. doi:10.1016/0005-7967(66)90082-9
Coon, J. T., & Miguel, C. F. (2012). The role of increased exposure to transfer-of-stimulus-control
procedures on the acquisition of intraverbal behavior. Journal of Applied Behavior Analysis,
45(4), 657–666. doi:10.1901/jaba.2012.45-657
Corcoran, R., Mercer, G. and Frith, C. D. (1995) Schizophrenia, symptomatology and social
inference: Investigating theory of mind in people with schizophrenia. Schizophrenia
Research, 17, 5–13.
Corral, D., & Jones, M. (2014). The effects of relational structure on analogical learning.
Cognition, 132, 280–300.
Corwin, A., & Greer, R. D. (2017). Effect of the establishment of Naming on learning from
teacher-demonstrations. Manuscript submitted for publication.
Cossairt, A., Hall, R. V., & Hopkins, B. L. (1973). The effects of experimenter’s instructions,
feedback, and praise on teacher praise and student attending behavior. Journal of Applied
Behavior Analysis, 6, 89–100. doi:10.1901/jaba.1973.6-89
Costa, A., & Pelaez, M. (2014). Implementing intensive tact instruction to increase frequency of
spontaneous mands and tacts in typically developing children. Behavioral Development
Bulletin, 19(1), 19–24. doi:10.1037/h0100569
Cote, D. L., Jones, V. L., Barnett, C., Pavelek, K., Nguyen, H., & Sparks, S. L. (2014). Teaching
problem solving skills to elementary age students with autism. Education and Training in
Autism and Developmental Disabilities, 49, 189–199.
Cowley, B. J., Green, G., & Braunling-McMorrow, D. (1992). Using stimulus equivalence
procedures to teach name-face matching to adults with brain injuries. Journal of Applied
Behavior Analysis, 25, 461–475.
Creel, R. (1980). Radical epiphenomenalism: BF Skinner’s account of private events.
Behaviorism, 8(1), 31–53.
Critchfield, T. S. (2014). Online equivalence-based instruction about statistical inference using
written explanation instead of match-to-sample training. Journal of Applied Behavior
Analysis, 47, 606–611.
Critchfield, T. S., & Fienup, D. M. (2010). Using stimulus equivalence technology to teach
statistical inference in a group setting. Journal of Applied Behavior Analysis, 43, 763–768.
doi:10.1901/jaba.2010.43-763
Cullen, C., & Barnes-Holmes, D. (2008). Implicit pride and prejudice: A heterosexual
phenomenon? In M.A. Morrison & T.G. Morrison (Eds.), Modern Prejudice (pp. 195–223).
New York, NY: Nova Science.
Cullen, C., Barnes-Holmes, D., Barnes-Holmes, Y., & Stewart, I. (2009). The Implicit Relational
Assessment Procedure (IRAP) and the malleability of ageist attitudes. The Psychological
Record, 59, 591–620.
Cunningham-Williams, R. M., Grucza, R. A., Cottler, L. B., Womack, S. B., Books, S. J.,
Przybeck, T. R., . . . Cloninger, C. R. (2005). Prevalence and predictors of pathological
gambling: Results from the St. Louis personality, health and lifestyle (SLPHL) study.
Journal of Psychiatric Research, 39, 377–390. doi:10.1016/j.jpsychires.2004.09.002
Daar, J. H., Negrelli, S., & Dixon, M. R. (2015). Derived emergence of wh-question–answers in
children with autism. Research in Autism Spectrum Disorders, 19, 59–71.
doi:10.1016/j.rasd.2015.06.004
D’Amato, M. R., & Colombo, M. (1985). Auditory matching-to-sample in monkeys (Cebus
apella). Animal Learning and Behavior, 13, 375–382.
Darwin, C. R. (1871). The descent of man, and selection in relation to sex. Volume 2. London,
UK: John Murray.
Darwin, C. R. (1877). A biographical sketch of an infant. Mind. A Quarterly Review of
Psychology and Philosophy, 2, 285–294. http://www.jstor.org/stable/2246907
Darwin, C. R. (1974). Metaphysics, materialism, and the evolution of mind: Early writings of
Charles Darwin. Transcribed and annotated by Paul H. Barrett with a commentary by
Howard E. Gruber. Chicago, IL: University of Chicago Press.
Darwin, C. R. (2013). On the origin of species. New York, NY: Sheba Blake Publishing. (Original
work published 1859)
de Rose, J. C., de Souza, E. G., & Hanna, E. S. (1996). Teaching reading and spelling: Exclusion
and stimulus equivalence. Journal of Applied Behavior Analysis, 29, 451–469.
de Souza, A. A., Akers, J. S., & Fisher, W. W. (2017). Empirical application of Skinner’s Verbal
Behavior to interventions for children with autism: A review. The Analysis of Verbal
Behavior, 33(2), 229–259. doi:10.1007/s40616-017-0093-7
de Souza, D. G., de Rose, J. C., & Domeniconi, C. (2009). Applying relational operants to
reading and spelling. In R. A. Rehfeldt & Y. Barnes-Holmes (Eds.), Derived relational
responding: Applications for learners with autism and other developmental disabilities (pp.
173–207). Oakland, CA: New Harbinger.
de Souza, A. A., & Rehfeldt, R. A. (2013). Effects of dictation-taking and match-to-sample
training on listing and spelling responses in adults with intellectual disabilities. Journal of
Applied Behavior Analysis, 46(4), 792–804. doi:10.1002/jaba.75
Deguchi, H. (1984). Observational learning from a radical-behavioristic viewpoint. The Behavior
Analyst, 7, 83–95.
Delgado, D., & Hayes, L. J. (2013). The integration of learning paradigms by way of a non-causal
analysis of behavioral events. Conductual, 1(2), 39–54.
Delgado, D., & Hayes, L. J. (2014). An integrative approach to learning processes: Revisiting
substitution of functions. The Psychological Record, 64(3), 625–637. doi:10.1007/s40732-
014-0071-6
Delgado, J. P., & Oblak, M. (2007). The effects of daily tact instruction on the emission of pure
mands and tacts in non-instructional settings by three preschool children with developmental
delays. Journal of Early Intensive Behavioral Interventions, 4(2), 392–41.
doi:10.1037/h0100381
Delgado, M., Erickson, N., & Johnson, K. (2015). Best practices in vocabulary acquisition with
both corrective decoding students and middle school students performing independent
research. A paper read at the 41st convention of the Association for Behavior Analysis
International, San Antonio, TX.
Delgado, M., & Johnson, K. (2010) Vocabulary instruction: Teacher-generated definitions vs.
student-generated definitions. A paper read at the 36th convention of the Association for
Behavior Analysis International, San Antonio, TX.
Delprato, D. J., & Midgley, B. D. (1992). Some fundamentals of B. F. Skinner’s behaviorism.
American Psychologist, 47(11), 1507–1520. doi:10.1037/0003-066X.47.11.1507
Deneault, J., & Ricard, M. (2006). The assessment of children’s understanding of inclusion
relations: Transitivity, asymmetry, and quantification. Journal of Cognition and
Development, 7. 551–570. doi:10.1207/s15327647jcd0704_6
Devany, J. M., Hayes, S. C., & Nelson, R. O. (1986). Equivalence class formation in language-
able and language-disabled children. Journal of the experimental analysis of behavior, 46,
 243–257.
Devine, B., Carp, C. L., Hiett, K. A., & Petursdottir, A. I. (2016). Emergence of intraverbal
responding following tact instruction with compound stimuli. The Analysis of Verbal
Behavior, 32(2), 154–170. doi:10.1007/s40616-016-0062-6
Dewey, J. (1933). How we think. In J. Dewey (1986), The later works of John Dewey, 1925–1953
(Vol. 8). Carbondale, IL: Southern Illinois University Press.
Dickes, N. R., & Kodak, T. (2015). Evaluating the emergence of reverse intraverbals following
intraverbal training in young children with autism spectrum disorder. Behavioral
Interventions, 30(3), 169–190. doi:10.1002/bin.1412
Dinsmoor, J. A. (1995a). Stimulus control: Part I. The Behavior Analyst, 18, 51–68.
doi:10.1007/BF03392691
Dinsmoor, J. A. (1995b). Stimulus control: Part II. The Behavior Analyst, 18, 253–269.
doi:10.1007/BF03392712
Dixon, M. R. (2000). Manipulating the illusion of control: Variations in gambling as a function of
perceived control over chance outcomes. The Psychological Record, 50, 705–719.
doi:10.1007/bf03395379
Dixon, M. R. (2014a). The PEAK Relational Training System: Direct Training Module.
Carbondale, IL: Shawnee Scientific Press.
Dixon, M. R. (2014b). The PEAK Relational Training System: Generalization Module.
Carbondale, IL: Shawnee Scientific Press.
Dixon, M. R. (2015). The PEAK Relational Training System: Equivalence Module. Carbondale,
IL: Shawnee Scientific Press.
Dixon, M. R. (2016). PEAK Relational Training System: Transformation module. Carbondale, IL:
Shawnee Scientific Press.
Dixon, M. R., Belisle, J., Munoz, B. E., Stanley, C. R., & Rowsey, K. E. (2017). Teaching
metaphorical extensions of private events through rival-model observation to children with
autism. Journal of Applied Behavior Analysis, 50(4), 744–749. doi:10.1002/jaba.418
Dixon, M. R., Belisle, J., Stanley, C. R., Munoz, B. E., & Speelman, R. C. (2017). Establishing
derived coordinated symmetrical and transitive gustatory-visual-auditory relations in
children with autism and related intellectual disabilities using the PEAK-E curriculum.
Journal of Contextual Behavioral Science, 6, 91–95. doi:10.1016/j.jcbs.2016.11.001
Dixon, M. R., Belisle, J., Stanley, C. R., Speelman, R. C., Rowsey, K. E., Kime, D., & Daar, J. H.
(2017). Establishing derived categorical responding in children with disabilities using the
PEAK-E curriculum. Journal of Applied Behavior Analysis, 50, 134–145.
doi:10.1002/jaba.355
Dixon, M. R., Buono, F. D., & Belisle, J (2016). Contrived motivating operations alter delay-
discounting values of disordered gamblers. Journal of Applied Behavior Analysis, 49, 986–
990. doi:10.1002/jaba.335
Dixon, M. R., & Holton, B. (2009). Altering the magnitude of delay discounting by pathological
gamblers. Journal of Applied Behavior Analysis, 42, 269–275. doi:10.1901/jaba.2009.42-269
Dixon, M., Rehfeldt, R. A., Zlomke, K.M., & Robinson, A. (2006). Exploring the development
and dismantling of equivalence classes involving terrorist stimuli. The Psychological
Record, 56, 83–103.
Dixon, M. R., Small, S. L., & Rosales, R. (2007). Extended analysis of empirical citations with
Skinner’s Verbal Behavior: 1984–2004. The Behavior Analyst, 30(2), 197–209.
doi:10.1007/bf03392155
Dixon, M. R., & Zlomke, K. M. (2005). Using the precursor to the relational evaluation procedure
(PREP) to establish the relational frames of sameness, opposition, and distinction. Revista
Latinoamericana de Psicologia, 37, 305–316.
Dollard, J., & Miller, N. E. (1950). Personality and psychotherapy; an analysis in terms of
learning, thinking, and culture. New York, NY: McGraw-Hill.
Domjan, M. (1983). Biological constraints on instrumental and classical conditioning:
Implications for general process theory. The Psychology of Learning and Motivation, 17,
215–277. doi:10.1016/S0079-7421(08)60100-0
Domjan, M. (2005). Pavlovian conditioning: A functional perspective. Annual Review of
Psychology, 56, 179–206. doi:10.1146/annurev.psych.55.090902.141409
Domjan, M. (2016). Elicited versus emitted behavior: Time to abandon the distinction. Journal of
the Experimental Analysis of Behavior, 105(2), 231–245. doi:10.1002/jeab.197
Donadeli, J. M., & Strapasson, B. A. (2015). Effects of monitoring and social reprimands on
instruction-following in undergraduate students. The Psychological Record, 65, 177–188.
doi:10.1007/s40732-014-0099-7
Donahoe, J. W. (1996). On the relation between behavior analysis and biology. The Behavior
Analyst, 19(1), 71–73. doi:10.1007/BF03392740
Donahoe, J. W. (2012). Reflections on behavior analysis and evolutionary biology: A selective
review of Evolution Since Darwin—The First 150 Years. Edited by M. A. Bell, D. J.
Futuyama, W. F. Eanes, & J. S. Levinton. Journal of the Experimental Analysis of Behavior,
97(2), 249–260. doi:10.1901/jeab.2012.97-249
Donley, C. R., & Greer, R. D. (1993). Setting events controlling social verbal exchanges between
students with developmental delays. Journal of Behavioral Education, 3(4), 387–401.
doi:10.1007/BF00961542
Dougher, M. J. (2013). Behaviorisms and private events. The Behavior Analyst, 36(2), 223–227.
Dougher, M. J., Hamilton, D. A., Fink, B. C., & Harrington, J. (2007). Transformation of the
discriminative and eliciting functions of generalized relational stimuli. Journal of the
Experimental Analysis of Behavior, 88, 179–197.
Dougher, M., Perkins, D. R., Greenway, D., Koons, A., & Chiasson, C. (2002). Contextual
control of equivalence-based transformation of functions. Journal of the Experimental
Analysis of Behavior, 78, 63–93.
Drake, C. E., Kellum, K. K., Wilson, K. G., Luoma, J. B., Weinstein, J. H., & Adams, C. H.
(2010). Examining the Implicit Relational Assessment Procedure: Four preliminary studies.
The Psychological Record, 60, 81–86.
Drake, C. E., Kramer, S., Sain, T., Swiatek, R., Kohn, K., & Murphy, M. (2015). Exploring the
reliability and convergent validity of implicit racial evaluations. Behavior and Social Issues,
 24, 68–87. doi:10.5210/bsi.v.24i0.5496
Duker, P., Kraaykamp, M., & Visser, E. (1994). A stimulus control procedure to increase
requesting with individuals who are severely/profoundly intellectually disabled. Journal of
Intellectual Disability Research, 38(2), 177–186. doi:10.1111/j.1365-2788.1994.tb00372.x
Dunne, S., Foody, M., Barnes-Holmes, Y., Barnes-Holmes, D., & Murphy, C. (2014). Facilitating
repertoires of coordination, opposition distinction, and comparison in young children with
autism. Behavioral Development Bulletin, 19, 37–47. doi:10.1037/h0100576
Dymond, S., May, R. J., Munnelly, A., & Hoon, A. E. (2010). Evaluating the evidence base for
relational frame theory: A citation analysis. The Behavior Analyst, 33, 97–117.
Dymond, S., O’Hora, D., Whelan, R., & O’Donovan, A. (2006). Citation analysis of Skinner’s
Verbal Behavior: 1984–2004. The Behavior Analyst, 29(1), 75–88. doi:10.1007/bf03392118
Dymond, S., & Rehfeldt, R. A. (2000). Understanding complex behavior: The transformation of
stimulus functions. The Behavior Analyst, 23, 239–254.
Dymond, S., & Roche, B. (2009). A contemporary behavior analysis of anxiety and avoidance.
The Behavior Analyst, 32(1), 7–27.
Dymond, S., & Roche, B. (2013). Advances in relational frame theory: Research and application.
Oakland, CA: Context Press.
Dymond, S., Roche, B., Forsyth, J. P., Whelan, R., & Rhoden, J. (2007). Transformation of
avoidance response functions in accordance with same and opposite relational frames.
Journal of the Experimental Analysis of Behavior, 88, 249–262.
Eby, C. M., & Greer, R. D. (2017). Effects of social reinforcement on the emission of tacts by
preschoolers. Behavioral Development Bulletin, 22(1), 23–43. doi:10.1037/bdb0000043
Edrisinha, C., O’Reilly, M., Sigafoos, J., Lancioni, G., & Choi, H. Y. (2011). Influence of
motivating operations and discriminative stimuli on challenging behavior maintained by
positive reinforcement. Research in Developmental Disabilities, 32, 836–845.
doi:10.1016/j.ridd.2010.10.006
Egan, C. E., & Barnes-Holmes, D. (2010). Establishing mand emergence: The effects of three
training procedures and modified antecedent conditions. The Psychological Record, 60(3),
473–488.
Eid, A. M., Alhaqbani, O. A., Asfahani, S. M., Alaql, M., AlSaud, A. N., Mohtasib, R. S., . . .
Fryling, M. J. (2017). Learning by doing and learning by observing: Training parents in
Saudi Arabia to implement the Natural Language Paradigm. Journal of Developmental and
Physical Disabilities, 29, 557–565. doi:10.1007/s10882-017-9544-2
Eikeseth, S., & Smith, D. P. (2013). An analysis of verbal stimulus control in intraverbal
behavior: Implications for practice and applied research. The Analysis of Verbal Behavior,
29(1), 125–135. doi:10.1007/bf03393130
Eisenberg, N. (2000). Emotion, regulation, and moral development. Annual review of psychology,
51(1), 665–697.
Ekman, P. (2003) Emotions revealed: Recognizing faces and feelings to improve communication
and emotional life. New York, NY: Henry Holt.
Endo, K., Ando, S., Shimodera, S., Yamasaki, S., Usami, S., Okazaki, Y., . . . Nishida, A. (2017).
Preference for solitude, social isolation, suicidal ideation, and self-harm in adolescents.
Journal of Adolescent Health, 61, 189–191.
Engelmann, S., & Carnine, D. W. (1982). Theory of instruction: Principles and applications.
Eugene, OR: ADI Press.
Epstein, R. (1985). The spontaneous interconnection of three repertoires. The Psychological
Record, 35 (2), 131–141.
Epstein, R. (1993). Generativity theory and education. Educational Technology, 33(10) 40–45.
Epstein, R. (1999). Generativity theory. In M. A. Runco and S. R. Pritzker (Eds.), Encyclopedia
of creativity (Vol. 1, pp. 759–766), Cambridge, MA: Academic Press.
Epstein, R., Kirshnit, R., Lanza, R., & Rubin, R. (1984). “Insight” in the pigeon: Antecedents and
determinants of an intelligent performance. Nature, 308, 61–62.
Estes, W. K., & Skinner, B. F. (1941). Some quantitative properties of anxiety. Journal of
Experimental Psychology, 29(5), 390–400.
Evans, S. S., & Evans, W. H. (1985). Frequencies that ensure skill competency. Journal of
Precision Teaching, 6 (2), 25–30.
Exposito, P. M., Lopez, M. H., & Valverde, M. R. (2015). Assessment of implicit anti-fat and pro-
slim attitudes in young women using the Implicit Relational Assessment Procedure.
International Journal of Psychology and Psychological Therapy, 15, 17–32.
Eysenck, H. J. (Ed.). (1960). Behaviour therapy and the neuroses: Readings in modern methods
of treatment derived from learning theory. New York, NY: Pergamon.
Eysenck, H. J. (1964). Experiments in behaviour therapy. New York, NY: Macmillan.
Fabrizio, M. A., & Pahl, S. (2007). An experimental analysis of two error correction procedures
used to improve the textual behavior of a student with autism. The Behavior Analyst Today,
8(3), 260–272. doi:10.1037/h0100618
Farrell, C. (2017). An investigation into the speaker-as-own-listener repertoire and reverse
intraverbal responding (Doctoral dissertation, Order No. 10278214). Available from
ProQuest Dissertations & Theses Global. (1899486948). doi:10.7916/D8P279JG
Farrell, L., Cochrane, A., & McHugh, L. (2015). Exploring attitudes towards gender and science:
The advantages of an IRAP approach versus the IAT. Journal of Contextual Behavioral
Science, 4(2), 121–128. doi:10.1016/j.jcbs.2015.04.002
Fawcett, S. B., Mathews, R. M., & Fletcher, R. K. (1980). Some promising dimensions for
behavioral community technology. Journal of Applied Behavior Analysis, 13, 505–518.
Feliciano, G. M. (2006). Multiple exemplar instruction and the listener half of naming in children
with limited speaker abilities (Doctoral dissertation, Order No. 3213505). Available from
ProQuest Dissertations & Theses Global. (305360301). Retrieved from http://eduproxy.tc-
library.org/?url=/docview/305360301?accountid=14258
Feng, H., Chou, W. C., & Lee, G. T. (2017). Effects of tact prompts on acquisition and
maintenance of divergent intraverbal responses by a child with autism. Focus on Autism and
Other Developmental Disabilities, 32(2), 133–141. doi:10.1177/1088357615610540
Fenske, E. C., Krantz, P. J., & McClannahan, L. E. (2001). Incidental teaching: A not-discrete-
trial teaching procedure. In C. Maurice, G. Green, & R. M. Foxx (Eds.), Making a
Difference: Behavioral intervention for autism. Austin, TX: PRO-ED.
Ferster, C. B., Culbertson, S., & Boren, M. C. P. (1975). Behavior principles (2nd ed.).
Englewood Cliffs, NJ: Prentice Hall.
Ferster, C. B., & Skinner, B. F. (1957). Schedules of reinforcement. New York, NY: Appleton-
Century-Crofts.
Fields, L., Garruto, M., & Watanabe, M. (2010). Varieties of stimulus control in a matching-to-
sample: A kernel analysis. The Psychological Record, 60, 3–26.
Fields, L., Reeve, K. F., Adams, B. J., Brown, J. L., & Verhave, T. (1997). Predicting the
extension of equivalence classes from primary generalization gradients: The merger of
equivalence classes and perceptual classes. Journal of the Experimental Analysis of
Behavior, 68, 67–91.
Fields, L., Reeve, K. F., Varelas, A., Rosen, D., & Belanich, J. (1997). Equivalence class
formation using stimulus-pairing and yes-no responding. The Psychological Record, 47,
661–686.
Fields, L., Travis, R., Roy, D., Yadlovker, E., Aguiar-Rocha, L, & Sturmey, P. (2009).
Equivalence class formation: A method for teaching statistical interactions. Journal of
Applied Behavior Analysis, 42, 575–593.
Fienup, D. M., & Brodsky, J. (2017). Effects of mastery criterion on the emergence of derived
equivalence relations. Advance online copy. Journal of Applied Behavior Analysis.
doi:10.1002/jaba.416
Fienup, D. M., Covey, D. P., & Critchfield, T. S. (2010). Teaching brain-behavior relations
economically with stimulus equivalence technology. Journal of Applied Behavior Analysis,
43, 19–33. doi:10.1901/jaba.2010.43-19
Fienup, D. M., & Critchfield, T. S. (2011). Transportability of equivalence-based programmed
instruction: Efficacy and efficiency in a college classroom. Journal of Applied Behavior
Analysis, 44, 435–450.
Fienup, D. M., Hamelin, J., Reyes-Giordano, K., & Falcomata, T. S. (2011). College-level
instruction: Derived relations and programmed instruction. Journal of Applied Behavior
Analysis, 44, 413–416.
Fienup, D. M., Mylan, S. E., Brodsky, J., & Pytte, C. (2016). From the laboratory to the
classroom: The effects of equivalence-based instruction on neuroanatomy competencies.
Journal of Behavioral Education, 25, 143–165.
Fienup, D. M., Wright, N. A., & Fields, L. (2015). Optimizing equivalence-based instruction:
Effects of training protocols on equivalence class formation. Journal of Applied Behavior
Analysis, 48, 613–631.
Figley, C. (1995). Compassion fatigue: Coping with secondary traumatic stress disorder in those
who treat the traumatized. London, UK: Brunner-Routledge
Finkel, A. S., & Williams, R. L. (2001). A comparison of textual and echoic prompts on the
acquisition of intraverbal behavior in a six-year-old boy with autism. The Analysis of Verbal
Behavior, 18(1), 61–70. doi:10.1007/bf03392971
Finn, H. E., Miguel, C. F., & Ahearn, W. H. (2012). The emergence of untrained mands and tacts
in children with autism. Journal of Applied Behavior Analysis, 45(2), 265–280.
doi:10.1901/jaba.2012.45-265
Finn, M., Barnes-Holmes, D., Hussey, I., & Graddy, J. (2016). Exploring the behavioral dynamics
of the implicit relational assessment procedure: The impact of three types of introductory
rules. The Psychological Record, 66(2), 309–321. doi:10.1007/s40732-016-0173-4
Finn, M., Barnes-Holmes, D., & McEnteggart, C. (2018). Exploring the single-trial-type-
dominance-effect on the IRAP: Developing a differential arbitrarily applicable relational
responding effects (DAARRE) model. The Psychological Record, 68, 11–25.
doi:10.1007/s40732-017-0262-z
Fiorile, C. A., & Greer, R. D. (2007). The induction of naming in children with no prior tact
responses as a function of multiple exemplar histories of instruction. The Analysis of Verbal
Behavior, 23(1), 71–87. doi:10.1007/BF03393048
Floody, M., Barnes-Holmes, Y., Barnes-Holmes, D., & Luciano, C. (2013). An empirical
investigation of hierarchical versus distinction relations in a self-based ACT exercise.
International Journal of Psychology and Psychological Therapy, 13, 373–385.
Franks, G. J., & Lattal, K. A. (1976). Antecedent reinforcement schedule training and operant
response reinstatement in rats. Animal Learning & Behavior, 4, 374–378.
doi:10.3758/BF03214424
Frias, F. A. (2017). How stimulus relations accrue for the names of things in preschoolers
(Doctoral dissertation, Order No. 10274980). Available from ProQuest Dissertations &
Theses Global. (1896618774). doi:10.7916/D8PZ5N6B
Frost, L., & Bondy, A. (2002). The Picture Exchange Communication System (PECS) training
manual (2nd ed.). Newark, DE: Pyramid Publications.
Fryling, M. J. (2017). The functional independence of Skinner’s verbal operants: Conceptual and
applied implications. Behavioral Interventions, 32(1), 70–78. doi:10.1002/bin.1462
Fryling, M. J., & Hayes, L. J. (2015). Similarities and differences among alternatives to Skinner’s
analysis of private events. The Psychological Record, 65(3), 579–587. doi:10.1007/s40732-
015-0130-7
Fryling, M. J., Johnston, C., & Hayes, L. J. (2011) Understanding observational learning: An
interbehavioral approach. The Analysis of Verbal Behavior, 27, 191–203.
Furrer, S.D., & Younger, B.A. (2008). The impact of specific prior experiences on infants’
extension of animal properties. Developmental Science, 11, 5, 712–21. doi:10.1111/j.1467-
7687.2008.00721.x
Galizio, M. (1979). Contingency-shaped and rule-governed behavior: Instructional control of
human loss avoidance. Journal of the Experimental Analysis of Behavior, 31(1), 53–70.
Galton, F. (1889). Natural inheritance. London, England: Macmillan.
Gamba, J., Goyos, C., & Petursdottir, A. I. (2015). The functional independence of mands and
tacts: Has it been demonstrated empirically? The Analysis of verbal behavior, 31(1), 10–38.
doi:10.1007/s40616-014-0026-7
Garaigordobil, M., & Maganto, C. (2011). Empathy and conflict resolution during infancy and
adolescence. Revista Latinoamericana de Psicología, 43(2), 255-266.
Geissmann, T. (2002). Duet-splitting and the evolution of gibbon songs. Biological Reviews, 77,
57–76. doi:10.1017/S1464793101005826
Gelman, S. A. (1988). The development of induction within natural kind and artefact properties.
Cognitive Psychology, 20, 65–95.
Gentner, D. (1983). Structure mapping: A theoretical framework for analogy. Cognitive Science,
7, 155–170.
Gentner, D., Holyoak, K. J., & Kokinov, B. N. (2001). The analogical mind: Perspectives from
cognitive science. Cambridge, MA: MIT Press.
Gerdes, K. E., Segal, E. A., & Lietz, C. A. (2010). Conceptualising and measuring empathy.
British Journal of Social Work, 40(7), 2326–2343.
Ghezzi, P. M. (1999). Toward a definition of social behavior. In P. M. Ghezzi, W. L. Williams, &
J. E. Carr (Eds.), Autism: Behavior analytic perspectives (pp. 212–220). Reno, NV: Context
Press.
Ghezzi, P. M., & Bijou, S. W. (1994). El entrenamiento de habilidades sociales en ninos solitarios
medianamente retardados (Social skills training for withdrawn mildly retarded children). In
L. J. Hayes, E. Ribes, & F. Lopez-Valadez (Eds), Psicologia interconductual:
Contribuciones en honor J.R. Kantor (pp. 91–110). Guadalajara, Mexico: Universidad de
Guadalajara.
Ghezzi, P. M., Bijou, S. W., & Chao, C. C. (1987, May). A comparison of two intervention
strategies for improving mildly retarded children’s conversation with nonretarded
schoolmates. Presented at the meeting of the Association for Behavior Analysis, Nashville,
TN.
Ghezzi, P. M., Bijou, S. W., & Chao, C. C. (1991). A manual for training raters to identify and
analyze referential interactions. The Psychological Record, 41, 473–486.
Ghezzi, P. M., Bijou, S. W., Umbreit, J., & Chao, C. C. (1987). Influence of age of listener on
preadolescents’ linguistic behavior. The Psychological Record, 37, 109–126.
Ghezzi, P. M., & Lyons, C. A. (1997). Specializations in the language sciences. In L.J . Hayes &
P. M. Ghezzi (Eds), Investigations in behavioral epistemology (pp. 203–214). Reno, NV:
Context Press.
Ghezzi, P. M., Robles, E., & Bijou, S. W. (1992). Relaciones sociometricas en platicar y jugar en
ninos primaria (Sociometric relationships between talk and play among elementary school
children). Revista Sonorense de Psicologia, 6, 15–22.
Gil, E., Luciano, C., Ruiz, F. J., & Valdivia-Salas, S. (2012). A preliminary demonstration of
transformation of functions through hierarchical relations. International Journal of
Psychology & Psychological Therapy, 12, 1–19.
Gil, E., Luciano, C., Ruiz, F. J., & Valdivia-Salas, S. (2014). A further experimental step in the
analysis of hierarchical responding. International Journal of Psychology & Psychological
Therapy, 14, 137–153.
Gil-Luciano B., Ruiz F. J., Valdivia-Salas S., & Suarez-Falcon, J. C. (2016). Promoting
psychological flexibility on tolerance tasks: framing behavior through deictic/hierarchical
relations and specifying augmental functions. The Psychological Record, 67(1), 1–9.
Gilbert, P. (2010). Compassion focused therapy. New York, NY: Routledge Press
Gilbert, T, (1962a). Mathetics: The technology of education. Journal of Mathetics, 1(1), 7–74.
Gilbert, T. (1962b). Mathetics II: The design of teaching exercises. Journal of Mathetics, 1(2), 7–
56.
Gilic, L., & Greer, R. D. (2009). Establishing Naming in typically developing two-year-old
children as a function of multiple exemplar speaker and listener experiences. The Analysis of
Verbal Behavior, 27(1), 157–177.
Gilliam, A., Weil, T. M., & Miltenberger, R. G. (2013). Effects of preference on the emergence of
untrained verbal operants. Journal of Applied Behavior Analysis, 46(2), 523–527.
doi:10.1002/jaba.34
Gladstone, B. W., & Cooley, J. (1975). Behavioral similarity as a reinforcer for preschool
children. Journal of the Experimental Analysis of Behavior, 23 (3), 357–368.
Glynn, E. L., Thomas, J. D., & Shee, S. M. (1973). Behavioral self-control of on-task behavior in
an elementary classroom. Journal of Applied Behavior Analysis, 6, 105–113.
doi:10.1901/jaba.1973.6-105
Goetz, J. L., Keltner, D., & Simon-Thomas, E. (2010). Compassion: An evolutionary analysis and
empirical review. Psychological Bulletin, 136(3), 351–374.
Golan, O., & Baron-Cohen, S. (2006). Systemizing empathy: Teaching adults with Asperger
syndrome or high-functioning autism to recognize complex emotions using interactive
multimedia. Development and psychopathology, 18(2), 591–617.
Golden, J. (2009). The etiology and treatment of significant societal problems: Behaviorists
diving into murky waters with strangers. International Journal of Behavioral Consultation &
Therapy, 5(1), 1–11.
Goldsmith, T. R., LeBlanc, L. A., & Sautter, R. A. (2007). Teaching intraverbal behavior to
children with autism. Research in Autism Spectrum Disorders, 1(1), 1–13.
doi:10.1016/j.rasd.2006.07.001
Goleman, D. (1998). Working with emotional intelligence. New York, NY: Bantam Books.
Gore, N. J., Barnes-Holmes, Y., & Murphy, G. (2010). The relationship between intellectual
functioning and relational perspective-taking. International Journal of Psychology and
Psychological Therapy, 10.
Gortner, E. T., Gollan, J. K., Dobson, K. S., & Jacobson, N. S. (1998). Cognitive-behavioral
treatment for depression: Relapse prevention. Journal of Consulting and Clinical
Psychology, 66, 377–384.
Goswami, U., & Brown, A. (1989). Melting chocolate and melting snowmen: Analogical
reasoning and causal relations. Cognition, 35, 69-95.
Goswami, U., & Brown, A. (1990). Higher-order structure and relational reasoning: Contrasting
analogical and thematic relations. Cognition, 36, 207–226.
Gould, E., Dixon, D. R., Najdowski, A. C., Smith, M. N., & Tarbox, J. (2011). A review of
assessments for determining the content of early intensive behavioral intervention programs
for autism spectrum disorders. Research in Autism Spectrum Disorders, 5(3), 990–1002.
doi:10.1016/j.rasd.2011.01.012
Gould, E., Tarbox, J., & Coyne, L. (2017). Evaluating the effects of Acceptance and Commitment
Training on the overt behavior of parents of children with autism. Journal of Contextual
Behavioral Science, 7, 81–88. doi:10.1016/j.jcbs.2017.06.003
Grannan, L., & Rehfeldt, R. A. (2012). Emergent intraverbal responses via tact and match-to-
sample instruction. Journal of Applied Behavior Analysis, 45(3), 601–605.
doi:10.1901/jaba.2012.45-601
Green, G. (2001). Behavior analytic instruction for learners with autism: Advances in stimulus
control technology. Focus on Autism and Other Developmental Disabilities, 16, 72–85.
doi:10.1177/108835760101600203
Green, D., Charman, T., Pickles, A., Chandler, S., Loucas, T., Simonoff, E., & Baird, G., (2009).
Impairment in movement skills of children with autistic spectrum disorders. Developmental
Medicine and Child Neurology, 51, 311–316. doi:10.1111/j.1469-8749.2008.03242.x
Greenberg, M. S., & Beck, A. T. (1989). Depression versus anxiety: A test of the content-
specificity hypothesis. Journal of Abnormal Psychology, 98(1), 9–13.
Greene, T. R. (1994). What kindergartners know about class containment hierarchies. Journal of
Experimental Child Psychology, 57, 72–88. doi:10.1006/jecp.1994.1004
Greer, R. D. (2008). The ontogenetic selection of verbal capabilities: Contributions of Skinner’s
verbal behavior theory to a more comprehensive understanding of language. International
Journal of Psychology and Psychological Therapy, 8, 363–386.
http://oai.redalyc.org/articulo.oa?id=56080310
Greer, R. D., Corwin, A., & Buttigieg (2011). The effects of the verbal developmental capability
of Naming on how children can be taught. Acta de Investigacion Psicologia, 1(1), 23–54.
Retrieved from http://scielo.unam.mx/pdf/aip/v1n1/v1n1la5.pdf
Greer, R. D., & Du, L. (2010). Generic instruction versus intensive tact instruction and the
emission of spontaneous speech. The Journal of Speech-Language Pathology and Applied
Behavior Analysis, 5(1), 1–19. doi:10.1037/h0100261 IN 1932-4731
Greer, R. D., & Du, L. (2015a). Experience and the onset of the capability to learn names
incidentally by exclusion. The Psychological Record, 65, 355–373. doi:10.1007/s40732-014-
0111-2
Greer, R. D., & Du, L. (2015b). Identification and establishment of reinforcers that make the
development of complex social language possible. International Journal of Behavior
Analysis and Autism Spectrum Disorders, 1(1), 13–34.
Greer, R. D., Pohl, P., Du, L., & Moschella, J. L. (2017). The separate development of children’s
listener and speaker behavior and the intercept as behavioral metamorphosis. Journal of
Behavioral and Brain Science, 7, 674–704. doi:10.4236/jbbs.2017.713045
Greer, R. D., & Ross, D. E. (2008). Verbal behavior analysis: Inducing and expanding complex
communication in children with severe language delays. Boston, MA: Allyn & Bacon.
Greer, R. D., & Singer-Dudek, J. (2008). The emergence of conditioned reinforcement from
observation. Journal of the Experimental Analysis of Behavior, 89, 15–29.
doi:10.1901/jeab.2008.89-15
Greer, R. D., Singer-Dudek, J., & Gautreaux, G. (2006). Observational learning. International
Journal of Psychology, 41, 486–499. doi:10.1080/00207590500492435
Greer, R. D., Singer-Dudek, J., Longano, J., & Zrinzo, M. (2008). The emergence of praise as
conditioned reinforcement as a function of observation in preschool and school age children.
Revista Mexicana de Psicologia, 25(1), 5–26. Retrieved from
http://www.redalyc.org/articulo.oa?id=243016300001
Greer, R. D., & Speckman, J. (2009). The integration of speaker and listener responses: A theory
of verbal development. The Psychological Record, 59(3), 449. doi:10.1007/bf03395674
Greer, R. D., Stolfi, L., Chavez-Brown, M., & Rivera-Valdez, C. (2005). The emergence of the
listener to speaker component of Naming in children as a function of multiple exemplar
instruction. The Analysis of Verbal Behavior, 21, 123–134. doi:10.1007/BF03393014
Greer, R. D., Yaun, L., & Gautreaux, G. (2005). Novel dictation and intraverbal responses as a
function of a multiple exemplar instructional history. The Analysis of Verbal Behavior, 21(1),
99–116. doi:10.1007/bf03393012
Gregory, M. K., DeLeon, I. G., & Richman, D. M. (2009). The influence of matching and motor-
imitation abilities on rapid acquisition of manual signg and exhange-based communicative
responses. Journal of Applied Behavior Analysis, 42(2), 399–404. doi:10.1901/jaba.2009.42-
399
Greville, W. J., Dymond, S., & Newton, P. M. (2016). The student experience of applied
equivalence-based instruction for neuroanatomy teaching. Journal of Educational
Evaluation for Health Professions, 13, 1–8.
Griffee, K., & Dougher, M. J. (2002). Contextual control of stimulus generalisation and stimulus
equivalence in hierarchical categorization. Journal of Experimental Analysis of Behavior, 78,
3, 433–447. doi:10.1901/jeab.2002.78-433.
Grossberg, J. M. (1964). Behavior therapy: A review. Psychological Bulletin, 62(2), 73–88.
Groves, P. M., & Thompson, R. F. (1970). Habituation: A dual-process theory. Psychological
Review, 77, 419–450. doi:10.1037/h0029810
Grow, L. L., & Kodak, T. (2010). Recent research on emergent verbal behavior: Clinical
applications and future directions. Journal of Applied Behavior Analysis, 43(4), 775–778.
doi:10.1901/jaba.2010.43-775
Guercio, J. M., Podolska-Schroeder, H., & Rehfeldt, R. A. (2004). Using stimulus equivalence
technology to teach emotion recognition to adults with acquired brain injury. Brain Injury,
18, 593–601.
Guerrero, M., Alós, F. J., & Moriana, J. A. (2015). Emergent relations with compound stimuli in
conditional and simple discriminations: An experimental application in children. The
Psychological Record, 65(3), 475–486. doi:10.1007/s40732-015-0123-6
Guralnick, M., J. (2001). Social competence with peers and early childhood inclusion. In M. J.
Guralnick (Ed.), Early childhood inclusion: Focus on change (pp. 481–502). Baltimore,
MD: Brookes.
Haaga, D. A., Dyck, M. J., & Ernst, D. (1991). Empirical status of cognitive theory of depression.
Psychological Bulletin, 110(2), 215–236.
Haegele, K. M., McComas, J. J., Dixon, M., & Burns, M. K. (2011). Using a stimulus
equivalence paradigm to teach numerals, English words, and Native American words to
 preschool-age children. Journal of Behavioral Education, 20, 283–296. doi:10.1007/s10864-
011-9134-9
Haggar, J., Ingvarsson, E. T., & Braun, E. C. (2018). Further evaluation of blocked trials to teach
intraverbal responses under complex stimulus control: Effects of criterion-level probes.
Learning and Motivation, 62, 29–40. doi:10.1016/j.lmot.2017.02.006
Halford, G. S., Andrews, G., & Jensen, I. (2002). Integration of category induction and
hierarchical classification: One paradigm at two levels of complexity. Journal of Cognition
and Development, 3, 2, 143–177. doi:10.1207/S15327647JCD0302_2.
Hall, G. & Sundberg, M. L. (1987). Teaching mands by manipulating conditioned establishing
operations. The Analysis of Verbal Behavior, 5(1), 41–53. doi:10.1007/BF03392819
Hall, S. S., DeBernardis, G. M., & Reiss, A. L. (2006). The acquisition of stimulus equivalence in
individuals with fragile X syndrome. Journal of Intellectual Disability Research, 50, 643–
651.
Hanna, E. S., de Souza, D. G., Rose, J. C., & Fonseca, M. (2004). Effects of delayed constructed-
response identity matching on spelling of dictated words. Journal of Applied Behavior
Analysis, 37(2), 223–227. doi:10.1901/jaba.2004.37-223
Hanson, N. R. (1958). Patterns of discovery: An inquiry into the conceptual foundations of
science. Cambridge, England: Cambridge University Press.
Haq, S. S., & Kodak, T. (2015). Evaluating the effects of massed and distributed practice on
acquisition and maintenance of tacts and textual behavior with typically developing children.
Journal of Applied Behavior Analysis, 48(1), 85–95. doi:10.1002/jaba.178
Haq, S. S., Zemantic, P. K., Kodak, T., LeBlanc, B., & Ruppert, T. E. (2017). Examination of
variables that affect the efficacy of instructive feedback. Behavioral Interventions, 32(3),
206–216. doi:10.1002/bin.1470
Harmon, K., Strong, R., & Pasnak, R. (1982). Relational responses in tests of transposition with
rhesus monkeys. Learning And Motivation, 13, 495–504. doi:10.1016/0023-9690(82)90006-
6
Harris, S. L. (1975). Teaching language to nonverbal children with emphasis on problems of
generalization. Psychological Bulletin, 82, 565–580. doi:10.1037/h0076903.
Haughton, E. (1972). Aims: Growing and sharing. In J. B. Jordon and L.S. Robbins (Eds.), Let’s
try doing something different kind of thing. (pp. 20–39). Arlington, VA: Council for
Exceptional Children.
Haughton, E. (1980). Practicing practices: Learning by activity. Journal of Precision Teaching, 1,
3–20.
Hausman, N. L., Borrero, J. C., Fisher, A., & Kahng, S. (2014). Improving accuracy of portion-
size estimations through a stimulus equivalence paradigm. Journal of Applied Behavior
Analysis, 47, 485–499.
Hausman, N. L., Borrero, J. C., Fisher, A., & Kahng, S. (2017). Teaching young children to make
accurate portion size estimations using a stimulus equivalence paradigm. Behavioral
Interventions, 32, 121–132. doi:10.1002/bin.1466
Hayes, L. J. (1991). Substitution and reference. In L. J. Hayes & P. N. Chase (Eds.), Dialogues on
verbal behavior: The first international institute on verbal relations (pp. 3–14). Reno, NV:
 Context Press.
Hayes, L. J. (1992). Equivalence as process. In S. C. Hayes & L. J. Hayes (Eds.), Understanding
verbal relations (pp. 97–108). Reno, NV: Context Press.
Hayes, L. J., & Chase, P. N. (1991). Dialogues on verbal behavior: The First International
Institute on Verbal Relations. Reno, NV: Context Press.
Hayes, L. J., & Fryling, M. J. (2009). Overcoming the pseudo-problem of private events in the
analysis of behavior. Behavior and Philosophy, 39–57.
Hayes, L. J., & Fryling, M. J. (2014). Motivation in behavior analysis: A critique. The
Psychological Record, 64(2), 339–347. doi:10.1007/s40732-014-0025-z
Hayes, L. J., & Fryling, M. J. (2015). A historical perspective on the future of behavior science.
The Behavior Analyst, 38(2), 149–161. doi:10.1007/s40614-015-0030-9
Hayes, S. C. (1984). Making sense of spirituality. Behaviorism, 12, 99–110.
Hayes, S. C. (1989). Nonhumans have not yet shown stimulus equivalence. Journal of the
Experimental Analysis of Behavior, 51(3), 385–392.
Hayes, S. C. (1996). Developing a theory of derived stimulus relations. Journal of the
Experimental Analysis of Behavior, 65, 309–311.
Hayes, S. C. (2004). Acceptance and commitment therapy, relational frame theory, and the third
wave of behavioral and cognitive therapies. Behavior Therapy, 35(4), 639–665.
Hayes, S. C., Barnes-Holmes, D., & Roche, B. (Eds.). (2001). Relational frame theory: A post-
Skinnerian account of human language and cognition. New York, NY: Plenum Press.
Hayes, S. C., Barnes-Holmes, D., & Wilson, K. G. (2012). Contextual behavioral science:
Creating a science more adequate to the challenge of the human condition. Journal of
Contextual Behavioral Science, 1(1), 1–16. doi:10.1016/j.jcbs.2012.09.004
Hayes, S. C., Brownstein, A. J., Haas, J. R., & Greenway, D. E. (1986). Instructions, multiple
schedules, and extinction: Distinguishing rule-governed from schedule-controlled behavior.
Journal of the Experimental Analysis of Behavior, 46, 137–147. doi:10.1901/jeab.1986.46-
137
Hayes, S. C., Brownstein, A. J., Zettle, R. D., Rosenfarb, I., & Korn, Z. (1986). Rule-governed
behavior and sensitivity to changing consequences of responding. Journal of the
Experimental Analysis of Behavior, 45, 237–256. doi:10.1901/jeab.1986.45-237
Hayes, S. C., Gifford, E. V., Townsend, R. C., & Barnes-Holmes, D. (2001). Thinking, problem-
solving, and pragmatic verbal analysis. In S. C. Hayes, D. Barnes-Holmes, & B. Roche
(Eds.), Relational frame theory: A post-Skinnerian account of human language and
cognition (pp. 87–101). New York, NY: Kluwer Academic/Plenum Publishers.
Hayes, S. C, & Hayes, L. J. (1989). The verbal action of the listener as a basis for rule-
governance. In S. C. Hayes (Ed.), Rule-governed behavior: Cognition, contingencies, and
instructional control (pp. 171–173). New York, NY: Plenum Press.
Hayes, S. C., Hayes, L. J., & Reese, H. W. (1988). Finding the philosophical core: A review of
Stephen C. Pepper’s world hypotheses: A study in evidence. Journal of the Experimental
Analysis of Behavior, 50(1), 97–111. doi:10.1901/jeab.1988.50-97
Hayes, S. C., Kohlenberg, B., & Hayes, L. J. (1991). The transfer of specific and general
consequential functions through simple and conditional equivalence relations. Journal of the
Experimental Analysis of Behavior, 56, 119–137.
Hayes, S. C., Levin, M. E., Plumb-Vilardaga, J., Villatte, J. L., & Pistorello, J. (2013). Acceptance
and commitment therapy and contextual behavioral science: Examining the progress of a
distinctive model of behavioral and cognitive therapy. Behavior Therapy, 44(2), 180–198.
doi:10.1016/j.beth.2009.08.002
Hayes, S. C., Luoma, J., Bond, F., Masuda, A., & Lillis, J. (2006). Acceptance and commitment
therapy: Model, processes, and outcomes. Behavior Research and Therapy, 44, 1–25.
Hayes, S. C., Pistorello, J., & Levin, M. E. (2012). Acceptance and commitment therapy as a
unified model of behavior change. The Counseling Psychologist, 40, 976–1002.
doi:10.1177/0011000012460836
Hayes, S. C., & Sanford, B. T. (2014). Cooperation came first: Evolution and human cognition.
Journal of Experimental Analysis of Behavior, 101, 112–129. doi:10.1002/jeab.64
Hayes, S. C., & Sanford, B. T. (2015). Modern psychotherapy as a multidimensional multilevel
evolutionary process. Current Opinion in Psychology, 2, 16–20.
doi:10.1016/j.copsyc.2015.01.009
Hayes, S. C., Sanford, B. T., & Chin, F. T. (2017). Carrying the baton: Evolution science and a
contextual behavioral analysis of language and cognition. Journal of Contextual Behavioral
Science. doi:10.1016/j.jcbs.2017.01.002
Hayes, S. C., Strosahl K. D., Bunting K., Twohig M. P., & Wilson K. G. (2004). What is
acceptance and commitment therapy? In Hayes S. C., Strosahl K. D. (Eds.), A practical
guide to acceptance and commitment therapy (pp. 1–30). New York, NY: Guilford Press.
Hayes, S. C., Strosahl, K. D., & Wilson, K. G. (1999). Acceptance and commitment therapy: An
experiential approach to behavior change. New York, NY: Guilford Press.
Hayes, S. C., Strosahl, K. D., & Wilson, K. G. (2012). Acceptance and commitment therapy: The
process and practice of mindful change (2nd ed.). New York, NY: Guilford Press.
Hayes, S. C., Strosahl, K., Wilson, K. G., Bissett, R. T., Pistorello, J., Toarmino, D., . . . Stewart,
S. H. (2004). Measuring experiential avoidance: A preliminary test of a working model. The
Psychological Record, 54(4), 553–578.
Hayes, S. C., & Wilson, K. G. (1994). Acceptance and commitment therapy: Altering the verbal
support for experiential avoidance. The Behavior Analyst, 17(2), 289–303.
Hayes, S. C., Zettle, R. D., & Rosenfarb, I. (1989). Rule-following. In S. Hayes (Ed.), Rule-
governed behavior: Cognition, contingencies, and instructional control (pp. 191–220).
Reno, NV: Context Press.
Heagle, A. I., & Rehfeldt, R. A. (2006) Teaching perspective-taking skills to typically developing
children through derived relational responding. Journal of Early and Intensive Behavior
Intervention, 3, 1–34.
Healy, O., Barnes, D., & Smeets, P. M. (1998). Derived relational responding as an operant: The
effects of between-session feedback. The Psychological Record, 48, 511–536.
Healy, O., Barnes-Holmes, D., & Smeets, P. M. (2000). Derived relational responding as
generalized operant behavior. Journal of the Experimental Analysis of Behavior, 74, 207–
 227. doi:10.1901/jeab.2000.74-207
Heffner, H. E., & Heffner, R. S. (2007). Hearing ranges of laboratory animals. Journal of the
American Association for Laboratory Animal Science, 46, 11–13.
Heider, N., Spruyt, A., & De Houwer, J. (2015). Implicit beliefs about ideal body image predict
body image dissatisfaction. Frontiers in Psychology, 6, 1402. doi:10.3389/fpsyg.2015.01402
Henrich, J. (2004). Cultural group selection, coevolutionary processes and large-scale
cooperation. Journal of Economic Behavior & Organization, 53, 5–35.
Hermann, J. A., Montes, A. I. D., Domínguez, B., Montes, F., & Hopkins, B. L. (1973). Effects of
bonuses for punctuality on the tardiness of industrial workers. Journal of Applied Behavior
Analysis, 6, 563–570. doi:10.1901/jaba.1973.6-563
Herrnstein, R. J. (1990). Levels of stimulus control: A functional approach. Cognition, 37(1),
133–166.
Herrnstein, R. J., Loveland, D. H., & Cable, C. (1979). Natural concepts in pigeons. Journal of
Experimental Psychology: Animal Behavior Processes, 2, 285–302. doi:10.1037/0097-
7403.2.4.285
Hinton, E. C., Dymond, S., Von Hecker, U., & Evans, C. J. (2010). Neural correlates of relational
reasoning and the symbolic distance effect: Involvement of parietal cortex. Neuroscience,
168, 138–148.
Hobbs, T. R., & Holt, M. (1976). The effects of token reinforcement on the behavior of
delinquents in cottage settings. Journal of Applied Behavior Analysis, 9, 189–198.
doi:10.1901/jaba.1976.9-189
Hoffman, M. L. (1981). Is altruism part of human nature? Journal of Personality and Social
Psychology, 40, 121–137.
Holt, G. L. (1971). Effect of reinforcement contingencies in increasing programmed reading and
mathematics behaviors in first-grade children. Journal of Experimental Child Psychology,
12, 362–369. doi:10.1016/0022-0965(71)90031-2
Holyoak, K. J., & Thagard, P. R. (1989). Analogical mapping by constraint satisfaction. Cognitive
Science, 13, 295–355.
Holz, W. C., & Azrin, N. H. (1961). Discriminative properties of punishment. Journal of the
Experimental Analysis of Behavior, 4, 225–232. doi:10.1901/jeab.1961.4-225
Honig, W. K., & Urcuioli, P. J. (1981). The legacy of Guttman and Kalish (1956): 25 years of
research on stimulus generalization. Journal of the Experimental Analysis of Behavior, 36,
405–445. doi:10.1901/jeab.1981.36-405
Hood, S. A., Luczynski, K. C., & Mitteer, D. R. (2017). Toward meaningful outcomes in teaching
conversation and greeting skills to with individuals with autism spectrum disorder. Journal
of Applied Behavior Analysis, 50, 459–486.
Horne, P. J., & Lowe, C. F. (1996). On the origins of naming and other symbolic behavior.
Journal of the Experimental Analysis of Behavior, 65, 185–241. doi:10.1901/jeab.1996.65-
185
Horner, R. H., & Day, H. M. (1991). The effects of response efficiency on functionally equivalent
competing behaviors. Journal of Applied Behavior Analysis, 24(4), 719–732.
 doi:10.1901/jaba.1991.24-719
Horowitz, F. D. (1987). Exploring developmental theories: Toward a structural/behavioral model
of development. Hillsdale, NJ: Lawrence Erlbaum.
Howlin, P., Baron-Cohen, S., & Hadwin, J. (1999). Teaching children with autism to mind-read: A
practical guide for teachers and parents. Chichester, UK: Wiley.
Hübner, M. M. C., Austin, J., & Miguel, C. F. (2008). The effects of praising qualifying
autoclitics on the frequency of reading. The Analysis of Verbal Behavior, 24(1), 55–62.
doi:10.1007/BF03393056
Hughes, C. (1992). Teaching self-instruction utilizing multiple exemplars to produce generalized
problem-solving among individuals with severe mental retardation. American Journal on
Mental Retardation, 97, 302–314.
Hughes, S., Barnes-Holmes, D., & Smyth, S. (2017). Implicit cross-community biases revisited:
Evidence for ingroup favoritism in the absence of outgroup derogation in Northern Ireland.
The Psychological Record, 67, 97–107. doi:10.1007/s40732-016-0210-3
Hughes, S., Barnes-Holmes, D., & Vahey, N. (2012). Holding on to our functional roots when
exploring new intellectual islands: A voyage through implicit cognition research. Journal of
Contextual Behavioral Science, 1(1), 17–38. doi:10.1016/j.jcbs.2012.09.003
Hussey, I., & Barnes-Holmes, D. (2012). The implicit relational assessment procedure as a
measure of implicit depression and the role of psychological flexibility. Cognitive and
Behavioral Practice, 19(4), 573–582. doi:10.1016/j.cbpra.2012.03.002
Ilardi, S. S., & Craighead, W. E. (1994). The role of nonspecific factors in cognitive-behavior
therapy for depression. Clinical Psychology: Science and Practice, 1(2), 138–155.
Imada, T., Zhang, Y., Cheour, M., Taulu, S., Ahonen, A., & Kuhl, P. K. (2006). Infant speech
perception activates Broca’s area: A developmental magnetoencephalography study.
Neuroreport, 17(10), 957–962. doi:10.1097/01.wnr.0000223387.51704.89
Imam, A. A., & Warner, T. A. (2014). Test order effects in simultaneous protocols. Learning &
Behavior, 42, 93–103.
Immordino-Yang, M. H., McColl, A., Damasio, H., Damasio, A. (May 2009). Neural correlates of
admiration and compassion. Proceedings of the National Academy of Sciences, 106, 19,
8021–8026.
Ingvarsson, E. T., & Hollobaugh, T. (2011). A comparison of prompting tactics to establish
intraverbals in children with autism. Journal of Applied Behavior Analysis, 44(3), 659–664.
doi:10.1901/jaba.2011.44-659
Ingvarsson, E. T., Kramer, R. L., Carp, C. L., Pétursdóttir, A. I., & Macias, H. (2016). Evaluation
of a blocked-trials procedure to establish complex stimulus control over intraverbal
responses in children with autism. The Analysis of Verbal Behavior, 32(2), 205–224.
doi:10.1007/s40616-016-0071-5
Inhelder, B., & Piaget, J. (1964). The early growth of logic in the child. London: Routledge &
Kegan Paul.
Jackson, M. L., Mendoza, D. R., & Adams, A. N. (2014). Teaching a deictic relational repertoire
to children with autism. The Psychological Record, 64, 791–802.
Jacobs, K. W., Morford, Z. H., King, J. E., & Hayes, L. J. (2017). Predicting the effects of
interventions: A tutorial on the disequilibrium model. Behavior Analysis in Practice.
doi:10.1007/s40617-017-0176-x
Jacobson, N. S., Dobson, K. S., Truax, R. A., Addis, M. E., Koerner, K., Gollan, J. K., . . . Prince,
S. E. (1996). A component analysis of cognitive-behavioral treatment for depression.
Journal of Consulting and Clinical Psychology, 64, 295–304.
Johnson, K. (2015). Behavioral education in the 21st century. Journal of Organizational Behavior
Management, 35, 135–150. doi:10.1080/01608061.2015.1036152
Johnson, K., & Layng, T. V. J. (1992). Breaking the structuralist barrier: Literacy and numeracy
with fluency. American Psychologist, 47, 1475–1498.
Johnson, K., & Layng, T. V. J. (1994). The Morningside Model of Generative Instruction. In R.
Gardner, D. Sainato, J. Cooper, T. Heron, W. Heward, J. Eshleman, & T. Grossi (Eds.),
Behavior analysis in education: Focus on measurably superior instruction (pp. 173–197).
Belmont, CA: Brooks-Cole.
Johnson, K., & Street, E. M. (2004). The Morningside model of generative instruction: What it
means to leave no child behind. Concord, MA: Cambridge Center for Behavioral Studies.
Johnson, K. J., & Street, E. M. (2012). From the laboratory to the field and back again:
Morningside Academy’s 32 years of improving students’ academic performance. The
Behavior Analyst Today, 13, 20–40.
Johnson, K., & Street, E. M. (2013). Response to intervention and precision teaching: Creating
synergy in the classroom. New York, NY: Guilford Publications
Johnston, J. (1996). Distinguishing between applied research and practice. The Behavior Analyst,
19, 35–47.
Jonah, B. A., & Grant, B. A. (1985). Long-term effectiveness of selective traffic enforcement
programs for increasing seat belt use. Journal of Applied Psychology, 70, 257–263.
doi:10.1037/0021-9010.70.2.257
Joyce, B. G., Joyce, J. H., & Wellington, B. (1993). Using stimulus equivalence procedures to
teach relationships between English and Spanish words. Education & Treatment of Children,
16, 48–65.
Juarascio, A. S., Forman, E. M., Timko, C. A., & Herbert, J. D. (2011). Implicit internalization of
the thin ideal as a predictor of increases in weight, body dissatisfaction, and disordered
eating. Eating Behaviors, 12, 207–213. doi:10.1016/j.eatbeh.2011.04.004
Kalish, C. W., & Gelman, S. A. (1992). On wooden pillows: Multiple classification and children’s
category-based inductions. Child Development, 63, 6, 1536–1557.
Kamin, L. J. (1969). Predictability, surprise, attention, and conditioning. In B. A. Campbell & R.
M. Church (Eds.), Punishment and aversive behavior (pp. 279–296). New York, NY:
Appleton-Century-Crofts.
Kanfer, F. H. (1959). Verbal rate, content, and adjustment ratings in experimentally structured
interviews. Journal of Abnormal Psychology, 58(3), 305–311.
Kanfer, F. H. (1961). Comments on learning in psychotherapy. Psychological Reports, 9(3), 681–
699.
Kanfer, F. H., & McBrearty, J. F. (1962). Minimal social reinforcement and interview content.
Journal of Clinical Psychology, 18, 210–215.
Kanfer, F. H., Phillips, J. S., Matarazzo, J. D., & Saslow, G. (1960). Experimental modification of
interviewer content in standardized interviews. Journal of Consulting Psychology, 24, 528–
536.
Kantor, J. R. (1921). An objective interpretation of meanings. American Journal of Psychology,
32, 231–248.
Kantor, J. R. (1922). An analysis of language data. Psychological Review, 29, 267–309.
Kantor, J. R. (1924). Principles of psychology (vol. 1). New York, NY: Alfred Knopf.
Kantor, J. R. (1926). Principles of psychology. (vol. II). New York, NY: Alfred Knoph.
Kantor, J. R. (1928). Can psychology contribute to the study of linguistics? Monist, 26, 630–648.
Kantor, J. R. (1929). Language as behavior and as symbolism. Journal of Philosophy, 26, 150–
159.
Kantor, J. R. (1933). A survey of the science of psychology. Chicago, IL: Principia Press.
Kantor, J. R. (1936). An objective psychology of grammar. Chicago, IL: The Principia Press.
Kantor, J. R. (1938). The role of language in logic and science. Journal of Philosophy, 35, 449–
463.
Kantor, J. R. (1953). The logic of modern science. Chicago, IL: Principia Press.
Kantor, J. R. (1958). Interbehavioral psychology. Chicago, IL: Principia Press.
Kantor, J. R. (1963), Behaviorism: Whose image? The Psychological Record, 13, 499–512.
Kantor, J. R. (1975). Psychological linguistics. Mexican Journal of Behavior Analysis, 1, 249–
268.
Kantor, J. R. (1977). Psychological linguistics. Chicago, IL: Principia Press.
Kantor, J. R. (1981). Reflections on speech and language. Mexican Journal of Behavior Analysis,
7, 91–106.
Kantor, J. R. (1982). Cultural psychology. Chicago, IL: The Principia Press.
Kantor, J. R., & Smith, N. W. (1975). The science of psychology: An interbehavioral survey.
Chicago, IL: Principia Press.
Karmali, I., Greer, R. D., Nuzzolo-Gomez, R., Ross, D. E., & Rivera-Valdes, C. (2005). Reducing
palilalia by presenting tact corrections to young children with autism. The Analysis of Verbal
Behavior, 21(1), 145–153. doi:10.1007/BF03393016
Kato, O. M., de Rose, J. C., & Faleiros, P. B. (2008). Topography of responses in conditional
discrimination influences formation of equivalence classes. The Psychological Record, 58,
245–267.
Kaufman, A., Baron, A., & Kopp, R. E. (1966). Some effects of instructions on human operant
behavior. Psychonomic Monograph Supplements, 1, 243–250.
Kazdin, A. E. (1979). Fictions, factions, and functions of behavior therapy. Behavior Therapy,
10(5), 629–654. doi:10.1016/S0005-7894(79)80066-0
Keane, M. T. (1997). What makes an analogy difficult? The effects of order and causal structure
on analogical mapping. Journal of Experimental Psychology: Learning, Memory, and
Cognition, 23, 4, 946–967.
Keintz, K. S., Miguel, C. F., Kao, B., & Finn, H. E. (2011). Using conditional discrimination
training to produce emergent relations between coins and their values in children with
autism. Journal of Applied Behavior Analysis, 44, 909–913. doi:10.1901/jaba.2011.44-909
Keller, F. S. (1968). “Good-bye, teacher…” Journal of Applied Behavior Analysis, 1, 79–89.
doi:10.1901/jaba.1968.1-79
Keohane, D. D., Pereira Delgado, J. A., & Greer, R. D. (2009). Observing responses: Foundations
of higher-order verbal operants. In R. A. Rehfeldt & Y. Barnes-Holmes (Eds.), Derived
relational responding: Applications for learners with autism and other developmental
disabilities (pp. 41–62). Oakland, CA: New Harbinger Publications.
Keuleers, E., Diependaele, K., & Brysbaert, M. (2010). Practice effects in large-scale visual word
recognition studies: A lexical decision study on 14,000 Dutch mono-and disyllabic words
and nonwords. Frontiers in Psychology, 1, 174.
Killeen, P. R., & Fetterman, J. G. (1988). A behavioral theory of timing. Psychological Review,
95, 274–295. doi:10.1037/0033-295X.95.2.274
Killeen, P. R., & Jacobs, K. W. (2017a). Coal is not black, snow is not white, food is not a
reinforcer: the roles of affordances and dispositions in the analysis of behavior. The Behavior
Analyst, 40(1), 17–38. doi:10.1007/s40614-016-0080-7
Killeen, P. R., & Jacobs, K. W. (2017b). The modulated contingency. The Behavior Analyst,
online first. doi:10.1007/s40614-017-0101-1
King, C. M., & Quigley, S. P. (1985). Reading and deafness. San Diego, CA: College Hill Press.
Kinloch, J. M., McEwan, J. S. A., & Foster, T. M. (2013). Matching-to-sample and stimulus-
pairing-observation procedures in stimulus equivalence: The effects of number of trials and
stimulus arrangement. The Psychological Record, 63, 157–174.
Kisamore, A. N., Carr, J. E., & LeBlanc, L. A. (2011). Training preschool children to use visual
imagining as a problem-solving strategy for complex categorization tasks. Journal of
Applied Behavior Analysis, 44(2), 255–278. doi:10.1901/jaba.2011.44-255
Kisamore, A. N., Karsten, A. M., & Mann, C. C. (2016). Teaching multiply controlled
intraverbals to children and adolescents with autism spectrum disorders. Journal of Applied
Behavior Analysis, 49(4), 826–847. doi:10.1002/jaba.344
Kitchener, R. F. (1977). Behavior and behaviorism. Behaviorism, 5, 11–71.
Kleinert, K. (2018). A comparison of bidirectional naming for familiar and non-familiar stimuli
and the effects of a repeated probe procedure for first grade students (Order No. 10784944).
Available from ProQuest Dissertations & Theses Global. (2031071957). Retrieved from
https://tc.idm.oclc.org/login?url=https://search-proquest-
com.tc.idm.oclc.org/docview/2031071957?accountid=14258
Klintwall, L., & Eikeseth, S. (2014). Early and intensive behavioral intervention (EIBI) in autism.
In Patel, V. B., Preedy, V., R., & Mortin, C. R. (Eds.), Comprehensive Guide to Autism (pp.
 117–137). New York, NY: Springer.
Kodak, T., Campbell, V., Bergmann, S., LeBlanc, B., Kurtz-Nelson, E., Cariveau, T., . . . Mahon,
J. (2016). Examination of efficacious, efficient, and socially valid error-correction
procedures to teach sight words and prepositions to children with autism spectrum disorder.
Journal of Applied Behavior Analysis, 49(3), 532–547. doi:10.1002/jaba.310
Kodak, T., Clements, A., & Ninness, C. (2009). Acquisition of mands and tacts with concurrent
echoic training. Journal of Applied Behavior Analysis, 42(4), 839–843.
doi:10.1901/jaba.2009.42-839
Kodak, T., Fuchtman, R., & Paden, A. R. (2012). A comparison of intraverbal training procedures
for children with autism. Journal of Applied Behavior Analysis, 45(1), 155–160.
doi:10.1901/jaba.2012.45-155
Kohlenberg, R. J., Bolling, M. Y., Kanter, J. W., & Parker, C. R. (2002). Clinical behavior
analysis: Where it went wrong, how it was made good again, and why its future is so bright.
The Behavior Analyst Today, 3(3), 248–253.
Kohlenberg, R. J., & Tsai, M. (1991). Functional analytic psychotherapy. Creating intense and
curative therapeutic relationships. New York, NY: Springer.
Kohlenberg, R., & Tsai, M. (1995). Functional analytic psychotherapy: A behavioral approach to
intensive treatment. Washington, DC: American Psychological Association.
Kohlenberg, R. J., Tsai, M., & Dougher, M. J. (1993). The dimensions of clinical behavior
analysis. The Behavior Analyst, 16(2), 271–282.
Kooistra, E. T., Buchmeier, A. L., & Klatt, K. P. (2012). The effect of motivating operations on
the transfer from tacts to mands for children diagnosed with autism. Research in Autism
Spectrum Disorders, 6(1), 109–114. doi:10.1016/j.rasd.2011.03.010
Kraemer, K. M., Luberto, C. M., O’Bryan, E. M., Mysinger, E., & Cotton, S. (2016). Mind-body
skills training to improve distress tolerance in Medical students: A pilot study. Teaching
Learning Medicine, 28(2), 219–28.
Krantz, P. J., Land, S. E., & McClannahan, L. E. (1989). Conversational skills for autistic
adolescents: An autistic peer as prompter. Behavioral Interventions, 4(3), 171–189.
doi:10.1002/bin.2360040303
Krantz, P. J., & McClannahan, L. E. (1993). Teaching children with autism to initiate to peers:
Effects of a script-fading procedure. Journal of Applied Behavior Analysis, 26(1), 121–132.
doi:10.1901/jaba.1993.26-121
Krasner, L. (1963). Reinforcement, verbal behavior and psychotherapy. The American Journal of
Orthopsychiatry, 33, 601–613.
Kubina, R. M., & Yurich, K. K. L. (2012). The precision teaching book. Lemont, PA: Greatness
Achieved.
Kuhl, P. (2003). Human speech and birdsong: Communication and the social brain. Proceedings
of the National Academy of Sciences, 100, 9645–9646.
http://www.pnas.org/content/100/17/9645.short
Kuhl, P. (2004). Early language acquisition: Cracking the speech code. Nature Reviews
Neuroscience, 5, 831–843. doi:10.1038/nrn1533
Kuo, Z. Y. (1967). The dynamics of behavior development: An epigenetic view. New York, NY:
Random House.
Lamarre, J., & Holland, J. G. (1985). The functional independence of mands and tacts. Journal of
the Experimental Analysis of Behavior, 43(1), 5–19. doi:10.1901/jeab.1985.43-5
Landa, R. K., Hansen, B., & Shillingsburg, A. (2017). Teaching mands for information using
‘when’ to children with autism. Journal of Applied Behavior Analysis, 50(3), 538–551.
doi:10.1002/jaba.387
Lane, S., & Critchfield, T. S. (1998). Classification of vowels and consonants by individuals with
moderate mental retardation: Development of arbitrary relations via match-to-sample
training with compound stimuli. Journal of Applied Behavior Analysis, 31, 21–41.
Lang, R., Rispoli, M., Sigafoos, J., Lancioni, G., Andrews, A., & Ortega, L. (2011). Effects of
language of instruction on response accuracy and challenging behavior in a child with
autism. Journal of Behavioral Education, 20, 252–259. doi:10.1007/s10864-011-9130-0
Lanzoni, S. (2015). A short history of empathy. The Atlantic. Retrieved from
https://www.theatlantic.com/health/archive/2015/10/a-short-history-of-empathy/409912
Laraway, S., Snycerski, S., Michael, J., & Poling, A. (2003). Motivating operations and terms to
describe them: Some further refinements. Journal of Applied Behavior Analysis, 36(3), 407–
414. doi:10.1901/jaba.2003.36-407
LaRue, R. H., Pepa, L., Delmolino, L., Sloman, K. N., Fiske, K., Hansford, A., . . . Weiss, M. J.
(2016). A brief assessment for selecting communication modalities for individuals with
autism spectrum disorders. Evidence-Based Communication Assessment and Intervention,
10(1), 32–43. doi:10.1080/17489539.2016.1204767
Lattal, K. A., Cançado, C. R., Cook, J. E., Kincaid, S. L., Nighbor, T. D., & Oliver, A. C. (2017).
On defining resurgence. Behavioural Processes, 141, 85–91.
doi:10.1016/j.beproc.2017.04.018
Lawson, T. R., & Walsh, D. (2007). The effects of observational training on the acquisition of
reinforcement for listening. The Journal of Early and Intensive Behavioral Intervention,
4(2), 430. doi:10.1037/h0100383
Layng, T. V. J., Twyman, J. S., & Stikeleather, G. (2004a). Engineering discovery learning: The
contingency adduction of some precursors of textual responding in a beginning reading
program. Analysis of Verbal Behavior, 20, 99–109.
Layng, T. V. J., Twyman, J. S., & Stikeleather, G. (2004b) Selected for success: How Headsprout
Reading Basics teaches beginning reading. In D. J. Moran and R. Malott (Eds.), Evidence
based educational methods (pp. 171–197). St. Louis, MO: Elsevier Science/Academic Press.
Leader, G., & Barnes-Holmes, D. (2001a). Establishing fraction-decimal equivalence using a
respondent-type training procedure. The Psychological Record, 51, 151–165.
Leader, G., & Barnes-Holmes, D. (2001b). Matching to sample and respondent-type training as
methods for producing equivalence relations: Isolating the critical variable. The
Psychological Record, 51, 429–444.
Leaf, J. B., Oppenheim-Leaf, M. L., Leaf, R., Courtemanche, A. B., Taubman, M., McEachin, J., .
. . Sherman, J. A. (2012). Observational effects on the preferences of children with autism.
Journal of Applied Behavior Analysis, 45, 473-483. doi:10.1901/jaba.2012.45-473
Leaf, J. B., Oppenheim-Leaf, M. L., Townley-Cochran, D., Leaf, J. A., Alcalay, A., Milne, C., . . .
McEachin, J. (2016). Changing preference from tangible to social activities through an
observation procedure. Journal of Applied Behavior Analysis, 49, 49–57.
doi:10.1002/jaba.276
LeBlanc, L. A., Coates, A. M., Daneshvar, S., Charlop-Christy, M. H., Morris, C., & Lancaster,
B. M. (2003). Using video modeling and reinforcement to teach perspective-taking skills to
children with autism. Journal of Applied Behavior Analysis, 36, 253–257.
LeBlanc, L. A., Miguel, C. F., Cummings, A. R., Goldsmith, T. R., & Carr, J. E. (2003). The
effects of three stimulus-equivalence testing conditions on emergent US geography relations
of children diagnosed with autism. Behavioral Interventions, 18, 279–289.
Lechago, S. A., Carr, J. E., Kisamore, A. N., & Grow, L. L. (2015). The effects of multiple
exemplar instruction on the relation between listener and intraverbal categorization
repertoires. The Analysis of Verbal Behavior, 31(1), 76–95. doi:10.1007/s40616-015-0027-1
Lechago, S. A., & Low, A. (2015). A review of the mand-for-information training research
literature. International Journal of Behavior Analysis & Autism Spectrum Disorders, 1(1),
35–54.
Lee, G. T., Chou, W. C., & Feng, H. (2017). Using intraverbal prompts to increase divergent
intraverbal responses by a child with autism. Behavioral Interventions, 32(4), 334–344.
doi:10.1002/bin.1496
Lee, R., & Sturmey, P. (2006). The effects of lag schedules and preferred materials on variable
responding in students with autism. Journal of Autism and Developmental Disorders, 36(3),
421–428. doi:10.1007/s10803-006-0080-7
Lee, R., & Sturmey, P. (2014). The effects of script-fading and a Lag-1 schedule on varied social
responding in children with autism. Research in Autism Spectrum Disorders, 8, 440–448.
doi:10.1016/j.rasd.2014.01.003
Leech, A., Barnes-Holmes, D., & McEnteggart, C. (2017). Spider fear and avoidance: A
preliminary study of the impact of two verbal rehearsal tasks on a behavior–behavior relation
and its implications for an experimental analysis of defusion. The Psychological Record,
67(3), 387–398. doi:10.1007/s40732-017-0230-7
Leon, A. L., & Rosales, R. (2017). Effects of bilingual instruction for a child with communication
disorder. Journal of Behavioral Edcuation, printed online first. doi:10.1007/s10864-017-
9272-9
Lerman, D. C., Parten, M., Addison, L. R., Vorndran, C. M., Volkert, V. M., & Kodak, T. (2005).
A methodology for assessing the functions of emerging speech in children with
developmental disabilities. Journal of Applied Behavior Analysis, 38(3), 303–316.
doi:10.1901/jaba.2005.106-04
Leslie, J. C., Tierney, K.J., Robinson, C.P., Keenan, M., Watt., A., & Barnes, D. (1993).
Differences between clinically anxious and non-anxious subjects in a stimulus equivalence
training task involving threat words. The Psychological Record, 43, 153–161.
Levin, M. E., Hayes, S. C., & Vilardaga, R. (2012). Acceptance and commitment therapy:
Applying an iterative translational research strategy in behavior analysis. APA Handbook of
Behavior Analysis, 2, 455–480.
Lewon, A. B. (2019). Referential behavior and interactions of young children with ASD and their
mothers in an EIBI program: A longitudinal study. Manuscript in preparation.
Lewon, M., & Hayes, L. J. (2014). Toward an analysis of emotions as products of motivating
operations. The Psychological Record, 64, 813–825. doi:10.1007/s40732-014-0046-7
Lin, E. L., & Murphy, G. L. (2001). Thematic relations in adults’ concepts. Journal of
Experimental Psychology, 130, 1, 3–28.
Linden, M., & Whimbey, A. (1990). Why Johnny can’t write. Hillsdale, NJ: Erlbaum.
Lindsley, O. R. (1960). Characteristics of the behavior of chronic psychotics as revealed by free-
operant conditioning methods. Diseases of the Nervous System, 21, monograph supplement,
66–78.
Lindsley, O. R. (1992). Precision teaching: Discoveries and effects. Journal of Applied Behavior
Analysis, 25, 51–57. doi:10.1901/jaba.1992.25-51
Linehan, M. M. (1987). Dialectical behavior therapy for borderline personality disorder: Theory
and method. Bulletin of the Menninger Clinic, 51(3), 261–276.
Lipkens, R., & Hayes, S. C. (2009). Producing and recognising analogical relations. Journal of
Experimental Analysis of Behaviour, 91, 105–126.
Lipkens, R., Hayes, S. C., & Hayes, L. J. (1993). Longitudinal study of the development of
derived relations in an infant. Journal of Experimental Child Psychology, 56, 201–239.
doi:10.1006/jecp.1993.1032
Lo, C. (2016). How the listener half of Naming leads to multiple stimulus control (Doctoral
dissertation, Order No. 10100022). Available from ProQuest Dissertations & Theses Global.
(1784011549). doi:10.7916/D8C82988
Lodhi, S., & Greer, R. D. (1989). The speaker as listener. Journal of the Experimental Analysis of
Behavior, 51, 353–559. doi:10.1901/jeab.1989.51-353
Longano, J. M., & Greer, R. D. (2014). Is the source of Naming multiple conditioned reinforcers
for observing responses? The Analysis of Verbal Behavior, 31, 96–117. doi:10.1007/s40616-
014-0022-y
López, J. C., & Luciano, C. (2017). An experimental analysis of defusion interactions based on
deictic and hierarchical framings and their impact on cognitive performance. The
Psychological Record, 67, 485–497.
Lorah, E. R., Parnell, A., Whitby, P. S., & Hantula, D. (2015). A systematic review of tablet
computers and portable media players as speech generating devices for individuals with
autism spectrum disorder. Journal of Autism and Developmental Disorders, 45(12), 3792–
3804. doi:10.1007/s10803-014-2314-4
Lorenz, K. (1965). Evolution and modification of behavior. London, England: Methuen.
Lotfizadeh, A. D., Edwards, T. L., Redner, R., & Poling, A. (2012). Motivating operations affect
stimulus control: A largely overlooked phenomenon in discrimination learning. The
Behavior Analyst, 35, 89-100. doi:10.1007/BF03392268
Lovett, S., Rehfeldt, R. A., Garcia, Y., & Dunning, J. (2011). Comparison of a stimulus
equivalence protocol and traditional lecture for teaching single-subject designs. Journal of
Applied Behavior Analysis, 44, 819–833.
Lowenkron, B. (1989). Instructional control of generalized relational matching to sample in
children. Journal of the Experimental Analysis of Behavior, 52, 293–309.
Lu, H., Chen, D., & Holyoak, K. J. (2012). Bayesian analogy with relational transformations.
Psychological Review, 119, 3, 617–648.
Lubinski, D., & Thompson, T. (1987). An animal model of the interpersonal communication of
interoceptive (private) states. Journal of the Experimental Analysis of Behavior, 48, 1–15.
doi:10.1901/jeab.1987.48-1
Luciano, C., Ruiz, F. J., Vizcaíno-Torres, R., Sánchez, V., Gutiérrez-Martínez, O., & López-
López, J. C. (2011). A relational frame analysis of defusion interactions in ACT. A
preliminary and quasi-experimental study with at-risk adolescents. International Journal of
Psychology and Psychological Therapy, 11, 165–182.
Luciano, C., Valdivia-Salas, S., Cabello, F., & Hernández, M. (2009). Developing self-directed
rules. In R. A. Rehfeldt & Y. Barnes-Holmes (Eds.), Derived relational responding:
Applications for learners with autism and other developmental disabilities (pp. 335–352).
Oakland, CA: New Harbinger.
Luciano, C., Valdivia, S., & Ruiz, F. J. (2012). The self as the context for rule-governed behavior.
In L. McHugh & I. Stewart (Eds.), The self and perspective taking: Research and
applications (pp. 143–160). Oakland, CA: Context Press.
Luciano, M. C., Herruzo, J., & Barnes-Holmes, D. (2001). Generalization of say-do
correspondence. The Psychological Record, 51, 111–130. doi:10.1007/bf03395389
Luo, Q., Perry, C., Peng, D., Jin, Z., Xu, D., Ding, G., & Xu, S. (2003). The neural substrate of
analogical reasoning: An fMRI study. Cognitive Brain Research, 17, 527–534.
Lynch, D. C., & Cuvo, A. J. (1995). Stimulus equivalence instruction of fraction-decimal
relations. Journal of Applied Behavior Analysis, 28, 115–126.
Lyons, C. A., & Williamson, P. N. (1988). Contributions of Kantor’s “Psychological Linguistics”
to understanding psychotic speech. Behavior Analysis, 23(3), 110–113.
MacDonald, J., & Ahearn, W. H. (2015). Teaching observational learning to children with autism.
Journal of Applied Behavior Analysis, 48, 800–816. doi:10.1002/jaba.257
MacDonald, R. P. F., Dixon, L. S., & LeBlanc, J. M. (1986). Stimulus class formation following
observational learning. Analysis and Intervention in Developmental Disabilities, 6, 73– 87.
Mackintosh, N. J. (1976). Overshadowing and stimulus intensity. Animal Learning & Behavior, 4,
186–192. doi
Mackintosh, N. J. (1977). Stimulus control: Attentional factors. In W. K. Honig & J. E. R.
Staddon (Eds.), Handbook of operant behavior (pp. 481–513). Englewood Cliffs, NJ:
Prentice-Hall.
Mahoney, M. J. (1974). Cognition and behavior modification. Cambridge, MA: Ballinger.
Malcolm, N. (1963). Knowledge and certainty. Englewood Cliffs, NJ: Prentice-Hall.
Maloney, E., & Barnes-Holmes, D. (2016). Exploring the behavioral dynamics of the Implicit
Relational Assessment Procedure: The role of relational contextual cues versus relational
coherence indicators as response options. The Psychological Record, 66(3), 395–403.
doi:10.1007/s40732-016-0180-5
Maloney, M., Desjardins, A., & Broad, P. (1990). Teach your children well. Journal of Precision
Teaching, 7(2), 36–58.
Marchand-Martella, N. E., Slocum, T. A., & Martella, R. C. (Eds.). (2004). Introduction to direct
instruction. Boston, MA: Pearson.
Marchese, N. V., Carr, J. E., LeBlanc, L. A., Rosati, T. C., & Conroy, S. A. (2012). The effects of
the question “What is it?” on tact training outcomes of children with autism. Journal of
Applied Behavior Analysis, 45(3), 539–547. doi:10.1901/jaba.2012.45-539
Marcus, A., & Wilder, D. A. (2009). A comparison of peer video modeling and self-video
modeling to teach textual responses in children with autism. Journal of Applied Behavior
Analysis, 42(2), 335–341. doi:10.1901/jaba.2009.42-335
Markman, E. M. (1989). Categorization and naming in children: Problems of induction.
Cambridge, MA: The MIT Press.
Markman, E. M., & Seibert, J. (1976). Classes and collections: Internal organization and resulting
holistic properties. Cognitive Psychology, 8, 516–577.
Marler, P. (1970a). Birdsong and speech development: Could there be parallels? American
Scientist, 58, 669–673. Retrieved from http://www.jstor.org/stable/27829317
Marler, P. (1970b). A comparative approach to vocal learning: Song development in white-
crowned sparrows. Journal of Comparative and Physiological Psychology, 71, 1-25.
Retrieved from http://psycnet.apa.org/doi/10.1037/h0029144
Marschark, M. (1993). Psychological development of deaf children. New York, NY: Oxford
University Press.
Marschark, M., Mourandian, V., & Halas, M. (1994). Discourse rules in the language productions
of deaf and hearing children. Journal of Experimental Child Psychology, 57, 89–107.
Masia, C. L., & Chase, P. N. (1997). Vicarious learning revisited: A contemporary behavior
analytic interpretation. Journal of Behavior Therapy & Experimental Psychiatry, 28, 41–51.
Mathews, R. M., & Dix, M. (1992). Behavior change in the funny papers: Feedback to cartoonists
on safety belt use. Journal of Applied Behavior Analysis, 25, 769–775.
doi:10.1901/jaba.1992.25-769
Matter, A. L., & Zarcone, J. R. (2017). A comparison of existing and novel communication
responses used during functional communication training. Behavioral Interventions, 32(3),
217–224. doi:10.1002/bin.1481
Matthews, B. A., Catania, A. C., & Shimoff, E. (1985). Effects of uninstructed verbal behavior on
nonverbal responding: Contingency descriptions versus performance descriptions. Journal of
the Experimental Analysis of Behavior, 43, 155–164. doi:10.1901/jeab.1985.43-155
May, R. J., Hawkins, E., & Dymond, S. (2013). Brief report: Effects of tact training on emergent
intraverbal vocal responses in adolescents with autism. Journal of Autism and
Developmental Disorders, 43(4), 996–1004. doi:10.1007/s10803-012-1632-7
McAuliffe, D., Hughes, S., & Barnes-Holmes, D. (2014). The dark-side of rule governed
behavior: An experimental analysis of problematic rule-following in an adolescent
population with depressive symptomatology. Behavior Modification, 38, 587–613.
doi:10.1177/0145445514521630
McClay, L., Schafer, M. C. M., van der Meer, L., Couper, L., McKenzie, E., O’Reilly, M. F., . . .
Sutherland, D. (2016). Acquisition, preference, and follow-up comparison across two
modalities taught to two children with autism spectrum disorder. International Journal of
Disability, Development and Education, 64(2), 117–130.
doi:10.1080/1034912X.2016.1188892
McDowell, E. E. (1968). A programed method of reading instruction for use with kindergarten
children. The Psychological Record, 18, 233–239. doi:10.1007/BF03393766
McHugh, L., Barnes-Holmes, Y., & Barnes-Holmes, D. (2004). A relational frame account of the
development of complex cognitive phenomena: Perspective-taking, false belief
understanding, and deception. International Journal of Psychology and Psychological
Therapy, 4(2), 303–324.
McHugh, L., Barnes-Holmes, Y., Barnes-Holmes, D., Stewart, I., & Dymond, S. (2006). False
belief as generalized operant behavior. The Psychological Record, 56, 341–364.
McHugh, L., Stewart, I., & Hooper, N. (2012). A contemporary functional analytic account of
perspective taking. In L. McHugh & I. Stewart (Eds.), The self and perspective taking (pp.
55–72). Oakland, CA: Context Press.
McLay, L. K., Sutherland, D., Church, J., & Tyler-Merrick, G. (2013). The formation of
equivalence classes in individuals with autism spectrum disorder: A review of literature.
Research in Autism Spectrum Disorders, 7, 418–431.
McMorrow, M. J., Foxx, R. M., Faw, G. D., & Bittle, R. G. (1987). Cues-pause-point language
training: Teaching echolalics functional use of their verbal labeling repertoires. Journal of
Applied Behavior Analysis, 20(1), 11–22. doi:10.1901/jaba.1987.20-11
McSweeney, F. K., & Murphy, E. S. (2009). Sensitization and habituation regulate reinforcer
efficacy. Neurobiology of Learning and Memory, 92, 189–198.
doi:10.1016/j.nlm.2008.07.002
Meichenbaum, D. (1977). Cognitive-behaviour modification: An integrative approach. New York,
NY: Plenum.
Mellor, J. R., Barnes, C. S., & Rehfeldt, R. A. (2015). The effects of auditory tact and auditory
imagining instructions on the emergence of novel intraverbals. The Analysis of Verbal
Behavior, 31(2), 236–254. doi:10.1007/s40616-015-0036-0
Mercer, C. D., Mercer, A. R., & Evans, S. (1982). The use of frequency in establishing
instructional aims. Journal of Precision Teaching, 3, 57–59.
Merwin, I. M., & Wilson, K. G. (2005). Preliminary findings on the effects of self-referring and
evaluative stimuli on stimulus equivalence class formation. The Psychological Record, 55,
561–575.
Metzner, R. (1963). Learning theory and the therapy of neurosis. Cambridge University Press.
Michael, J. (1982). Distinguishing between discriminative and motivational functions of stimuli.
Journal of Experimental Analysis of Behavior, 37, 149–155. doi:10.1901/jeab.1982.37-149
Michael, J. (1985). Two kinds of verbal behavior plus a possible third. The Analysis of Verbal
Behavior, 3(1), 1–4. doi:10.1007/BF03392802
Michael, J. (1988). Establising operations and the mand. The Analysis of Verbal Behavior, 6(1),
3–9. doi:10.1007/BF03392824
Michael, J. (1993). Establishing operations. The Behavior Analyst, 16, 191–206.
doi:10.1007/BF03392623
Michael, J., Palmer, D. C., & Sundberg, M. L. (2011). The multiple control of verbal behavior.
The Analysis of Verbal Behavior, 27(1), 3–22. doi:10.1007/bf03393089
Miguel, C. (2016). Common and intraverbal bidirectional Naming. The Analysis of Verbal
Behavior, 32, 125–138. doi:10.1007/s40616-016-0066-2
Miguel, C. F., Yang, H. G., Finn, H. E., & Ahearn, W. H. (2009). Establishing derived textual
control in activity schedules with children with autism. Journal of Applied Behavior
Analysis, 42(3), 703–709. doi:10.1002/bin.1365
Ming, S., Mulhern, T., Stewart, I., Moran, L., & Bynum, K. (2018). Testing and training class
inclusion in typically developing young children and individuals with autism. Journal of
Applied Behavior Analysis, 51, 1, 53–60.
Mizael, T. M., de Almeida, J. H., Silveira, C. C., & de Rose, J. C. (2016). Changing racial bias by
transfer of functions in equivalence classes. The Psychological Record, 66(3), 451–462.
doi:10.1007/s40732-016-0185-0
Monclus, B., Uribio, G., & Szabo, T. G. (2016). Deriving rules from context: An RFT approach to
teaching problem-solving skills to children with autism. Paper presented at the Annual
Convention of the Association for Behavior Analysis International, Chicago, IL.
Montoya, M. M., Molina-Cobos, F.J., & McHugh, L. (2017). A review of relational frame theory
research into deictic relational responding. The Psychological Record, 67, 569–579.
Moore, J. (1980). On behaviorism and private events. The Psychological Record, 30(4), 459–475.
Moore, J. (2009). Why the radical behaviorist conception of private events is interesting, relevant,
and important. Behavior and Philosophy, 37, 21–37.
Mori, A., & Cigala, A. (2016). Perspective taking: Training procedures in developmentally
typical preschoolers. Different intervention methods and their effectiveness. Educational
Psychology Review, 28, 267–294.
Morris, E. K. (1982). Some relationships between interbehavioral psychology and radical
behaviorism. Behaviorism, 10, 187–216.
Morris, E. K. (1984). Interbehavioral psychology and radical behaviorism: Some similarities and
differences. The Behavior Analyst, 7, 197–204. doi:10.1007/BF03391903
Mueller, M. M., Palkovic, C. M., & Maynard, C. S. (2007). Errorless learning: Review and
practical application for teaching children with pervasive developmental disorders.
Psychology in the Schools, 44(7), 691–700. doi:10.1002/pits.20258
Mulhern, T., Stewart, I., & Elwee, J. M. (2017). Investigating relational framing of categorization
in young children. The Psychological Record, 67(4), 519–536. doi:10.1007/s40732-017-
0255-y
Mulhern, T., Stewart, I., & McElwee, J. (2018). Facilitating relational framing of classification in
young children. Journal of Contextual Behavioral Science, 8, 55–68.
Murayama, T., & Tobayama, T. (1997). Preliminary study on stimulus equivalence in beluga
(Delphinapterus leucas). Japanese Journal of Animal Psychology, 47, 79–89.
doi:10.2502/janip.47.79
Murphy, C., & Barnes-Holmes, D. (2009a). Derived more–less relational mands in children
diagnosed with autism. Journal of Applied Behavior Analysis, 42(2), 253–268.
doi:10.1901/jaba.2009.42-253
Murphy, C., & Barnes-Holmes, D. (2009b). Establishing derived manding for specific amounts
with three children: An attempt at synthesizing Skinner’s Verbal Behavior with relational
frame theory. The Psychological Record, 59(1), 75–92.
Murphy, C., & Barnes-Holmes, D. (2010). Establishing five dervied mands in three adolescent
boys with autism. Journal of Applied Behavior Analysis, 43(3), 537–541.
doi:10.1901/jaba.2010.43-537
Murphy, C., Barnes-Holmes, D., & Barnes-Holmes, Y. (2005). Derived manding with seven
children diagnosed with autism: Synthesizing Skinner’s Verbal Behavior with relational
frame theory. Journal of Applied Behavior Analysis, 38(4), 445–462.
doi:10.1901/jaba.2005.97-04
Murphy, C., Hussey, T., Barnes-Holmes, D., & Kelly, M. (2015). The Implicit Relational
Assessment Procedure (IRAP) and attractiveness bias. Journal of Contextual Behavioral
Science, 4(4), 292–299. doi:10.1016/j.jcbs.2015.08.001
Murphy, C., MacCarthaigh, S., & Barnes-Holmes, D. (2014). Implicit Relational Assessment
Procedure and attractiveness bias: Directionality of bias and influence of gender of
participants. International Journal of Psychology and Psychological Therapy, 14(3), 333–
351.
Najdowski, A. C. (2017). Flexible and focused: Teaching executive function skills to individuals
with autism and attention disorders. New York, NY: Elsevier.
National Governors Association Center for Best Practices, & Council of Chief State School
Officers. (2010). Common Core State Standards. Retrieved from
http://www.corestandards.org
Ninness, C., Barnes-Holmes, D., Rumph, R., McCuller, G., Ford, Payne, R., . . . Elliot, M. P.
(2006). Transformations of mathematical and stimulus functions. Journal of Applied
Behavior Analysis, 39, 299–321.
Ninness, C., Dixon, M., Barnes-Holmes, D., Rehfeldt, R. A., Rumph, R., McCuller, G., &
McGinty, J. (2009). Constructing and deriving reciprocal trigonometric relations: A
functional analytic approach. Journal of Applied Behavior Analysis, 42, 191–208.
Nolan, J., Murphy, C., & Barnes-Holmes, D. (2013). Implicit Relational Assessment Procedure
and body-weight bias: Influence of gender of participants and targets. The Psychological
Record, 63, 467–488. doi:10.11133/j.tpr.2013.63.3.005
NoRo, F. (2005). Using stimulus equivalence procedures to teach receptive emotional labeling to
a child with autistic disorder. The Japanese Journal of Special Education, 42, 483–496.
Nuzzolo-Gomez, R., & Greer, R. D. (2004). Emergence of untaught mands or tacts of novel
adjective-object pairs as a fucntion of instructional history. The Analysis of Verbal Behavior,
20(1), 63–76. doi:10.1007/BF03392995
O’Connor, M., Farrell, L., Munnelly, A., & McHugh, L. (2017). Citation analysis of relational
frame theory: 2009–2016. Journal of Contextual Behavioral Science, 6, 152–158.
http://dx.doi.org/10.1016/j.jcbs.2017.04.009
O’Hare, F. (1973). Sentence combining: Improving student writing without formal grammar
instruction. Urbana, IL: National Council of Teachers of English.
O’Hora, D., Roche, B., Barnes-Holmes, D., & Smeets, P. M. (2002). Response latencies to
multiple derived stimulus relations: Testing two predictions of relational frame theory. The
Psychological Record, 52, 51–75.
O’Neill, J., & Rehfeldt, R. A. (2017). Computerized behavioral skills training with selection-
based instruction and lag reinforcement schedules for responses to interview questions.
Behavior Analysis: Research and Practice, 17(1), 42–54. doi:10.1037/bar0000043
O’Neill, J., Rehfeldt, R. A., Ninness, C., Munoz, B. E., & Mellor, J. (2015). Learning Skinner’s
verbal operants: Comparing an online stimulus equivalence procedure to an assigned
reading. Analysis of Verbal Behavior, 31, 255–266.
O’Neill, J., & Weil, T. M. (2014). Training deictic relational responding in people diagnosed with
schizophrenia. The Psychological Record, 64, 301–310.
O’Reilly, A., Roche, B., Ruiz, M., Tyndall, I., & Gavin, A. (2012). The Function Speed
Acquisition Test (FAST): A behavior analytic implicit test for assessing stimulus relations.
The Psychological Record, 62, 507–528. doi:10.1007/BF03395817
O’Reilly, M. F. (1997). Functional analysis of episodic self-injury correlated with recurrent otitis
media. Journal of Applied Behavior Analysis, 30, 165–167. doi:10.1901/jaba.1997.30-165
O’Reilly, M., Aguilar, J., Fragale, C., Lang, R., Edrisinha, C., Sigafoos, J., . . . Didden, R. (2012).
Effects of a motivating operation manipulation on the maintenance of mands. Journal of
Applied Behavior Analysis, 45(2), 443–447.
O’Reilly, M. F., Lacey, C., & Lancioni, G. E. (2000). Assessment of the influence of background
noise on escape-maintained problem behavior and pain behavior in a child with Williams
Syndrome. Journal of Applied Behavior Analysis, 33, 511–514. doi:10.1901/jaba.2000.33-
511
Observer. (1970). Wanted: A better direction for linguistic psychology. The Psychological
Record, 20, 263–265.
Observer. (1971a). Comments and queries: Words, words, words. The Psychological Record, 21,
269–271.
Observer. (1971b). Comments and queries: Toward an improved linguistic model for science. The
Psychological Record, 21, 429–434.
Observer. (1976). Comments and queries: What meaning means in linguistics. The Psychological
Record, 26, 441–445.
Observer. (1983a). Comments and queries: Meanings as events and as constructions in
psychology and linguistics. The Psychological Record, 33, 433–440.
Observer. (1983b). Comments and queries: The role of tabula metaphors in psychology and in
language study. The Psychological Record, 33, 279–284.
Oden, D. L., Thompson, R. K. R., & Premack, D. (1988). Spontaneous transfer of matching by
infant chimpanzees. Journal of Experimental Psychology: Animal Behavior Processes, 14,
140–145.
Ogawa, A., Yamazaki, Y., Ueno, K., Cheng, K., & Iriki, A. (2010). Neural correlates of species-
typical illogical cognitive bias in human inference. Journal of Cognitive Neuroscience, 22,
2120–2130. doi:10.1162/jocn.2009.21330
Oppenheim, R. W. (1980). Metamorphosis and adaptation in the behavior of developing
organisms. Developmental Psychobiology, 13(4), 353–356. doi:10.1002/dev.420130402
Overton, D. A. (1984). State dependent learning and drug discriminations. In L. L. Iversen et al.
(Eds.), Drugs, neurotransmitters, and behavior (pp. 59–127). New York, NY: Springer.
Pack, A. A., Herman, L.M., & Roitblat, H. L. (1991). Generalization of visual matching and
delayed matching by a California sea lion (Zalophus califomianus). Animal Learning and
Behavior, 19, 37–48.
Padilla Dalmau, Y. C., Wacker, D., Harding, J., Berg, W., Schieltz, K., Lee, J., . . . Kramer, A. R.
(2011). A preliminary evaluation of functional communication training effectiveness and
language preference when Spanish and English are manipulated. Journal of Behavioral
Education,20(4), 233–251. doi:10.1007/s10864-011-9131-z.
Palmer, D. C. (1991). A behavioral interpretation of memory. In L. J. Hayes & P. N. Chase (Eds.),
Dialogues on verbal behavior (pp. 261–279). Reno, NV: Context Press.
Palmer, D. C. (2009). The role of private events in the interpretation of complex behavior.
Behavior and Philosophy, 37, 3–19.
Palmer, D. C. (2011). Consideration of private events is required in a comprehensive science of
behavior. The Behavior Analyst, 34(2), 201–207.
Palmer, D. C. (2012). The role of atomic repertoires in complex behavior. The Behavior Analyst,
35, 59–73.
Palmer, D. C. (2016). On intraverbal control and the definition of the intraverbal. The Analysis of
Verbal Behavior, 32(2), 96–106. doi:10.1007/s40616-016-0061-7
Paniagua, F.A. (1990). A procedural analysis of correspondence training techniques. The
Behavior Analyst, 13, 107–119.
Paniagua, F. A., & Baer, D. M. (1982). The analysis of correspondence training as a chain
reinforceable at any point. Child Development, 53, 786–798. doi:10.2307/1129393
Parker, L. H., Cataldo, M. F., Bourland, G., Emurian, C. S., Corbin, R. J., & Page, J. M. (1984).
Operant treatment of orofacial dysfunction in neuromuscular disorders. Journal of Applied
Behavior Analysis, 17, 413–427. doi:10.1901/jaba.1984.17-413
Parrott, L. J. (1983). Systemic foundations for the concept of “private events.” In N. W. Smith, P.
T. Mountjoy, & D. H. Ruben (Eds.), Reassessment in psychology: The Interbehavioral
alternative (pp. 251–268). Lanham, MD: University Press of America.
Parrott, L. J. (1984a). J. R. Kantor’s contributions to psychology and philosophy: A guide to
further study. The Behaviour Analyst, 7, 169–181.
Parrott, L. J. (1984b). Listening and understanding. The Behavior Analyst, 7, 29.
doi:10.1007/bf03391883
Parrott, L. J. (1986). On the role of postulation in the analysis of inapparent events. In H. W.
Reese & L. J. Parrott (Eds.), Behavior science: Philosophical, methodological, and
empirical advances (pp. 35–60). Hillsdale, NJ: Erlbaum.
Parrott, L. J. (1987). Rule-governed behavior: An implicit analysis of reference. In S. Modgil &
C. Modgil (Eds.), B. F. Skinner: Consensus and controversy (pp. 265–276). New York, NY:
Falmer Press.
Partington, J. W. (2008). The assessment of basics language and learning skills-revised: Scoring
instructions and IEP development guide (2nd ed). Pleasant Hill, CA: Behavior Analysts.
Partington, J. W., Sundberg, M. L., Newhouse, L., & Spengler, S. M. (1994). Overcoming an
autistic child’s failure to acquire a tact repertoire. Journal of Applied Behavior Analysis,
27(4), 733–734. doi:10.1901/jaba.1994.27-733
Pavlov, I. P. (1927). Conditioned reflexes: an investigation of the physiological activity of the
cerebral cortex. Oxford, England: Oxford Univ. Press.
Pedersen, R. (2009). Empirical research on empathy in medicine—A critical review. Patient
Education and Counseling, 76(3), 307–322.
Peláez, M. (2013). Dimensions of rules and their correspondence to rule-governed behavior.
European Journal of Behavior Analysis, 14, 259–270.
Pennington, R. C. (2016). Write on! Using assistive technology and systematic instruction to
teach sentence writing to students with moderate to severe disability. Journal of Special
Education Technology, 31(1), 50–57. doi:10.1177/0162643416633336
Pennington, R. C., & Delano, M. E. (2012). Writing instruction for students with autism spectrum
disorders: A review of literature. Focus on Autism and Other Developmental Disabilities,
27(3), 158–167. doi:10.1177/1088357612451318
Pennington, R. C., Foreman, L. H., & Gurney, B. N. (2018). An evaluation of procedures for
teaching students with moderate to severe disabilities to write sentences. Remedial and
Special Education, 39(1), 27–38. doi:10.1177/0741932517708428
Pereira Delgado, J. A., & Greer, R. D. (2009). The effects of peer monitoring training on the
emergence of the capability to learn from observing instruction received by peers. The
Psychological Record, 59, 407–434.
Pérez-González, L. A., & Alonso-Álvarez, B. (2008). Common control by compound samples in
conditional discriminations. Journal of the Experimental Analysis of Behavior, 90(1), 81–
101. doi:10.1901/jeab.2008.90-81
Pérez-González, L. A., & Carnerero, J. J. (2014). Emerging tacts and selections from previous
learned skills: A comparison between two types of naming. The Analysis of Verbal Behavior,
30, 184–192. doi:10.1007/s40616-014-0011-1
Pérez-González, L. A., Garcia-Asenjo, L., Williams, G., & Carnerero, J. J. (2007). Emergence of
intraverbal antonyms in children with pervasive development disorder. Journal of Applied
Behavior Analysis, 40, 697–701. doi:10.1901/jaba.2007.697-701
Perkins, D. D., & Zimmerman, M. A. (1995). Empowerment theory, research, and application.
American Journal of Community Psychology, 23, 569–579.
Perkins, F. T. (1931). A further study of configurational learning in the goldfish. Journal of
Experimental Psychology, 14, 508–538.
Persicke, A., Tarbox, J., Ranick, J., & St. Clair, M. (2012). Establishing metaphorical reasoning in
children with autism. Research in Autism Spectrum Disorders, 6, 913–920.
doi:10.1016/j.rasd.2011.12.007
Peters, L. C., & Thompson, R. H. (2015). Teaching children with autism to respond to
conversational partners’ interests. Journal of Applied Behavior Analysis, 48, 544–562.
Petursdottir, A. I., Carp, C. L., Matthies, D. W., & Esch, B. E. (2011). Analyzing stimulus-
stimulus pairing effects on preferences for speech sounds. The Analysis of Verbal Behavior,
27(1), 45–60. doi:10.1007/BF03393091
Petursdottir, A., & Devine, B. (2017). The impact of verbal behavior on the scholarly literature
from 2005 to 2016. The Analysis of Verbal Behavior, 33(2), 212–228. doi:10.1007/s40616-
017-0089-3
Petursdottir, A., Carr, J. E., Lechago, S. A., & Almason, S. M. (2008). An evaluation of
intraverbal training for teaching categorization skills. Journal of Applied Behavior Analysis,
41(1), 53–68. doi:10.1901/jaba.2008.41-53
Phillips, J. S., Matarazzo, R. G., Matarazzo, J. D., Saslow, G., & Kanfer, F. H. (1961).
Relationships between descriptive content and interaction behavior in interviews. Journal of
Consulting Psychology, 25, 260–266.
Piaget, J. (1970). Piaget’s theory. In P. H. Mussen (Ed.), Carmichael’s manual of child psychology
(pp. 703–732). New York, NY: Wiley.
Piaget, J., Montangero, J., & Billeter, J. (1977). Les correlates [Correlations]. In J. Piaget (Ed.),
L’abstraction refléchissante (pp.115–129). Paris: Presses Universitaires de France.
Pierce, W. D., & Cheney, C. D. (2013). Behavior analysis and learning (4th ed.). New York, NY:
Psychology Press.
Pistoljevic, N., & Greer, R. D. (2006). The effects of daily intensive tact instruction on preschool
students’ emission of pure tacts and mands in non-instructional setting. Journal of Early and
Intensive Behavioral Interventions, 3, 103–120. Retrieved from
http://psycnet.apa.org/doi/10.1037/h0100325
Pituch, K. A., Green, V. A., Didden, R., Lang, R., O’Reiley, M. F., Lancioni, G. E., & Sigafoos, J.
(2011). Parent reported treatment priorities for children with autism spectrum disorders.
Research in Autism Spectrum Disorders, 5, 135–143.
Pohl, P. (1983). Central auditory processing. V. Ear advantages for acoustic stimuli in baboons.
Brain and Language, 20, 44–53. doi:10.1016/0093-934X(83)90031-7
Pohl, P. (1984). Ear advantages for temporal resolution in baboons. Brain and Cognition, 3, 438–
444. doi:10.1016/0278-2626(84)90033-2
Pohl, P., Greer, R. D., Du, L., & Moschella, J. L. (2018). Verbal development, behavioral
metamorphosis, and the evolution of language. Perspectives on Behavior Science, 1–18.
doi:10.1007/s40614-018-00180-0
Pohl, P., Grubmüller, H. G., & Grubmüller, R. (1984). Developmental changes in dichotic right
ear advantage (REA). Neuropediatrics, 15, 139–144. doi:10.1055/s-2008-1052357
Polya, G. (1954). Mathematics and Plausible Reasoning: Volume 1. Induction and Analogy in
Mathematics. Princeton, NJ: Princeton University Press.
Power, P. M., Harte, C., Barnes-Holmes, D., Barnes-Holmes, Y. (2017a). Exploring racial bias in
a country with a recent history of immigration of black Africans. The Psychological Record,
67(3), 365–375 doi.org/10.1007/s40732-017-0223-6
Power, P. M., Harte, C., Barnes-Holmes, D., Barnes-Holmes, Y. (2017b). Combining the Implicit
Relational Assessment Procedure and the recording of event related potentials in the analysis
of racial bias: A preliminary study. The Psychological Record, 67(4), 499–506.
doi:10.1007/s40732-017-0252-1
Power, P., Barnes-Holmes, D., Barnes-Holmes, Y., & Stewart, I. (2009). The Implicit Relational
Assessment Procedure (IRAP) as a measure of implicit relative preferences: A first study.
The Psychological Record, 59, 621–640. doi:10.1007/BF03395684
Powers, M. B., Vörding, M. B. Z. V. S., & Emmelkamp, P. M. (2009). Acceptance and
commitment therapy: A meta-analytic review. Psychotherapy and Psychosomatics, 78(2),
73–80.
Premack, D. (1959). Toward empirical behavior laws: I. Positive reinforcement. Psychological
Review, 66(4), 219–233. doi:10.1037/ h0040891
Premack, D. G., & Woodruff, G. (1978). Does the chimpanzee have a “theory of mind”?
Behavioral and Brain Sciences, 1, 515–526.
Proffitt, J. B., Coley, J. D., & Medin, D. L. (2000). Expertise and category-based induction.
Journal of Experimental Psychology: Learning, Memory & Cognition, 26, 4, 811–828.
doi:10.10371/0278-7393.26.4.811
Pytte, C. L., & Fienup, D. M. (2012). Using equivalence-based instruction to increase efficiency
in teaching neuroanatomy. The Journal of Undergraduate Neuroscience Education, 10,
A125–A131.
Rabelo, L. Z., Bortoloti, R., & Souza, D. H. (2014). Dolls are for girls and not for boys:
Evaluating the appropriateness of the Implicit Relational Assessment Procedure for school-
age children. The Psychological Record, 64, 71–77. doi:10.1007/s40732-014-0006-2
Rachlin, H. (1984). Mental, yes. Private, no. Behavioral and Brain Sciences, 7(4), 566–567.
Rachlin, H. (2011). Baum’s private thoughts. The Behavior Analyst, 34(2), 209–212.
Rachman, S. (1967). Systematic desensitization. Psychological Bulletin, 67(2), 93–103.
doi:10.1037/h0024212
Ramirez, J., & Rehfeldt, R. A. (2009). Observational learning and the emergence of symmetry
relations in teaching Spanish vocabulary words to typically developing children. Journal of
Applied Behavior Analysis, 42, 801–805. doi:10.1901/jaba.2009.42-801.
Reese, H. W. (1994). Cognitive and behavioral approaches to problem solving. In S. C. Hayes, L.
J. Hayes, M. Sato, & K. Ono (Eds.), Behavior analysis of language and cognition (197–
258). Reno, NV: Context Press.
Rehfeldt, R. A. (2011). Toward a technology of derived stimulus relations: An analysis of articles
published in the Journal of Applied Behavior Analysis, 1992–2009. Journal of Applied
Behavior Analysis, 44, 109–119. doi:10.1901/jaba.2011.44-109
Rehfeldt, R. A., & Barnes-Holmes, Y. (Eds.). (2009). Derived relational responding: Applications
for learners with autism and other developmental disabilities: A progressive guide to
change. Oakland, CA: New Harbinger Publications.
Rehfeldt, R. A., Dillen, J. E., Ziomek, M. M., & Kowalchuk, R. K. (2007). Assessing relational
learning deficits in perspective-taking in children with high-functioning autism spectrum
disorder. The Psychological Record, 57, 23.
Rehfeldt, R. A., & Hayes, L. J. (1998). The operant-respondent distinction revisited: Toward an
understanding of stimulus equivalence. The Psychological Record, 48, 187–210.
doi:10.1007/BF03395266
Rehfeldt, R. A., Latimore, D., & Stromer, R. (2003). Observational learning and the formation of
classes of reading skills by individuals with autism and other developmental disabilities.
Research in Developmental Disabilities, 24, 333–358. doi:10.1016/S0891-4222(03)00059-3
Rehfeldt, R. A., & Root, S. L. (2005). Establishing derived requesting skills in adults with severe
developmental disabilities. Journal of Applied Behavior Analysis, 38(1), 101–105.
doi:10.1901/jaba.2005.106-03
Rescorla, R. A. (1967). Pavlovian conditioning and its proper control procedures. Psychological
Review, 74, 71–80. doi:10.1037/h0024109
Rescorla, R. A. (1988). Pavlovian conditioning: It’s not what you think it is. American
Psychologist, 43, 151–160. doi
Rescorla, R. A., & Wagner, A. R. (1972). A theory of Pavlovian conditioning: Variations in the
effectiveness of reinforcement and nonreinforcement. In A. H. Black & W. F. Prokasy (Eds.),
Classical conditioning II: Current research and theory (pp. 64–99). East Norwalk, CT:
Appleton-Century-Crofts.
Reyes-Giordano, K., & Fienup, D. M. (2015). Emergence of topographical responding following
equivalence-based neuroanatomy instruction. The Psychological Record, 65, 495–507.
Reynolds, G. S. (1961). Attention in the pigeon. Journal of the Experimental Analysis of
Behavior, 4, 203–208. doi:10.1901/jeab.1961.4-203
Ribeiro, D. M., Elias, N. C., Goyos, C., & Miguel, C. F. (2010). The effects of listener training on
the emergence of tact and mand signs by individuals with intellectual disabilities. The
Analysis of Verbal Behavior, 26(1), 65–72. doi:10.1007/BF03393084
Ribeiro, D. M., Miguel, C. F., & Goyos, C. (2015). The effects of listener training on
discriminative control by elements of compound stimuli in children with disabilities. Journal
of the Experimental Analysis of Behavior, 104(1), 48–62. doi:10.1002/jeab.161
Ribeiro, T. A., Gallano, T. P., Souza, D. de H., & de Souza, D. das G. (2017). Responding and
learning by exclusion in 2-year-olds: The case of adjectives. Psychological Record, 67, 293–
314.
Ribes, E. (1977). Relationship between behavior theory, experimental research and behavior
modification techniques. The Psychological Record, 2, 417–424.
Ribes, E. (1981). Reflexiones sobre el concepto de inteligencia y su desarrollo. Revista Mexicana
de Análisis de la Conducta, 7, 107–116.
Ribes, E. (1989). La inteligencia como comportamiento un análisis conceptual. Revista Mexicana
de Análisis de la Conducta, 15, 51–67.
Ribes, E. (1993). Behavior as the functional content of language. In S. C. Hayes, L. J. Hayes, T.
R. Sarbin, & H. W. Reese (Eds.), Varieties of contextualism (pp. 283–297). Reno, NV:
Context Press.
Ribes, E. (1997). Causality and contingency: Some conceptual considerations. The Psychological
Record, 47, 619–639.
Ribes, E. (1999). Teoría del condicionamiento y lenguaje. Un análisis histórico-Conceptual.
México: Taurus.
Ribes, E. (2001). Functional dimensions of social behavior: Theoretical considerations and some
preliminary data. Mexican Journal of Behavior Analysis, 27 (2), 285–306.
Ribes, E. (2003). Concepts and theories: Relation to scientific categories. In K. A. Lattal & P.
Chase (Eds.), Behavior theory and philosophy (pp. 147–164). New York, NY: Kluver-
Plenum.
Ribes, E. (2004a). Behavior is abstraction, not ostension: Conceptual and historical remarks on
the nature of psychology. Behavior & Philosophy, 32, 55–68.
Ribes, E. (2004b). Theory, scientific research, and technical applications: How related in operant
psychology? In J. E. Burgos & E. Ribes (Eds.), Theory, basic and applied research, and
technological applications in behavior science: Conceptual and methodological issues (pp.
19–44). Guadalajara, México: Universidad de Guadalajara.
Ribes, E. (2006a). Human Behavior as language: Some thoughts on Wittgenstein. Behavior &
Philosophy, 34, 109–121.
Ribes, E. (2006b). Which should be the contribution of psychology to education? In Q. Jing, M.
Rosenzweig, G. D’Ydewalle, H. Zhang, H. Chih, & K. Zhang (Eds), Progress in science
around the world (pp. 23–34). Hove, UK: Psychology Press.
Ribes, E. (2007). On two functional meanings of “knowing.” In J. E. Burgos & E. Ribes (Eds.),
Knowledge, cognition, and behavior (pp. 139–150). Guadalajara, México: Universidad de
Guadalajara.
Ribes, E. (2008). Educación básica, desarrollo psicológico y planeación de competencias. Revista
Mexicana de Psicología, 25, 193–207.
Ribes, E. (2011). El concepto de competencia: Su pertinencia en el desarrollo psicológico y la
educación. Bordón: Revista de Pedagogía, 63, 33–45.
Ribes, E. (2018). El estudio científico de la conducta individual: introducción a la teoríade la
psicología. Ciudad de México: El Manual Moderno.
Ribes, E., Pulido, L., Rangel, N., & Sánchez-Gatell, E. (2016). Sociopsicología: instituciones
yrelaciones interindividuales. Madrid: La Catarata.
Ricard, M. (2015). Altruism: The power of compassion to change yourself and the world. London,
UK: Atlantic Books Ltd.
Richards, M., Mossey, J., & Robins, D. L. (2016). Parents’ concerns as they relate to their child’s
development and later diagnosis of autism spectrum disorder. Journal of Developmental and
Behavioral Pediatrics, 37(7), 532–540.
Rilling, J. K., Gutman, D. A., Zeh, T. R., Pagnoni, G., Berns, G. S., & Kilts, C. D. (2002). A
neural basis for social cooperation. Neuron, 35(2), 395–405.
Rilling, M. (1977). Stimulus control and inhibitory processes. In W. K. Honig & J. E. R. Staddon
(Eds.), Handbook of operant behavior (pp. 432–480). Englewood Cliffs, NJ: Prentice-Hall.
Ringdahl, J. E., Falcomata, T. S., Christensen, T. J., Bass-Ringdahl, S. M., Lentz, A., Dutt, A., &
Schuh-Claus, J. (2009). Evaluation of a pre-treatment assessment to select mand
 topographies for functional communication training. Research in Developmental
Disabilities, 30(2), 330–341. doi:10.1016/j.ridd.2008.06.002
Risley, T. R., & Hart, B. (1968). Developing correspondence between the non-verbal and verbal
behavior of preschool children. Journal of Applied Behavior Analysis, 1, 267–281.
doi:10.1901/jaba.1968.1-267
Ritzert, T. R., Forsyth, J. P., Berghoff, C. R., Barnes-Holmes, D., & Nicholson, E. (2015). The
impact of a cognitive defusion intervention on behavioral and psychological flexibility: An
experimental evaluation in a spider fearful non-clinical sample. Journal of Contextual
Behavioral Science, 4(2), 112–120. doi:10.1016/j.jcbs.2015.04.001
Robbins, J. K. (2011). Problem solving and analytical thinking in a classroom environment. The
Behavior Analyst Today, 12, 40–47.
Robbins, J. K. (2014). Learn to Reason with TAPS: A Problem Solving Approach [Curriculum
program]. Available at http://www.talkaloudproblemsolving.com.
Robbins, J. K., Layng, T. V. J., & Jackson, P. J. (1995). Fluent thinking skills. Seattle, WA:
Robbins/Layng & Associates.
Robertson, D. L., & Peláez , M. (2016). Behavior analytic concepts and change in a large
metropolitan research university: The Graduation Success Initiative. Journal of
Organizational Behavior Management, 36, 123–153. doi:10.1080/01608061.2016.1200513
Robinson, G. E., & Barron, A. B. (2017). Epigenetics and the evolution of instincts: Instincts
evolve from learning and share the same cellular and molecular mechanisms. Science,
365(6333), 26–27. doi:10.1126/science.aam6142
Roche, B., & Barnes, D. (1997). A transformation of respondently conditioned stimulus function
in accordance with arbitrarily applicable relations. Journal of the Experimental Analysis of
Behavior, 67, 275–301.
Roche, B., Linehan, C., Ward, T. E., Dymond, S., & Rehfeldt, R. A. (2004). The unfolding of the
relational operant: A real-time analysis using electroencephalography and reaction time
measures. International Journal of Psychology and Psychological Therapy, 4, 587–603.
Roddy, S., Stewart, I., & Barnes-Holmes, D. (2010). Anti-fat, pro-slim, or both? Using two
reaction-time based measures to assess implicit attitudes to the slim and overweight. Journal
of Health Psychology, 15, 416–425. doi:10.1177/1359105309350232
Roddy, S., Stewart, I., & Barnes-Holmes, D. (2011). Facial reactions reveal that slim is good but
fat is not bad: Implicit and explicit measures of body-size bias. European Journal of Social
Psychology, 41, 688–694. doi:10.1002/ejsp.839
Rogers-Warren, A., & Baer, D. M. (1976). Correspondence between saying and doing: Teaching
children to share and praise. Journal of Applied Behavior Analysis, 9, 335–354.
doi:10.1901/jaba.1976.9-335
Romanes, G. J. (1888). Mental evolution in man: Origins of human faculty. London, England:
Keegan Paul.
Ronspies, J., Schmidt, A. F., Melnikova, A., Krumova, R., Zolfagari, A., & Banse, R. (2015).
Indirect measurement of sexual orientation: Comparison of the Implicit Relational
Assessment Procedure, viewing time, and choice reaction time. Archives of Sexual Behavior,
44(5), 1483–1492. doi:10.1007/s10508-014-0473-1
Rosales-Ruiz, J., & Baer, D. M. (1997). Behavioral cusps: A developmental and pragmatic
concept for behavior analysis. Journal of Applied Behavior Analysis, 30, 533–544.
doi:10.1901/jaba.1997.30-533
Rosales, R., & Rehfeldt, R. A. (2007). Contriving transitive conditioned establishing operations to
establish derived manding skills in adults with severe developmental disabilities. Journal of
Applied Behavior Analysis, 40(1), 105–121. doi:10.1901/jaba.2007.117-05
Rosales, R., Rehfeldt, R. A., & Huffman, N. (2012). Examining the utility of the stimulus pairing
observation procedure with preschool children learning a second language. Journal of
Applied Behavior Analysis, 45, 173–177. doi:10.1901/jaba.2012.45-173
Rosales, R., Rehfeldt, R. A., & Lovett, S. (2011). Effects of multiple exemplar training on the
emergence of derived relations in preschool children learning a second language. The
Analysis of Verbal Behavior, 27, 61–74.
Ross, S. M., & Proctor, S. (1973). Frequency and duration of hierarchy item exposure in a
systematic desensitization analogue. Behaviour Research and Therapy, 11(3), 303–312.
Ruiz, F. J. (2010). A review of acceptance and commitment therapy (ACT) empirical evidence:
Correlational, experimental psychopathology, component and outcome studies. International
Journal of Psychology and Psychological Therapy, 10, 125–162.
Ruiz, F. J. (2012). Acceptance and commitment therapy versus traditional cognitive behavioral
therapy: A systematic review and meta-analysis of current empirical evidence. International
Journal of Psychology and Psychological Therapy, 12(3), 333–357.
Ruiz, F. J., & Luciano, C. (2011). Cross-domain analogies as relating derived relations among two
separate relational networks. Journal of the Experimental Analysis of Behaviour, 95(3), 369–
385.
Ruiz, F. J., & Luciano, C. (2015). Common physical properties among relational networks
improve analogy aptness. Journal of the Experimental Analysis of Behaviour, 103, 3, 498–
510.
Ruiz, F. J., Riaóo, D., Suárez, J., & Luciano, C. (2016). Effect of a one-session ACT protocol in
disrupting repetitive negative thinking: A randomized multiple-baseline design.
International Journal of Psychology and Psychological Therapy, 16(3), 213–233.
Ryle, G. (1949). The concept of mind. New York, NY: Barnes & Noble.
Ryle, G. (1962). Dilemmas. Cambridge, England: Cambridge University Press.
Salkovskis, P. M., Wroe, A. L., Gledhill, A., Morrison, N., Forrester, E., Richards, C., . . . Thorpe,
S. (2000). Responsibility attitudes and interpretations are characteristic of obsessive
compulsive disorder. Behaviour Research and Therapy, 38(4), 347–372.
Sandoz, E. K., & Anderson, R. S. (2015). Building awareness, openness, and action: Values work
in behavior therapy. The Behavior Therapist, 38, 60–69.
Sandoz, E. K., & Hebert, E. R. (2016). Using derived relational responding to model statistics
learning across participants with varying degrees of statistics anxiety. European Journal of
Behavior Analysis, 1–19.
Saunders, R. R., Drake, K. M., & Spradlin, J. E. (1999). Equivalence class establishment,
expansion, and modification in preschool children. Journal of the Experimental Analysis of
Behavior, 71, 195–214.
Saunders, R. R., Wachter, J. A., & Spradlin, J. E. (1988). Establishing auditory stimulus control
over an eight-member equivalence class via conditional discrimination procedures. Journal
of the Experimental Analysis of Behavior, 49, 95–115.
Sautter, R. A., & LeBlanc, L. A. (2006). Empirical applications of Skinner’s analysis of Verbal
Behavior with humans. The Analysis of Verbal Behavior, 22(1), 35–48.
doi:10.1007/BF03393025
Sautter, R. A., LeBlanc, L. A., Jay, A. A., Goldsmith, T. R., & Carr, J. E. (2011). The role of
problem solving in complex intraverbal repertoires. Journal of Applied Behavior Analysis,
44(2), 227–244. doi:10.1901/jaba.2011.44-227
Schauffler, G., & Greer, R. D. (2006). The effects of intensive tact instruction on audience-
accurate tacts and conversational units. Journal of Early and Intensive Behavioral
Interventions, 3, 120–132. doi:10.1037/h0100326
Scheel, M. H., Fischer, L. A., McMahon, A. J., Wolf, J. E. (2011). The Implicit Relational
Assessment Procedure (IRAP) as a measure of women’s stereotype about gay men. Current
Research in Social Psychology, 18(2), 11–23.
Scheel, M. H., Roscoe, B. H., Scaewe, V. G., & Yarbrough, C. S. (2014). Attitudes towards
Muslims are more favorable on a survey than on an Implicit Relational Assessment
Procedure (IRAP). Current Research in Social Psychology, 22(3), 22–32.
Schlinger, H. D. (2008). Consciousness is nothing but a word. Skeptic, 13(4), 58–63.
Schlinger, H. D. (2017). Reflections on Verbal Behavior at 60. The Analysis of Verbal Behavior,
33(2), 179–190. doi:10.1007/s40616-017-0087-5
Schlund, M. W., Cataldo, M. F., & Hoehn-Saric, R. (2008). Neural correlates of derived relational
responding on tests of stimulus equivalence. Behavioral and Brain Functions, 4, 6.
Schlund, M. W., Hoehn-Saric, R., & Cataldo, M. F. (2007). New knowledge derived from learned
knowledge: Functional-anatomic correlates of stimulus equivalence. Journal of the
Experimental Analysis of Behavior, 87, 287–307.
Schmelzkopf, J., Greer, R. D., Singer-Dudek, J., & Du, L. (2017). Experiences that establish
preschoolers’ interest in speaking and listening to others. Behavioral Development Bulletin,
22, 44. doi:10.1037/bdb0000026
Schutte, R. C., & Hopkins, B. L. (1970). The effects of teacher attention on following instructions
in a kindergarten class. Journal of Applied Behavior Analysis, 3, 117–122.
doi:10.1901/jaba.1970.3-117
Scott, J. P., & Nagy, Z. M. (1980) Behavioral metamorphosis in mammalian development. In E.
C. Simmel (Ed.), Early experiences and early behavior: Implications for social development
(pp. 15–37). New York, NY: Academic Press.
Seaver, J. L., & Bourret, J. C. (2014). An evaluation of response prompts for teaching behavior
chains. Journal of Applied Behavior Analysis, 47(4), 777–792. doi:10.1002/jaba.159
Sella, A. C., Ribeiro, D. M., & White, G. W. (2014). Effects of an online stimulus equivalence
teaching procedure on research design open-ended questions performance of international
undergraduate students. The Psychological Record, 64, 89–103. doi:10.1007/s40732-014-
0007-1
Sellers, T. P., Kelley, K., Higbee, T. S., & Wolfe, K. (2016). Effects of simultaneous script
training on use of varied mand frames by preschoolers with autism. The Analysis of Verbal
Behavior, 32(1), 15–26. doi:10.1007/s40616-015-0049-8
Sevdalis, V., & Raab, M. (2014). Empathy in sports, exercise, and the performing arts.
Psychology of sport and Exercise, 15(2), 173–179.
Shafer, E. (1993). Teaching topography-based and selection-based verbal behavior to
developmentally disabled individuals: Some considerations. The Analysis of Verbal
Behavior, 11(1), 117–133. doi:10.1007/BF03392892
Shatz, M., & Gelman, R. (1973). The development of communication skills: Modifications in the
speech of young children as a function of listener. Monographs of the Society for Research in
Child Development, 30, 1–38.
Shevrin, H., & Dickman, S. (1980). The psychological unconscious: A necessary assumption for
all psychological theory? American Psychologist, 35(5), 421–434.
Shillingsburg, M. A., Frampton, S. E., Cleveland, S. A., & Cariveau, T. (2018). A clinical
application of procedures to promote the emergence of untrained intraverbal relations with
children with autism. Learning and Motivation, 62, 51–66. doi:10.1016/j.lmot.2017.02.003
Shillingsburg, M. A., Hollander, D. L., Yosick, R. N., Bowen, C., & Muskat, L. R. (2015).
Stimulus-stimulus pairing to increase vocalizations in children with language delays: A
review. Analysis of Verbal Behavior, 31(2), 215–235. doi:10.1007/s40616-015-0042-2
Shimoff, E., Catania, A. C., & Matthews, B. A. (1981). Uninstructed human responding:
Sensitivity of low-rate performance to schedule contingencies. Journal of the Experimental
Analysis of Behavior, 36, 207–220. doi:10.1901/jeab.1981.36-207
Shoenfeld, W. N. (1969). J. R. Kantor’s Objective Psychology of Grammar and Psychology and
Logic: A retrospective appreciation. Journal of the Experimental Analysis of Behavior, 12,
329–347.
Sidman, M. (1971). Reading and auditory-visual equivalences. Journal of Speech & Hearing
Research, 14, 5–13. doi:10.1044/jshr.1401.05
Sidman, M. (1994). Equivalence relations and behavior: A research story. Boston, MA: Authors
Cooperative.
Sidman, M., & Cresson, O. (1973). Reading and crossmodal transfer of stimulus equivalences in
severe retardation. American Journal of Mental Deficiency, 77, 515–523.
Sidman, M., & Tailby, W. (1982). Conditional discrimination vs. matching to sample: An
expansion of the testing paradigm. Journal of the Experimental Analysis of Behavior, 37, 5–
22. doi:10.1901/jeab.1982.37-5
Siegler, R. S. (1983). Information processing approaches to development. In P. H. Mussen (Ed.),
Handbook of child psychology: Volume I. History, theory, and methods (pp. 129–211). New
York, NY: Wiley.
Sigafoos, J., Kerr, M., Roberts, D., & Couzens, D. (1994). Increasing opportunities for requesting
in classrooms serving children with developmental disabilities. Journal of Autism and
Developmental Disorders, 24(5), 631–645. doi:10.1007/BF02172143
Silver, M., & Oakes, P. (2001). Evaluation of a new computer intervention to teach people with
autism or Asperger syndrome to recognize and predict emotions in others. Autism, 5(3), 299–
 316.
Simic, J., & Bucher, B. (1980). Development of spontaneous manding in language deficient
children. Journal of Applied Behavior Analysis, 13(3), 523–528. doi:10.1901/jaba.1980.13-
523
Sinclair, S., Hack, T. F., McClement, S., & Raffin-Bouchal, S. (2016). Peeling the onion: Patients’
perspectives and experiences of the similarities and differences of sympathy, empathy and
compassion. Journal of Pain and Symptom Management, 52, 147–148.
Singer-Dudek, J., & Oblak, M. (2013). Peer presence and the emergence of conditioned
reinforcement from observation. Journal of Applied Behavior Analysis, 46, 592–602.
doi:10.1002/jaba.72
Singer-Dudek, J., Oblak, M., & Greer, R. D. (2011). Establishing books as conditioned reinforcers
for preschool children as a function of an observational intervention. Journal of Applied
Behavior Analysis, 44, 421–434.
Skinner, B. F. (1938). The behavior of organisms: An experimental analysis. New York, NY:
Appleton-Century Crofts.
Skinner, B. F. (1945). The operational analysis of psychological terms. Psychological Review,
52(5), 270–277. doi:10.1037/h0062535
Skinner, B. F. (1953). Science and human behavior. New York, NY: The Free Press.
Skinner, B. F. (1956). A case history in scientific method. American Psychologist, 11(5), 221–
233. doi:10.1037/h0047662
Skinner, B. F. (1957). Verbal behavior. New York, NY: Appleton-Century-Crofts.
Skinner, B. F. (1963). Operant behavior. American Psychologist, 18(8), 503–515. doi
Skinner, B. F. (1966a). The phylogeny and ontogeny of behavior. Science, 153 (3741), 1205–
1213. doi:10.1126/science.153.3741.1205
Skinner, B. F. (1966b). What is the experimental analysis of behavior? Journal of the
Experimental Analysis of Behavior, 9, 213–218. doi:10.1901/jeab.1966.9-213
Skinner, B. F. (1969) Contingencies of reinforcement: A theoretical analysis. New York, NY:
Appleton-Century-Crofts.
Skinner, B. F. (1974a). About behaviorism. New York, NY: Alfred A. Knopf.
Skinner, B. F. (1974b). Walden two. New York, NY: Macmillan.
Skinner, B. F. (1981). Selection by consequences. Science, 213(4507), 501–504.
doi:10.1126/science.7244649
Skinner, B. F. (1989). The behavior of the listener. In S. Hayes (Ed.), Rule-governed behavior:
Cognition, contingencies, and instructional control (pp. 85–96). Reno, NV: Context Press.
Slattery, B., & Stewart, I. (2014). Hierarchical classification as relational framing. Journal of the
Experimental Analysis of Behavior, 101, 1, 61–75. doi:10.1002/jeab.63.
Slattery, B., Stewart, I., & O’Hora, D. (2011). Testing for transitive class containment as a feature
of hierarchical classification. Journal of the Experimental Analysis of Behavior, 96, 2, 243–
260. doi:10.1901/jeab.2011.96-243.
Smeets, P. M., Leader, G., & Barnes, D. (1997). Establishing stimulus classes in adults and
children using a respondent-type training procedure: A follow-up study. The Psychological
Record, 47, 285–308.
Smith, D. P., Eikeseth, S., Fletcher, S. E., Montebelli, L., Smith, H. R., & Taylor, J. C. (2016).
Emergent intraverbal forms may occur as a result of listener training for children with
autism. The Analysis of Verbal Behavior, 32(1), 27–37. doi:10.1007/s40616-016-0057-3
Smith, N. W. (2016). The myth of mind: A challenge to mainstream psychology and its imposed
constructs. St. Petersburg, FL: BookLocker.com, Inc.
Smyth, S., Barnes-Holmes, D., & Forsyth, J. P. (2006). A derived transfer of simple
discrimination and self-reported arousal functions in spider fearful and non-fearful
participants. Journal of the Experimental Analysis of Behavior, 85, 223-246.
doi:10.1901/jeab.2006.02-05
Soha, J. A., & Peters, S. (2015). Vocal learning in songbirds and humans: A retrospective in
honor of Peter Marler. Ethology, 121, 933–945. doi:10.1111/eth.12415
Spradlin, J. E., & Brady, N. (2008). A behavior analytic interpretation of theory of mind.
International Journal of Psychology and Psychological Therapy, 8(3), 335–350.
Sprinkle, E. C., & Miguel, C. F. (2013). Establishing derived textual activity schedules in children
with autism. Behavioral Interventions, 28(3), 185–202. doi:10.1002/bin.1365
Stanley, C. R., Belisle, J., & Dixon, M. R. (2018). Equivalence-based instruction of academic
skills: Application to adolescents with autism. Journal of Applied Behavior Analysis, 51,
352–359. doi:10.1002/jaba.446
Stauch, T., LaLonde, K., Plavnick, J. B., Bak, M. S., & Gatewood, K. (2017). Intraverbal training
for individuals with autism: The current status of multiple control. The Analysis of Verbal
Behavior, 33(1), 98–116. doi:10.1007/s40616-017-0079-5
Steele, D., & Hayes, S. C. (1991). Stimulus equivalence and arbitrarily applicable relational
responding. Journal of the Experimental Analysis of Behavior, 56, 519–555.
doi:10.1901/jeab.1991.56-519
Sterkin, V. L. (2012). The effects of the social listener reinforcement protocol on the audience
control of stereotypy and social operants for students with developmental delays. (Doctoral
dissertation, Columbia University). doi:10.7916/D8DF6Z94
Sternberg, R. J. (1977). Component processes in analogical reasoning. Psychological Review, 84,
353–378.
Stevens, M. J., & Pfost, K. S. (1983). The role of cognition in cognitive behavior therapy.
Psychology: A Journal of Human Behavior, 20(3–4), 20–24.
Stewart, I., Barnes-Holmes, D., & Roche, B. (2004). A functional-analytic model of analogy
using the relational evaluation procedure. Psychological Record, 54, 531–552.
Stewart, I., Barnes-Holmes, D., Roche, B., & Smeets, P. (2001). Generating derived relational
networks via the abstraction of common physical properties: A possible model of analogical
reasoning. The Psychological Record, 51, 381–408.
Stewart, I., Barnes-Holmes, D., Roche, B., & Smeets, P. (2002). A functional-analytic model of
analogy: A relational frame analysis. Journal of the Experimental Analysis of Behavior,
78(3), 375–396.
Stewart, I., McElwee, J., & Ming, S. (2013a). Erratum to: Language generativity, response
generalization, and derived relational responding. The Analysis of Verbal Behavior, 29(1),
137–155. doi:10.1007/s40616-016-0060-8
Stewart, I., McElwee, J., & Ming, S. (2013b). Language generativiy, response generalization, and
derived relational responding. The Analysis of Verbal Behavior, 29(1), 137–155.
doi:10.1007/BF03393131
Stewart, I., Slattery, B., Chambers, M., & Dymond, S. (2018). An empirical investigation of part-
whole hierarchical relations. European Journal of Behaviour Analysis, 19, 1, 105–124.
doi:10.1080/15021149.2017.1416525
Still, K., Rehfeldt, R. A., Whelan, R., May, R., & Dymond, S. (2014). Facilitating requesting
skills using high-tech augmentative and alternative communication devices with individuals
with autism spectrum disorders: A systematic review. Research in Autism Spectrum
Disorders, 8(9), 1184–1199. doi:10.1016/j.rasd.2014.06.003
Stockwell, F. M. J., Hopkins, L. S., & Walker, D. J. (2017). Implicit and explicit attitudes toward
mainstream and BDSM sexual practices and their relation to interviewer behavior: An
analogue study. The Psychological Record, 67, 435-446. doi:10.1007/s40732-017-0225-4
Stockwell, F. M. J., Walker, D. J., & Eshleman, J. W. (2010). Measures of implicit and explicit
attitudes toward mainstream and BDSM sexual terms using the IRAP and questionnaire with
BDSM/fetish and student participants. The Psychological Record, 60, 307–324.
doi:10.1007/BF03395709
Stokes, T. F., & Baer, D. M. (1977). An implicit technology of generalization. Journal of Applied
Behavior Analysis, 10, 349–367. doi:10.1901/jaba.1977.10-349
Stolfi, L. (2005). The induction of observational learning repertoires in preschool children with
developmental disabilities as a function of peer-yoked contingencies (Doctoral dissertation,
Order No. 3174899). Available from ProQuest Dissertations & Theses Global. (305016342).
Retrieved from http://eduproxy.tc-library.org/?url=/docview/305016342?accountid=14258
Storlie, J. L., Rehfeldt, R. A., & Aguirre, A. A. (2015). Observational learning across three verbal
operants in a child with autism. International Journal of Behavior Analysis & Autism
Spectrum Disorders, 1, 113–121.
Street, E. M., & Johnson, K. (2014). The sciences of learning, instruction, and assessment as
underpinnings of the Morningside Model of Generative Instruction. Acta de Investigación
Psicológica, 4, 1773–1793.
Stromer, R., & Mackay, H. A. (1992). Spelling and emergent picture-printed word relations
established with delayed identity matching to complex samples. Journal of Applied Behavior
Analysis, 25, 893–904.
Strosahl, K. D., Robinson, P. J., & Gustavsson, T. (2015). Inside this moment. Oakland, CA.
Context Press.
Sullivan, J., & Barner, D. (2014). The development of structural analogy in number-line
estimation. Journal of Experimental Child Psychology, 128, 171–189.
Sundberg, M. L. (2008). VB-MAPP: Verbal Behavior Milestones Assessment and Placement
Program: A language and social skills assessment program for children with autism or other
developmental disabilities. Concord, CA: AVB Press.
Sundberg, M. L. (2013). Thirty points about motivation from Skinner’s book Verbal Behavior.
The Analysis of Verbal Behavior, 29, 13–40. doi:10.1007/BF03393120
Sundberg, M. L. (2016). Verbal stimulus control and the intraverbal relation. The Analysis of
Verbal Behavior, 32(2), 107–124. doi:10.1007/s40616-016-0065-3
Sundberg, M. L., Endicott, K., & Eigenheer, P. (2000). Using intraverbal prompts to establish
tacts for children with autism. The Analysis of Verbal Behavior, 17(1), 89–104.
doi:10.1007/BF03392958
Sundberg, M. L., & Michael, J. (2001). The benefits of Skinner’s analysis of verbal behavior for
children with autism. Behavior Modification, 25(5), 698–724.
doi:10.1177/0145445501255003
Sundberg, M. L., Michael, J., Partington, J. W., & Sundberg, C. A. (1996). The role of automatic
reinforcement in early language acquisition. The Analysis of Verbal Behavior, 13(1), 21–37.
doi:10.1007/BF03392904
Sundberg, M. L., & Partington, J. W. (2013). Teaching language to children with autism and
other developmental disabilities (2nd ed.) Pleasant Hill, CA: Behavior Analysts, Inc.
Sundberg, M. L., & Sundberg, C. A. (2011). Intraverbal behavior and verbal conditional
discriminations in typically developing children and children with autism. The Analysis of
Verbal Behavior, 27(1), 23–44. doi:10.1007/bf03393090
Tajfel, H. (1981). Human groups and social categories: Studies in Social Psychology. Cambridge,
UK: Cambridge University Press.
Takahashi, K., Yamamoto, J., & Noro, F. (2011). Stimulus pairing training in children with autism
spectrum disorder. Research in Autism Spectrum Disorders, 5, 547–553. doi:
10.1016/j.rasd.2010.06.021
Tarbox, J., Tarbox, R., & O’Hora, D. (2009). Nonrelational and relational instructional control. In
R. A. Rehfeldt & Y. Barnes-Holmes (Eds.), Derived relational responding: Applications for
learners with autism and other developmental disabilities: A progressive guide to change
(pp. 111–128). Oakland, CA: New Harbinger Publications.
Tarbox, J., Zuckerman, C. K., Bishop, M. R., Olive, M. L., & O’Hora, D. P. (2011). Rule-
governed behavior: Teaching a preliminary repertoire of rule-following to children with
autism. The Analysis of Verbal Behavior, 27, 125–139. doi:10.1007/bf03393096
Taylor, B. A., & DeQuinzio, J. A. (2012). Observational learning and children with autism.
Behavior Modification, 36, 341–360. doi:10.1177/0145445512443981
Taylor, B. A., DeQuinzio, J. A., & Stine, J. (2012). Increasing observational learning of children
with autism: A preliminary analysis. Journal of Applied Behavior Analysis, 45, 815–820.
doi:10.1901/jaba.2012.45-815
Taylor, B. A., Levin, L., & Jasper, S. (1999). Increasing play-related statements in children with
autism toward their siblings: Effects of video modeling. Journal of Developmental and
Physical Disabilities, 11, 253–264. doi:10.1023/A:1021800716392
Taylor, I., & O’Reilly, M. F. (1997). Toward a functional analysis of private verbal self-
regulation. Journal of Applied Behavior Analysis, 30, 43-58.
Taylor, I., & O’Reilly, M. F. (2000). Generalization of supermarket shopping skills for individuals
with mild intellectual disabilities using stimulus equivalence training. The Psychological
 Record, 50, 49–62.
Terrace, H. S. (1966). Stimulus control. In W. K. Honig (Ed.), Operant behavior: Areas of
research and application (pp. 271–344). New York, NY: Appleton-Century-Crofts.
Thomas, B. R., Lafasakis, M., & Sturmey, P. (2010). The effects of prompting, fading, and
differential reinforcement on vocal mands in non-verbal preschool children with autism
spectrum disorders. Behavioral Interventions, 25(2), 157–168. doi:10.1002/bin.300
Thorndike, E. L. (1911). Animal intelligence. New York, NY: The MacMillan Co.
Tiemann, P., & Markle, S. (1990). Analyzing instructional content: A guide to instruction and
evaluation. (4th Ed.) Champaign, IL: Stipes.
Timberlake, W. (1988). The behavior of organisms: Purposive behavior as a type of reflex.
Journal of the Experimental Analysis of Behavior, 50(2), 305–317.
doi:10.1901/jeab.1988.50-305
Timberlake, W. (2004). Is the operant contingency enough for a science of purposive behavior?
Behavior and Philosophy, 32(1), 197–229.
Timberlake, W., & Allison, J. (1974). Response deprivation: An empirical approach to
instrumental performance. Psychological Review, 81(2), 146–164. doi:10.1037/h0036101
Timberlake, W., & Farmer-Dougan, V. A. (1991). Reinforcement in applied settings: Figuring out
ahead of time what will work. Psychological Bulletin, 110(3), 379–391. doi:10.1037/0033-
2909.110.3.379
Timmins, L., Barnes-Holmes, D., & Cullen, C. (2016). Measuring implicit sexual response biases
to nude male and female pictures in androphilc and gynephilic men. Archives of Sexual
Behavior, 45, 829–841. doi:10.1007/s10508-016-0725-3
Tincani, M. (2004). Comparing the picture exchange communication system and sign language
training for children with autism. Focus on Autism and Other Developmental Disabilities,
19(3), 152–163. doi:10.1177/10883576040190030301
Tincani, M., & Devis, K. (2011). Quantitative synthesis and component analysis of single-
participant studies on the picture exchange communication system. Remedial and Special
Education, 32(6), 458–470. doi:10.1177/0741932510362494
Tomasello, M. (2008). Origins of human communication. Cambridge, MA: MIT Press.
Tomonaga, M., Matzusawa, T., Fujita, K., & Yamamoto, J. (1991). Emergence of symmetry in a
visual conditional discrimination by chimpanzees (Pan troglodytes). Psychological Reports,
68, 51–60.
Torelli, J. N., Lambert, J. M., Da Fonte, M. A., Denham, K. N., Jedrzynski, T. M., & Houchins-
Juarez, N. J. (2016). Assessing acquisition of and preference for mand topographies during
functional communication training. Behavior Analysis in Practice, 9(2), 165–168.
doi:10.1007/s40617-015-0083-y
Törneke, N. (2010). Learning RFT: An introduction to relational frame theory and its clinical
application. Oakland, CA: New Harbinger.
Törneke, N., Luciano, C., Barnes-Holmes, Y., & Bond, F. (2016). RFT for clinical practice: Three
core strategies in understanding and treating human suffering. In R. D. Zettle, S. C. Hayes,
 D. Barnes-Holmes, & T. Biglan (Eds.), Handbook of Contextual Behavioral Science (pp.
254–272). Chichester, UK: Wiley Blackwell.
Toussaint, K. A., & Tiger, J. H. (2010). Teaching early braille literacy skills within a stimulus
equivalence paradigm to children with degenerative visual impairments. Journal of Applied
Behavior Analysis, 43, 181–194. doi:10.1901/jaba.2010.43-181
Towe, A. L. (1954). A study of figural equivalence in the pigeon. Journal of Comparative and
Physiological Psychology, 47, 283–287.
Townley-Cochran, D., Leaf, J. B., Taubman, M., Leaf, R., & McEachin, J. (2015). Observational
learning for students diagnosed with autism: A review paper. Review Journal of Autism and
Developmental Disorders, 2, 262–272. doi:10.1007/s40489-015-0050-0
Trucil, L. M., Vladescu, J. C., Reeve, K. F., DeBar, R. M., & Schnell, L. K. (2015). Improving
portion-size estimation using equivalence-based instruction. The Psychological Record, 65,
761–770
Tullis, C. A., Frampton, S. E., Delfs, C. H., & Shillingsburg, M. A. (2017). Teaching problem
explanations using instructive feedback. The Analysis of Verbal Behavior, 33(1), 64–79.
doi:10.1007/s40616-016-0075-1
Twyman, J. S. (1996). The functional independence of impure mands and tacts of abstract
stimulus properties. The Analysis of Verbal Behavior, 13(1), 1–19. doi:10.1007/BF03392903
U.S. Department of Education, National Center for Education Statistics. (2017). The Condition of
Education 2017 (NCES 2017–144).
Vahey, N., Boles, S., & Barnes-Holmes, D. (2010). Measuring adolescents’ smoking-related
social identity preferences with the Implicit Relational Assessment Procedure (IRAP) for the
first time: A starting point that explains later IRAP evolutions. International Journal of
Psychology and Psychological Therapy, 10(3), 453–474.
Valdivia-Salas, S., Luciano, C., Gutiérrez, O., & Visdómine, C. (2009). Establishing empathy. In
R. A. Rehfeldt & Y. Barnes-Holmes (Eds.), Derived relational responding applications for
learners with autism and other developmental disabilities: A progressive guide to change
(pp. 301–312). Oakland, CA: New Harbinger.
Valdovinos, M. G., & Kennedy, C. H. (2004). A behavior-analytic conceptualization of the side
effects of psychotropic medication. The Behavior Analyst, 27, 231–238.
doi:10.1007/BF03393182
Valentino, A. L., Conine, D. E., Delfs, C. H., & Furlow, C. M. (2015). Use of a modified chaining
procedure with textual prompts to establish intraverbal storytelling. The Analysis of Verbal
Behavior, 31(1), 39–58. doi:10.1007/s40616-014-0023-x
Vallinger-Brown, M., & Rosales, R. (2014). An investigation of stimulus pairing and listener
training to establish emergent intraverbals in children with autism. The Analysis of Verbal
Behavior, 30(2), 148–159. doi:10.1007/s40616-014-0014-y
Valverde, M. R., Luciano, C., & Barnes-Holmes, D. (2009). Transfer of aversive respondent
elicitation in accordance with equivalence relations. Journal of the Experimental Analysis of
Behavior, 92, 85–111.
van der Meer, L., Sutherland, D., O’Reilly, M. F., Lancioni, G. E., & Sigafoos, J. (2012). A
further comparison of manual sign, picture exchange, and speech generating devices as
 communication modes for children with autism spectrum disorder. Research in Autism
Spectrum Disorders, 6, 1247–1257. doi:10.1016/j.rasd.2012.04.005
Van Houten, R. (1980). Learning through feedback: A systematic approach for improving
academic performance. New York, NY: Human Sciences Press.
Varelas, A., & Fields, L. (2017). Equivalence based instruction by group based clicker training
and sorting tests. The Psychological Record, 1–10.
Vaughan, M. E., & Michael, J. L. (1982). Automatic reinforcement: An important but ignored
concept. Behaviorism, 10(2), 217–227.
Vedora, J., & Conant, E. (2015). A comparison of prompting tactics for teaching intraverbals to
young adults with autism. The Analysis of Verbal Behavior, 31(2), 267–276.
doi:10.1007/s40616-015-0030-6
Vilardaga, R. (2009). A relational frame theory account of empathy. International Journal of
Behavioral Consultation and Therapy, 5, 2, 178–184.
Vilardaga, R., Estévez, A., Levin, M. E., & Hayes, S. C. (2012). Deictic relational responding,
empathy, and experiential avoidance as predictors of social anhedonia: Further contributions
from relational frame theory. The Psychological Record, 62, 409–432.
Villatte, J. L, Villatte, M., & Hayes, S. C. (2012). A naturalistic approach to transcendence:
Deictic framing, spirituality, and prosociality. In L. McHugh & I. Stewart (Eds.), The self
and perspective-taking (pp. 199-216). Oakland, CA: New Harbinger.
Villatte, M., Monestès, J. L., McHugh, L., Freixa i Baqué, E., & Loas, G. (2008) Assessing
deictic relational responding in social anhedonia: A functional approach to the development
of theory of mind impairments. International Journal Behavioral Consultation and Therapy,
4 (4), 360–373.
Villatte, M., Monestès, J. L., McHugh, L., Freixa i Baqué, E., & Loas, G. (2010). Assessing
perspective taking in schizophrenia using relational frame theory. The Psychological Record,
60, 413.
Villatte, M., Vilardaga, R., & Monestès, J. L. (2012). How the self relates to others when
perspective-taking is impaired. In L. McHugh & I. Stewart (Eds.), The self and perspective-
taking (pp. 109–124). Oakland, CA: New Harbinger.
Villatte, M., Villatte, J. L., & Hayes, S. C. (2016). Mastering the clinical conversation. Language
as intervention. New York, NY: The Guilford Press.
Vitale, A., Barnes-Holmes, Y., Barnes-Holmes, D., & Campbell, C. (2008). Facilitating
responding in accordance with the relational frame of comparison: Systematic empirical
analyses. The Psychological Record, 58, 365.
Vosniadou, S., & Ortony, A. (1989). Similarity and analogical reasoning: A synthesis. In S.
Vosniadou & A. Ortony (Eds.), Similarity and Analogical Reasoning (pp. 1–18). Cambridge,
UK: Cambridge University Press.
Vygotsky, L. S. (1967). Play and its role in the mental development of the child. Soviet
Psychology, 5(3), 6–18.
Walker, B. D., & Rehfeldt, R. A. (2012). An evaluation of the stimulus equivalence paradigm to
teach single-subject design to distance education students via Blackboard. Journal of
Applied Behavior Analysis, 45, 329–344.
Walker, B. D., Rehfeldt, R. A., & Ninness, C. (2010). Using the stimulus equivalence paradigm to
teach course material in an undergraduate rehabilitation course. Journal of Applied Behavior
Analysis, 43, 615–633. doi:10.1901/jaba.2010.43-615
Wallace, M. D., Iwata, B. A., & Hanley, G. P. (2006). Establishment of mands following tact
training as a function of reinforcer strength. Journal of Applied Behavior Analysis, 39(1),
17–24. doi:10.1901/jaba.2006.119-04
Wang, J. M. (2015). Preference-for-solitude and depressive symptoms in Chinese adolescents.
Personality and Individual Differences, 100, 151–156.
Ward, W. D., & Stare, S. W. (1990). The role of subject verbalization in generalized
correspondence. Journal of Applied Behavior Analysis, 23, 129–136.
doi:10.1901/jaba.1990.23-129
Watkins, C. L., Pack-Teixeira, L., & Howard, J. S. (1989). Teaching intraverbal behavior to
severely retarded children. The Analysis of Verbal Behavior, 7(1), 69–81.
doi:10.1007/bf03392838
Watson, J. B. (1913). Psychology as the behaviorist views it. Psychological Review, 20(2), 158–
177. doi:10.1037/h0074428
Watt, A. W., Keenan, M., Barnes, D., & Cairns, E. (1991). Social categorization and stimulus
equivalence. The Psychological Record, 41, 371–388.
Watts, A. C., Wilder, D. A., Gregory, M. K., Leon, Y., & Ditzian, K. (2013). The effect of rules on
differential reinforcement of other behavior. Journal of Applied Behavior Analysis, 46, 680–
684. doi:10.1002/jaba.53
Wehmeyer, M. L., Palmer, S. B., Agran, M., Mithaug, D. E., & Martin, J. (2000). Teaching
students to become causal agents in their lives: The self-determining learning model of
instruction. Exceptional Children, 66, 439–453.
Weil, T. M., Hayes, S. C., & Capurro, P. (2011) Establishing a deictic relational repertoire in
young children. The Psychological Record, 61, 371–390.
Wells, A. (2000). Emotional disorders and metacognition: Innovative cognitive therapy.
Chichester, UK: Wiley.
Wells, A., & Matthews, G. (1994). Attention and emotion: A clinical perspective. Hove, UK:
Erlbaum.
West-Eberhard, M. J. (2003). Developmental plasticity and evolution. Oxford, UK: Oxford
University Press.
Whelan, R., & Barnes-Holmes, D. (2004). The transformation of consequential functions in
accordance with the relational frames of same and opposite. Journal of the Experimental
Analysis of Behavior, 82, 177–195.
Whelan, R., Barnes-Holmes, D., & Dymond, S. (2006). The transformation of consequential
functions in accordance with the relational frames of more-than and less-than. Journal of the
Experimental Analysis of Behavior, 86, 317–335. doi:10.1901/jeab.2006.113-04
Whimbey, A. (1990). Analytical reading and reasoning. (2nd ed.). Cary, NC: Innovative
Sciences.
Whimbey, A. (1995). Mastering reading through reasoning. Cary, NC: Innovative Sciences.
Whimbey, A., & Jenkins, E. L. (1987). Analyze. Organize. Write. (rev. ed.). Hillsdale, NJ:
Erlbaum.
Whimbey, A., Johnson, M. H., Williams, E., Sr., & Linden, M. (2017). Blueprint for educational
change: Improving reasoning, literacies, and science achievement with cooperative learning.
Seattle, WA: Morningside. (Original work published in 1993 in Washington, DC: Right
Combination, Inc.).
Whimbey, A., & Lochhead, J. (1999). Problem solving and comprehension. (6th ed.). Mahwah,
NJ: Erlbaum.
Williams, B. A. (1987). The other psychology of animal learning: A review of Mackintosh’s
Conditioning and Associative Learning. Journal of the Experimental Analysis, 48, 175–186.
doi:10.1901/jeab.1987.48-175
Williams, B. A. (1994). Conditioned reinforcement: Experimental and theoretical issues. The
Behavior Analyst, 17, 261–285. doi:10.1007/BF03392675
Willmoth, V., Lewon, A. B., Taylor, E., & Ghezzi, P. M. (May, 2017). Teaching conversational
interactions to a young child with autism. Presented at the annual meeting of the Association
for Behavior Analysis International, Denver, CO.
Wilson, D. S., Hayes, S. C., Biglan, A., & Embry, D. D. (2014). Evolving the future: Toward a
science of intentional change. Behavioral and Brain Sciences, 37(4), 395–416.
doi:10.1017/S0140525X13001593
Wilson, G. T., & Davison, G. C. (1971). Processes of fear reduction in systematic desensitization:
Animal studies. Psychological Bulletin, 76(1), 1–14.
Wilson, K. G., Bordieri, M., & Whiteman, K. (2012). The self and mindfulness. In L. McHugh &
I. Stewart (Eds.), The self and perspective-taking (pp. 181–198). Oakland, CA: New
Harbinger.
Wilson, K. G., & Sandoz, E. K. (2008). Mindfulness, values, and the therapeutic relationship in
Acceptance and Commitment Therapy. In S. F. Hick & T. Bein (Eds.), Mindfulness and the
therapeutic relationship (pp. 89–106). New York, NY, Guilford Press.
Wilson, K. G., Sandoz, E. K., Kitchens, J., & Roberts, M. (2010). The Valued Living
Questionnaire: Defining and measuring valued action within a behavioral framework. The
Psychological Record, 60(2), 249–272.
Winborn-Kemmerer, L., Ringdahl, J. E., Wacker, D. P., & Kitsukawa, K. (2009). A demonstration
of individual preference for novel mands during functional communication training. Journal
of Applied Behavior Analysis, 42(1), 185–189. doi:10.1901/jaba.2009.42-185
Wiskow, K. M., Matter, A. L., & Donaldson, J. M. (2018). An evaluation of lag schedules and
prompting methods to increase variability of naming category items in children with autism
spectrum disorder. The Analysis of Verbal Behavior, 34(1–2), 100–123. doi:10.1007/s40616-
018-0102-5
Wispé L. (1986) The distinction between sympathy and empathy: To call forth a concept, a word
is needed. Journal of Personality and Social Psychology, 50, 314–321.
Wittgenstein, L. (1953). Philosophical investigations. Oxford, England: Basil & Blackwell.
Wittgenstein, L. (1969). On certainty. Oxford, England: Basil & Blackwell.
Wittgenstein, L. (1980). Remarks on the philosophy of psychology (Vol. 2). Oxford, England:
Basil & Blackwell.
Wolf, M. M. (1978). Social validity: The case for subjective measurement or how behavior
analysis is finding its heart. Journal of Applied Behavior Analysis, 11, 203–214.
Wolf, M. M., Phillips, E. L., Fixsen, D. L., Braukmann, C. J., Kirigin, K. A., Willner, A. G., &
Schumaker, J. (1976). Achievement place: The teaching-family model. Child Care
Quarterly, 5(2), 92–103.
Wolpe, J. (1958). Psychotherapy by reciprocal inhibition. Palo Alto, CA: Stanford University
Press.
Wolpe, J. (1961). The systematic desensitization treatment of neuroses. The Journal of Nervous
and Mental Disease, 132(3), 189–203.
Wolpe, J. (1976a). Behavior therapy and its malcontents—I. Denial of its bases and
psychodynamic fusionism. Journal of Behavior Therapy and Experimental Psychiatry, 7(1),
1–5. doi:10.1016/0005-7916(76)90032-X
Wolpe, J. (1976b). Behavior therapy and its malcontents—II. Multimodal eclecticism, cognitive
exclusivism and “exposure” empiricism. Journal of Behavior Therapy and Experimental
Psychiatry, 7(2), 109–116. doi:10.1016/0005-7916(76)90066-5
Wolpe, J. (1978). Cognition and causation in human behavior and its therapy. American
Psychologist, 33(5), 437–446. doi:10.1037/0003-066X.33.5.437
World Health Organization. (2016). Health statistics and information systems. Estimates for
2000–2012. World Health Organization.
Wulfert, E. (2002). Can contextual therapies save clinical behavior analysis? The Behavior
Analyst Today, 3(3), 254–258. doi:10.1037/h0099984
Wulfert, E., Greenway, D. E., Farkas, P., Hayes, S. C., & Dougher, M. J. (1994). Correlation
between self-reported rigidity and rule-governed insensitivity to operant contingencies.
Journal of Applied Behavior Analysis, 27, 659–671. doi:10.1901/jaba.1994.27-659
Wymer, S. C., Tarbox, J., Beavers, G. A., & Tullis, C. A. (2016). Teaching children with autism to
follow rules specifying a behavior and consequence. The Analysis of Verbal Behavior, 32,
265–274. doi:10.1007/s40616-016-0059-1
Wynne, L. (2004). The missing theory: Why behavior analysis has little impact on voluntary adult
outpatient services. Ethical Human Psychology And Psychiatry: An International Journal of
Critical Inquiry, 6(2), 135–146.
Yamamoto, J., & Asano, T. (1995). Stimulus equivalence in a chimpanzee (Pan troglodytes). The
Psychological Record, 45, 3–21.
Yoon, S. Y., & Bennett, G. M. (2000). Effects of a stimulus-pairing procedure on conditioning
vocal sounds as reinforcers. The Analysis of Verbal Behvior, 17(1), 75–88.
doi:10.1007/BF03392957
Yorio, A., Tabullo, Á., Wainselboim, A., Barttfeld, P., & Segura, E. (2008). Event-related
potential correlates of perceptual and functional categories: Comparison between stimuli
matching by identity and equivalence. Neuroscience Letters, 443, 113–118.
Zaki, J., & Ochsner, K. N. (2012). The neuroscience of empathy: Progress, pitfalls and promise.
Nature Neuroscience, 15(5), 675.
Zax, M., & Klein, A. (1960). Measurement of personality and behavior changes following
psychotherapy. Psychological Bulletin, 57(5), 435–448.
Zentall, T. R., Galizio, M., & Critchfield, T. S. (2002). Categorization, concept learning, and
behavior analysis: An introduction. Journal of the Experimental Analysis of Behavior, 78,
237–248. doi:10.1901/jeab.2002.78-237
Zettle, R. D., & Hayes, S. C. (1982). Rule-governed behavior: A potential theoretical framework
for cognitive-behavior therapy. Advances in cognitive-behavioral research and therapy, 1,
73–118.
Zettle, R. D., & Hayes, S. C. (1986). Dysfunctional control by client verbal behavior: The context
of reason-giving. The Analysis of Verbal Behavior, 4, 30–38. doi:10.1007/bf03392813
Zettle, R.D., Hayes, S. C., Barnes-Holmes, D., & Biglan, A. (Eds.) (2016). The Wiley handbook
of contextual behavioral science. UK: Wiley, Blackwell.
Ziomek, M. M., & Rehfeldt, R.A. (2008). Investigating the acquisition, generalization, and
emergence of untrained verbal operants for mands acquired using the Picture Exchange
Communication System in adults with severe developmental disabilities. The Analysis of
Verbal Behavior, 24(10), 15–30. doi:10.1007/BF03393054
Zuriff, G. E. (1979). Ten inner causes. Behaviorism, 7, 1–8.
Editor Mitch Fryling, PhD, BCBA-D, is associate professor and chair of
the division of special education and counseling at California State
University, Los Angeles. He has authored and coauthored many scholarly
publications, primarily in the area of behavioral theory and philosophy,
especially as it pertains to complex human behavior and system
development in behavior analysis. He is current editor of The Psychological
Record.
Editor Ruth Anne Rehfeldt, PhD, BCBA, is professor of behavior analysis
and therapy at Southern Illinois University in Carbondale, IL. She holds
doctoral and master’s degrees in psychology from the University of
Nevada, Reno; and a bachelor’s degree in psychology from the University
of Puget Sound. She is also a board-certified behavior analyst. Ruth Anne is
a fellow of the Association for Behavior Analysis International, and
previous editor of The Psychological Record. She has published numerous
articles in the area of verbal behavior and derived relational responding for
individuals with autism spectrum disorder (ASD) and other developmental
disabilities.
Editor Jonathan Tarbox, PhD, BCBA-D, is director of the master of
science in applied behavior analysis program at the University of Southern
California, as well as director of research at FirstSteps for Kids. He is editor
in chief of the journal, Behavior Analysis in Practice, and serves on the
editorial boards of several scientific journals related to autism and behavior
analysis. He has published four books on autism treatment; is series editor
of the Elsevier book series, Critical Specialties in Treating Autism and
Other Behavioral Challenges; and author of well over seventy peer-
reviewed journal articles and chapters in scientific texts. His research
focuses on behavioral interventions for teaching complex skills to
individuals with autism, applications of acceptance and commitment
therapy (ACT) training inside of applied behavior analysis, and applications
of applied behavior analysis to issues of diversity and social justice.
Editor Linda J. Hayes, PhD, is professor of psychology, and founder and
director of the behavior analysis program at the University of Nevada,
Reno.
 Index

A
ABA renewal effect, 13
abolishing operations (AOs), 17, 25, 301
about this book, 5
abstract relational responding, 264
acceptance and commitment therapy (ACT), 231, 256–260, 301; defusion interactions in, 289;
general description of, 256; interdependent processes in, 257–260; scientific strategy and support
for, 256–257
accompanying mediative interactions, 77
accuracy of rules, 219
acquisition of behavior, 108
ACT. See acceptance and commitment therapy
active listening, 149
active modes, 95–96
adduction, 134
age-related biases, 270–271
alive language, 95, 112
alteration contacts, 103
analogies, 199–205; child development and, 203–204; cross-domain, 202; future directions for
studying, 211; key features of, 200–201; nonarbitrary properties in, 201
analytic hierarchy, 206, 209
Analytical Reading and Reasoning (Whimbey), 147
Analyze. Organize. Write (Whimbey & Jenkins), 147
antecedent control, 14–16
antecedents of behavior, 8
applied behavior analysis (ABA), 6, 7, 91, 104, 260, 301
aptitude criteria, 110
arbitrarily applicable relational responding (AARR), 176, 177–178, 301
assessment: IRAP procedure for, 265, 266–278, 288; of observational learning, 120–121
asymmetrical class containment, 206–207, 208
atomic repertoires, 118
attention, 121
augmenting, repertoire of, 220–221
autism spectrum disorder (ASD), 25, 39, 231–232, 287, 301
automatic reinforcement, 23

B
Baires, Natalia, 115
Bandura, Albert, 116–117
Barbero, Adrián, 281
Barnes-Holmes, Dermot, 264
behavior: challenges in studying, 2–3; contingency-shaped, 216; language and human, 95–105;
maladaptive, 224, 227–228; metaphor of acquiring, 108; psychological, 98–100, 101–102;
referential, 72–92; rule-governed, 5, 214–232; social, 56–57, 264–280; three-term contingency
and, 8–9. See also verbal behavior
behavior analysis, 20; clinical, 251–252; developmental verbal, 69; future directions for, 261–262;
issues challenging, 250; observational learning and, 117–124; private events and, 251, 261; verbal
behavior and, 21–34, 182–183
behavior chains, 168
behavior intervention plans, 225–226
behavior therapy, 7, 252–256; clinical behavior analysis and, 251–252; first wave of, 252–253;
implications of evolution of, 260–262; second wave of, 253–254; third wave of, 254–256
behavioral developmental cusps, 56, 70
behavioral evolution, 152–153
behavioral metamorphosis, 65–69
behavioral skills training (BST), 49, 301
behavior-altering effect, 16
behaviorism, 53–54
bidirectional naming (BiN), 301; explanation of, 61–62; types of, 63–65
bidirectional verbal operants, 56–70; behavioral metamorphosis and, 65–69; identifying and
establishing, 58–65
Biographical Sketch of an Infant (Darwin), 66
biological behavior, 99
bistimulation, 75
blend of behaviors, 136, 139
blocked-trial procedures, 45–46
Blueprint for Educational Change (Whimbey et al.), 145
body image biases, 271
Boullion, Gina Q., 250
brief and immediate relational responses (BIRRs), 272, 273, 302
Brodsky, Julia, 157

C
Campbell, Vincent, 214
candidate elements, 8–9
candidate response, 139
categorization tasks, 239
Cfunc and Crel properties, 274, 275, 276
chaining procedures, 40
children: analogical reasoning in, 203–204; hierarchical relational responding in, 209–211;
referential behavior study with, 81
class inclusion responding, 210
classes of behavior, 10
classical conditioning, 8, 18
classification hierarchy, 206–207, 208–209
clinical behavior analysis, 251–252
codics, 21
cognitive defusion, 259
cognitive problem solving, 235–236
combinatorial entailment, 179–180
combined stimulus procedure, 138
committed action, 260
Common Core, 171
comparison: functional contact of, 103; relational frame of, 190–192
comparison-as-node training structure, 162
comparison stimuli, 125
compassion, 294–296; definition and description of, 294–295; distinguishing empathy from, 296;
functional contextual approach to, 296–297; perspective taking related to, 282–283, 298; research
studies on, 295–296. See also empathy
compassion fatigue, 296
competence and competencies, 110–111, 112
complete interactions, 80, 89, 90
complex verbal operants, 39–51; dictation taking, 49, 50; intraverbals, 40–49; textual behavior, 49–
50
complexity of rules, 219
complex-to-simple (CTS) protocol, 164, 302
component behaviors, 136, 140, 154
composition, concept of, 132
compound stimulus control, 41, 42–43
compound tact instruction, 40
concept formation, 158, 169
concept training, 15
conceptual hierarchy, 206–207, 208, 211
conditional discrimination, 41, 45–46, 118
conditioned MOs, 17
conditioned reinforcers, 123–124
conditioned response (CR), 7, 302
conditioned stimulus (CS), 7, 302
conditioning: observational, 59–60; operant, 6–7, 8–9; respondent, 6, 7–8
conduct and behavior, 98
consequences of behavior, 8
consequential control, 9, 14
constative knowing, 109
constructivist problem solving, 235–236
contextual behavioral science (CBS), 7, 256–257, 302
contextual cues, 208, 209
contingency adduction, 133–156; behavioral evolution and, 152–153; case study in, 143–145;
classroom use of, 143; collateral emotional effects of, 151–152; critical features of, 135; designing
instruction to produce, 138; examples of, 134–143; future directions in, 154–155; generativity
and, 133–134; variations of, 136–137
contingency adduction moment, 155
contingency adduction probe, 134–135
contingency-shaped behavior, 216
contingent relations, 217
control: antecedent, 14–16; consequential, 9, 14; convergent, 42, 45–47; divergent, 42, 43–45;
intraverbal, 40; stimulus, 14, 157
convergent control, 42, 45–47
conversational units, 57
coordination, frame of, 185–186
copying text, 49, 50
coupling, 103
cross-domain analogies, 202
cultural behavior, 77

D
DAARRE model, 266, 275–278
Darwin, Charles, 66–67, 100
dead language, 95
deictic framing, 193–194, 286; empathy development and, 292–294; empirical research on, 287–
290; perspective taking as, 285–287, 288, 290
delayed gratification, 242
delayed imitation, 121
delayed MTS procedures, 161, 162
delayed performance, 116–117
delayed prompting, 155
Delgado, Marianne, 143, 145
denotative language, 101
derived intraverbals, 47–48
derived mands, 33–34
derived relational responding (DRR), 174, 255, 302; clinical behavior analysis and, 255; as
generalized operant behavior, 181–182; language related to, 175
derived stimulus relations, 5; description of emergent or, 174; observational learning research in,
124–129
derived vs. instructed rules, 222
Descent of Man, The (Darwin), 67
detachment, concept of, 103
developmental evolutionary biology, 69
Developmental Plasticity and Evolution (West-Eberhard), 66
developmental verbal behavior analysis, 69
dialectical behavior therapy (DBT), 254, 302
dictation taking, 49, 50
differential arbitrarily applicable relational responding effects (DAARRE) model, 266, 275–278, 302
differential observing response (DOR), 40, 303
differential reinforcement of low rate (DRL), 215, 303
differential reinforcement of other behavior (DRO), 225, 303
discrimination: conditional, 41, 45–46, 118; observational, 121; simple verbal, 41–42; stimulus, 14;
verbal conditional, 42–43
discrimination training, 14
discriminative stimulus, 14, 42, 157, 216, 303
dishabituation, 8
distinction, frame of, 186–188
distributed trials, 40
divergent control, 42, 43–45
Dixon, Mark R., 174
do-say correspondence, 63
Du, Lin, 55
duplics, 21

E
ear advantages, 67–68
early intensive behavioral intervention (EIBI), 84, 88, 90, 303
EBI. See equivalence-based instruction
echoics, 21, 23–25; definition of, 23; instruction using, 23–24; stimulus-stimulus pairing and, 24–25
education: contingency adduction in, 133–156; equivalence-based instruction in, 157–172; flaws
inherent in models of, 111–112; functional competence in, 110–111; generative instruction in, 132;
process of institutional, 105–106; psychological dimensions in, 106; rule-governed behavior in,
225. See also instruction
electroencephalograms (EEGs), 195
emergent relations, 32, 174
emotional intelligence, 291
emotions: contingency adduction and, 151–152; empathy development and, 293
empathy, 291–294; definition and description of, 291; distinguishing compassion from, 296;
functional contextual approach to, 296–297; perspective taking related to, 282–283, 291, 298;
relational framing account of, 292–294; research studies on, 291–292. See also compassion
English language learner (ELL), 35, 303
Epstein, Robert, 133
equivalence class formation, 161, 164, 167, 169
equivalence procedure, 138
equivalence relations, 137, 141–143, 159, 200, 218
equivalence-based instruction (EBI), 157–172, 303; applications for, 166–169; explanatory overview
of, 158; future directions for, 169–172; instructional design choices in, 160–166; match-to-sample
procedures in, 161–162; research study typifying, 159–160; response generalization in, 165–166;
training protocols for, 163–165; training structures for, 162–163
equivalence-equivalence responding, 201–202
Erickson, Nicole, 145
errorless learning, 16, 44
establishing operations (EOs), 17, 18, 25, 303
event-related potentials (ERPs), 202
evolutionary perspective, 295–296
exclusion procedure, 138, 140
experiential acceptance, 258–259
explicitness of rules, 219
extended and elaborated relational responses (EERRs), 272, 273, 303
extension contacts, 103
extinction, 12–13
extinction burst, 12–13
Extrinsic Affective Simon Task (EAST), 267, 303

F
fading process, 16
Ferster, Charles, 153
Fienup, Daniel M., 157
first-wave behavior therapies, 252–253
fixed ratio (FR) reinforcement schedule, 227, 303
flexible responding, 34
framing: definition of, 285; repertoires of, 242
frequency building, 150–151
Fryling, Mitch, 115
Function Acquisition Speed Test (FAST), 267
functional analysis, 22
functional analytic psychotherapy (FAP), 256, 285, 304
functional aptitude, 110
functional classes, 119
functional competence, 110
functional contextualism (FC), 7, 256–257, 304
functional independence, 30–33
functional MRIs (fMRIs), 195

G
gambling behavior, 223–224
Garcia, Yors A., 20, 38
Garcia-Zambrano, Sebastian, 20, 38
gender bias, 269–270
generalization processes, 119, 165–166
generalized echoic responding, 23
generalized operants, 181–182
generative instruction, 132, 154, 155
generative repertoires, 145–149
generative responding, 132, 155
generativity: contingency adduction and, 133–134; explanation of, 131–132
Ghezzi, Patrick M., 4, 72
Gil-Luciano, Bárbara, 281
Greer, R. Douglas, 55

H
habituation, 7
Harte, Colin, 264
Haughton, Eric, 150
Hayes, Linda J., 1, 6
hierarchical classification, 205
hierarchical part-whole analysis, 205
hierarchical relations, 192–193
hierarchies, 205–211; analytic, 206, 209; child development and, 209–211; classification, 206–207,
208–209; contextual cues for, 208, 209; future directions for studying, 211–212; relational frame
of, 192–193
hybrid training protocol, 164, 165

I
I-Here-Now perspective, 286, 290, 293, 298
imitative behaviors, 118, 121
Implicit Association Test (IAT), 267, 304
implicit relational assessment procedure. See IRAP
implicit social cognition, 265, 266–267; DAARRE model and, 275–278; future directions for
studying, 278–279; IRAP research and, 267–272; REC model and, 272–275
incidental learning, 61–62
incomplete interactions, 80, 89, 90
institutional behavior, 99
instructed vs. derived rules, 222
instruction: echoic, 23–24; equivalence-based, 157–172; generative, 132; intraverbal, 43–49; mand,
27–29; tact, 30, 46–47. See also education
instructive feedback (IF), 40, 43–44, 304
intelligence, concept of, 106, 107
intensive tact protocol, 59
interactions: complete vs. incomplete, 80, 89, 90; defusion, 289; mediative, 77; referential, 73, 76–
77, 79–82
interbehavioral psychology, 7, 92, 119
interbehavioral theory, 107
interbehaviorism, 119
interrupted behavior chains procedure, 28
interrupted chained task procedure, 34
inter-stimulus interval (ISI), 126, 304
interventions: early intensive behavioral, 84, 88, 90, 303; peer-mediated, 40, 48–49
intraverbal control, 40
intraverbal operants, 40, 239
intraverbal prompting, 44
intraverbals, 21, 40–49, 239; convergent control and, 45–47; derived, 47–48; divergent control and,
43–45; instructional procedures, 40–41, 43–49; peer-mediated interventions, 48–49; prerequisites
for teaching, 43; reverse, 47
IRAP (implicit relational assessment procedure), 265, 304; implicit social cognition and, 266–267;
REC model related to, 272–275; research on deictic framing using, 288; RFT perspective on
effects of, 272–278; social research using, 267–272
IRAP effect, 266–267

J
Jacobs, Kenneth W., 6
Johnson, Kent, 131

K
Kantor, J. R., 4–5, 7, 72, 73–74, 92, 95, 99, 101
Kirsten, E. B., 198
knowledge or knowing, 106, 107, 109–110
Kuhl, Patricia, 69

L
lag schedules of reinforcement, 29, 40, 44
language: dead vs. alive, 95; derived relational responding and, 175; human behavior and, 95–105;
Kantor’s study of, 73–75; operant behavior and, 182; ordinary vs. scientific, 94; psychological use
of, 101; referential nature of, 100; relational framing and, 195; RFT account of, 183; scientific vs.
ordinary, 101; Skinnerian account of, 182–183; social human practice and, 94–95; understanding
related to, 97–98; verbal behavior distinguished from, 57. See also verbal behavior
Learn to Reason with TAPS (Robbins), 148
learning: behavioral, 107, 108–109; errorless, 16, 44; incidental, 61–62; instructional, 106, 112;
observational, 115–130; one-trial, 117; types of, 109–110; vocal, 68
Lewon, Matthew, 6
linear series (LS) training structure, 162, 163, 304
linguistic behavior, 95–96, 97, 98
linguistic response systems, 96–97
listener training, 30
logical model, 101, 103
longitudinal studies, 195
Luciano, Carmen, 281

M
maladaptive behaviors, 224, 227–228
mand modality, 25
mands, 21, 25–29; definition of, 25; derived, 33–34; instruction using, 27–29; selection-based
systems of, 25–26; topography-based systems of, 26–27
many-to-one (MTO) training structure, 162–163, 304
Marler, Peter, 68
Mastering Reading Through Reasoning (Whimbey), 147
match-to-sample (MTS), 304; procedures, 158, 161–162, 168; tasks, 124–125, 126, 127, 168;
training and testing, 201
McEnteggart, Ciara, 264
McLoughlin, Shane, 198
measurement categories, 102
mediating responses, 46
mediation, concept of, 103
mediative referential interactions, 76–77
mental terms and expressions, 99, 100
mentalism, 261
metacognitive therapy, 254–255
metamorphosis, behavioral, 65–69
metaphorical reasoning, 242–244
mid-level terms, 257
mind processes, 100
Ming, Siri, 198
Molina-Cobos, Francisco, 281
Morningside Academy, 132, 149
Moschella, Jennifer Lee, 55
motivating operations (MOs), 16–17, 18, 25, 304
Mulhern, Teresa, 198
multiple exemplar instruction (MEI), 30, 32–33, 40, 61, 127, 304–305
multiple exemplar training (MET), 202, 305
multiple tact instruction, 46–47
mutual entailment, 178–179

N
naming: bidirectional, 61–62, 63–65; unidirectional, 61
narrative referential interactions, 76
national identity bias, 267–268
negative punishment, 11, 12
negative reinforcement, 10–11
neutral stimulus, 8
nonarbitrary relational responding, 175
nonreferential behavior, 76
novel behavior, 134, 138, 139, 152
 O
Objective Psychology of Grammar, An (Kantor), 74
observational conditioning, 59–60
observational learning (OL), 115–130, 305; assessment of, 120–121; Bandura’s theory of, 116–117;
behavior analysis and, 117–124; conceptual analysis of, 116–120; derived stimulus relations and,
124–129; developing repertoires of, 121–123; emergence of conditioned reinforcers through, 123–
124; explanatory overview of, 115–116; extended implications of, 128; future directions in, 129;
prerequisite skills for, 121; research studies on, 120–129; stimulus equivalence and, 124–125
observing response, 159
On the Origin of Species (Darwin), 66
one-to-many (OTM) training structure, 162–163, 305
one-trial learning, 117
operant conditioning, 6–7, 8–9, 18
operant variability, 13
operational categories, 102, 104
opposition, frame of, 188–190
overshadowing, 16

P
Pavlov, Ivan, 6
PEAK Relational Training System, 170, 194–195, 305
peer-mediated interventions (PMIs), 40, 48–49, 305
performative knowing, 109
perspective taking (PT), 283–290; definition and description of, 283; as deictic framing, 285–290;
evaluating skill in, 182; functional approach to, 284–285; intellectual functioning and, 194; related
to empathy and compassion, 282–283, 298; speaker-as-own-listener and, 51; traditional approach
to, 284
Picture Exchange Communication System (PECS), 25, 305
Pio, Savannah, 214
pliance, repertoire of, 220, 228
Pohl, Peter, 55
positive punishment, 11–12
positive reinforcement, 10–11
pragmatic verbal analysis, 240–244, 245, 247
preceding mediative interactions, 77
precision teaching, 40
precurrent behavior, 146
prejudicial behaviors, 264–265
prerequisite skills: for observational learning, 121; for rule-governed behavior, 229; for teaching
intraverbals, 43
present moment awareness, 258
Principles of Psychology (Kantor), 73
private events: ACT processes and, 257, 258, 259; behavior analysis and, 251, 261; third-wave
behavior therapies and, 256
problem solving, 234–249; behavioral studies on, 236–240; conceptual approaches to, 235–236;
examples of teaching, 237–240; explanatory overview of, 234–235; future directions for studying,
245–246; practice guidelines for, 246–248; pragmatic verbal analysis for, 240–244, 245; qualities
required for, 148–149; relevance of rule deriving for, 244–245; RFT perspective on, 240–244;
strategic vs. valuative, 241–242; thinking aloud for, 148
Problem Solving and Comprehension (Whimbey & Lochhead), 148
Promoting the Emergence of Advanced Knowledge (PEAK), 170, 194–195, 305
prompt delay, 44
proprioceptive feedback, 247
psychological behavior: explanation of, 98–100; field model for analyzing, 101–102
psychological flexibility, 256, 257, 258, 259, 261, 289
psychological inflexibility, 256, 258, 259
Psychological Linguistics (Kantor), 74, 75
psychological suffering, 260
psychological terms, 94
psychopathology, 251–252, 256
public health issues, 224
punishment, 9, 11–12

R
Rachal, D. Owen, 250
racial bias, 269, 273, 277, 278
radical behaviorism, 5, 6, 251
random interval (RI) reinforcement schedule, 222, 305
random ratio (RR) reinforcement schedule, 222, 305
randomized controlled trials (RCTs), 195, 305
rate of responding, 9
reactive modes, 95–96
REC model, 272–275
recruitment, 135
referent, 75
referential behavior, 72–92; detailed analysis of, 75–77; development of, 87–90; explanatory
overview of, 72–73; future directions for studying, 90–92; interaction analysis and, 79–82;
ordinary language as, 100; origins of analysis of, 73–75; symbolizing behavior vs., 78–79;
teaching, 82–87; word analysis and, 77–78
referential interactions, 73, 76–77
reflexivity, 161, 170
Rehfeldt, Ruth Anne, 20, 38
reinforcement, 9, 10–11; automatic, 23; conditioned, 123–124; differential, 215, 225; lag schedules
of, 29, 40, 44; positive vs. negative, 10–11; random ratio and interval schedules of, 222; social
listener, 58–59
reinstatement, 13
relapse effects, 13
relating, behavior of, 177
relational coherence indicators (RCIs), 275, 306
relational elaboration and coherence (REC) model, 272–275, 306
relational evaluation procedure (REP), 202
relational frame families, 184–194; of comparison, 190–192; of coordination, 185–186; of deictic
responding, 193–194; of distinction, 186–188; of hierarchy, 192–193; of opposition, 188–190
relational frame theory (RFT), 4, 33–34, 174–196, 306; families of relations in, 184–194;
fundamental concepts in, 177–178, 198–199; future directions in, 194–196; hope provided by,
183–184; IRAP derived from, 265; origin of relational frames and, 174–177; problem solving
informed by, 235, 240–244; research providing foundation of, 255; rule-governed behavior and,
218
Relational Frame Theory (Hayes, Barnes-Holmes, & Roche), 183
relational frames: families of, 184–194; generalized operants and, 181–182; language emergence
and, 195; origin of, 174–177; properties of, 178–181
relational responding, 174–176, 177, 184
religious bias, 268
remembering talk, 100
renewal, 13
representational categories, 102
respondent behavior, 7
respondent conditioning, 6, 7–8, 18
respondent-type training, 126
response bias, 266–267
response classes, 9–10
response recovery, 13
resurgence, 13
reverse intraverbals, 47
RFT. See relational frame theory
RGB. See rule-governed behavior
Ribes-Iñesta, Emilio, 94
rigid rule following, 227–228, 231–232
Robbins, Joanne, 148
Romanes, John, 100
Rosales, Rocío, 20, 38, 115
rule deriving, 244–245
rule following: establishing repertoires of, 226–227; functional units of, 220–221
rule-governed behavior (RGB), 5, 214–232, 306; applied research on, 223–229; basic research on,
221–223; contemporary accounts of, 217–218; definition of rules and, 216; establishing
repertoires of, 229–230; explanatory overview of, 214–215; future directions for studying, 231–
232; gambling behavior and, 223–224; higher education and, 225; maladaptive or rigid, 227–228,
231–232; public health issues and, 224; RFT informed by research on, 255; rule following
repertoires and, 226–227; say-do correspondence and, 228–229; teaching rules for flexible, 230–
231; types of rule following in, 220–221
rules: behavior intervention plans and, 225–226; contemporary accounts of, 217–218; deictic
framing and formation of, 290; describing behavior vs. contingencies, 222–223; dimensional
characteristics of, 219–220, 230; instructed vs. derived, 222; teaching for flexible RGB, 230–231;
technical definition of, 216

S
sample stimulus, 125
sample-as-node training structure, 162
Sandoz, Emily K., 250
say-do correspondence, 60, 228–229
schedule thinning, 44
schedules of reinforcement: lag, 29, 40, 44; random ratio and interval, 222
schooling. See education
scientific language, 101
scripted problem solving, 246–247
script-fading procedures, 29
S-delta, 14, 306
second-wave behavior therapies, 253–254
selection by consequences, 10
selection-based mand systems, 25–26
selection-based verbal behavior, 22
self-as-context, 259–260
self-awareness, 283, 284–285
self-compassion, 296, 297
self-determined learning model of instruction (SDLMI), 237–238, 306
self-talk, 60–61, 62–63
sensitivity to changes, 221–222
sequence of behaviors, 136, 138
setting conditions, 75, 76
sexuality-related biases, 270
Shapes-and-Colors IRAP, 274, 275–278
Shoenfeld, Nate, 74
Sidman, Murray, 158, 174, 217
silent dog method, 245
simple stimulus control, 41
simple verbal discrimination, 41–42
simple-to-complex (STC) protocol, 164, 165, 306
simultaneous MTS procedures, 161, 162
simultaneous (SIM) protocol, 163–164, 165, 306
single-trial-type-dominance-effect, 274
skill-acquisition programming, 35
Skinner, B. F.: criticisms of work of, 3–5; on death of behaviorism, 53–54; operant concept of, 1–2,
7; problem solving analysis of, 234, 235, 240, 244; radical behaviorism of, 5, 6; role of private
events for, 251; on rule-governed behavior, 214, 215; on the self and self-knowledge, 283, 284–
285; verbal behavior analysis of, 21–34
smoker status biases, 271–272
SOAR (Stress on Analytical Reasoning) program, 148
social behavior: implicit cognition and, 264–280; verbal behavior as, 56–57
social categorization, 265, 266
social cognition. See implicit social cognition
social listener reinforcement (SLR), 58–59, 307
social research, 267–272; on age, 270–271; on body image, 271; on gender, 269–270; on national
identity, 267–268; on race, 269; on religion, 268; on sexuality, 270; on smoker status, 271–272
social validity, 91
solving problems. See problem solving
source of a rule, 219
speaker-as-own-listener, 51
speech perception, 67
speech-generating devices, 25
spontaneous recovery, 13
Standard Celeration Charts, 150
Stanley, Caleb R., 174
Stewart, Ian, 198
stimulus classes, 15
stimulus control, 14, 157
stimulus discrimination, 14
stimulus disparity, 15
stimulus equivalence, 4, 158; concept formation and, 158; contingency adduction and, 141–143;
equivalence-based instruction and, 158; observational learning and, 124–125; relational frames
and, 175; RFT informed by research on, 255; rule-governed behavior and, 217–218; trajectory of
research on, 158–159
stimulus generalization, 15–16, 157, 158, 247
stimulus pairing, 40
stimulus salience, 15
stimulus substitution, 120
stimulus-pairing observation (SPO), 126–128, 307
stimulus-pairing procedures, 40, 48, 126–128
stimulus-stimulus pairing (SSP), 24–25, 307
strategic problem solving, 241, 247
Street, Elizabeth M., 131
strengthening of behavior, 9
Stress on Analytical Reasoning (SOAR) program, 148
substitute stimulation, 5
symbolizing behavior, 77, 78–79
symmetry, 159
Szabo, Thomas G., 234

T
tact prompts, 44
tacts, 21, 29–30; benefits of repertoire of, 29; instruction using, 30, 46–47; intensive protocol using,
59
talk aloud problem solving (TAPS), 148, 307
talking to oneself, 60–61
Tarbox, Jonathan, 214
taxonomic categories, 102–103
technical language, 94
temporal components of rules, 219–220
textual behavior, 21, 49–50
theory of mind (ToM), 284, 292, 307
think aloud pair problem solving (TAPPS), 146, 148, 307
third-wave behavior therapies, 254–256
three-term contingency, 8
tool skills, 141, 151
topography-based mand systems, 26–27
topography-based verbal behavior, 22
tracking, repertoire of, 220, 228
transfer of stimulus control (ToSC), 28, 39, 41, 44, 307
transfer of stimulus functions, 167–168
transformation of contingencies, 103
transformation of stimulus function, 64, 180–181
transitive class containment, 206, 207, 208
transitive conditioned establishing operations (CMO-T), 28, 307
transitivity, 159

U
unconditioned MOs, 17
unconditioned response (UR), 7, 307
unconditioned stimulus (US), 7, 307
understanding, process of, 97–98
unidirectional naming (UiN), 61, 307
unilateral property induction, 207, 208
units of analysis problem, 1
untaught responding, 132
 V
valuative problem solving, 241
value-altering effect, 16
valued living, 257–258
verbal behavior, 3; empirical research on, 22–30; language distinguished from, 57; problem solving
based on, 240–241; relational frame theory and, 33–34; rules on development of, 68; selection- vs.
topography-based, 22; Skinner’s analysis of, 21–34, 182–183; as social behavior, 56–57; talking
to oneself as, 60–61. See also language
verbal behavior developmental theory (VBDT), 56, 308
Verbal Behavior (Skinner), 3, 38, 74, 132, 182
verbal conditional discrimination, 42–43, 308
verbal operants, 21–33; bidirectional, 56–70; complex, 39–51; echoics, 21, 23–25; functional
independence of, 30–33; intraverbals, 21, 40–49, 239; mands, 21, 25–29; tacts, 21, 29–30; textual
behavior, 21
video models, 238–239
Villatte, J. L., 290
vocal learning, 68

W
Watson, John B., 6
weakening of behavior, 9
West-Eberhard, Mary Jane, 66
Whimbey Analytical Skill Inventory, 146
Whimbey, Arthur, 145–148
Whitehead, Alfred North, 53
Wittgenstein, Ludwig, 95
words: concepts distinguished from, 98; historical development of, 97; problem with studying, 77–
78. See also language
writing, advent of, 97
 1 Author’s Note: The author would like to acknowledge Ainsley Lewon
and Vanessa Willmoth for their assistance in the preparation of this chapter.
2 Readers familiar with Skinner’s (1957) Verbal Behavior will recognize
the parallel between “manding” and mediative interactions.
3 From a behavior analytic perspective, these factors are not cognitive in
the sense that they are something other than behavior; rather, these factors are
conceptualized as additional behavior to be accounted for in learning from
observation.
4 Importantly, Fryling et al. (2011) also consider nonhuman animal
demonstrations of OL and note that nonhuman animals are unlikely to engage
in this sort of rule stating. In other words, the fact that nonhuman animals
learn from observation further suggests that we should be careful in our
consideration of verbal behavior in accounting for OL. The finding that those
who describe their environment may learn more from it should only suggest
that careful attention to what is observed greatly enhances OL.
5 There are a number of topics that may be related to developing
observing responses more broadly. For example, Keohane, Pereira Delgado,
and Greer (2009) described a number of protocols that facilitate the
development of observing responses.
6 The authors would like to thank Marta Leon, Director of Editorial and
Instructional Design for Learning A–Z for her data analysis and helpful
suggestions related to Headsprout Reading.
7 Paper read at the 50th anniversary Conference of the Graduate School
of Education, the University of Pennsylvania, 1965. Reprinted in Ferster,
Culbertson, and Boren (1975, p. 565). Extracted pages available from the
first author of this chapter, a student of Ferster’s in the 1970s.
8 Morningside Academy is a third level, scientifically driven service
organization (Johnston, 1996) consisting of a laboratory school and current
outreach to 130 schools and agencies throughout the US and Canada through
Morningside Teachers’ Academy. It is not a level 2 organization, such as a
university, that conducts controlled research in applied settings. “The priority
of applied practice is to deliver effective service, not to answer experimental
questions” (Johnston, 1996). A comprehensive science is strongest when it
integrates basic research and applied research and modifies applied research
procedures to develop successful protocols of service delivery.
 9 Once a behavior has been recruited and reinforced, it becomes part of
the behaver’s history; thus, there is only one moment of contingency
adduction related to a particular S-R relation, and there is only one moment
of contingency adduction related to a particular behavior.
10 We are eager to hear from others who can tell us how they have
engineered other procedures.
11 After two failed contingency adduction probes, Headsprout Early
Reading provides direct instruction in consonant blends.
12 In precision teaching, the original meaning of “application,” the “A”
in the acronyms for fluency such as RAPS, REAPS, and RESSA, was clearly
about component skills facilitating composite behavior. The “A” has nothing
to do with performing a skill in a new context.
13 We should note that many investigations have demonstrated that
building component frequencies accelerates composite skill instruction and
performance with children and adults with disabilities. The first author did
significant work in designing instruction to teach activities of daily living
skills such as handwashing and teeth brushing to adults with developmental
disabilities in an institution. Teaching focused on practicing tool skill
components of these activities, such as reaching, touching, grasping, placing,
releasing, pushing, pulling, shaking, squeezing, and tapping; and informally
demonstrating their combination to students. Neither task analysis nor
chaining procedures were necessary. Even sales and customer service staff
have learned to engage in lengthy problem-solving conversations with
customers by focusing on product knowledge facts (Binder & Bloom, 1989;
Binder & Sweeney, 2002). Many of these investigations have been published;
however, we cannot describe them here. Instead, we will stick to our goal of
describing generative responding in academic skill development.
14 Paper read at the 50th anniversary Conference of the Graduate
School of Education, the University of Pennsylvania, 1965. Reprinted in
Ferster et al. (1975, p. 567). Extracted pages available from the first author.
15 Author Note: A portion of Daniel Fienup’s effort was conducted
while affiliated with Queens College, CUNY.
16 These standard numbers refer to American Common Core
educational standards, which can be found at http://www.corestandards.org.
 17 The first two authors contributed equally to the chapter.
18 According to RFT, many of the functions of stimuli that we
encounter in the natural environment may appear to be relatively basic or
simple but have acquired those properties due, at least in part, to a history of
relational framing. Even a simple tendency to orient more strongly toward
one stimulus rather than another in your visual field may be based on
relational framing. Identifying the name of your hometown or city from a
random list of place names may occur more quickly or strongly because it
coordinates with other stimuli that control strong orienting functions (e.g., the
many highly familiar stimuli that constitute your hometown). Such functions
may be defined as Cfunc properties because they are examples of specific
stimulus functions (i.e., orienting) that are acquired based on, but are separate
from, the entailed relations among the relevant stimuli; the latter are labeled
Crel properties (see Finn et al., 2018).
19 Figure 16.1 adapted from D. Barnes-Holmes, Finn, McEnteggart, &
Barnes-Holmes [2017]. Adapted by permission of Springer Nature.
 from

Your gateway to knowledge and culture. Accessible for everyone.

z-library.se singlelogin.re go-to-zlibrary.se single-login.ru

O cial Telegram channel

Z-Access

https://wikipedia.org/wiki/Z-Library
ffi