OM METHODS, ONTOLOGIES, AND REPRESENTATION
IN THE SOCIOLOGY OF SCIENCE:
WHERE DO WE STAND? ‘
We are weak today in ideal matters because intelligence is diverced from
aspiration. The bare force of.crcumstance compels us onwards in the daily
+ detait of our beliefs and acts, but our deeper thoughts and desires tun
backwards. When philosophy shall have co-operated with the cpurse of
‘events and made clear and coherent the meaning of the daily detail, science
ang emotion will interpenetrate, practice and imagination will embrace
Poetry and religious feeling wil be the unforced flowers of life
John Dewey, Reconstruction in Philosophy, 1920:212~13"
This is a time ‘of ferment, struggle, and change in the social sciences.
Various versions of postpositivism are making waves in the academy. In
anthropology, “postmodernist” agendas are rejecting earlier colonialist
views and representations of the “other” and experimenting with new forms
Of representation. to-feminist studies, Anglo-American “empiricist/pragma-
tistprogressivist” feminists debate French skepticistidealisUradical” fem.
nists (Draine 1989). th sociology, symbolic interactionists-are in dialogue
with “postmodernist” anthropologists and literary theorists about the possi-
bilities of social scienet, “while ethnomethodologists continue their fine.
grained analyses of everyday practices as ordinary people construct theip
‘own analyses of their society and its workings. Old boundavies Between
disciplines are breaking down in certain areas, and new synergies are being
constructed. :
In social studies of science, the name given to the kind of research I do, a
similar deconstruction and reanalysis is uriderway. Reflexivity, textual anal-
ysis, ethnotnethddology, and constructivist approaches have introduced
similat critiques of positivist analyses of:science and have raised concems
about how one should write or “tell” about science. These new perspec-
tives, in turn, have been criticized not only by realist Sociologists and
historians of science, but also, by some relativist sociologists .of science.
207 |
In David Maines (ed,), Social Organization and Social Process: Essays in Honor of Anselm L.
Strauss. Hawthorne, NY: Aldine de Gruyter, 1991, pp. 207-248.208 Joan H. Fujimura
In this paper, ! discuss several new perspectives in the Sociolégy of sci:
ence and. the relevance of Anselm Strauss’s work to understanding and
‘analyzing, these perspectives. The major protagonists in this discussion are
Michel Callon and Bruno Latour, with their actor-network approach, and
Sieve Woolgar, with his reflexivity program. They argue against both the
Zsetrong programme” tradition of sociology of science as well as the “rela-
tivist programme” of Harry Collins, finding both to be asymmetric in their
analysis of science and society.? They argue that those who do social studies
Gi science should be as critical of their own constructions as they are of
‘scientists’ constructions. In their view, the basic problem is whether ahd
how one should explain or tell about science without being positivist about
science oF society. | also discuss the debates on this issue between post-
modernist anthropologists and their feminist critics.
‘Why discuss this most recent “reflexive turn’ in a festschrift paper for
‘Anselm Strauss? Anselm, who has been both. friend and teacher to me,
provides powerful tools for doing sociology while preserving the integrity of
those we write about. These tools are his writings on identity, perspectives,
social worlds, and negotiated order, and his and Barney Glaser’s grounded-
theory methods. 'Our recent studies of sciencé have used Strauss’s work to
portray how scientific facts, theories, artifacts, rules, and procedures are
negotiated orders. A scientific representation or.procedure is itself a negoti-
ied order, that is, the organization of scientific work commitments and
perspectives at a particular time and place, Constructing scientific facts,
theories, artifacts, and procedures are temporally located collective pro-
cesses. ‘
In my view, the next step is to bring more and different, perspectives into
the negotiations. | suggest that Anselm Strauss's work on identity, social
worlds, negotiated order, and grounded theory allows us to retain. the
complexities of situations and provides ways to include new and diverse
perspectives in our studies of science- and society-in-the-making. Stfauss's
Work uses the tension between indeterminacy and structure to highlight
process, change, interaction, collision, and the constant presence of multi-
ble and different perspectives. At the same time, this cacophony of often
Contradictory voices is tied to specific situations, as viewed by the partici-
pants themselves. With respect to identity, there is also no simple one-to-
She mapping between perspective and community membership, a principle
Wwe can call the indeterminacy of perspectives. Thus, Strauss's work provides
the space for framing questions about the pursuit of social change, while his
sociology—with its emphasis on complexity, novelty, and the development
‘of new categories and coricepts—is itself open to change and is not lirhited
to traditional sociological terms.
Inva gesellschaft society, we find the ground beneath us shifting as we
stand on it. Since one’s position and frame of reference can change quicklyThe Sociology of Science: Where Do We Stand? 209
and frequently, we find ourselves struging to [e}construct ourselves as we
{relconstruct the other. Donna Haraway (1988, p. 578) self-mockingly calls
this “selF-induced multiple personality disorder,” Anselm would re oe the
gatement to read “‘selfisocialy-induced multiple personality disorder”
Strauss’s answer to the angst of modemity and gesellschatt society is to
emphasize the possibilities for novelty and “new values’:
es brecisel this continual necessityfor reassessment that permits the innovae
Shosne onetY of human ie, expectations were.atways failed if presece
nee eat evonts were exactly as anticipated from experiences with past
Snes-—then action would be thoroughly ritualistic, and eonceplions etemally
fined signation, in fact, rests upon ambiguous, confused, not wholly de,
net Stations. Out of ambiguity arises challenge and the discover’ ore
valtes: “lt is in the areas of ambiguity that transformations take place,
without such areas transformation would be impossible.” (1969 {1959],26)
Thus, Strauss sees the dynamics of the human condition as a melange of
angst, novelty, confusion, and opportunites as he moves us toward the
embrace of practice and imagination.
tn what follows, | fist lay out the problems and then, discuss Strauss
Conttibutions to the solutions.
Latour and Woolgar: From Laboratory Life to Reflexivity
and Associations
Studies in the sociology and history of science over the last two decades
have given us a variety of glimpses into how science is locally practiced ig
the production of knowledge. These studies‘show that there is no general
Jrethod that governs or describes. problem-solving efforts within any one
line of research, much less in a discipline or specialty” Instead, seiehting
link up heterogeneous elements in nonrational fashions." shor, science ie
done using many methods. Moreover, Science is done by-many people and
not just by “scientists.”
Some sociologists, histérians,
these studies to argue that,
je cannot compare scientif-
ic representations against each other, since each is a result of a different
method, a different set of negotiations among heterogeneous elements
Rather, as science analysts, we'should be agnostic in our teatment ofall eck,
entific representations,
Latour and Woolgar (1979) extended this argument to sociological repre-
Sentations in their early work on the construction of scientific facts.* They210 Joan H. Fujimura
ized tre strong programmers of the Edinburgh school of sociology of
fee Yor making scientific constructs problematic while simultaneously
king for 3ranted sociological constructs.” In their studies of the social
aereuctic. «of scientific knowledge, the strong programmers “explained”
Enguledge. claims dnd antfacs by pointing tothe social and cultural contin-
gencies uri der which they were produced, In thei arguments with svong
rogramnm='s, Latour and Woolgar (1986:281) contended that we should edit
Fhe word *” sociat” from ‘'soctal construction” (ct, Lynch 1982). “Society”
ipower,” “classes,” “interests,” “cultures,” nations,” “institutions,”
niprofessiors,” and other representations of the social are constructed by
sociologist > just as “black holes” are constructed by astronomers, “The
anthropole gist dobs not know the nature of the society under study, nor
wihere to ciraw the boundaries between the realms of technical, social,
vntfie, tural, and soon” (Latour and Woolgar 1986:278). This applica,
ton of syrmmetry to “science” and “society” has implications, they argued,
foc the methods and techniques adopted in the social studies of science:
‘The reviion of epistemological preconceptions about science raises awkward
Ghestiors aout the nature of social analysis. Can we go on being instrume
Sify teaist in our ove research practices while proclaiming the need to
sae yaify this tendency among natural scientists? Should we be vocal about
deve process of science, hitherto hidden from view, and yet silent about
the soc processes of our own research? The hesitant, ciferential response to
this deop-tooted isu partly accounts forthe proliferation of research perspec
ihis desich has*accompanied the feleaze (rom pre-Kubiian arthodoxy. Al
though penerally United in their disdain forthe tracitiona (received) view of
trowel Practitioners of the new social study of scientific knowledge difer
aereay in their methodological styles and preferences. (1986:275-76)
In their subsequent work, Latour and Woolgar have gone off in’ different
directions, but both continue to attempt to deal with this “problem of
fallibility,”/ which they define as “the argument that all forms of description,
report, observation and so on can always be undermined” —including our
ee (latour and Woalgar, 1986:282). One problem with this description of
fallibility is that it too can be undermined. | point out problems with this
description in a later section of this paper. First, however, | will discuss
Woolgar’s and Latour's recent work on the problem of fallibility.
Woolgar and Postmodernist Anthropologists on Reflexivity
Woolgar’s (1988a:94) response to the problem of falliblity has been to
argue that we should highlight it in our science studies work, Rather than
arBitinue to use the same ideology of representation in our studies of
science-making, he argues thatThe Sociology of Science: Where Do We Stand? an
We should accept the universal applicability of fallibilty Ithus, to our own
work as well] and find ways of coming to terms wth it. Instead of utlizing tin
2 merely critical role, the aim would be to retain and constantly draw atention
to the phenomenon ‘in the course of four} description and analysis. ... We
need to explore forms of literary expression whereby the monster can be
simultaneously kept at bay and allowed a position atthe heart of our enter=
prise, (Latour and Woolgar, 1986:263?
We should attempt new forms of representation that do not establish scien-
tists, scientific work, and laboratories as the “other’“the exoticithe object of
study and the social analyst as the observerisubject. “Reflexivity is thus a
‘way of reminding the reader that all texts are stories. This applies as much to
thesfacts of our scientists as to the fictions ‘through which’ we display their
work. The story-like quality of texts denotes the essential uncertainty of theit
interpretation: The reader can never ‘know for sure’ (Latour and Woolgar,
1986:284).*
In order to prevent our own representations from being read as truth
claims, Woolgar proposes that we write texts that deconstruct themselves, In
Knowledge and Reflexivity (1988b), he exemplifies this approach. A photo-
graph of the author writing a caption for the same photograph is decon-
structed in a following “reflexion” by a photograph of the text that includes
the earlier photograph. The caption for the second photograph states {he
dialogical inter-penetration of image.and caption.” Similarly, every state-
ment is followed by a second voice, the reflexive voice, which deconstructs
the original statement. “The idea is that this approach might modify existing
conventions and thereby provide new ways of interrogating representation.
The notion of interrogating representation contrasts with the aims of either
explaining representation . . . ot trying to escape it” Woolgar (1988a:95,
emphasis in original). While such reflexive ethnographies might not help us
to escape “the hegemony of scientific discourse,” of objectivism and real-
ism, they “may serve to keep alive the kinds of inversions suggested by the.
constitutive argument."" The ultimate goal, however, is to “develop a per-
spective which begins to provide adequate and effective resistance to the
rhetoric of realism without slipping back into realistic rhetoric in the course
of our own ‘tesearch.’ ” In this effort, Woolgar (1988a:109) calls upon us to
find ways of interrogating Selt [the analyst/observet, “the disregarded agent
of representation.”
| agree with Woolgar and his colleagues‘that the various sociologies could
do well with some “self-interrogation.” That is, theyhve could do well by
taking the analyst seriously as an actor. Indeed, in many qualitative research
approaches in:sociology, the reseatchers/writers are assumed to “make a
difference” (e.g., Mead 1938; Emerson 1983) in the situation and are
therefore actors in the scene. Most of us who have been "in the field” also
understand that each of us perceives the situation in different ways. Each212 Joan H. Fujimura
“ethnographer” might hear and sée (and smell, taste, touch) different things
and in different ways. Yet, many of us offen ignore our analyst-actor pres-
ence in the situation when we write our texts. This is exactly the point
postmodernist anthropologies and literary critics have been making for the
past several years.
With this point on self-interrogation taken, | want to discuss some con-
cers, questions, and comments about the reflexive program. How far are
‘we to take our self-interrogation and in which directions? That is, self
interrogation does not eliminate the choices and decisions made in con-
structing texts, On what grounds do we base these selections? If reflexivity is
a never-ending spiral (upward or downward or outward), do we ever get off?
If so, when do we get off and for which reasons? Woolgar is still struggling,
with these questions, and | await his answers. (I discuss my attempts to
answer these questions below.)
Itis clear that Woolgar intends to get off the spiral. His stated ultimate goal
Lis to “develop a perspective which begins to provide adequate and
effective resistance to” the rhetoric of realism without slipping back into
realistic rhetoric in the course of our own ‘research.’ (1988a:94-95)
However, reaching such a goal, developing such a perspective, { would
argue, is to reach a metadiscourse, which will be above representing itself as
a truth claim, That is, Woolgar wants to write narratives that will be read as
(reflexive) stories and not as truth claims.
Woolgaris attempting a kind of radical politics with this reflexive move. It
is a kind of second-order politics where he wants to make clear to the reader
that the text speaks for the author and not for the “other,” the object. As |
stated earlier, some postmodernist (or poststructuralist) anthropologists like
those represented in Clifford and Marcus (1986) and Marcus and Fischer
(1986) have also attempted to construct reflexive representations of other
cultures, to deconstruct representations, or to analyze representationat
forms—or, in Rabinow’s (1986) critique of Clifford's work, to construct a
representation of representations. Postmodernist anthropologists have at-
tempted new representational forms to signal to the reader the constructed,
invented, partial, and processual nature of the ethnographic text. “[AIII”
constructed truths are made possible by powerful ‘lies’ of exclusion and
rhetoric. Even the best ethnographic texts—serious, true fictions—are sys-
tems, or economies, of truth, Power and history work through them, in ways
their authors cannot fully control” (Clifford 1986:7).
Postmodernist anthropological efforts arose during the decolonialist, poste
war period—especially the radical and optimistic 1960s and 1970s—when
the dominance of the “Western world” was being challenged. These an-
thropologists pointed out the partisan: views presented by.earlier anthro-
pological texts in a world dominated by Western imperialism. WesternThe Sociology of Science: Where Do We Stand? 213
anthropologists not only presented biased perspectives of non-Western cul-
es they also colluded in the denigration, degradation, and deitruction of
these other cultures,"
In their effort to avoid the predator role, some anthropologists began to
Suestion and change their conventions of writing about other cultures. Their
efforts were situated among a number af other “‘postmodem’” efforts to
theorize about the possibilities and limits of representation. The concern
focused “less [on] how to speak well than fon] how to speak af all, and to
actmeaningfully, in the world of public cultural symbols” without misrepre-
Senting and therefore dominating the ‘other’ "(Clifford 198¢:11),
For some anthropologists, the patial answer to this question wae toclarify
the relations of production of the ethnographic tent “(The general trend
was] toward a specification of discourses in ethnography: who speaks? who
itera ten and where? with or to whom? under what insttstonal and
historical constraints?” (Clifford 1986:13). Their new anthropological style
aimed to make clear that the descriptions of other cultures were entirely the
Products of peoplé of @ particular culture and of a particular anthropologist.
These were not “true” representations of those cultures, bur merely repre-
sentations dependent ori the sifuations of the authors, Instead, truth and
relational, the outcome of discourses among subjects"
| agree with ard support many of the sentiments, goals, and efforts of
these “postmodernist” anthropologists, However, as Woolgar (1988a:100)
himself notes, postmodemist anthropological efforts have nt met with
Inuch success in dealing with the problem of falibilty, that, eth Ge effort
fo keep'“the monster... at bay... (while) simultaneously allowling it] a
Peete ae heart of our enterprise” (Latour and Woolgar 1986.285F The
Sharpest criticism is that postmodernist anthropology netlect te on repre.
barn ann teover, while postmodernist anthropology might have keen
boin of anticolonialist sentiments, itis presently located in different socio-
historical situ
colonial times and since the 1960s).!"
Rabinow uses Frederic Jameson's work to make an important Point about
the source of this lack of historical location in postmodemn culture, Ac-
cording to Rabinow (1986:252), Jameson argues that "the ost-modernist is
blind to her own situation and situatedness because, qua post-modemist,
she fs committed to a doctrine of partiality and flux for which a such
things as one’s own situation are so unstable, so without identity, that they214 Joan H. Fujimura
canr ot serve as objects of sustained reflection. Postmodernist pastiche is
both a critical position and a dimension of our.contemporary world.” This
poin ¢ has many ramifications, to which | will return. ,
Re binow (1986:252~53) then places postmodernist anthropology, and the
new ethnographic writing in the academy and argues that contemporaiy
proc iamations of anticolonialism are political moves within the academic
com “nunity. In the academy of the 1980s, proclamations of anticolonialism
are ‘basically beside the point." Instead, says Rabinow, “an anthropology
of arithropology” is needed. For example, an examination of postmodern
dialogics within the context of the “less glamorous, if more immediately
conssraining, conditions” of tenure and reputation-making would be of
cons derabie value. Do “longer, dispersive, multi-authored texts" yield
tenure? If not, who can afford to. write them? How does “corridor talk”
about fieldwork experiences make or break reputations and careers?
Rabiriow (1986:253-54) calls for an analysis of these conditions of prod-
uction:”
‘My wager is that looking at the conditions under which people are hired, given
tervure, published, awarded grants, and feted would repay the effort, How has
the “deconstructionist” wave differed from the other major trend in the
academy in the past decacie—feminism? How are careers made now? How are
careers destroyed now? What are the'boundaties of taste? Who established
and who enforces these civilities? Whatever else we know, we certainly know:
that the material conditions under which the textual movement has flourished
must include the university, its micropolitics, its trends. We know that this
level of power relations exists, affects us, influences our themes, forms, _
contents, audiences. We owe these issues attention—if only to establish their
relative weight. (pp. 253-54)
In a similar but harsher vein, feminists Hartsock (1987) and Mascia-Lees
and colleagues (1989:15) interpret two coincident events by attributing
negative motives to the postmodernist anthropologists’ reflexive turn:
{Hartsock] finds it 2urious that the postmodern claim that verbal constructs do
‘not correspond in a direct way to reality has arisen precisely when women and
non-Western peoples have begun to speak for themselves and, indeed, to
speak about global systems of power differentials. In fact, Hartsock suggests
thatthe postmodem view that truth and knowledge are contingent and multi.
ple may be seen to act as a truth claim itself, a claim that undermines the
ontological status of the subject at the very time when women tind non
Wester peoples have begun to claim themselves as subject. (Mascia-Lees et
al. 1989:15) .
Mascia-Lees and colleagues are also deeply concerned about who is
excluded from the discourse when postmodernist anthropological writing
becomes more arcane (although they are not always specific enough aboutThe Sociology of Science: Where Do We Stand? 215
hographies more obscure and, thus, dificult for anyone bor highly trained
Piialists to dispute” ip. 10). (Similarly, in thei crtigue ef Woolgar,
Collins and Yearley (1991) restate Trevor Pinch’s (1988) argument that “in
the hands of less subtle authors, multivocalt is merely» ernicious way of
imposing authority: by making it seem as though everyone he hace allowed
2.39; utile the author retains control over the voices of oprenetis “
Thomas and Thomas (1926:572) pointed out that “[f Ipeople} define
Fihations 2s real, they are real in their consequences.” While Twp that
Hartsock and Mascia-Lees and colleagues are too harch in attributing nega-
ty Personalipolitical motives to postmodernist writers, we con analyze
their tesponses a5 one set of contemporary consequences ef fan reflexive
anthrone etroPclogy.* Regardless of the politically radical nection ve
iranropology’s role in colonialism as the initial impetus for the ware dia-
logics, the contemporary sitiation is diferent. While the nog dialogicians
sey agaist forms of representation‘ the “other” that have bon oa
aed with colonialist anthopology, ather groups of “akere oe self.
eonsciusly choosing to use these same forms to advance their agendas,
rane Get of actors on the contemporary scene are feminists who fear wre the
ew dialogics wil determine the legitimated form of tes fo the future, thus
forcing them to write in ways they think undermine the authority of their
Should eee ensigns. Whether 1 agree with them or not, their reactions
should not be discounted or ignored. Textual innovators might just listen to
this audience and start a dialogue with them,
One reason forthe strong negative response of writers like Hartsock and
Mascig-Lees is likely the lack of such dialogue, the exclusion vt feminist
GuinoBrephies from texts lke Clifiord and Marcus's volume chive essen
Lally states that he and George Marcus had determined the agenda of the
advanced seminar and therefore of their volume. The Concerns, questions,
Gnd issues represented in the feminist anthropologiealliteature oa not fit
their agenda and were therefore excluded. Clifford acknowledges this omis-
sion: :
In the case of our seminar and volume, by stréssing textual form and by
privileging textual theory, we focused the topic in ways that excluded certain
forms of ethnographic innovation. This fact emerged in the seminar discus-
sions, during which it became clear that concrete institutional forces—tenure
Patterns, Canons, the influence of disciplinary author!
‘of power—could not be evaded. From this perspective,ne Joan H. Fujimura
their g ceatest impact. Clearly our sharp separation of form from content—and
aur fetishizing of form—yas, and is, contestable. (1986:21),
Ciiford then excuses their exclusion by claiming that the seminar partici-
pan ts we re not “truly postmodern” yet that theirs was a “modernist” failure
Yet | wonder whether, for some, the postmodernist preoccupation with
text..a/ presentation will blind them t6 content. When does one qualify as
“truly” postmodern? I return to Rabinow’s point: can.posimademist tn
logics be separated from their contexts of production in the academy? think
not. i
With reference to this last question, Rabinow’s (1986) discussion of
anth ropological feminism helps to complexify the question if not answer ft
Clifford argues that feminist ethnography has focused on setting the record,
straight about women or on revising anthropological categories (for exam
ple, che natureculture opposition). Ithas not produced either unconvention,
al fo-ms of writing ot a developed reflection on ethnogranhic textuality so
such ” (1986:21). Strathern (quoted in Rabinow 1986:255~6, see also Strat-
bern 19872,b) argues that the reason for this lies in different goals: “Perot,
nism, for Strathern, proceeds from the initial and unassimilable fact of
domination.” Anthropological feminists, argues Strathern, arent interested
in advancing the discipline but in shitting the discourse. Anthropological
femiriism, So
“alters the nature of the audience, the range of readership and the kinds of
interactions between author and reader, and alters the Zubject matter o
corwersation in the way it allows others to speak—what is talked abou arch
whom one stalking o.” trathéa is hot seeking invent a new synthe, a
to strngihen diference.:.!Stratheen’s anthropological feminist insists von
not losing sight of fundamental differences, power telationships, hieracheel
dornination. She seeks to articulate a communal identity onthe bos of
conflict, separation, and antagonism: partially as a defense against ihe thee
of encompassment by a paradigm of love, mutuality, and understanding i
which she sees other motives and structures; partially as a device to preseg
‘meaningful difference per se asa distinctive value. (Rabinow 1986-258. 40)
Similarly, Haraway, while in dialogue with constructivist studies of sci
ence as well as with deconstruction and semiology, chooses to construct »
feminist theory of situated knowledges that is based on what she calls “the
privilege of partial perspective”: : :
We seek knowledges} ruled by panial sight and limited voice-—not pitiality
for its own sake but, rather, for the sake of connections anid uneepectey
‘openings situated knowledges make possible. situated knowledges mre abera
Communities, not about isolated individuals. The only way to find a lator
vision isto be somewhere in pancular. The science question in fonsnig eeThe Sociology of Science: Where Do We Stand? 27
ts images are not the products of
paral pansrendence of limit (the viel from above) borates joining
if partial views and halting voices into’ collarons subject position that
Finite send i800 ofthe means of ongoing finite embodine sy living within
wines pnd Comadctions—of views ftom somewhen (1988:590 fermphasis
Again, whether we agree with Strathern and Haraway or not, we have to
Goanize that not all writers have the same goals as ostmoclemist dialogi-
seer agtlexive writers. How we write depends on wharnn want to
accomplish, 4
‘My point here is that not all writers choose to Separate their work fam
social moorings. Some might choose to write in fornia that'resemble un-self-
Conscious represeritation but that are quite self-consciously constructed,
Writers rhight have different purposes and differenr audiences at different
times. tt seems that texts (and seminars) like Clifford and Marcus's appear to
be promotiag only one means, that of ‘unconventional, innovative forms of
writing text. Why should textual innovation be the primary criterion for,
inclusion in academic discourses?
Zo bring the discussion back to Woolgar as Promised, Woolgar’s view of
jhe reflexive text asthe best form of representation ig Consequence of his’
framing of the problem of fllibilty, defined as “tie argtiment that all forms
of description, report, observation and so on can always be undermined”
(Latour and Woolgar 1986:282). This framing of te problem -however, has
Probleris of its own. ‘the demonstration that a clan arises as part of a
Socially organized praxis does'nat implicate ts corecinon or fallibility—it
simply situates the claim. f Woolgar takes seriously the viéw that knowl.
edge is contingent and multiple, then-why does he interpret the situating of a
Claim as a discounting of it? Pragmatist philosophers” argued long ago that
truth and knowledge are contingent and multiple and thar all representations
Personal'experierice, which is the temporal infersection f
tional, national, and international actions and events
Tres bragmatsts have also argued, however, that there oc other alterna-
tive. Different perspectives ate boin of different experiences, Th
bith and knowledge are relative’ matters, malleable and ‘dyna
claims—and claims not to be trth—are realized) in shen Consequences, to
Paraphrase Thomas and Thomas (1928),
Moreover, if we grant that textual innovation is vital, we should also be’
reflexive about it. would be interesting to ask, for example, who gets to
choose what qualifies as “good text worthy of inclusion and attention? That
is: H the signifier is freed from a concert with fs tletion wae external218 Joan H. Fujimura
referent, it does not float free of any referentiality at all; rather its referent
becomes gther texts, other images” (Rabinow 1986:250). We need then to
look at how other texts and images become referents.
Again, let me stress that | support new, experimental forms of writing. In_
their attempts to beat back the positivist visions of colonial days, the new
ethnographies have also provided a means for new kinds of stories to be
told. | agree that positivist, linear forms of writing constrain the presentation
of multiple voices and contingencies. agree with the efforts to invent new
forms of representation and to question un-self-conscious representation. A
fine example of an experimental ethnography is Dorinne Kondo's (1990)
Crafting Selves. Kondo is a Japanese-American anthropologist who studied
small, family-owned company workers in a “working-class” part of Tokyo.
{mn her first chapter especially, she provides an exquisite example of present.
ing herself as the author of the text. “Herself” and “author’ here are the key
terms. Kondo manages to convey her multiple selves to us as they were,
Gonstructed in interaction with her coworkers.in two small companies and,
in the family wha,took her into their home on her arrival in Japan. Her text
lets this reader “feel” the changes she experiences as she interacted with
others at work and at home. It shows how she changed them and how they
changed her, of how situations developed as a result of these interactions.
She makes clear her presence and participation in the situation. There is no
stranger looking on here. She is a participant in more ways than traditional
anthropologists usually acknowledge, and:her writing makes this perfectly
clear. Yet, Konco also manages :o tell a fascinating story about gender,
power, class, and the construction of multiple selves.
In my view, the next step in the social study of science is to examine this
malleability and dynamism of truth matters by studying the processes of its
production in different situations and from different viewpoints. While we
will never be able fully to understand and represent the views of the other,
this does not mean that we should not even attempt to “explain” science.
‘Woolgar argues that we should refrain from constructing explanations of
science or society in order to avoid naive realism. But there are explanations
and there are “explanations.” While | do not disagree with Woolgar on
fundamental assumptions about the nature of knowledge or representations,
| strongly disagree with him about the possibility of representing knowledge
construction and its consequences,
Where Actor-Networks Meet and Diverge from Pragmatist-Symbolic
Interactionist Sociology of Science
Latour (1987, 1988a,b,c, 1989), takes a path Somewhat, different from
Woolgar's. {While | primarily use Latour’s name in my text since | haveThe Sociology of Science: Where Do We Stand? 219
relied on more of his recent texts, the actor-network approach (or the -
science of associations) was constructed in collaboration with his colleagues
Michel Callon (1986, 1987; see also Callon and Latour 1981, 1991) and
John Law (1986; Law and Callon 1988). Latour argues that we can eK
plain” how science is constructed and insists that we should produce more
interesting, lively, rich, dense, and suggestive stories about this in order to
convince our readers to believe us. For Latour, the problem is not tha
readers believe our texts, but that they do not believe. Here he argues thal
Woolgar is wrong to assume that readers are naive enough to believe our
stories.
In order to construct dense, rich, detailed, more credible stories, Latourl
(1988b:148) argues that “we give voice to those whose support is necessary
to us." Here the support of the “voices” of microbes and the social hygien-
ists (among others) is necessary, argues Latour, for the understanding of
Pasteur's “scientific” success. However, when we “listen’” to these voices,
Latour also suggests that we be as agnostic about society as we are about
nature and instead examine the coproduction of society and nature. Only
when we drop our “sociological” understandings will we clearly “hear”
these voices. When studying Pasteur, for example, he argues that we should
use Pasteur’s “theory of society,”” which includes agents and agencies that
no sociologist would a priori imagine to be part of the social and historical .
situation.
Latour’s method is to use three sets of journals and histories of medicine
and France to “follow” Pasteur as he goes about constructing strong associa-
tions between heterogeneous elements to build strong (or weak) networks,
which are the makings of both the Sciences and society,, He argues that
Pasteur boldly linked together microbes, the hygienist movement, army
dsictors, farmers, surgeons, brewers—to define and redefine microbes, dis-
ease, fermentation, the practice of army doctors, French military success,
hygienic measures used throughout, France, theology, and French societ
Post-Pasteurian France was entirely different from pre-Pasteurian France.
According to Latour, Pasteur’s genius was located not in his microbiology
but in his strategy of movernent, in his linking of scattered things, in his
“placing his weak forces in all the places where immense social movernents
showed passionate interest in a problem” (1988b:70-71). At the same time,
he cut the link between his allies and his microbe and claimed that his
success came from fundamental research in his laboratory, that is from
“science.” Thus, Pasteur recruited allies in these social movements and
then negated their role in his success. Again,, the important lesson fot
sociologists of science, argues Latour, is that we.too must not assume that
we know what society is and is not and that we should study instead how,
society is constructed by” actors: 720 Joan H. Fujimura
Social studies of science are learning what society is from scientific practices,
‘not teaching scientists what really isthe social structure inside which they are
the unaware passive "dopes. . ..” {cf. Garfinkel (1967) on the “cultural
dope!'| No actor enters in a society without determining it itself, and does not
tundergo any influence or force without mapping out their nature, ontology,
circulation and limits, .. . [Wle have to learn from the actors our sociology,
history, chronology, theology, psychology and oftology. .. . Each new topic,
each new subject requires a diferent metalanguage, a diferent interpretative
toolbox. (1989:17—19)
Accrediting “voice” and actor status to nonhumans is alsb part of Callon
and Latour’s sociology of association. Microbes, scallops, batteries, doors,
acid rain, and other nonhumans are part of the (re)construction of society
and science. Microbés have reconstructed social and cultural praetices and
our “natural” landscapes and are now part of our everyday maps of the
world as we go about our lives. “Society is not only made up just of
{peoplel, for everywhere micrabes intervene and act. . . . We cannot re-
duce the action of the microbe to a sociological éxplatiation, since the
action of the microbe redefined not only society but also nature and the
whole caboodle (Latour 1988b:35-38)..
Callon and Latour’s inclusion of nonhumars as actors has raised some
controversy in science studies. For example,'Collins and Yearley (1991)
argue that Callon and Latour actually take the word of the technoscientists
about the activities of nonhuman actors and therefore are conveying double
agency to scientists by allowing them to speak as scientists and as nonhu-
mans. The conflict rises for two reasons. First, Callon and Latour's use of
the word “actor” conveys a more commonsense understanding of agency to
the nonhumans, which equals that of human agency. In one sense, they
appear to be giving the status of personhood to nonhumans. In fact, they are
leveling humans and nonhumans but only in their roles as “actants” rather
than “actors.’"* As actants, humans and nonhumans are equally enrolled
and disciplined—or resisting enrollment and discipline—for the purposes of
actors like Pasteur. Another way to speak of nonhumans is to expand on the
human perspectives on nonhuman actors. Who else attempts to construct
nonhumans as agents’or actants? How do these other actors construct and
attribute agency to nonhuman actors? Who cares about which nonhumans
and why (see, e.g., Clarke 1987; Fujimura 1991; Clarke and Fujimura
19927
However, speaking of nonhumans from the point of view of hiiman actors,
and their activities is precisely what Callon and Latour are attempting to
avoid. They are explicating a néw ontology, and this is the second source of
misunderstanding and cohiflict over nonhuman actors.” Using the sociology
of associations or actor-networks, they wish to explain not only science but
also the world. They do not want to stand in society and look aut at science.
Latour wants to explain the construction and reconstruction of science,The Sociology of Science: Where Do We Stand? 221
and te ORY, Society, polity, and history. As Latour argues, science
and fociety are the things to be. explained, not the exrlancoes, In this
redefined ontology, Boyle's pump and Pasteur’s microbes are actants in the
Feconstruction of society as are human actanty, Perhaps the clearest pre-
and Hobbes, the Church decided matters of truth and face After Boyle and
Hobbes, gentlemen scientists decided matters of natoray representation and
um of the social contract... The very divide
sahering us into the modem werd was made forthe) purpose Martin now
the represen rod "see double” and make no direct Connection benno
the representation of nonhumans and the representation of humans, between
the artificiality of the facts and the artificiality of the Body Politic. (Latour
1990, p. 147; emphasis in original) 4
redistribute “politics. . ., will, lability, Tespect, humanity, soul as well” to
nonhumans. Hobbes ani Boyle, as the two. founding fathers of this English
theo one econspite to make one and the same innovation in peltier}
theory: to science the representation of nonhumans and we possibility of
influence by or appeal to politic; to politics the représentaticn ot citizens
ith. influence by o: relation to the nonhumans produced and malvieny
by science and technology” (Latour 1990:159). “According to Latour, now222 Joan H. Fujimura
that we understand how this dual construction dichotomized our views of
the world, we should recognize that the modern world has never existed.
Therefore, we should return to a whole new (or old) ontology where things
and people are equal actors in the (re)construction of the world. Thus, wes
‘cannot use social context to explain scientific knowledge, since “neither of
them exist before Boyle and Hobbes achieve their respect goals and settle
their disputes.”
While | agree with Callon and Latour’s new (old) ontology, like the
pragmatists | follow, | take that ontology for granted. In the early part of this
century, pragmatist philosophers rejected the subject-object -or knowing-
known dichotomy, and symbolic-interactionist sociologists constructed
their sociology on that bass, n simplistic terms, they would argue similarly
that humans and nonhumans interactively (relconstruct the social and natu-
ral worlds, As sociologists, their focus was more on humans (and human
constructions) with nonhumans in the background, Callon and Latour
choose to keep humans and nonhumans equally in the foreground. Addi-
tionally, Latour wishes to change the situation and abolish the seventeenth-
century dichotomy. Here | would argue that he gives short shrift to the
consequences of-human action. However much we might disagree with
Boyle's and Hobbes’s seventeenth-century dichotomy, its consequences
exist in our contemporary worlds and cannot be wilisked away by a philo~
sophical argument., That is, while Callon and Latour might be philosophi-
cally correct about the constructed nature of the science-society dichotomy
(who represents nonhumans versus who represents humans), the conse-
quences of that construction are important. By now these consequences
include scientific institutions, scientific professional societies, scientific
journals, science granting agencies, academic career ladders and promo-
tions, scientific advisory panels, scientific “expertise,” scientific disciplines,
and more. These are consequences of designating and organizing science as
2 kind of work and enterprise different from, say, legal work. Where Callon
and Latour argue that we should study science and society as outcomes, |
argue that we should study science and society as"both constitutive of and
consequences of action. :
| want to examine the practices, activities, concerns, and trajectories ofall
the different participants—including nonhumans—in scientific work. In
contrast to Latour, | ain still sociologically interested it undérstanding why
and how some human perspectives win over others in the construction of
technologies and truths, why and how some human actors will go along
with the will of other actors, and,why and how some human actors resist
being enrolled. While | am clearly interested in nonhumans (especially
since technologies are human constructions), my interest is organized by the
himans who care about, fight over, and! commit resources to these things.The Sociology of Science: Where Do We Stand? 23
‘An additional difference between Latour’s agenda and mine is that he wants
to collapse dichotomies, while | want to construct concepts and theories
(my tools) to help some people win over others, again in the pragmatist
tradition. As an analyst acting—not just observing—in the world, 1 want to
take sides, to take stands. While Callon and Latour in their debate with the
Bath and Edinburgh schools emphasize that sociology is “no more scien-
tific,” | want to argue that our sociology is no less scientific than the science
of molecular biologists or computer engineers. Their negative framing dis-
empowers us in our debates with scientists. 1 still, want. to argue with
scientists that, for example, there is no reason for us to sequence the human
genome at this time. I still want to argue with the engineers that they need to
consult more than their computers when making decisions abaut what kind
of technology to produce. | still want to argue with soctobiologists that
gender roles are not determined by reproductive organs. And I want to win,
‘As Callon and Latour have so clearly described to us, winning requires all
‘kinds of machinations and manipulations. Most importantly, it requires that
\we-first-get people's attention and manage to get others to agree with our
arguments. Again, as Latour (19886) so eloquently tells Woolgar, the con-
cern is not that others will believe us but that they will not even listen to us.
‘Where postmodern anthropologists might have qualms about representing
other humans, | want to have at feast as much power to represent them as
does Walter Gilbert, a molecular biologist who wants to represent human
beings (e.g., “Joe. Blow") by putting each person’s entire genome (DNA)
sequence on a compact disk.
But here ! need to stop and explain what ! mean when | say that | accept
Callon and Latour’s ontology as taken for granted. In order to explain this, |
have to discuss the similarities between Latour's approach to science studies
‘and mine. | being with a comparison of Latour’s approach with that of my
teachers—pragmatist philosophers and symbolic-interactionist sociologists,
especially Strauss.
Latour’s “theory of associations’ is similar to early symbolic-interactionist
studies of actors trying to define their situations." These studies emphasize
change, diversities of goals and motives, the different perspectives held by
different actors (individual and collective), efforts to define (and redefine)
the situation on the part of every actor on the scene, coniflict among actor's,
definitions and perspectives, negotiation, and negotiated order. Everett
Hughes (1971), for example, wanted to understand how members of differ-
ent immigrant groups cooperated in the workplace while holding very
different definitions of the situation. He focused on the conflict, struggle,
and negotiations over which set of conventions should govern how mem-
bers of these different groups were to work together. Notice the similarities
between Hughes’ and Latour’s framin;224 Joan H. Fujimura
Actors producing thei own sociology, thir own meting resourses Ae
rot." is
Sore core the individual actor is only one of the possible
20k vn aAD( acon. (Al the actors ael engaged in making others play
senate eosin norms of behaviour. They are aso all engaged) in building,
roles and flict others will be forced ot coerced or seduced into Pog
ae ee expected toe. Pasteur invents not only 2 ype of inka with,
ya eyes an France, he also invents ways f0 insure that his own tdefini-
tions last. (1989, p. 25; emphasis in original)
similarly, Latour and Woolgar's views of scientific facts have much in
common with Strauss’s concept of negotiated order. Latout and Woolgar
(lov) portrayed "fact" as pockets Forder that last fora shor time in the
sor of laboratory activities. They can be overthrown af any 1 by the
nas Of disorder or bythe creation of new pockets of ord in the same or
aor gboratoy, Straus (et al. 1963, 1964; see also Maines 1977, 1982)
ate very similarly about order as established in organization’ his work
soenenotated ordes, on sentimental order and awareness of ons (Glaser
a eeeniss 1965, 1967), and on negotiations (strauss 19783) Briefly,
Sreanization as negotiated order is temporal and local” % constructed
through the negotiations among the actors and grouPs of actors on and off
the seene, Every instance, kind, and scale of order is achieved through.
rmmplex daily interactions conflicts, agreements, abdicatonsy force, ne-
foliations, manipulation, persuasion; education anong many actors
wrorking wnder specific focal conditions. Even routines are orders to be
eplaned and not assumed. | argue that we can use Strauss ‘work to show
Pant elentifc facts, theories, artifacts, rules, and procedures a negotiated
ents. A scientific representation or procedure i itself @ negotiated order,
Chat is the organization of scientific work commitments af 2 particular time
ry place. Constructing scientific facts, theovies, artifacts and procedures
are emporaly located collective processes. These approaches the study
ae aerymuch orders are negotiated and such situations defined have been
ord to study science by some of Strauss's Sudents.
ae ase co worked under Everett Hughes's assumption that under
different conditions, "it could have been otherwise": ‘
The pecullatty ofthe social scietisti hot that there i 53m toward which
The eerie attude, forall true academic people have this attitude toward
fe as abut rater that we have this atitude toward socal aranBOTS0
Son ee ines the social scientist o the extent that he clits and acts
arcs mandate to think that any social arrangement igh be otherwise, is
Fon ao uaaian, in that he can conceive the underdog being oF fe
Are rate waitor, since Re Wis Yo understand the enemy and hats to
Pe fy that the enemy] might have a case; and the ultimate ‘conservative, in
Imply trata nt easly espouse new social dctrines. (1971:552)The Sociology of Science: Where Do We Stand? 225,
This is similar to Latour’s argument against a traditional soctalogy of knowl-
edge approach. He argues that sociological understandings of things “so-
cial” are very limited, If not static and reified. Sociologists do not “know’"
any better than the scientists we study what the relevant “social” and
““histarical" arrangements are. ‘We simply have to suspend for a’moment
the idea, that we know what society is made of” (Latour 1989:11). "There
aré no resources but the explicit work and words of the actors themselves,
and we-can't like pre-Garfinkel sociologists used to do, arrive in anew
setting with our conceptions about what society is and how it works” (p.
20). He asks us. to adopt Garfinkel’s (1967) ethnomethodological, indif-
ference to sociological theory (cf. Lynch 19912,b).” In the saree vein, Latour
argues that “historians in their anti-theoretical and ant-intellectual ways [do
rnot,assume that they know beforehand what the world is made-ofl, and
{thus} they are in general always ten times more intezesting than the most
astute sociologists” (Latour 1989:19)." “Instead of explaining everything
with the same cause and framework, and instead of abstaining from expla-y
nation in fear-of breaking the reflexive garne, we shall provide a one-off
explanation, using a tailor-made cause. | am all for throwaway causes and
for one-off explanations” (Latour 1988b:174)
| agree entirely with Latour’s “indifference” to sociological theory, which
for him means static, simplistic, reductionist, and reified models of science
or society. For example, in his arguments against the use of the term
“ideplogy,"" Latour states that there-is no static “society” consisting of
‘groups who hold particular and homogeneous “ideologies”: “There is no
‘overall social structure, with a fixed composition of transhistorical objects
(institutions, ideologies, forces, interests, etc.) with which to account for
actors’ moves. Actors are building theic own society, their own context and
asking’their own questions” (1989:17).
However, | argue that we can sometimes benefit by using sociological
concepts and categories, as long as we do not allow our analyses to be
limited to, or constrained by, them. We can use concepts like hierarchy aod
stratification taken from sociology and apply therp carefully to the situation
of interest. That is, we can use them as sensitizing concepts, as long as we
remain “sensitive” to possible misfits during the research process. These
two concepts, for example, can he useful in analyzing the stratification of
perspectives jn,the construction of “truths.” We can examine the estab-
lished truth in any particular situation as.a resolution, of negotiations or
conflicts among several or many different perspectives (Méad 1938). This
resolution sometimes takes the form of a hierarchy, which we can examine
to see whose perspective “won’ and Jor how long a period of time. In the
case of my sesearch, for example, { attempted to “explain” why malecular-
bialogical approaches for explaining and representing cancer were winning26 Joan H, Fujimura
in the early 1980s by examining the articulation’of perspectives (Fujimura
1986).
| also agree with Latour’s premise that “each new topic, each new subject
requires a different metalanguage, a different interpretative toolbox.”
Strauss (1987; Strauss and Corbin 1990) and Glaser (1977; Glaser and
Strauss 1967), for example, have developed methods for helping us to
achieve, as closely as possible, this state of relative freedom from the
limitations of constraining metalanguages of grand theories. Glaser and
Strauss’s (1967) book was actually a polemic against grand theories. They
would probably agree with Latour’s (1989:15) statement that “as a rule what
actors say is always more interesting and more revealing than what analysts
say." But, of course, “analysts” are also “actors.” This perspective is the
impetus behind Glaser and Strauss’s development of in vivo codes, the use
‘of terms that our respondents use to represent their viewpoints. For example,
in my research (Fujimura 1986, 1987, 1988), “doable problems,” “band:
wagons,”” and “packages” were terms used by the scientists and technicians
Linterviewed. Yet, they also make cleat that our selection and use of fn vivo
codes to develop conceptual schemes is no more than our, the researchers’,
selection and use. Indeed, Strauss and Glaser specifically argue that the
same data, when reexamined by another researcher (or by the same re-
searcher ‘at a different point in time), might yield entirely different codes,
concepts; and theories. Further, for contemporary research, two observers
in the same situation (e.g., laboratory) might see, hear, and note down
entirely different kinds of activities. This method emphasizes process, plu-
rality, flexibility, novelty, the generation of many theories and concepts,
and attention to local situations, and thus prevents us fram falling into the
rigid steuctural frameworks of funetionalism and Marxism. Strauss argues
that phenomena are so rich that we should mine them for more stories, more
concepts, more "grounded theories,” more ways of "'seeing” phenomena
rather than limit ourselves to one set of concepts, theory, or way of seeing.
However, Strauss and Glaser also argue for the generation of afalytic
concepts and theories that take us beyond each local situation and beyond
the flattened equivalence of reflexivist stories. They argue for the generation
of more formal theories, which can “explain several substantive situations.
Concept and theory generation is achieved through the use of the constant
comparative method, where examples and situations are compared with
each other to generaie new codes and concepts. These examples and
situations are taken from the research in process, from the literature (soc
logical, anthropological, educational, nursing, etc.) as well as from the
researchers’ own personal experiences.
Here again, however, the differences between our two approaches are”
not as great as they might seem. One could argue that Latour and his
colleagues have also constructed a formal theory, “theory of associations, ””The Sociology of Science: Where Do We Stand? 2
with concepts like translation, “obligatory passage point,” and “network”
taniuting the theoretical framework. Later (1987), especially, has used
this approach and its concepts to explain act assimilate a vast array of cases
in the study of science into his approach. The latest example of assimilation
it Latour’ (1990) review of Shapin and Schaflase (1985) Leviathan and the
Air-Pump.
Woe tect, then, there are many affinities and compatibilities between
Callon and Latour’s approach and the Pragmatist-interactionist approach. in
argument the sociology of associations is not fan from the ppragmatist
argument that the “logic of inquiry”
RINATIVES TO THE REFLEXIVE PROGRAM
(AND UN-SELF-CONSCIOUS REPRESENTATION)?
4 ie make two suggestions for future work in the sociology of science
and discuss the advantages of Anselm Strauss’s work for these new direc.
tions. First, rather than focusing on finding the ultimate form of representa-
tion or undermining our Tepresentations, we should encourage new voices
mate 20 ake ther own represontations, ans sag representations
made by others. We have opened up the dialogue among scientific practi-
toners, and between these practitioners and trey Practices, to other audi-
nstance, feminists who argue that there arg
people, concems, and ways of doing science wholwhich have been ignored
by, excluded from, or defined by science ® We need to shift our discourse in
au nah © quote Rabinow quoting Suathem "gc the nature of the
Gugience, the range of readership and the kinds ci interactions between
author and reader, and alters the subject matterne Conversation ithe way it
allows others to speak—what is talked about oy whom one is talking to’
(abinow 1906:225; see also Haraway 1988), Tae Suggestion is not new,
but it needs to be made again.
Further, since we do not have one fundamental method to be exalted over
a thers, | would like to encourage and generate profusion and diversity
of methods, theopes, and “facts.” That is | would like to see other perspec.
src Mew Perspectives, the sociology of selene tne also need to regard
pantetests” as empowering these voices, | use Degen W'S Concept of interests
here:228 Joan #1, Fujimura
Individuality in a social and moral sense is something to be wrought out. tt
‘means initiative, inventiveness, vatied resourcefulness, assumption of respon-
sibiliy in choice of belief anid conduct. These are not gilts, but achievements
As achievements, they are not absolute but relative (0 the use that is to be
made of them. And this use varies with the environment.
The import ofthis conception comes out in considering the fortunes ofthe
idea of self-interest... . Interests are specific and dynamic; they are the
natural terms,of any concrete social thinking, But they are damned beyond
recovery when they are identified with the things of a petty selfishness, They
can be empliyed as vital terms only when the self is seen to be in protess, and
interest to be a name for whatever is concemed in furthering its movement,
(1920:194-95)
IF truth is negotiated order (temporally and spatially located), we need to
understand which perspectives are not included in the final product and
how they were eliminated during the processes of negotiation. Maki ng
scientific representations and artifacts is a collective process, as our studies
have shown, We.need to hear from the other human (as well as nonhuman)
actors who have not been included if the process. Who has not had, the
opportunity to speak? Who has been actively ignored or glossed over?
Second, while I agree with-Woolgar that we have no universal criteria
upon which to ground our stories, I do not agree that we should have to limit
ourselves to writing agnostic, reflexive texts.-Instead, we can pursue local
criteria in making our statements (and jadgments) of scientific, anthropologi-
cal, and sociological work, Local criteria are grounded in personal, histori-
cal, institutional; national, intemational, and other specific contexts, For
example; on reading an article comparing assimilation and accommodation
attitudes among three generations of Japanese-Americans in the American
Sociology Review, | was astoufided to find that the authors-had left out the
internment of Japanese-Americans during World War Il as a relevant vati-
able. Given my knowledge of the different images of self and membership in
American society of nisei Japanese-Americans before and after their intem-
‘ment, | was certain that the internment experience should have had some
impact on attitudes towards accommodation and assimilation..* One might
argue that | have chosen an obviously poor piece of work, the very kind of
sociology Woolgar and Latour are decrying, to make my point. However,
this poor piece of work was given a position of authority in a respectable
sociologital journal. To ignore this article, the journal, or this kind of
sociology entirely (as we go about analyzing ourselves analyzing our repre-
sentations) is to allow others to control what is being said about a group
about which I care very much.
My point is that we can make decisions about what is a better or worse
science and sociglogy in more finite terms. These decisions should be
clearly grounded in our positions as people in the world, hat is, we write
our representations (and choose our representational styles) based on explic-The Sociology of Science: Where Do We Stand? 29
it political grounds. | agree with’ Dewey (1927) and more zecently Feyera-
bbend (1975) that scientific and political discourse is continuous. | agree with
Rorty that the divisions made by Kant between science, morality, and art
were the first mistake. These divisions should not be taken seriously. Rather
than acting like “an isolated order of priests” trying to finally decide on the
ultimate metanarrative, we can speak from Where we stan as people who
act in various ways in the world.”
‘We do not have to choose between telling or accepting flattened (equal-
ized, neutlelralized) relativist texts, on the one hand, and claiming final
truth, on the other. We can choose to commit our work and resources to the
methods, science, and sociology that most suit our aspirations, to use
Dewey's word. Dewey saw the hermeneutic turn as
an opportunity to grow up, to be free to male {ourselves rather than seeking
direction from some imagined outside source. . . . His experimentalism ask
us 10 see knowledge-claims as proposals about what actions to ty out next.
[T]he recent reaction in favor of hermeneutical social sciences . . . has
taken for granted that if we don’t want something like Parsons, we have to take
something like Foucault; e., that overcoming the deficiencies of Weberian
Zweckrationalitit requires going all the way, repudiating the “will to truth.”
‘What Dewey suggested was that we keep the will to truth and the optimism
that goes with it, but free them from the behaviorist notion that Behaviorese is
Nature's Own Language and from the notion of man as “transcendental or
enduring subject." For, in Dewey's hands, the will to truth is not the urge to
dominate but the urge to create, to “attain working hatmony among diverse
desires.”
This may sound too pat, too good to be true. | suggest thatthe reason we
find it so is that we are convinced that liberalism requires the notion of a
common human nature, oF a common set of moral principles which binds us
all, or some other descendent of the Christian notion of the Brotherhood of
Man. So we have to come to see liberal social hope—such as Dewey’s—as
inherently self deceptive and philosophically naive. . .. But there seems no
particular reason why. . . we have to keep on repeating all the nasty things
about bourgeois liberalism which [Mara taught us to say, There is no inferen-
tial connection between the disappearance of the transcendental subject—of
man” as something having a natuye which sotiely can repress oF
tunderstand—and the disappearance of human solidarity. Bourgeois liberalism
seems to me-the best example of this solidarity we have yet achieved, and
Deweyan pragmatism the best articulation of il...
‘Although Foucault and Dewey are trying to do the saine thing, Dewey
seems to me to have done it beter, gimply because his vocabulary allows
room for unjustifiable hope, and an Ungroundable tut vital sense of human
solidarity. (Rorty 1982:206-7; emphasis added)
Using this “outmoded” version of human solidarity, we can attempt as best
2s we can to align our sociologies with the ways we want to actin the world
and with the kinds of consequences we want our actions to have.” Pragma-
tists and symbolic interactionists made it their “reform’" business to do this.”230 Joan H. Fujimura
| return to my earlier questions: Given that seltinterrogation does not
eliminate the choices and decisions made in constructing texts, on what
rounds do we base these selections? In other words, how do we come to
ur aspirations? Who do we want to be? And who is “we? This is where
Anselm Strauss's work on the complexities.of identities, perspectives, and
aspirations and the grounded-theory method are most relevant. While
Strauss does not provide answers, his contributions provide tools for com.
plexifying our work on the problem,
Indeterminancy, Identity, Perspectives, and Sense-Making
Anselm Strauss argues that historical and social locations (what he would
call “social worlds”) and identity are interactively constructed, and that
Perspectives and aspirations emerge dynamically from this interaction.
Since meanings are both culturally created and mediated, all inierpretations
r perspectives are based in communities or-sociél worlds. However, this
emergence is a complicated affair, since communities are not clearly cir.
cumscribed and defined and individuals participate in many different come
munities simultaneously. Moreover individuals also participate in the’ on.
Boing construction of social worlds. Thus, identities and perspectives are
‘multiple, processual, and.cialectical. Ther® is no simple, ont-to-one map-
ping between perspective and community membership, a principle that we
Can call the indeterininancy of perspectives (in line with the pragmatists’
notion of the indeterminancy of situations).” Strathem, for example,
chooses to identify herself as a feminist in her work and construct her
discourse and analyses from the perspectives of difference and domination,
which she considers ta be fundamental facts. While these facts dominate her
perspective, Strathern supposedly had (and has) other possible alternative
Perspectives from which to choose, since we all have access to multiple
selves and communities (or social worlds), Nevertheless, Strathern has
chosen to stand somewhere in particular, as has Haraway, 4s earlier noted
Many postmodémnists, on the other hand, do not want to use their situations
and situatedness to construct their perspectives. They want their representa.
tions to be open to many interpretations, just as their identities are open to
many interpretations. However, not choosing is also choosing: “IWle can-
not solve [the crisis of representation| by ignoring the relations of representa-
tional forms-and social practices. . . . If we attempt to eliminate social
referentiality, other referents [e.g., other texts) will occupy the voided
positions” (Rabinow 1986:250). :
This analysis of the tension between indeterminaitcy and structure, identi-
ty and sociahistorical location, and political choicés and collective memory
is the beauty of Strauss’s work on identity and social psychology. RatherThe Sociology of Science: Where Do We Stand? 231
than focus on indeterminancy of al representations like the postniodemnists
Gyan the partial or naive view of the structural determinism of Marxists,
Straus focuses on the multiplicity, fluidity, temporality, and processual
cate Or identity, perspectives, and social worlds, He provides no neat
sees NO Certain way to classify people, ideas, actions
Perhaps more importantly, Strauss gives us tools wit Which to work with
this tension in our constructions of interpretating statements, concepts, and
theories. | discuss three of these tools here First, Strauss’s (1969 [1959))
Mitrors and Masks: The Search for Identiy, develops an approach to the
Study of identity and mutiple perspectives, Our Obligation as analysts, he
reminds us, is to remember that we are studying actors who are likely to
iinporang etsPectives and their interactions, We will miss some and, more
spuotant Our interpretations will be guided by our ow perspectives, But
iain, there is no other viable alteative, Nong.of us eon ea making an
are fuzzy and fluid, and in all
trauss 1978a,b, 1982, '1984),
tions is a multiplicity, even a
of perspectives in a complexly
ther hand, he also accepts the responsibilities ofthe analyst. That is, as
analysts, we make interpretive statements, whether they are “descriptive” or
{fconceptual.” While Strauss attempts to include many vdices (and interac-
the elit Bis interpretations, he also makes it clear that his te analyst’) is
the ultimate interpretive perspectivels) .
Sauss's Mirrors and Masks: The Search for dentivy (1969) discusses the
basic assumptions that underlay his approach te this tension between inde-
{Giminartcy and structure in the construction of identities sod perspectives, If
Fi serene this work closely, we find that he also presented way for us,
his students to tudy sciencesmakingthat helpous toes the “interests” and
om actives of various social worlds in a complex ways in cans discussion of
how to do science studies. 1 will walk through his approach, weaving
See232 Joan H. Fujimura
together his ideas on identity construction with the problem of sense- and
science-making. The two main points are (1) a highly abstract concept like
identity is not only possible, but is also as rich, dense, dynamic, and
Unreified as Latour-could:want and as grounded and explicit as Woolgar
‘would want; and (2) sociologists of science can use this interweaving of
identity, perspectives, and social worlds in our study of sense-making—both
‘ours and the scientists'—and anyone else's for that matter.
Scientists, sociologists, and anthropologists make sense-making their pri-
mary work activity. We construct descriptions, explanations, categories,
and classifications for events and phenomena. We regularly try to make
sense of events and phenomena by naming and classifying them as Strauss
reminds us:
The philosophers, John Dewey and Arthur Bentley, int Knowing and the
Known, have argued that to narne is to know, and that the extent of knowing is
dependent upon the extent ofthe naming. By this they do not mean to sugges},
_anything magical about the act olsnaming, but to make that act central to any
“huiian’s cognition of his world... . To name; then, is not only to inclicate; it
is to identify an object as sorne kind of object. An act of identification requires
that the thing referred to be placed within a category. .
should be noted, however, that any particular object can be named, and
thus located, in countless ways. ... The nature or essence of an object does
not reside mysteriously within the object itself but is dependent upon how itis
defined by the namer. (1969, pp., 19-20)
At times, our experiences are interrupted (Dewey 1938; Mead 1938)
when we come upon events and phenomena that we cannot readily place
within our existing categories and classifications. In some cases, we are
forced by other events to fit (however badly) an unnamed eventiphe-
tomenon inlo a category to construct a new category. At other times, we
gan leave them unnamed and unexplained until later events force us to deal
with them (Law and Lynch 1990; Star and Gerson 1986).
Because,an object can be named and classified in many different ways,
its classification reveals more about the perspectives, of the classifier or
scientist than about the object. As Strauss has argued, perspective is both
inevitable and consequential:
[To name or designate is always to do this from some point of view. Frpm a
single identical perspective otherwise seemingly different things can be
classed together. Justification lies in the perspegtive, not in the things, If you
do not agree with your neighbbr’s classification, this nay only signify that you
have a somewhat or wholly different basis for drawing symbolic circles around
things. . . . The way in which things are classed together reveals, graphically
as well as symbolically, the perspectives of the classifier. (1969:20)The Sociology of Science: Where Do We Stand? 233
\Weeleain to name and classify events and phenomena throagh experience
and“action, by doing:
Sirice values are not in objects but are evaluations of objects, it follows that
Fags han eit OWN experiencing in order jo do they one oe
is aw upon meee intioduced to new terminalogy, the best wouna ay
aw upon possibly analogous experiences ard heey may of may not lead
rea eanate conceptions... Everyone has at some tine Kee introduced to
faperlences ourselves... You yourself must do, seer veg undergo—to use
Joh! Dewey's terms, (pp! 24-35)
Perhaps the most important point is that‘naming, sense-making, and the
inten aht Beneration of perspectives are activities conducted through
inicraction (Strauss 1969, Chapter 3). People (including sc ientists) engage
Sapee-making with other people and things {nonhutrons) fee example;
rather than decide a priori that scientists construct schemes for organizing
and tepiesenting nature in their heads, we can focec on their interactions
and the work activities through'which these schemes sive in practice. In my
research, for instdnce, | analyze the collective cdnstruction cp the "proto-
sg eene," a norihal cellular gene that is claimed to cause cancer, through
Audying the interactions of instuments, materials, tech iques, technicians,
funding agencies, entrepreneurial firms, investors, Congress, the public!
and many different types of scientists Fujimura 1987, 1988, 1991). From
155, 1962) de-
fined social worlds as the shared perspectives of reference groups, which
prganize their collective action, Strauss (1978) and Becker (1974, 1982)
Focused on the common activities.” Since we learn to name and classify
develo te er, Becker and Strauss proposed that shared pererecier
evelep through common, activities, Thus, while partcigame might come
from differgnt reference groups, theie perspectives can change in interaction
focused on a common activity (like researching and writing) with members
father reference groups. A social world then is a group whos “members!”
‘commit themselves to, or orient themselves. around, a common activity or
Set of activities. When we use-common activities to guide our investigations234 Joan H. Fujimura ~
and analyses, all kinds of new groups—groups notipreviously discussed in
the sociological literature—and new representations are possible. In addi-
tion, although focused on a common activity, members of a social world are
not necessarily always in agreement. There are fierce debates about how
their activities should be carried out, as well as fiercely’shared and defended
ideologies. The point ig that these ideologies and debates are constructed
through interactions among members of social worlds and form the focal
points (arenas) of social change (see, e.g., Strauss et al. 1964, 1978),
The organization of commitments around particular ways of seeing, of
representing nature, consists of these commitments and constitutes the
robustness of any theory, concept, or classification. In Strauss's terms, after
naming events and phenomena and classifying them into categories, we
tend to organize our actions in jerms of those classifications. We treat the
event or phenomenon in the same manner that we treat other members of
those categories. Classifying a phenomenon, then, has consequences for
our future action (see also Duster 1990).* However, again these conse~
quences are neither predetermined nor necessarily endurifig. That is, our
future action can also force us to redefine or reorganize our classifications.
Negotiations occur because people and groups often have different view-
points, In order to convince others to join in one’s actions, one needs to
persuade others of one’s perspective, one’s way of construing phenomena.
Negotiations between perspectives are the means by which facts are settled.
Here Strauss has argued forcefully that these are political act
Since classification and evaluation are not merely private acts but are usially,
if not predominantly, public concerns, problematic situations and issues are
foci of public contention as well as private debate. At the very minimum two
perspectives will appear. . . . Contention for terminological prizes isnot mere
squabbling over words, for words are mandates for action, and sometimes
classficatory decisions involve a mater of fife and death. At the very leak
Ipeople’s} interests are deeply involved. (1969, pp. 26~27; emphasis added)
Strauss (1978a) develops this point about negotiations even further in a later
work. Among fhe “line items” in his list of dimensions to, consider in
analyzing any negotiation is the point that these different viewpoints in
constructing reality rarely involve parties with equal or balaticed redources
In his analysis of change dnd transformation, Strauss makes significant
contributions. In the study of identity, he points to the continual transforma~
tions of identities of people as adults as they “move in and out of, and up
and down within, social structures” and, le would now likely add, s0%
worlds:
{[Cloming to new terms a person becomes something other than s/he once was,
Terminological shifts necessitate, but also signalize, new evaluations; of sellThe Sociology of Science: Where Do We Stand? 235
iatbers of events, acts, and object; and the ransformation of erception
of demerit having changed, theres no going back leneeor eit case
sagetlonl- One can look back, but she can waluete ole histher new
Status. (1969:88)
Bense-making and selfmaking, he argues, are continual processes that
never end:
ven without direct new experience, something novel may be learned
shout an object... As long as laming. connie weeny concepts,
takes es and a5 long as revision takes place, rearfaniecnen of behavior
takes plac
sy ontming of identifying of things is, then, a continual problem never
really over and done with. .. . [S}ome portion of one's classiticatory terminol.
Bye symbolic screen through which the world isordesed eno Banized,is
severed tr stain-or just has been—or wil be. Ceonge Hey (who
must baa (ssfications are really hypotheses) swold Sy necessarily
Ton ep fom tho very nature of action which brings inte wave reconstruc
on of past experience and the aisingof new abject fp 9 ‘emphasis added)
However, even in the most equable of situations, change is arduous:
Emergent valuation is hardly a serene process, for the ‘reassessment of experi-
that you thas sti testul. When some object or event is pnned dan so
{hatyou think you know wiat itis, then youiave the ae that tis portion
iL {e2s of your world is known. ut et the object ston a of character and it
esse a eaent that whatever it was, it no longer is These
define it, act toward i,‘and for what purposes; and’ vieny
ate you, yourself, in relation to itt tp. 265
This last quote frames thy igst in another way. However, as
Sailer noted, Strauss's answer to the angst of modernity sd Besellschatt
However, as Strauss points out especial ly in his work on statys passages,
we do not transform our identities individually apart from social interaction.
worlds, is pertinent. Since we learn to
aad feat Classify events and phenomena through experience and acta
and through interaction with others, identities evoles throygh continual
interaction with people, events, activities both within and outside one’s
Sie rls. ‘Since social worlds are not static entities Ibull wearer
ally being reconstituted” (Shibutani 1955:567) and since an individual
belongs to many and diferent social worlds, an individual hac conn and
abundant opportunites for changing and shifing perspecties
identities and, social worlds are continuously being constructed and re-
Constructed or maintained through interactions with others in concrete
|
|
{
|
|
}
}
}
!236 Joan H. Fujimura
contexts. These concrete contexts include what Strauss (Strauss and Corbin
1990) calls “structural conditions" (aiid what | have long-windedly called
“longer-lived institutionalized lines of action”). However, the use of struc-
ture here is neither simplistic nor static nor deterministic. Structures are
translated through perspnal and collective experiences into meanings that
shape individual identities andl actions. Individual and collective action also.
n turn shape social structures (Thomas and Znaniecki 1918). If we are
empirical, we must ask, How do the actors in the situation construct them?
Perhaps most important is the point that personal identity is tightly tied to
membership in social groups through history, collective identity, and collec-
live memory:
{Plersonal identity is meshed with group idegtty, which itself ests upon an
historical past... A social psychology without full focus upon history is a
blind psychology. A concem with personal styles, strategies, careers—in
short, with personal identities—requites a parallel concern with shared, oF
collective identities, viewed through time. (Strauss 1969:173-75)
Collective identities and collective memories of social groups and socigl
worlds are crucial for understanding that the continuous reconstruction does
not allow for, independent reconstructions of individual identities or per-
spectives. Others will remember what we might choose to forget and thus
force us to remember and be held accountable.” Even in a gesellschaft
society, we find that the ties that bind are not so easily cut. Whether
collective memories include the oppression of women by men, of the Serbs
by the Austrians, Hungarians, and Turks, or of the Croats by the Serbs
(Strauss 1969:169-74), the Holocaust (Maier 1988), or the internment expe-
riences of Japanese-Americans, these memories and histories inform and
help define their identities, perspectives, and present and future actions.
While our final representations depend on our perspectives, Strauss’s
work off the complexities of identities, perspectives, and aspirations and his
grounded-theory method helps us as analysts to attempt to construct com-
plex ahd responsible representations. His concept of identity fully incorpo
rates the multiple and processual nature of group memberships and their
complex interactions with perspective formation and action, while his social
world perspective moves us beyond a narrow focus on formal sociological
groups.
Herbert Blumer (1931:526), argued that conception shapes perception,
that sensual perception is shaped and organized by conception. Perceptual
experience does hot exist without interpretation. What science does is to
construct concepts to shape and organize perception in three ways: “(1) it
introdicés a new orientatior or point of view; (2) it serves as a tool, or as a
means of transacting business with one’s environment; (3) it makes possible
deductive reasoning and so the anticipation of new experience.” StraussThe Sociology of Science: Where Do We Stand? 237
and Glaser constructed a set of methods for systematically inventing and
deploying Blumer’s (1954) sensitizing concepts (as opposed to definitive
Concepts) to guide subsequent perception...Sensitizing concepts are ideas
that suggest potential avenues for exploration. They are not rigid, static,
reified objects. Taken together, Blumer’s sensitizing concepts, Strauss’
cacophony of voices and interactions, and Strauss and Glaser’s grounded-
theory methods make it possible for analysts to introduce new perspectives,
to change the world in some fashion and to some degree, and to anticipate
new ways of being.
Conclusions
One lesson from Strauss’s treatment of identity for the sociology of science
is that our constructions as sociologists both are and have consequences.
Scientific and social scientific classifications and constructions are out.
Comes of negotiations between different and oftentimes conflicting view:
Points in specific situations. The outcomes of these situated negotiations,
our constructs, also have consequences. They involve “people's interests”
(read “perspectives if not matters of “tifé or death.” If facts and truths are
egotiated orders,we need to bring other perspectives into the negotiating
arena and empower them with authority. We need to ask, To whom are we
committed? when we arialyze and write our stories. And we need to do so
with full focus on “our” and "their" histories, collective memories, and
aspirations for collective action.
There is no finally stable ground on which any of us—feminists, sociolo-
gists of science, people of color——base our stories. As a feminist, a wontian of
color, and a sociologist of science (to name just a few of my multiple
identities), construct (and reconstruct) my meanings, histories, stories, and
aspirations collectively with others. When | write, however, + must take
responsibility and hold myself accountable for the final perspective. The
point is to make explicit to myself and to my'audiences just where I stand,
‘my operating perspective, and the ground on which my, concepts are
‘constructed.
Acknowledgments
| am indebted to several colleagues for their critical and editorial comments on
earlier drafts of this paper. They include Alberto Cambrosio, Adele Clarke, Kjell
Doksum, Michael Fortun, Dominique Lestel, Michael Lynch, Annemarie Mol, Leigh
Star, Anselm Strauss; Kathleen Uno, and the editor of this volume, David Maines.
The paper has also benefited from discussions over coffee and beer atthe conference238 Joan H. Fujinmura
on Rediscovering Skill in Science, Technology and Medici organized by the
Oe sadies Centre atthe University of Bath in September 1990, However, | take
responsibility forthe final representation.
Notes
1. Roty (1985) prompted me to took again at Dewey's Reconsiucton 1
Philosophy (1920).
My now, many strong progransmets are no lone Som programmers in
henna eeent wrngs. Compare Pickering 1990), fF eames ‘with Pickering
tige4) Collins, however, i stil very much Collins: For 9 particularly strong protest
(roti the new dtecions of Clfon Latour, and WoRlEaY Set Collins and Yearley’s
ae ey seein (ur, Calon and Latout (1991) and Woolga’ 991) for tele
aavfesto Cling and Veatey. There are ath elleryis Se studies, whom |
eRe seuss here. See, for example, Mualkay (1985) Aebnore (1989), and other
ve te Woolgar (1980b); cf. Woolgar and Paviluch (10°
Hot i eo Feyerabend (1973), Fleck (1979), and Kun (270) Kranakis (1989)
and Trowele (1980) also describe national and cultual fren tit problem
Construction and -solving efforts
Sato condensing 3 lot of work as well as several diferent approaches"here in
orks se hang ttl the history af the “ew” sucioloby 9 NT this
per For examples ofthe heterogeneous Constuclon of scientific knowledge, se
Bene tore) am symmetry, allo (1986) on translation and building networks,
aot and Keating (1980) on tacit knowledge, Clarke (19906) on political
Cambrosie a ins (1985) on relativism aed scientific contre Fujimura on
conte tignment (1987) and an defining situations VN, ‘Knort-Cetina (1979)
anticline (1987, 1988) on envoling allies and War networks, Law
on tie ferogencous engineering, Law and Willams U0 putting facts
together, Lynch (1985) on shoptal Sar (1985) on trangulation effect, Star and
Griesemer (1989) on boundary objects
esr {Oe exaanpl, Harding (1966) on the work done by werner laborato-
Heenan Cee nthe Taboratory’ prosucts (8. fownal ‘antcles) nor in
tes that is Mente work, Star (1991) onthe work done by wore and Aricon
seco ineteenth-century taxidermy, and Shapin (1989) on technicians’ work in
fRenteenth-century and modem laboratories
Bee algo Woolgar and Pawluch (1985).
Sif. group of studies began to emerge n the early 197%, especially with the
wo tof blogs (1976) and Barnes (1977) and inches stiles Py ‘Mackenzie (1981),
works oe o08), Shapin (1979, 1982), and Shapin and Schaffer (1985).
sen fe the quotes in this paragraph are taken from (he L#.Out and Woolgar
vwlbang, Vr restate Woolgar’s argument in his 1983 ace,
are re with his last statement. Just because a tex nT for does
pot nea tk ts interpretation fs uncertain. But that’s another argument.
mean het te genealogy as presented by many postmodernist ‘anthropologist
see farneson (1983) for an effort to tistoricize the emersence ‘of "postmodert
culture
HU. example ofthis representational form isthe Wale collection (see
ciiford 1986), However, even in the Walker collection, one ight ask wy the text