Epistemological Chicken

10

Epistemological Chicken
H. M. Collins and Steven Yearley

The game of "chicken" involves dashing across the road in front of

speeding cars. The object of the game is to be the last person to cross.
Only this person can avoid the charge of being cowardly. An early
crosser is a "chicken" (noun), that is, a person who is "chicken"
(adjective).
Alternation
In his book Invitation to Sociology, Peter Berger describes the pro-
cess of "alternation." Sociologists, he explains, develop the ability
to switch between different frames of reference. They learn how to
take on the ways of being in the world that are characteristic of the
groups they study. In doing this they learn that their own taken-for-
granted-reality, including their most deeply held beliefs, are but one
set of beliefs among many. This ability, which seems to the sociolo-
gist to be little more than applied common sense, is surprisingly
narrowly distributed within society. It is a skill that good sociolo-
gists (also anthropologists and some philosophersl acquire through
their training. Most academic training narrows the viewpoint and
reinforces the single way of seeing that is the trademark of the dis-
cipline. Just as Christians know that Christ was the Savior and Mus-
Copyright © 1992. University of Chicago Press. All rights reserved.
lims know that Muhammad was the Prophet, so economists learn

that the world is a set of rationally interlocking self-interests, phy-
sicists learn that everything can be predicted mathematically so
long as it can be isolated (and so long as it is not indeterministic or
chaoticl, biologists learn that everything can be explained by some
variant of evolutionary theory, and neurochemists know that in-
dividuals are fundamentally controlled by the flow of juices in
the brain.
Sociologists don't know anything in quite this way; they only
know how it is to know. The sociologist is promiscuous, experienc-
ing many loves without ever falling in love. This is neither a happy
nor an endearing state. But while promiscuity is not a recipe for
301
Science as Practice and Culture, edited by Andrew Pickering, University of Chicago Press, 1992. ProQuest Ebook Central,
http://ebookcentral.proquest.com/lib/brandeis-ebooks/detail.action?docID=625217.
Created from brandeis-ebooks on 2021-12-21 11:59:29.
ARGUMENTS
love, it is for education. A well-educated person is not just a faithful

specialist but one who knows how to take another's point of view-
even to invade another's world of knowledge.
The achievement of the sociology of scientific knowledge (SSK)
can be understood as an extension of this ability to "alternate."
Where, for example, sociologists must understand the culture of re-
ligious believers and of worldly atheists, SSKers must be ready to be
convinced by geological uniformitarianism and catastrophism; now
they must know that the universe is filled with gravity waves, now
that it is not; in one location they grasp the reality of psychokinesis,
in another they must be sure that the universe is exhausted by the
four known forces. But SSK also makes us specially aware of sharp
differences not only in the content but also in the quality of knowl-
edge. What Mr. John Doe knows as he rides the Clapham omnibus
he knows with much more certainty than what Professor Jonathon
Doe knows as he reads the Philosophical Investigations. Typically,
as we have remarked, the sociologist knows less than the natural
scientist, while the sociologist of science knows still less. Those
engaged from day to day with the problem of reflexivity would, if
they could achieve their aims, know nothing at all. We might say
that SSK has opened up new ways of knowing nothing.
In spite of this achievement, all of us, however sophisticated, can
switch to modes of knowing that allow us to catch buses and hold
mortgages. We all engage as a matter of fact in what we might call
"meta-alternation." Our argument here is that social studies of sci-
ence ought to erect meta-alternation as a principle, not treat it as a
failing. To treat it as a failing is to invite participation in an escala-
tion of skepticism which we liken to the game of chicken; in this
case the game is epistemological chicken.
In what follows we examine two recent tendencies or schools in
science studies which play the game in different ways. The most
straightforward case is the escalation from relativism through dis-

course analysis and "new literary forms" to reflexivity. In the end,
the relativist regress leads us to have nothing to say.
A more complicated case is the French school. A clear prescrip-
tion that emerged from relativism was, as Bloor (1973, 1976) put it
so well, symmetry. That is, sociologists of scientific knowledge
should treat correct science and false science equally; they should
analyze what are taken by most scientists to be true claims about
the natural world and what are treated by most as mistaken claims
in the same way. The idea was that the construction of the boundary
between the true and the false would become the topic rather than
the starting point as in existing sociologies of science. The French
302
H. M. COLLINS AND STEVEN YEARLEY
have adopted a semiotic approach which gives rise to a much more

radical symmetrism. Not only is the boundary between true and
false seen as a construction, but from now on, so are all dichoto-
mies- poulet epistemologique. 1 The net effect of treating the whole
world as a system of signs is, however, to come full circle back to
the prosaic world of John Doe. As we will explain, the philosophy
may be radical, but the implications are conservative. Where there
are no differences except the differences between words there are no
surprises left-no purchase for skeptical levers to shift the world on
its axis. If anything moves it is the world as a whole. It slides un-
noticed; nothing is realigned, nothing trembles, nothing falls. The
two types of epistemological escalation-reflexive and semiotic-
seem very different, yet in each case the result is impotence.
The Relativist Regress
Relativism, discourse analysis, reflexivity

Since the mid-1970s each new variant of SSK has tended to be a little
more radical than the one before it. Each new variant has stood
longer on the relativist road. Each group in tum has made the same
mistake about its foundational role. In the initial flourish, some so-
ciologists of scientific knowledge believed they were solving epis-
temological puzzles that had defeated philosophers. They believed
that the close study of the activity of scientists showed that even
in the hardest case of knowledge "truth [was] a socially organized
upshot of contingent courses of linguistic, conceptual and social be-
haviour" (McHugh, quoted in Collins 1975, 205j. But epistemologi-
cal problems are not resolved by empirical discoveries. 2 Collins
should have looked more carefully at the words he quoted. McHugh
actually says that "truth is only conceivable as a socially organized
upshot": (1971, 329; emphasis addedj. Being inconceivable, "real"

truth is hardly susceptible to empirical study. The importance of
the philosophical arguments about relativism in the 1970s was, in
retrospect, not that they showed that relativism was true but that
it was tenable and therefore could be used as a methodology for
the study of science. This was in a context where many philoso-
phers argued that relativism was untenable and therefore could lead
only to flawed research. Proponents of SSK used in-principle argu-
1. In French the term poulet has the not-inappropriate connotation of policeman.

2. Though what counts as doing epistemology may be radically affected by the
developments of new styles of research on the problem of knowledge.
303
ARGUMENTS
ments to show the necessity of epistemological agnosticism, while

their case studies revealed the appropriateness of methodological
relativism. 3
Within the first few nanoseconds of the relativist big bang, nearly
everyone realized that the negative levers were equally applicable to
the work of the sociologists and historians themselves. 4 Under ap-
propriate scrutiny their work too looked like a "socially organized
upshot," but this did not seem very interesting; the new-style analy-
sis of science, with its implications for the relationship between
science and the rest of the world of knowledge, seemed to be the
exciting thing. Before long, however, Mulkay, Potter, and Yearley
(1983) announced the program of "discourse analysis."
Discourse analysis refused to accept the evidence marshaled by
SSK case studies as an unproblematic representation of scientists'
activities. We might say that it closely examined the way SSK
practitioners used their evidence and revealed the constructional
work involved in making social science out of interview material. 5
Discourse analysis revealed some features of the way people orga-
nize their talk that were of interest (see Gilbert and Mulkay 1984,
90-111; Mulkay and Gilbert 1982; Yearley 1985). The exercise
would have been unexceptionable but for its proponents' insis-
tence that discourse analysis itself was epistemologically founda-
tional. They thought that discourse analysis comprised a critique of
SSK while itself being invulnerable to the same critique. In short,
they believed that because SSK was vulnerable to deconstruction it
was methodologically flawed, whereas discourse analysis touched
bedrock.
The discourse analysts, it seemed, had failed to appreciate the
universal vulnerability of knowledge making to the methods of SSK.
They also failed to understand that social analysis of science does
not show the science to be wrong; by the same token, discourse
analysis does not show that something other than SSK is right. In
all cases validity is the outcome of social negotiation; the absence
of social negotiations is not a condition of validity. Articles critical
3. The philosophical arguments (not the case studies) which supported episte-
mological agnosticism were not without epistemological force, however. They are
fatal for those who claim authority in virtue of the epistemological high ground. They
level out the epistemological terrain just as the empirical studies level out the sci-
entific terrain.
4. The nanosecond joke is due to Steve Shapin.
5. In fact, first-flush discourse analysis did not examine such work very closely;
rather it looked at published papers and sought to mount in-principle, methodological
arguments; see, for example, Gilbert and Mulkay 1980.
304
of discourse analysis soon appeared (Collins 1983; Shapin 1984) and

discourse analysis has been largely abandoned within SSK. 6
Discourse analysis paved the way for more radical deconstruction
which goes under the title of "reflexivity." 7 Those of the reflexive
persuasion have noted the omnipervasiveness of the social con-
struction of truth-it is, as McHugh noted, a conceptual rather than
an empirical point-and have set out to sidestep all truth-making
conventions. "New literary forms" are an important part of this
movement, since single authorship is taken to be one conventional
method of constructing certainty. Reflexive texts often use "multi-
vocality" to avoid authority. Witty authors can write as though they
were more than one person so as to prevent a text from reaching a
conclusion; to each argument or rhetorical gambit there is always
an answer or rhetorical counterplay. Eventually the dialogue peters
out without coming to any conclusion. The absence of convergent
argument draws ironic attention to the devices that are normally
used to make conclusions come about. Indeed, as one of the Pinches
(Pinch and Pinch 1988) points out, in the hands of less subtle au-
thors, multivocality is merely a pernicious way of imposing author-
ity, by making it seem as though everyone has been allowed a say,
while the author retains control over the voices of opponents.
Subtle reflexivists realize that their work leads nowhere. For ex-
ample, Woolgar has said that getting nowhere should be seen as an
accomplishment, not a failure. 8 Nevertheless, on occasion reflex-
ivists appear to claim that their work is associated with progress
within the sociology of science: thus Woolgar and Ashmore (1988, 7)
state that "The exploration of reflexivity is the next natural de-
velopment of the relativist-constructivist perspective in the social
study of science," and they supply a diagrammatic three-stage his-
tory of the field to reinforce their case. While they satirize their own
"progressivist" depiction of the field with blithe headings and cap-
tions such as "The Next Step" (1988, 7) and by the observation that
these authors would not "be quite so stupid as to introduce a defini-
tive list of aims and arguments in the volume without their being
some clever reflexive point to it all" (1988, 2), this is not taken so
far as to eliminate all academic defence. One argument invokes con-
sistency: "the growing confidence with which scholars have argued
6. But is still interesting to social psychologists; witness the success of Potter and
Wetherell 1987.
7. For the most recent, full-blown, witty work in the reflexivist genre, see Ash-
more 1989.
8. This was his comment from the floor las recalled by Collins) at a conference
on rhetoric and the strong program in Iowa in 1987.
305
,ARGUMENTS
that natural scientific knowledge is a social construct, is now ac-

companied by growing interest in the consequences of applying this
same argument to knowledge generated by the social sciences"
(1988, 1; emphasis in originall. The interest in the social condition-
ing of knowledge can evidently be extended to our own knowledge
claims; we have no right to exclude ourselves.
Of course such equivalence was also recommended by Bloor. But
among reflexivists this assertion is accompanied by a second argu-
ment: that reflexivity brings extra dividends. Unlike Bloor, who
would have science studies resemble natural science, reflexivists see
the study of knowledge making as "an occasion for exploring new
ways of addressing longstanding questions of knowledge and episte-
mology" (Woolgar and Ashmore 1988, 1-21. And, lest this be taken
as one of the disingenuously listed "aims and arguments," we may
note that Woolgar had earlier asserted that SSK "can only be a dis-
traction from any attempt to come to terms with the fundamentals
of knowledge production" (1983, 2631. More recently the same au-
thor has noted that while SSK usefully challenges defenders of "'sci-
entific method' . . . it leaves the distinct impression that deeper,
more fundamental questions remain unanswered" (1988a, 981.
As we have implied, the consistency argument is unobjection-
able. Woolgar (1983, 2451 nicely makes the point that SSK has
equivocated about the extent to which lithe social world [is] re-
garded as fully a substitute for the natural world in knowledge de-
termination." Outside of scientific Marxism, no social scientist
would expect sociology to tell us the proper apperception of nature.
When we talk of access to knowledge of nature we must mean ac-
cess through the sciences. Of course insofar as self-appointed advo-
cates of science seek to attribute the success of science to the
pursuit of some set of rules or systematic behaviors, SSK challenges
these accounts. As we have just seen, even Woolgar is happy with
this "use" of SSK. These are, so to speak, the powers and limits of
SSK. If SSK is turned on social science in general and on itself in
particular, these powers and limitations turn with it. Just as SSK has
no direct, unmediated route to nature, so reflexive study can expect
no immediate access to the truths of the social world. Equally, the
powers remain the same: Woolgar concludes his recent book-length
study, Science: The Very Idea, by criticizing those who seek to pre-
scribe rules of sOciological inquiry by reference to natural scientific
method (1988a, 106-81. If the call for reflexivity is simply-not
"merely" but "simply"-a call for consistency, then this is all we
can get from it: the powers and limitations of SSK applied to our
306
claims to knowledge of the social world. How can it do anything

more fundamental?
This prospect is, as we have said, unobjectionable. But it is not
consequential. Even if, in Woolgar's words, we are left with the "dis-
tinct impression that deeper, more fundamental questions remain
unanswered" (1988a, 98), reflexive analysis cannot answer these
questions. Reflexivists do claim consequentiality for their work; in
their writings they seek to derive it from a sort of phenomenological
jump into the study not of social scientific knowledge but of know-
ing per se. Woolgar poses this issue in terms of the "Problem"
(1983, 240-2; 1988a, 31-37). For Woolgar, the Problem concerns
the revocable, indefinite, and nondemonstrable link which binds the
objects of knowledge to our statements about those objects. How do
we know that the chromatograph represents the chemical, that the
respondent's answer represents his or her attitude, that the image in
the telescope represents the world imaged through it? The Problem
is thus "a general and irresolvable problem" of knowing (1983,240).
What reflexivity shows is the ubiquity of this problem. It occurs in
epistemological theorizing but also at the laboratory bench and in
the argumentational practice of social scientists.
According to Woolgar, the Problem is perceived by members as a
series of "methodological horrors." He lists strategies by which
practitioners can sidestep them (1988a, 33-36). For example, the
Problem may be deemed a problem only for less sophisticated sci-
ences (psychiatry suffers, but not medicine), or it may be construed
as a technical problem whose solution will come when one's own
science becomes more sophisticated (psychologists will soon be able
to measure attitudes better). It may be recognized only as it applies
to others, or it may be relegated to the status of a "made up" or
"academic" problem.
Woolgar's identification of the Problem is very persuasive. But
what are, we to make of it; how does it benefit us to know of it?

Clearly the Problem cannot be solved. Neither can it be bypassed by
adopting a phenomenological policy of introspection. In 1983 it was,
however, recommended (cautiously) that we "celebrate" it (1983,
263). Why, though, should reflexive analyses which draw attention
to the Problem be valued or at least prized any more highly than
other well-known skeptical philosophies? While other scientists
and lay inquirers proceed by merely "managing" the Problem (and
we can disclose this managing work; see Woolgar's recent emphasis
on ethnography [1988bIJ, reflexive analyses face up to it. But facing
up to it can only lead nowhere. It is hard to construe this as taking
307
ARGUMENTS
us, in Woolgar's words, into the fundamentals of knowledge produc-

tion. Woolgar and the other reflexivists may stand steadfast in the
path of the traffic, but a heroic description of the experience is mis-
placed. The traffic passes and they survive unscathed-but only be-
cause they have fallen down a hole in the road.
In sum, following the lead of the relativists, each new fashion in
SSK has been more epistemologically daring, the reflexivists coming
closest to self-destruction. Each group has made the same mistake
at firstj they have become so enamored of the power of their nega-
tive levers on the existing structures as to believe they rest on bed-
rock. But this is not the case. Though each level can prick misplaced
epistemological pretentions, they stand in the same relationship to
each other as parallel culturesj no level has priority and each is a
flimsy building on the plain. Accepting this we can freely use what-
ever epistemological "natural attitude" is appropriate for the pur-
pose at handj we can alternate between them as we will. That is
what methodological relativism is about-the rejection of any kind
of foundationalism and its replacement, not by permanent revolu-
tion but by permanent insecurity. To reverse the vertical scale of the
metaphor, while SSK showed that science did not occupy the high
ground of culture, the newer developments must be taken to dem-
onstrate not the failure of SSK, but that there simply is no high
ground.
In the absence of decisive epistemological arguments, how do we
choose our epistemological stance? The answer is to ask not for the
meaning but for the use. 9 Natural scientists, working at the bench,
should be naive realists-that is what will get the work done. Soci-
ologists, historians, scientists away from the bench, and the rest of
the general public should be social realists. Social realists must ex-
perience the social world in a naive way, as the day-to-day founda-
tion of reality (as natural scientists naively experience the natural
world).l0 That is the way to understand the relationship between

science and the rest of our cultural activities. Close description of
9. We mean this in the Wittgensteinian sense. We mean that the endlessly agoniz-
ing search for essential meanings is senseless, since the meaning of something equals
its use in a form of life. Meaning and use are but two sides of the same coin. When
we go on to analyze the French school, we will take advantage of this idea to examine
the use/meaning of the semiotic approach. In other words, we will not engage in a
scholarly exercise to uncover the roots of semiotic ideas; we will ask what the whole-
sale adoption of these ideas would mean for the practice of the sociology and history
of scientific knowledge. Whatever the roots, that is the meaning of the semiotic ap-
proach for SSK.
10. For an analysis of the use of social realism and its relation to recent debates
in the philosophy of social science, see Yearley 1988.
308
the human activity makes science look like any other kind of prac-
tical work. Detailed description dissolves epistemological mystery
and wonder. This makes science one with our other cultural en-
deavors without making it necessary to deny that scientists have
more skill, experience, and wisdom than others in the matters they
deal with.
Notice that granting skill, experience, and wisdom is not quite
the same as granting the authority that would follow from ac-
cepting natural realism as an epistemological foundation. The dif-
ference is important for several reasons. Social realism in studies of
science discourages the sterile emulation of canonical models of sci-
entific activity by social scientists and others. It discourages ordi-
nary people from judging science against a criterion of infallibility;
since science cannot deliver infallibility, to judge it thus is to risk
widespread disillusion. Making science a continuous part with the
rest of our culture should make us less intimidated and more ready
to appreciate its beauty and accomplishments. It should make us
more ready to use it for what it is, to value its insights and wisdom
within rather than without the political and cultural process.
What use are the other epistemological levels that the game of
chicken has opened out? First, their very existence has led to the
end of epistemology. No one can take epistemological foundation-
alism seriously any more. The effect of meta-alternation is similar
to that of cultural alternation in its broadening and liberalizing ef-
fect. More directly, discourse analysis is useful for analyzing dis-
course, and clearly there is a living and some joy in reflexivity
(Collins 1989). The concerns of reflexivists are close to those of rhe-
toricians; complete skepticism regarding the very matter of argu-
mentation is a good starting point for the analysis of the force of
argument. We believe, however, that the big job of sorting out the
relationship between cultural enterprises has to be done from the
level of social realism. The work can be done from no other level.
The French School

One of the most attractive approaches in the study of science to
have appeared in very recent years has been that of "I'ecole" de
l'Ecole de Mines in Paris. The principal actors are Michel CalIon and
Bruno Latour; Latour's sparkling writings having been especially in-
fluential in the English-speaking world. Both the French and the re-
flexivists have been attracted by the lure of consistency. In the
French case it is a matter of consistency in the treatment of all di-
chotomies. For example, natural"actants" and human actors must
309
ARGUMENTS
be treated symmetrically. This strategy too requires great daring,

but like full-blown reflexivity it can only bring annihilation or re-
treat-retreat from the traffic or retreat from the world. What does
the French work do for us? What is its use?ll
In spite of early coincidences of interest, the starting point of the
French approach is quite different from that of SSK. Whereas the
Anglo-American tradition of epistemological thought is concerned
with how we represent reality (what the relationship is between the
world and our representational devices), the Continental tradition
more naturally asks how anything can represent anything else. This
has led some reviewers to consider the French work to be the most
progressively radical approach in science studies (see, for example,
de Vries 1988, 8). Whereas Bloor's tenet of symmetry seems radical
within the Anglo-American tradition, the French approach extends
symmetry of treatment to many more dimensions. Treating natural
"actants" and human actors symmetrically comes naturally from
the semiotic starting point. Crucially, however, whereas the Anglo-
American approach, with its concern with our representations of
the natural world, is essentially human-centered, the semiotic ex-
tension of the question about representation has no center. From
this viewpoint it is just as natural to ask how natural objects repre-
sent us as to ask how we represent natural objects. A cardboard cut-
out of a traffic policeman is a representation of US;12 its power, as an
actant, to control us is just as suitable a case for analysis as the
power with which the mapmaker "dominates the world" (Latour
1987,224).
The effect of SSK has been to show that the apparent independent
power of the natural world is granted by human beings in social
negotiation. Because the special power and authority of natural
scientists comes from their privileged access to an independent
realm, putting humans at the center removes the special authority.
In the work of the French school, symmetry between all kinds of

actants once more removes humans from the pivotal role. This is
the key to understanding the "uses" of the French approach. It ex-
plains why the case studies which are emerging from this approach
look prosaic when viewed from SSK. While the approach seems, on
the one hand, radical and shocking, on the other hand it does not
provide any countercommonsensical surprises. Where it is novel
vis-a-vis earlier SSK approaches it is essentially more conservative.
11. See also some recent reviews of Latour's books !Yearley 1987; Shapin 1988).
12. In Europe, cardboard policemen are sometimes used to warn of danger on the
roads.
310
As we will see, the extension of the principle of symmetry to other

dimensions makes it less easy to be symmetrical in the original
Bloorian sense. Symmetry of treatment between the true and the
false requires a human-centered universe.
Laboratory life
In retrospect, the divergence between the human-centered and the
French approach can be traced back to Laboratory Life (Latour and
Woolgar 1979). A central idea in this book is the "inscription de-
vice." Latour saw the biological laboratory filled with devices for
making inscriptions-inscriptions that were then transformed and
combined with other transcriptions and eventually published. These
representations were the reality-stuff of science-gaining truth pre-
cisely as they became separate from the messy activity of the labo-
ratory. The scientific trick was to transform the to and fro of daily
life in the laboratory into a paper transcription which could move
outside the laboratory, creating the reality of the phenomenon under
investigation. The representation had a power that the activity in
the laboratory did not have.
The idea of the force of an inscription comes from the method-
ology of the study. The method espoused by Latour was observation
informed by the perspective of the estranged visitor. Participation,
the method of verstehende sociology, to which the controversy stud-
ies (Collins 1981) aspired, played no part in Laboratory Life. Latour
may have worked in the Salk Institute, but he proudly proclaimed
his failure to understand what he was doing. It is this that makes
inscription devices appear autonomous. That is exactly how they
are meant to look to those who are less than expert. The point is
that in controversial areas it is experts who are the first consumers
of inscriptions, not outsiders; to experts, everything is mutable. The
participant is struck by thi& mutability whereas the stranger sees

immutable forces.
Those who took an interpretative approach to the study of scien-
tific contr~versies and who tried to grasp the exp.etes· viewpoint
were, then, less enthusiastic about granting to inscriptions the de-
gree of autonomy implied in Laborutory Life. For example, Travis
looked at a controversy within the biological sciences-over the
chemical transfer of learned behavior-and found that the inscrip'
tions produced by mass spectrometers in biological laboratories
were not universally accepted as representing reality. A mass spec-
trometer, like all other technologies, requires skilled interpretation.
According to Travis's study, spectrometry traces that did not fit with
311
ARGUMENTS
favored hypotheses of experts outside the source laboratory were re-

jected by them.
The difference, then, between the approach of the controversy
studies and that of Laboratory Life was in the extent to which the
former saw the granting of authority to inscriptions and other repre-
sentations as an active and revocable process-one which could be
halted at any time. Laboratory Life stressed the "once and for all"
production and release of an inscription with autonomous power.
In the early days, we could see this as a difference in emphasis
rather than principle. Furthermore, in recent passages Latour talks
of the amount of work that would be required to challenge inscrip-
tions. He leaves the door open to their deconstruction at the hands
of sufficiently determined humans. On such occasions he could still
be interpreted as differing from SSK only in emphasis (with the in-
scriptions growing steadily more powerful). But there is substantial
ambiguity around this point. Certainly the idea that representations
have power is clearly expressed in Latour's later formulation-the
"immutable mobile" (1987, 227). Here again, agency is granted to
pieces of paper-the first step on the road to granting agency to
things. Admittedly the stress is on the granting of agency-the in-
scriptions have to be inscribed and the immutable mobiles have to
be created and to gain assent. But once the agency has been granted,
the inscriptions gain a degree of autonomy. They can compel further
assent. It is in this that their immutability lies.
Scallops and things

So far we have discussed the agency of things that have been created
by humans-as in the Salk Institute. But French-style radical sym-
metry draws no boundary between objects that have been created
and those that occur naturally. Here we run into some difficulty of
interpretation, for if the whole subject matter is signs and represen-

tations, it is hard to know exactly how we should make the distinc-
tion between that which occurs naturally and that which is made.
Does this distinction too lie only in the way things are taken to
represent one another? In the face of these complex problems we
must fall back once more on asking not for the real meaning but for
the use. What does the semiotic approach do for our understanding
of the world?
CalIon's (1986) paper on the scallops of St. Brieuc Bay is a good
starting point. Consider first CalIon's remarks on the nature of the
boundary between the natural and the social:
312
The observer must abandon all a priori distinctions between natu-

ral and social events. He must reject the hypothesis of a definite
boundary which separates the two. These divisions are considered to
be conflictual, for they are the result of analysis rather than its point
of departure. Further, the observer must consider that the repertoire
of categories which he uses, the entities which are mobilized, and the
relationships between these are all topics for actors' discussions. In-
stead of imposing a pre-established grid of analysis upon these, the
observer follows the actors in order to identify the manner in which
these define and associate the different elements by which they build
and explain their world, whether it be social or natural. (200-201).
The capacity of certain actors to get other actors whether they be

human beings, institutions or natural entities-to comply with them
depends upon a complex web of interrelations in which Society and
Nature are intertwined. (201)
In both of these quotations the principles of radical symmetry are

set out. 13 CalIon's analysis proceeds symmetrically, using his vo-
cabulary of problematization, interessement, enrollment, and mobi-
ization. Here is how he treats the mobilization of the scallops.
If the scallops are to be enrolled, they must first be willing to anchor
themselves to the collectors. But this anchorage is not easy to achieve.
In fact the three researchers will have to lead their longest and most
difficult negotiations with the scallops. (211)
The researchers are ready to make any kind of concession in order

to lure the larvae into their trap. What sort of substances do larvae
prefer to anchor on? Another series of transactions is necessary to
answer this question.
It was noted that the development of the scallops was slower with
collectors made of straw, broom, or vegetable horsehair. These types
of supports are too compressed and prevent water from circulating
correctly through the collector. (213)
13. Just how radical the symmetry is, however, is not entirely clear. Though as
we will see, the scallops of St. Brieuc Bay are to be treated as actors on a par with the
fishermen, the creation of symmetry is very much in the hands of the analysts. The
analysts remain in control the whole time, which makes their imposition of sym-
metry on the world seem something of a conceit. Would not complete symmetry
require an account from the point of view of the scallops? Would it be sensible to
think of the scallops enrolling the scallop researchers so as to given themselves a
better home and to protect their species form the ravages of the fishermen? Does the
fact that there is no Sociological Review Monograph series written by and for scallops
make a difference to the symmetry of the story? Fortunately we do not need an an-
swer to these questions before we continue our analysis.
313
ARGUMENTS
These passages are written in just the same style as the passage in
the next paragraph where CalIon discusses what the scientists will
count as signifying successful attachment of the scallops to the col-
lectors: "At what number can it be confirmed and accepted that
scallops, in general, do anchor themselves?" To answer this question
the researchers had to negotiate not with the scallops but with their
scientific colleagues. It turned out that they met little resistance to
their preferred definition of a "significant" degree of attachment be-
cause-and here the problem is highlighted-"of the negotiations
which were carried out with the scallops in order to increase the
interessement and of the acts of enticement which were used to re-
tain the larvae (horsehair rather than nylon, etc.)" (213).
We see in these passages some of the fruits of radical symmetry.
While the scientists had to negotiate with their colleagues over the
meaning of "successful anchorage," their success came about partly
because the scallops had already agreed to anchor themselves in rea-
sonable numbers. The scallops are full parties to the negotiations.
CalIon's account shows how all the parties-fishermen, scien-
tists, scallops-at first give their assent to the networking ambi-
tions of the research scientists. But catastrophe strikes in later
years. The fishermen suddenly dredge up all the experimental scal-
lops and sell them. The scallops too withdraw their cooperation:
The researchers place their nets but the collectors remain hope-
lessly empty. In principle the larvae anchor, in practice they refuse to
enter the collectors. The difficult negotiations which were successful
the first time fail in the following years .... The larvae detach them-
selves from the researchers' project and a crowd of other actors carry
them away. The scallops become dissidents. The larvae which com-
plied are betrayed by those they were thought to represent. (219-20)
The crucial final quotation is: "To establish ... that larvae anchor,
the complicity of the scallops is needed as much as that of the fish-
ermen" (222).
CalIon's account of the negotiations between the scientists and
the fishermen is a fine study in the relationships between tech-
nology and society. But as a social account of the making of knowl-
edge it is prosaic, because the story of the scallops themselves is an
asymmetrical old-fashioned scientific story. A symmetrical, SSK-
type account would analyze the way it came to be agreed, first, that
the scallops did anchor, and second-at a later date-that they
did not anchor. Into this analysis the question of whether or not
the scallops complied would not enter. The informing assumption
314
would be that whether there were more or fewer scallops anchoring

early and late in the story did not affect the extent to which scallops
were seen to be anchoring early and late. No SSK story could rely on
the complicity of the scallops; at best it could only rely on human-
centered accounts of the complicity of the scallops.
To look at the difficulty from a different angle, let us ask what
traditional SSK studies would gain from radical symmetrism and the
actor-network perspective. Should the Collins ian story of gravity
waves (1975, 1985) be retold granting agency to the gravity waves in
addition to the scientists? CalIon's description of scallops as actors
gains some of its rhetorical appeal from the fact that scallops are
living things. One can just about imagine them deciding whether or
not to anchor themselves to the horsehair collectors. Similarly, La-
tour is following a familiar figure of speech when he describes tides
and winds as "acting" on explorers' ships (1987, 221). It is more dif-
ficult in the case of gravity waves. But given the imaginative re-
sources, what does it add to our understanding to say that gravity
waves granted their allegiance to Joseph Weber but refused to grant
it to anyone else? Far from adding to our understanding it seems
to us that the resulting account would look just like the account
of a conventional historian of science-except that the historian
wouldn't talk of allegiances with gravity waves and failures of ne-
gotiations with gravity waves, but of discoveries and failures of ex-
perimental technique. The language changes, but the story remains
the same. 14
14. Is it just the vocabulary that is radical? Let us try the old trick of rewriting
some of the quotations from CalIon using the conventional language of the history
of science. "If the scallops are to be enrolled, they must first be willing to anchor
themselves to the collectors. But this anchorage is not easy to achieve. In fact the
three researchers will have to lead their longest and most difficult negotiations with
the scallops (211 )." History of science version: If the scallops are to be cultivated they
must anchor'on the collectors. But anchorage is not easy to achieve. In fact the three
researchers will have a lot of trouble developing appropriate techniques.
The researchers are ready to make any kind of concession in order to lure
the larvae into their trap. What sort of substances do larvae prefer to anchor
on? Another series of transactions is necessary to answer the question.
It was noted that the development of the scallops was slower with collec-
tors made of straw, broom, or vegetable horsehair. These types of supports are
too compressed and prevent water from circulating correctly through the col-
lector. (213)
History of science version: The researchers are willing to try anything. What sort
of substances do larvae prefer to anchor on? Another series of experiments is neces-
sary to answer the question. It was noted that the development of the scallops was
slower with collectors made of straw, broom, or vegetable horsehair. These types of
315
ARGUMENTS
In any case, the complicity of the scallops (or whateverl, if it is to

playa part in accounts of this sort, ought to be properly recorded.
How is the complicity of scallops to be measured? There is only one
way we know of measuring the complicity of scallops, and that is by
appropriate scientific research. If we are really to enter scallop be-
havior into our explanatory equations, then CalIon must demon-
strate his scientific credentials. He must show that he has a firm
grip on the nature of scallops. There is not the slightest reason for
us to accept his opinions on the nature of scallops if he is any less
of a scallop scientist than the researchers he describes. In fact, we
readers would prefer him to be more of a scallop expert than the
others if he is to speak authoritatively on the subject. Is he an au-
thority on scallops? Or did he merely report the scientists;' views on
the matter. If the latter, then we have two possibilities. One way of
using scientists' reports is in a typical SSK study, traditionally sym-
supports are too compressed and prevent water from circulating correctly through
the collector.
The researchers place their nets but the collectors remain hopelessly
empty. In principle the larvae anchor, in practice they refuse to enter the col-
lectors. The difficult negotiations which were successful the first time fail in
the following years .... The larvae detach themselves from the researchers'
project and a crowd of other actors carry them away. The scallops become dis-
sidents. The larvae which complied are betrayed by those they were thought
to represent. (219-20)
To establish ... that larvae anchor, the complicity of the scallops is needed
as much as that of the fishermen. (222)
The history of science version: The researchers place their nets, but the collectors
remain hopelessly empty. In theory the larvae ought to anchor, in practice they don't.
The difficult experiments which were successful the first time fail in the following
years .... The larvae fail to attach themselves and get carried away. The larvae seem
to have changed their nature-the first experiment apears to have succeeded under
unrepresentative experimental conditions. To establish that larvae anchor, the tests

must be safeguarded against the fishermen's interference, but this is only one factor
in the experiment.
The reason why this is so easy is that the relationship between the scientists and
the scallops in CalIon's account is just as is found in the old history of science.
Agency and responsibility are divided in the same way. In SSK, by contrast, agency is
much more in the domain of the humans. For example, consider this passage from
Collins's story of gravity waves (1985, 91): "It is not obvious how the credibility of
the high flux case fell so low. In fact, it was not the single uncriticized experiment
that was decisive; scientists rarely mentioned this in discussion. Obviously the sheer
weight of negative opinion was a factor, but given the tractability, as it were, of all
the negative evidence, it did not have to add up so decisively. There was a way of
assembling the evidence, noting the flaws in each grain, such that outright rejection
of the high flux claim was not the necessary inference."
316
metrical, and not in the least radically symmetrical. But if he really

means what he says about the importance of the scallops as actors,
and yet he persists with scientists' secondhand reports of scallops'
behavior rather than providing firsthand expertise, then he gives us
a pre-SSK study open to all the problems of asymmetry. Certainly
we do not have a study that can offer us any surprises about the
natural world, or one that clarifies the credibility and authority of
science.
Doors and grooms

The problem of radical symmetrism, as we hope to have revealed, is
that it must rest on routine methods of scientific research for that
part of its evidence concerned with the nonhuman actants. There is
no reason to suppose that sociologists are particularly good at gath-
ering such evidence. The problem of method is very evident in La-
tour's recent work. The piece we look at is entitled "Where are the
Missing Masses? Sociology of a Door."lS
Latour treats as actors (or actants-it makes no differencej a door
and an automatic door closer.l6 An automatic door closer is known
colloquially in French as a "groom/' by analogy with the human
groom who might once have been a doorman. This does lend the
door closer, like the scallops, a degree of anthropomorphic appeal,
but this is certainly not true of the door itself. In the paper Latour
sets out to explore the potency of doors (and other technological
artifactsj as actors. In this he has a very grand purpose in mind. He
intends to explain social order. His explanation of social order is
that the constraints that are missing within existing social and po-
litical analysis of order (the missing massesj are to be found in
things. Doors, and other artifacts, act as constant constraints upon
our behavior. Though they are physical mechanisms, their effect is
indistinguishable from normative or moral control: a door allows us

to walk through only at a certain speed and only in a certain place
in the wallj a cardboard policeman enforces the traffic code in a way
similar to a real policemanj a seat-belt warning signal enforces the
law on seat beltsj my desk computer forces me to write instructions
15. This article has a long history. A version of it was published once in a refereed
journal (Latour and Johnson 1988) and another version is due to appear in an edited
volume (Latour forthcoming). The version we refer to is an in-between draft presented
at the conference which spawned the edited volume. It is sometimes hard to know
when Latour is being completely serious, but we will assume two publications of
essentially the same article in refereed outlets is as good a sign as any.
16. Schaffer 1991 refers to the giving of life to inanimate objects as "hylozoism."
317
ARGUMENTS
to it in an exact syntactical form. The missing masses (analogous to

the missing mass that is required to explain the slow rate of expan-
sion of the universe) are to be found in the technological things that
surround us. We delegate actor status to these things and along the
way we delegate power to them.
Now as explained in respect to CalIon's analysis of scallops, im-
puting agency to things requires a method of analysis of their po-
tency. It is no good just talking about inscriptions and immutable
mobiles; I won't learn from a No Smoking sign why some people
obey it while many others ignore it. Furthermore, the method must
be something more than social analysis; it must be more than an
analysis of what we grant to things and their representations on a
moment-to-moment basis. If this was all there was to it, then, first,
radical symmetry would have been abandoned and we would be back
with an SSK-style human-centered, social-realist universe; and sec-
ond, Latour would not have discovered the missing masses. The
missing masses, if they are to count as a discovery, must lie in the
agency of things outside of ourselves.
Given that we want to understand the power and agency of doors
and door closers, the appropriate method, we would think, would
include something of mechanical engineering, something of mate-
rials science, and something of the more engineering-oriented parts
of architecture. Alternatively it might consist of detailed examina-
tion of the use of doors; some videotaping of people going in and
out of doors would be a good start. It is clear that the interpre-
tative method is unusable, since doors have no social life in which
we could participate. This fits well with the scientific "unity of
method" first offered in Laboratory Life; we may open doors a hun-
dred times a day, but we are all strangers in the world of things.
Interpretation is no good. What method is left? It ought to be sci-
ence, but Latour is not an expert in any of the fields of science that
would help him understand doors, nor does he have videotapes or

other close records of doors in use. How then does he convince us of
the potency of the agency of doors? The question is not where are
the missing masses? but where is the missing method? It turns out
to be the method of counterfactual hypotheses.
In order to understand the power of doors we are asked to imagine
how it would be if they were not there:
Now, draw two columns ... in the right column, list the work
people would have to do if they had no door; in the left column write
down the gentle pushing (or pulling) they have to do in order to fulfil
the same tasks. Compare the two columns: the enormous effort on
the right is balanced by the little one on the left, and this is all
318
thanks to hinges. I will define this transformation of a major effort

into a minor one, by the words displacement or translation or delega-
tion or shifting; I will say that we have delegated ... to the hinge the
work of reversibly solving the hole-wall dilemma .... everytime you
want to know what a non-human does, simply imagine what other
humans or other non-humans would have to do were this character
not present. (4-5)
Latour exploits the imaginative possibilities to the limit. If there
were no door in a wall, anyone wanting to enter would have to de-
molish part of the wall and rebuild it again after passing through.
On the other hand, if there were merely a hole, then cold winds
would enter causing the occupants to become ill. In the case of La
Villete, the draughts would get in and the historians' drafts would
not get out. One may use counterfactual method as such a rich
source of jests precisely because the writer is in complete control
over the counterfactual scenario. That is why it is a dangerous
method in history. What was the power of the exact contour of Cleo-
patra's nose? What was the historical contribution of Napoleon's
breakfast on the morning of the Battle of Waterloo? What is the
power, not of a door, but of a row of bricks in a wall; take the bricks
away and wall and roof will fall down killing everyone inside. This
row of bricks, it turns out, has the force and responsibility of the
row of strong men which would otherwise be required to support
the wall. Precisely the same number of strong men are required to
replace two rows of bricks, or half a row of bricksY The counterfac-
tual method is very exciting to use because the imagination is given
17. Actually a similar analysis of doors has already been done by a British come-
dian. Rik Mayall once had an act on BBC television in which he played "Kevin
'TUrvey-Investigative Reporter." Turvey, portrayed as an unprepossessing youth
from Birmingham in a horrible blue parka, sat in silhouette in a black revolving chair
turning dramatically to the camera as his report began. He would begin to describe
his investigation of some subject of moment, but almost immediately would lose
track, and his description would degenerate into a hilarious account of the inconse-
quentialities of moment-to-moment existence.
In one episode, "The Supernatural, /I 'TUrvey reports that he is in his kitchen when
he hears a ringing noise which he eventually identifies as someone at the back door:
So I got up-right? Went out into the hall. 'Cause, well, you've got to get out
into the hall to get to the back door you see. The only alternative is, like,
smashing down the wall next to the cooker. And I'm not gonna get involved in
all that again-right?
Anyway, I got to the back door-right? Opened it up and everything. That
was easy really-it's-Well, I've done it loads of times before. Just got to twist
the handle a bit and open up the door. Piece of piss really. Well it's made out of
wood ... etc.
319
ARGUMENTS
such free rein. It is like a child's toy clock facej the hands may be
set anywhere, but it is not of much use for telling the time.
The absence of methodological control over fantasy allows Latour
to develop his concept of "delegation" unhindered by traditional
problems. Using imaginative license to the full, he is able to tell
convincing stories about the way we delegate power to technological
artifacts. The lack of control over method allows control to be given
to things. This way he appears to resolve a major philosophical
problem-the distinction between action and behavior-but ap-
pears is the operative word. The distinction continues to trouble
those who actually try to make machines that can appropriate hu-
man responsibilities. Latour sometimes seems to realize that things
don't fulfill quite the same duties as humans: "Three rows of dele-
gated non-human act ants (hinges, springs and hydraulic pistons) re-
place, 90% of the time, either an undisciplined bell-boy ... or, for
the general public, the programme instructions that have to do with
remembering-to-close-the-door-when-it-is-cold" (11). This 90 per-
cent is about as far as Latour can go, and even that realization does
not come from the counterfactual method (which could produce any
percentage figure desired). Superficial analysis gives rise to the claim
that the door closer "substitutes for the actions of people, and is a
delegate that permanently occupies the position of a humanj ... and
it shapes human action by prescribing back what sort of people
should pass through the door" (12). But the author himself notes
that harder work is required to solve this kind of problem properly:
"Specialists of robotics have very much abandoned the pipe-dream
of total automationj they learned the hard way that many skills are
better delegated to humans than to non-humans, whereas others
may be moved away from incompetent humans" (25 n. 5 ).18
The phrase "the hard way" is exactly to the point. There are a
number of hard ways of finding out about our relationship to ma-
18. In a later version of the paper we find a note 9 which reads: "In this type of
analysis there is no effort to attribute forever certain competences to human and
others to non-humans. The attention is focussed on following how any set of com-
petences is distributed through various entities." This note describes the position
much better. To reiterate, however, if we wanted to discover the distribution of com-
petences among nonhumans, we would not normally go first to a sociologist or phi-
losopher. One critical reader of this paper has suggested that a more sympathetic
treatment of Latour would see him as trying to produce a description of the experi-
enced materiality of things. But we simply do not experience the materiality of doors
as having anything to do with saving us from the effort of knocking down walls,
whether they be those of La Villette or Kevin Turvey's kitchen (see n. 17). If we need
a description of the materiality of things-we thought this was what we already had
within the prosaic view of science-this is not a good way of going about it.
320
chines and other artifacts. Scientists, technologists, philosophers,

builders of intelligent machines, and social scientists all struggle in
their different ways to make progress along the route. The builders
of intelligent machines have one method: they try to model human
beings; the scientists, technologists, and philosophers work hand in
hand using another method: it is called natural science. What soci-
ology of scientific knowledge provides is a third method, no longer
subservient to accounts of the work of the scientists and technolo-
gists and the stories of philosophers but rooted in a special under-
standing of social life. It would be a tragedy to surrender this new
way of comprehending the world to the accounts of scientists, tech-
nologists, and philosophers. It would also be a tragedy not to exploit
the particular understanding of our relationship to machines that
this method allows us.
Latour's treatment of the technological world fails in three ways.
First, it enrolls the false ally of the counterfactual method. Second,
since the counterfactual method fails, the story of the power of
things can amount to no more than technologists' secondhand ~c
counts; it is sociologically prosaic. Third, the distinction between
human action and the behavior of things and its significance for the
automation of human skills is, ironically, one area where sociolo-
gists of scientific knowledge have the ability to speak authorita-
tivelyas scientists. If there were any areas where we might claim to
know something, this would certainly be among the foremost; 19 yet
it is just here that the French school passes the responsibility by.
Consider this third failing. SSK has put us in a position to under-
stand what knowledge is, what humans do as they make knowledge,
what the difference is between the social construction of knowl-
edge and the asocial operation of machines and other things, how
the operation of machines may seem to pass for knowledge, and
what the limits are on the replacement of human skills by artificial
means. NI these understandings have come through exploring the

difference between humans and things, the difference between ac-
tions and mere behavior, and through criticizing the mistaken at-
tempts-for example, by rationalist philosophers of science-to
describe human action as formula-bound behavior.20
19. Other areas would include science education and the public understanding of
science and of scientific authority. See for examples, Collins and Shapin 1989; Collins
1988; Wynne 1982; Yearley 1989.
20. Interestingly, the Latour-Thrvey (see n. 171 style of description of everyday
things can itself be used as part of a critique of the ability of machines to act as
humans. It can be used as a critique of artificial intelligence. One might think of
ilL_Til descriptions as ironic dramatizations of the commonsense, tacit knowledge
321
ARGUMENTS
The consequences of the semiotic method amount to a backward

step, leading us to embrace once more the very priority of techno-
logical, rule-bound description, adopted from scientists and tech-
nologists, that we once learned to ignore. This backward step has
happened as a consequence of the misconceived extension of sym-
metry that takes humans out of their pivotal role. If nonhumans are
actants, then we need a way of determining their power. This is the
business of scientists and technologists; it takes us directly back to
the scientists' conventional and prosaic accounts of the world from
which we escaped in the early 1970s.
Conclusion
In this paper we have reviewed two post-relativist positions in sci-
ence studies. One group has sought to penetrate to "foundational"
knowledge by turning constructionist tools on constructionism; the
other has aimed to generalize symmetry by treating all act ants that
are party to the scientific enterprise in the same manner. We have
shown the shortcomings of both approaches.
Despite their considerable differences, the two positions have one
important feature in common, a feature which distinguishes them
from relativism or constructionism. Both supply an elaborate vo-
cabulary for describing the means of knowledge making, but the
vocabulary does not allow explanations of why certain knowledge
claims are accepted and others are not. In the case of Wool gar, for
example, we have seen the labels which he gives to ways of manag-
ing the Problem in everyday life, such as treating it as merely a
technical difficulty (1988a, 34). The labels delineate the range of
strategies which may be adopted. Equally, CalIon (1986, 206) refers
to the strategy which scientists may use of trying to set up an
"obligatory passage point" through which the arguments of others
have to pass. And when describing the work of cartographers and
that humans have by virtue of their culture. These cannot be expressed in program·
ming rules, because of their potential to ramify explosively, and because it is never
clear how much needs to be said explicitly. Counterfactual hypotheses have just the
same elusiveness as common sense. The elusiveness of common sense was not no·
ticed while artificial intelligence was a dream; it has been discovered by the builders
of intelligent machines only with the failure of their project. Only now is it becoming
clear to them that nonhuman actants have no common sense. The counterfactual
method is still unfortunately bolstering the dreams of semioticians. Dreyfus 119761
first makes this point with respect to artificial intelligence. For a more recent discus·
sion of the problem utilizing some of the insights of SSK, see Collins 1990.
322
classification makers, Latour (1987, 223) talks of stages in grasping

the object of knowledge: making it mobile, rendering it stable, and
making it combinable. Thus, a map fixes the limits of, say, an island;
allows knowledge of the island to be taken away from the island's
shores; and permits this knowledge to be transmitted and used in
conjunction with other geographical expertise.
Both approaches allow suggestive names to be given to common
rhetorical-argumentational strategies. Other people can easily rec-
ognize things to which they too can give these names. (This may be
part of the current appeal of these authors' work.) But at best these
names identify common moves in knowledge making. When scien-
tists try to establish themselves or their technique as an obligatory
passage point, it will benefit the analyst to be able to put a well-
known name to the strategy. That, however, does not tell us why
only some actors have been able to get away with enforcing their
view of the world. We still must ask: why one map rather than an-
other? What is in the relationship between the island and the map
that makes it a successful map? Is there something more than the
social relations of the map maker with clients and colleagues? If
there is, what is it? If there is not, what is all the fuss about? We
have been there already and tried to explain as well as describe.
This is not to say that every piece of science studies must be
an explanation, but it must leave room for explanation. In their
emphasis on form, the reflexivity and actor-network theory ap-
proaches both exclude explanation in the descriptive languages they
provide. 2 !
In sum, the game of epistemological chicken as played by relativ-
ists and their successors has been destructive. The French actor-
network model is philosophically radical, but when we ask for its
use, it turns out to be essentially conservative-a poverty of method
making it subservient to a prosaic view of science and technology.
While the reflexivist players have escaped the fate of the foolhardy
by jumping into a hole in the road from which there is no escape,
the adherents of the actor network turn out to have crossed the road
well before the traffic was in sight, leaving only their ventrilo-
quist's voices echoing between the curbs. Listen and understand, but
do not follow too closely. Neither program is foundational. Meta-
alternation, not oblivion, is where these programs should lead.
21. Furthermore, both disavow their scientific (or naturalistic) credentials (Wool-
gar 1988a, 98; Latour 1988, 165). Unlike naturalistic SSK, they search for an alterna-
tive route which turns out not to be explanatory-not just in fact but in orientation
(see Woolgar 1988a, 108).
323
ARGUMENTS
The true relativist's world is a world without foundations. What

follows? The discovery that the foundations of physics are not as
secure as was once believed makes no difference to what it means to
be a good physicist-though it does mean that physics cannot claim
authority over competing knowledge claims in virtue of episte-
mology. Now we have the same lesson writ large. The discovery that
all foundations of knowledge are as fragile as those of the physicist
gives us the opportunity to alternate between worlds, but it makes
no difference to what it means to be a good sociologist or a good
historian of scientific knowledge. Of course we cannot claim epis-
temological authority either, as though we ever had a chance. We
can only compete on even terms for our share of the world with all
the usual weapons. In the relativist's world you have to decide what
you want to do; epistemology does not make the decision for you.
But once you have decided what to do, all there is left is to do it.
REFERENCES
Ashmore, M., 1989. The Reflexive Thesis: Wrighting Sociology of Scientific
Knowledge. Chicago: University of Chicago Press.
Berger, P. L. 1963. Invitation to Sociology. Garden City, N.Y.: Anchor Books.
Bloor, D. 1973. Wittgenstein and Mannheim on the Sociology of Mathemat-
ics. Studies in the History and Philosophy of Science 4: 173-91.
- - . 1976. Knowledge and Social Imagery. London: Routledge and Kegan
Paul.
CalIon, M. 1986. Some Elements of a Sociology of Translation: Domestica-
tion of the Scallops and the Fishermen of St. Brieuc Bay. In J. Law ed.,
Power, Action, and Belief: A New Sociology of Knowledge? London:
Routledge and Kegan Paul, 196-233.
Collins, H. M. 1975. The Seven Sexes: A Study in the Sociology of a
Phenomenon, or the Replication of Experiments in Physics. Sociology
9:205-24.
- - . 1983. An Empirical Relativist Programme in the Sociology of Sci-

entific Knowledge. In K. Knorr and M. J. Mulkayeds., Science Observed.
Beverly Hills: Sage, pp. 85-114.
- - . 1985. Changing Order: Replication and Induction in Scientific
Practice. Beverly Hills: Sage.
- - . 1988. Public Experiments and Displays of Virtuosity: The Core-Set
Revisited. Social Studies of Science 18: 725-48.
- - . 1989. Tu Quoque. Review of Malcolm Ashmore, The Reflexive
Thesis. THES. no. 904, 2 March 1990, 20.
- - . 1990. Artificial Experts: Social Knowledge and Intelligent Ma-
chines. Cambridge: MIT Press.
Collins, H. M., ed. 1981 Knowledge and Controversy: Studies of Modern
Natural Science. Special Issue of Social Studies of Science 11 (1).
324
Collins, H. M., and S. Shapin. 1989. Experiment, Science Teaching, and the
New History and Sociology of Science. In M. Shortland and A. Warwick,
eds., Teaching the History of Science. Oxford: Blackwell, 67-79.
De Vries, G. 1988. Transferring Knowledge: Science and Its Practices. Re-
view article including Latour's book. EASST Newsletter 7: 5-12.
Dreyfus, Hubert. 1979. What Computers Can't Do. New York: Harper
and Row.
Gilbert, G. N., and M. Mulkay. 1980. Contexts of Scientific Discourse: So-
cial Accounting in Experimental Papers. In K. Knorr Cetina, R. Krohn,
and R. Whitley, eds. The Social Process of Scientific Investigation. Dor-
drecht: Reidel, 269-94.
- - . 1984. Opening Pandora's Box: A Sociological Account of Scientists'
Discourse. Cambridge: Cambridge University Press.
Latour, B. 1987. Science in Action. Milton Keynes: Open University Press.
- - . 1988. The Politics of Explanation: An Alternative. In S. W. Wool-
gar, ed., Knowledge and Reflexivity: New Frontiers in the Sociology of
Knowledge. Beverly Hills: Sage, 155-76.
- - . Forthcoming. Where Are the Missing Masses? Sociology of a Few
Mundane Artifacts. In W. Bijker and J. Law, eds., Constructing Networks
and Systems. Cambridge: MIT Press.
Latour, B., and J. Johnson. 1988. Mixing Humans with Non-Humans: Soci-
ology of a Door-Opener. Social Problems 35: 298-310.
Latour, B., and S. Woolgar. 1979. Laboratory Life: The Social Construction
of Scientific Facts. Beverly Hills: Sage.
McHugh, P. 1971. On the Failure of Positivism. In J. D. Douglas, ed., Un-
derstanding Everyday Life. London: Routledge and Kegan Paul, 320-35.
Mulkay, M., and G. N. Gilbert. 1982. Joking Apart: Some Recommendations
concerning the Analysis of Scientific Culture. Social Studies of Science
12:585-615.
Mulkay, M., J. Potter, and S. Yearley. 1983. Why an Analysis of Scientific
Discourse Is Needed. In K. D. Knorr Cetina and M. Mulkay, eds., Science
Observed: Perspectives on the Social Study of Science. London: Sage,
171-203.
Pinch, T., and T. Pinch. 1988. Reservations about Reflexivity and New Lit-
erary Forms, or Why Let the Devil Have All the Good Tunes? In S. W.
Woolgar, ed., Knowledge and Reflexivity: New Frontiers in the Sociology
of Knowledge. London: Sage, 178-97.
Potter, J., and M. Wetherell. 1987. Discourse and Social Psychology: Be-
yond Attitudes and Behaviour. London: Sage.
Schaffer, S. 1991. The Eighteenth Brumaire of Bruno Latour. Studies in the
History and Philosophy of Science 22(1):174-92.
Shapin, S. 1984. Talking History: Reflections on Discourse Analysis. Isis
75:276,125-28.
- - . 1988. Following Scientists Around. Review of Latour 1987. Social
Studies of Science 18: 533-50.
325
ARGUMENTS
Travis, G. D. L. 1981. Replicating Replication? Aspects of the Social Con-

struction of Learning in Planarian Worms. In Collins 1981, 11-32.
Woolgar, S. W. 1983. Irony in the Social Study of Science. In K. D. Knorr and
M. Mulkay, eds., Science Observed. Beverly Hills: Sage, 239-66.
- - . 1988a. Science: The Very Idea. Chichester: Ellis Horwood and Lon-
don: Tavistock.
- - . 1988b. Reflexivity is the Ethnographer of the Text. In S. W. Woolgar,
ed., Knowledge and Reflexivity: New Frontiers in the Sociology of
Knowledge. Beverly Hills: Sage, 14-34.
Woolgar, S. W. and M. Ashmore. 1988. The Next Step: An Introduction to
the Reflexive Project. In S. W. Woolgar, ed., Knowledge and Reflexivity:
New Frontiers in the Sociology of Knowledge. Beverly Hills: Sage, 1-11.
Wynne, B. 1982. Rationality or Ritual? The Windscale Inquiry and Nuclear
Decisions In Britain. Chalfont St. Giles, Bucks.: British Soc, Hist. Sci.
Monograph.
Yearley, S. 1985. Representing Geology: Davy's Presentation of Science
at the Royal Institution. In T. Shinn and R. Whitley, eds., Expository
Science: Forms and Functions of Popularisation. Dordrecht: Reidel,
pp.79-101.
- - . 1987. The Two Faces of Science. Review of Latour 1987. Nature 326
(6115): 23-29 April, 754.
- - . 1988. Settling Accounts: Accounts, Action, and Sociological Expla-
nation. British Journal of Sociology 39: 578-99.
- - . 1989. Bog Standards: Science and Conservation at a Public Inquiry.
Social Studies of Science 19:421-38.
326

Epistemological Chicken

Uploaded by

Copyright:

Available Formats

Epistemological Chicken

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Epistemological Chicken

Uploaded by

Copyright:

Available Formats

10

The game of "chicken" involves dashing across the road in front of

lims know that Muhammad was the Prophet, so economists learn

love, it is for education. A well-educated person is not just a faithful

straightforward case is the escalation from relativism through dis-

have adopted a semiotic approach which gives rise to a much more

The Relativist Regress

Relativism, discourse analysis, reflexivity

upshot": (1971, 329; emphasis addedj. Being inconceivable, "real"

1. In French the term poulet has the not-inappropriate connotation of policeman.

ments to show the necessity of epistemological agnosticism, while

of discourse analysis soon appeared (Collins 1983; Shapin 1984) and

that natural scientific knowledge is a social construct, is now ac-

claims to knowledge of the social world. How can it do anything

what are, we to make of it; how does it benefit us to know of it?

us, in Woolgar's words, into the fundamentals of knowledge produc-

world).l0 That is the way to understand the relationship between

The French School

be treated symmetrically. This strategy too requires great daring,

In the work of the French school, symmetry between all kinds of

As we will see, the extension of the principle of symmetry to other

participant is struck by thi& mutability whereas the stranger sees

favored hypotheses of experts outside the source laboratory were re-

Scallops and things

interpretation, for if the whole subject matter is signs and represen-

The observer must abandon all a priori distinctions between natu-

The capacity of certain actors to get other actors whether they be

In both of these quotations the principles of radical symmetry are

The researchers are ready to make any kind of concession in order

correctly through the collector. (213)

would be that whether there were more or fewer scallops anchoring

In any case, the complicity of the scallops (or whateverl, if it is to

unrepresentative experimental conditions. To establish that larvae anchor, the tests

metrical, and not in the least radically symmetrical. But if he really

Doors and grooms

indistinguishable from normative or moral control: a door allows us

to it in an exact syntactical form. The missing masses (analogous to

would help him understand doors, nor does he have videotapes or

thanks to hinges. I will define this transformation of a major effort

chines and other artifacts. Scientists, technologists, philosophers,

means. NI these understandings have come through exploring the

The consequences of the semiotic method amount to a backward

have to pass. And when describing the work of cartographers and

classification makers, Latour (1987, 223) talks of stages in grasping

The true relativist's world is a world without foundations. What

- - . 1983. An Empirical Relativist Programme in the Sociology of Sci-

Travis, G. D. L. 1981. Replicating Replication? Aspects of the Social Con-

You might also like