David Bloor Remember the Strong Program

7/29/2019 David Bloor Remember the Strong Program

1/14

Remember the Strong Program?

David Bloor

Science, Technology, & Human Values, Vol. 22, No. 3. (Summer, 1997), pp. 373-385.

Stable URL:

http://links.jstor.org/sici?sici=0162-2439%28199722%2922%3A3%3C373%3ARTSP%3E2.0.CO%3B2-Z

Science, Technology, & Human Values is currently published by Sage Publications, Inc..

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available athttp://www.jstor.org/about/terms.html. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtainedprior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content inthe JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained athttp://www.jstor.org/journals/sage.html.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printedpage of such transmission.

The JSTOR Archive is a trusted digital repository providing for long-term preservation and access to leading academicjournals and scholarly literature from around the world. The Archive is supported by libraries, scholarly societies, publishers,and foundations. It is an initiative of JSTOR, a not-for-profit organization with a mission to help the scholarly community takeadvantage of advances in technology. For more information regarding JSTOR, please contact [email protected].

http://www.jstor.orgSat Nov 10 19:09:50 2007
http://links.jstor.org/sici?sici=0162-2439%28199722%2922%3A3%3C373%3ARTSP%3E2.0.CO%3B2-Zhttp://www.jstor.org/about/terms.htmlhttp://www.jstor.org/journals/sage.htmlhttp://www.jstor.org/journals/sage.htmlhttp://www.jstor.org/about/terms.htmlhttp://links.jstor.org/sici?sici=0162-2439%28199722%2922%3A3%3C373%3ARTSP%3E2.0.CO%3B2-Z


2/14

Remember the Strong Program?David Bloor

University of Edinburgh

Before arriving at the Science Studies Unit in 1967 I took a degree inexperim ental psychology at Cambridge. Richard Gregory and Donald B road-bent were just two of the fine group of experimentalists Cam bridge housedat that time. My exposu re to the Psychological Laboratory and the AppliedPsychology Unit w as intellectually exc iting, and I acquired a great respectfor the discipline. It provided me w ith an orientation I have never lost. I wantto make use of that orientation now, as a way of introducing a discussion ofthe sociology of knowledge and, in particular, of the so-called strong pro-gram.' This may seem an odd procedure because it straddles disciplinaryboundaries, but I hope I can sho w that such worries are misplaced.

Bartlett's LawsOne of the books I encountered as a student was F. C. Bartlett's classic

study Remembering. (The author, Sir Frederic Charle s Bartlett [1886-19691,had held the chair in Cambridge and was one of the first experimentalpsychologists to be elected a fellow of the Royal Society of London.)Remembering was first published in 1932 but was still required reading inthe 1960s. Th e book m ade a great impression on m e. Two points especiallyhave stayed in my m ind. I shall call them Bartlett's First Law and Bartlett'sSecond Law, though these are my labels and not how he presented them.Bartlett's Laws have been strikingly confirmed over the years by the re-sponses made to th e strong program. L et me tell you what these laws are andexplain how they hav e been confirmed.

Bartlett's First Law m ight be called the "Law of Complexity." It says thatthe complexity of a response is a function of the complexity of the respon dingorganism, not the complexity of the stimulus. Bartlett was interested inme mo ry and he was working in the shado w of the great pioneer Ebbingh aus.Ebb ingha us developed the use of simple, nonsense syllables as the materialScience,Technology,& Human Values, Vol. 22 No . 3, Summer 1997 373-3858 1997Sage Publications Inc.


3/14

374 Scienc e, Technology,& Human Values

for studyin g retention and forgetting. H e wanted to strip away the effect ofprior familiarity or any kind of special interest in the material to be remem -bered, s o he voided the stimu lus material of meaning. In this way he hope dto lay bare the m achinery of the memory itself, regardless of past exp erience ,pas t training, atten tion, motivation, and other variables. Bartlett rejected thisand to ok the study of mem ory out of the laboratory into a naturalistic setting.He argued that you do not produce simple or pure memory responses bypresenting apparently simp le stimuli, like nonsense syllables. It is still thewhole organism that is going to respond. This is why the complexity of aresponse is not a function of the complexity of the stimulus but of theresponding organism.Ho w is this law exemplified in responses to the strong program ? To m akethe con nection , I need to begin with the assertion that the strong program isitself really very sim ple. The basic idea is this: just as experimental psy cholo-gists turn their scientific curiosity onto the processes of individual cognitio n,s o we can generalize the enterprise and turn scientific curiosity onto collec-tive cognition. At the heart of our collective understanding of the world liesthe impressive institution of science, so let us se e what happens when w e tryto know science scientifically. Now, is this really a simp le idea? Am I perhapsplaying a trick? For exam ple, I speak of know ing science scientifically, butthat formu la, with its double use of the word "science," might seem convo-luted rather than simple. So me people will say it presupposes the very thingto be investigated, so either an investigation is not needed (because wealready know the answe r) or the investigation is impossible (because we d onot know the answer, s o we do not know how to proceed). I am confident inpredicting that, by virtue of their training and their intellectual predilections,many sociologists and philosophers would rapidly run through some suchthought process-and conclud e that the project of know ing science scientifi-cally w as highly suspect. In so me illegitimate way, the critics will think, it"privileges" science, or "privileges" so me particular accoun t of it. Here, then ,we can see som e of the complexity of the response that was generated.Le t m e now try to take this on e step further. Is it really so wrong and s omethodologically suspect to try to know science scientifically? Didn'tBartlett have to depend on his mem ory when he w rote his book on remem -bering? Do n't experimental psychologists have to use the very s am e cogni-tive and perceptual processes they are investigating in the course of theirinvestigations? Of course they do-and yet that does not, in practice, threatenthe interest of their findings or diminish their capacity to m ake progress. Ho wd o they m anage to pull off this amazing methodological trick? An d if theycan do it, why can't we? I think we can do it too. We need to s ee that the only"privileging" the psychologists are doing is to make practical use of the


4/14

Bloor I Remember the Strong Program? 375

faculties they are investigating-that is, they use the things them selv es, notany particular account or theory of them. No account or theory is "privi-leged." M y feeling was, and is, that we can do the sam e in the sociology ofscience, provided we have acquired, and can take for granted, certain "in-stincts" abou t how to proceed "scientifically." Of course, these are notliterally instincts; they are culturally acquired. The point is we need certainhabits of m ind, rather than an acco unt or theory. We need these to start us offand keep ourselves going; they are resources (just as our memories andperceptual sys tems are resources), but they d o not predetermine whe re theiruse will lead us. The re might be som e enveloping, metaphysical limitationsbuilt in to such a procedure, but there is nothing w e can d o about that, and alot that can be opened up and uncovered in the meantime.I naively expected this essentially simple aim would strike everyone asobviously desirable. I knew that som e resistance mig ht be predicted on thebasis of Durkheim's ideas about the untouchable, "sacred" basis of ourdeepest social rules, but I adm it that I did not think this would be m ore thana residual effect. I massively underestimated the way in which such a sim pleidea, depend ing only on w idespread and well-ingrained scientific "instincts,"would evoke a response embodying all the convoluted complexity of thesoc ial "organism" we are pleased to call "academ ic life."To explain why s o man y of the responses to the strong program had littleconnection w ith the words on the page , I need to go on to Bartlett's Seco ndLaw. This might be called the "Law of Conventionalization." Instead ofEbbinghaus's nonsen se syllables, Bartlett studied rememb ering by usingpieces of real, living culture. H e located folk stories and my ths belonging toone ethnic group and presented them to members of a second ethn ic group.Th e second group w as his experimental subjects. He studied what happenedto the mem ory of these stories by getting his subjects to recall them at a laterperiod, and then again ev en later, and so on. He found a gradual chang e inhow the story was remembered, but a change with interesting sociologicalproperties. First, the story was progressively simplified: it lost detail charac-teristic of the original culture, but lost them in suc h a way that the mismatchwith the seco nd culture was diminished. Then, the recollected version of thestory even began to have detail added to it, but of course, detail drawn fromthe second culture. S o once again, it was shifted in the direction of the cultureof the person who was remembering it. These selective and constructiveprocesses proved to Bartlett both the creative and the culturally shapedcharac ter of the mechanism s responsible for mem ory. He called this process"conventionalization."Le t me quickly give you an exam ple of conventionalization in the writingsof critics. Here is one from a recent article by Stephen C ole (1996), published


5/14

376 Science, Technology,& Human Values

in a book called The Flight from Science and Reason. In this article one ofmy own contributions to the strong program is identified as having a "neo-Marxist orientation" (p. 275). It is assimilated to Marxism despite the factthat neither Marx nor Marxism has received more than the merest mentionin anything I have ever written. Cole takes as his target an article called"Durkheim and Mauss Revisited" (Bloor 1982). As the title imp lies,Durkheim, not Marx, was my model, but Durkheim gets no mention by Cole.One of the examples I used in that article was based on historical work(mainly by M argaret Jacob and Jack Jacob) on what is called the "corpuscularphilosophy." The corpuscular philosophy was the theoretical basis of thewo rk of early modem scientists such as Robert Boyle and Isaac New ton. O neof its interesting features is the doctrine of the passivity of matter. Th e workof these historians suggests that part of the attraction of the idea that matteris passive lay in its social use. It allowed a theory of matter to be constructedwhich could be used as a model for how society should be ~ r g a n i z e d . ~hepassivity of matter also gets no mention by Cole. Instead, the thesis of thearticle is reconstructed as Bloor saying "that Boy le's law w as influenced byhis conservative political beliefs" (p. 275). The re was, however, no mentionof Boyle's law in the article. Nor, for that matter, did I characterize Bo yle as"conservative." (I mig ht add that Cole offers no alternative , or better, solutionto the problem addressed: of why B oyle and Newton insisted that matter waspassive.) So you can see how some details are dropped, while others areadded, all to m ake the thing remembered fit into salient cultural categories.Rather than my attempt being assimilated to Durkheimian anthropology, it isconventionalized into the polemically more convenient category of "Marx-ism." I have no great objection to such a label, though I would not use it ofmyself, and I should have expected a sociologist such as Cole to be sensitiveto the differences between M arx and Du rkheim. But labels apart, I think SirFrederic might have been proud of Professor Co le for providing such a niceillustration of his law.

Becoming ReflexiveIt might look as if I am merely using B artlett's laws to poke fun at critics,

saying they h ave misremembered or misunderstood what they are criticizing.But that would be a limited and one-sided use. It would also be against thespirit of the strong program itself to imply that others might be causallydetermined by such law s while we , the sociologists, are not. No: the point isthat if Bartlett's laws really are deep, then they must apply to ourselves, aswell a s our critics. We, too, must be producing com plicated and convoluted


6/14


responses to sim ple and direct things that are said to us. We, too, must bestereotyping, elaborating, and conventionalizing the work of our critics. Noone does this deliberately, and if it were deliberate, that would b e anotherprocess, not the one Bartlett was describing. We do these things unwittingly.We m ay feel that a question fails to do justice to our position, so instead of adirect answer we will explain why it is incorrectly posed. Seeing the questionas indicative of underlying attitudes, we may respond to these, rather than thequestion itself. In good conscience, and with the intention of getting to theheart of the matter, we will thus exemplify Bartlett's First Law. Given thecomplexity of our debates, this will lead us into necessary simplifications toexpose, as we think, the real logic of the issue, and thereby we obey Bartlett'sSecond Law.

W here does this leave rational discourse? If our basic cognitive machineryis selective and creative in the way Bartlett described, it is difficult to se e howwe can carry on rational arguments with one another. At this point somephilosophers proceed as if our minds are made up of two different parts, onlyone of which is subject to causal laws, while the other, more rational, part canbe mobilized s o as to overcome these deplorable tendencies. But then w e arecaugh t in a dilemma. On the one side is the despairing idea that we are doom edto misunderstand one another. On the other side there is a dualism and atranscendentalism, en ticing us away from the real world of cognition stud iedby B artlett and his scientific colleagues. If w e wish to be both scientific anddown to earth, and reasonable and rational beings, how are we to make sens eof the situation? Can w e escape the dilemm a?

One possible way out is as follows. Perhaps a large number of mecha-nisms, each with a characteristic individual "fault," can be made to interactwith one another to produce a collective mechanism which does not have thissam e fault. Perhaps the faults can be m ade to cancel out. In other words, wemust lift our gaze from psychological processes to sociological processes.Let m e give you an exam ple of a sociological move of this kind. The so urcemay com e as a surprise: it is drawn from the work of Karl Popper (1963)whois fam ous for his advocacy of what he called "critical fallibilism." Sc ienc eand rationality, he said, are driven along by criticism: dogma is bad andprotects error from the liberating effect of criticism. Science, therefore, mus tbe undogm atic. But, a critic might say, isn't this highly unrealistic? Don't weknow that people are dogmatic and lack m uch capacity for self-criticism? Itlooks as if Popper was hopelessly op timistic about human nature because hewanted to build a self-critical institution out of unself-critical individuals.Th is would be to underestimate Popper. It is logically possible for science asan institution to be se lf-critical, even if not a single scientist were cap able ofbeing self-critical. The secret lies in getting the right socia l arrangement. If


7/14


we can get one dogmatist to criticize another dogmatist, then dogma getscollectively criticized. Popper's idea was that we need a pluralistic socialsystem or, even better, a competitive social system. In this way we canachiev e self-criticism at the higher level even though it is absent at the low erlevel.

Poppe r's idea provides a model for transcending Bartlett's laws, but weneed to refine it and add to it. First, we need to assume something like anatural tendency, of a generalized kind, disposing us to monitor and s anctionand criticize one another and to be sensitive to such sanc tions by others. Inthis way individual, subjective tendencies to assimilate , construct, and "con-ventionalize" can be held in check. But, second, we must not place all theemphasis on negative sanctioning. If the end result is going to avoid astultifying atom ism or an anarchic individualism, w e must be ab le to explainthe em ergence of shared patterns of belief. Th e sanctioning and control m ustwork in a concerted way to defend or sustain something positive, and thissid e of the matter has not yet been provided for.

To meet this need, notice that Bartlett-like mechanisms may generateproblems under so me cond itions, but under others they are positive virtues.Th ese "faults" can work for us rather than against us, and usually do workfor us. If they did not, it is difficult to see how they could have evolved. T heirpositive contribution is that a group of organisms interacting on the basis ofcognitive machinery of this kind will gradually get more and mo re similar toone another. Their internal representations of the world will converge, so thatthe organis ms will become mo re cognitively alike. They w ill com e to sharewhat Bartlett called a "schema." The schema, or conventionalized repre-sentation , will become the baseline for the operation of the very process bywhich it was generated in the first place. If we combine this effect with ageneralized sanctioning tendency, of the kind just postulated, w e produce acollectiv e mechanism with an intriguing feature. It will generate somethin gvery like a so cial norm o r a social institution but, of course, a cog nitive normand a co gnitive institution. Bartlett mechanisms, therefore, make evo lution-ary sense because they a re wonderful for creating institutions. They "prime"the system-that is, set it going-and then sustain it once it has been created(see Barnes 1983; Haugeland 1990).I want to draw your attention to two important features of this argument.First, the idea that the Bartlett mech anism s work, as it were, for both goo dand ill is very characteristic of experimental psychology. Here is the pre-viously mentioned D onald Broadbent on this theme in his book In Defenceof Empirical Psychology (1973): "Despite years of psychological effort," hesaid, "it is still not widely realized in our culture that a man can see som ethingwhich did not happen, and that he does so precisely through the workin gs of


8/14


the system which in other cases makes him see accurately" (p. 63).There arenot two psychological mechanisms, one for veridical perception and one forerroneous percep tion, ust as for Bartlett there were not two m echanisms, onefor good conventionalizationswhich lead to soc ial solidarity and one fo r badconventionalizations which are divisive and lead to m isunderstand ing. Ingeneral, what is needed is a unitary causal account dealing, as we m ight say,"symmetrically" w ith both. Here, then, is one of the sources of the symmetrypostulate of the strong program; it can be seen as grounded in the practice ofCambridge experimental ps y ~ ho lo g y. ~The second thing to notice is that I have been dealing here with certainfundam ental featu res of socia l and cognitive order-namely, the precondi-tions for, and the emergence of, norms and institutions. Admittedly, mypresentation has been simple and schem atic, but whatever its defects thereis , I hope, no denying the subject area in which these remarks are located:they concern the basis of social life itself. I want to em phasize this becausethere is widespread misunderstanding on the question. It has become com-monplace am ong critics of the strong program to say that its advocates treat"nature," or the physical world, as problematic but treat society as unprob-l e m a t i ~ . ~am not pretending that what I have just said, about how normsmight be precipitated by Bartlett mechanisms, itself constitutes a sufficientanswer to these criticisms. But it can remind us that the real state of affairsis not how these critics portray it. These fundamental, sociological themeshave long been a preoccupation of advocates of the strong program. Thenature of social reality itself has frequently been a central issue. The need forsuch a concern is clear and, indeed, unavoidable. The knowledge of nature,being part of culture, is a facet of collective life. To treat knowledge asprob lematic is to treat social reality as problematic; and conversely, to treatsocial reality as problematic means sooner or later confronting the characterof knowledge. Them es of this kind were explored at aprofound level in BarryBarnes's article "Social Life as Bootstrapped Induction" (1983) and in hisbook The Nature of Powe r (1988). Similarly, Steven Shapin's thesis-thatthe problem of cognitive order is the problem of social order-has beenexemplified in detail in his A Social History of Truth(1994) and in Leviathanand the Air Pump (Shapin and Schaffer 1985).

Not a Single Example?"Conventionalization" means that certain stereotyped patterns of thinking,or "schemas," achieve a special status within a group. They com e to be"norms" sustained in the course of interaction. Ideas of this kind have played


9/14

380 Science, Technology,& HumanValues

a prominent role in empirical studies of the changes in scientific know ledge.What may be called "reception studies" conform closely to the Bartlettparadigm of taking an element of culture from one group and studying itstransformation as it is absorbed into the think ing of another. Arecent and fineexam ple of this gen re is Andrew Warw ick's (199 2, 19 93) technicallydetailed account of the reception of Einstein's relativity theory in Cambridge.Warwick argues that Einstein's 1905 article was understood in a quitedifferent way in Cam bridge from how it was understood in Germany. Indeed,there was not even a unitary understanding of it within Cambridge. Altho ughit was a minority interest, prior to 1905 there were two small but differentschools working in the area that later came to be known as "relativity theory."One of these groupings centered on the mathematician Cunningham, theother around the experimentalist Campbell. Cunningham, coming from thehighly specialized background of the mathematics Tripos, saw Einstein'swork as a contribution to an approach based on certain mathematical tech-niques and transformations. (These derived from a technique known as"inversion," a favorite theme in Tripos examinations.) Campbell, embracing theself-consciously different ethos of the Cavendish Laboratory, saw Einstein'sarticle in terms of its implications for his own version of relativity. Tha t wastailored to current lines of experimental work, used the model of Faradaytubes, and was fashioned with different, and simpler, mathematical tech-niques than those used by Cunningham. These two scientists and theircoworkers w ere not "misreading" Einstein's article; they were simply readingit. Their responses w ere not, and could not, be determined just by what w ason the pages before them. The requirements of intelligibility and comm uni-cation involve their bringing to bear on Einstein's article the full range ofinterpretive resources at their disposal, and those resources were different ineach case. Each of them "conventionalized" the meaning of the words,equations, and concepts they encountered in terms of the collectively heldschemas of the groups to which they belonged and in terms of the meaningsthat functioned as intellectual currency within it.Warwick's ability to ca ny through such an analysis at the mathematicaldep th called for in this case makes it an outstanding example of its kind. It isno criticism of the work to observe that the underlying explanatory model isnot, in itself, a novel one. In light of this fact it is interesting to co me back tothe article by Cole (1996), mentioned earlier, and reflect on his claim thatthere is not a single example he knows of which could count as a real, andwell-documented, case of the content of knowledge being significantlyinfluenced by social factors. I think Warwick's (1992, 1993) article refutesCole-as does all the othe r work in the sam e genre as Warwick's. Clearly,what looks to som e of us like a demonstration of the social determination of


10/14

Bloor I Remember the Strong Program ? 381

content must look to Co le like something else. He thinks those who believethat the content of knowledge is significantly influenced by social variableshave either a sloppy definition of "social" or a sloppy definition of "knowl-edge" or both. As he puts it, they have not clearly defined their independen tand dependent variables.Let us look carefully at Cole's (1996) methodo logical specifications andsee how they relate to historical work of the kind furnished by W anvick (1992,1993).To make their case, says Cole, the sociologist of know ledge must showthat "cognitive content (as it was accepted by the scientific community)turned out one way rather than another because of some social process"(p. 280). The emphasis on causality here is quite right. It is the conditiondealing with acceptance by the scientific community that needs examining.For Cole the sociologist must demonstrate a social influence on m ore thanthe "doing of science." The influence must be on what he calls a "know ledgeoutcome." This is defined as "a piece of science that has come to be acceptedas true by the scientific community and thereby entered the core know ledgeof that discipline" (p. 278). Cole wants to in troduce methodological exacti-tude into the proceedings to offset what he, rather unkindly, calls the rhetori-cal tricks (p. 280) of the constructivists. Methodo logical exactitude is cer-tainly desirable, but I do not think Cole's own formulations are as sharp inthis respect as they should be. He has specified a set of suficient conditionsfor the dem onstration of social influence but proceeds as if he has furnishednecessary conditions. In this I think he is wrong. I want to suggest that thereis no good reason at all for seeing Warwick's impressive article as any thingother than a demonstration of the influence of social variables on know ledgeoutcom es. The social variables in question are the local traditions of work,and the outcomes are the particular readings of what later came to beinstitu tionalized as relativity theory.'

How Experiments EndI can think of a reason why a critic of the socio logy of knowledge mightfollow Cole in mistaking sufficient conditions for necessary conditions.Suppose someone holds the fo llowing two be liefs. First, they believe "real"

knowledge is wholly determined by its object, so any residual, subjectivefactors would necessarily detract from its status. They think in terms of azero-sum game with "nature" and "society" trading off agains t one another.Every claim sociologists make about social influence is heard as a corre-sponding denial of the role of reality. Second, they believe that whenknowledge comes to be finally accepted by the scientific community it is


11/14

382 Science,Technology,& Human Values

almost certainly "real" knowledge, so it must have been shorn of any socialinfluences on it. When properly conducted experiments on some topic arefinally brought to an end by the experts reaching agreement, then, on thisargument, they prov ide us with the unmediated contact with reality charac-teristic of real knowledge. Such a critic would reason that any significantsociological claim must, therefore, be directed at such fixed and settled caseswhich a re the only examples of real knowledge.Clearly, such a procedure would stack the cards against the sociologist ofknowledge. Were anyone to point out some identifiablebias, prejudice, myth,or conventionalization at work in the experiments, that would merely showthat the necessary process for generating "real" knowledge had not yet runits course. Anything the sociologist could get a grip on would be immediatelydisqualified as an instance of the kind of thing they ought to be grapplingwith-it would not be real know ledge. The position of the crit ic is thus adisguised tautology. Perhaps even more insidiously, such a stance draws aboundary which inh ibits the flow of wholly reasonable inductive inferences.Of course, the solid core of present-day, esoteric, scientific knowledge isgoing to be enormously difficult to analyze sociologically, but there ought tobe no reason why reasonable, inductive conclusions about the character ofthat knowledge cannot be based on other cases accessible to, for exam ple,historians of science. The result of disallowing these, and treating them as ifthey fail to satisfy the necessary conditions of an adequate analysis, is that itblocks off inductive inference about "real" knowledge. Of course, this is whatsomeone subscribing o the two beliefs just described would want to do. Theywould believe they were protecting science against its detractors--exactly asCole does.I hope no one will think that, in spe lling out these assumptions,I am simplyfighting old battles. W ho now believes these things, apart from a few residual"positivists"? Well, to begin w ith, old or not, such battles haveflaredup againand basic principles need reasserting. I also suspect that many of those whofeel they have left old-fashioned "positivism" behind may be deludingthem selves. But my target is notjust Co le. Other critics of the strong program,com ing from within the approach broadly known as "constructivism," workwith exactly the same stereotypes.Those of you familiar with Bruno Latour'sattack on the strong program will recognize a remarkably similar strategy(see, in particular, Latour 1992).He also thinks that sociology is playing azero-sum game with the ideas of nature and society ; that is why he rejectswhat he understands as the symm etry principle of the strong program in favorof his own generalized version. It is, therefore, not surprising hat critics, suchas Cole, find Latour an ally in this regard. He reaffirms their suspicions about


12/14

Bloor I Remember the Strong Program? 38 3

the sociology of know ledge. They find their worst fears played back to themand confirmed.

ConclusionWhat is the culture against which it is best to understand the strongprogram ? What are the assumptions on which it is really based? I have triedto convey a sense of this-as far as my own work is concerned-by using

Bartlett's (1932) book and by showing you (quoting Broadbent 1973) hownaturally a causal and "symmetrical" stance emerges from the study ofcognitive processes in experimental psychology (at least, in its Cam bridgeversion). There are obvious differences between psychology and sociology,but often those differences are accentuated by unnecessary and whollycontingent disciplinary boundaries. The fate of Bartlett's classic book is acase in point. At one s tage of the argum ent, in the chapter called "Conven-tionalisation," he looked at the development-beginning in the World Warof 1914-1918--of devices used for the night detection of attacking aircraft.His interest was in the role of teamwork, in both the development and theoperation of such systems, and in the different national styles of the technol-ogy that resulted. The relevant section, which Bartlett described (ratherstrikingly in retrospect) as an exam ple of "social constructiveness" (p. 276),is no m ore than a page long, and it is, by admission, sketchy. Sketchy or not,the fact remains that back in 1932, Bartlett had actually started to do thesociology of sc ience and t e c h n ~ l o g ~ . ~ome aspects of Bartlett's concern withpractical, often military, problems of this kind were taken up within thediscipline of applied psychology, but "social constructiveness" was not. Inaccordance with Bartlett's own laws, his contribution was "conventional-ized." It w as remembered selectively and adjusted to fit the dominant schemaof the emerging discipline of experimental psychology-and, of course, his"social constructiveness" was the bit that was forgotten.

Notes1. In what follows, I shall explain the points of the program that are needed to make thediscussion self-contained . For a fuller account see B loor (1991) and Barnes, Bloor, and Henry

(19%).2. The significance of the Jacobs' argument was f m t made clear to me by Shapin (1980).3. By "symmetry" I mean the principle that the form of explan ation used by a sociolo gistshould not depend on the sociologist's own evaluation of the truth of the belief to be explained.


13/14


What we think of as atruth is just as sociologically problematic as what we thinkof as a falsehood.In both cases what is at issue are the local, contingent causes of credibility.4. Claims of t h ~ s ind are to be found in Bruno Latour's writing (such as his 1992 article"One M ore Tum after the Social Turn"). They have been echoed in, for example, Joseph Rouse'srecent book Engaging Science (1996).5. There is also no clear distinction between C ole's (1996) catego ries of "doing science" and"knowledge outcomes." Indeed, there are plenty of reasons drawn, for example, from thedow n-to-ea rth tradition of pragmatism, for doubting whether the distinction can be sustained ata fundamental level. Is it not reasonable to see knowledge as embedded in the doing, or thepractice, of science?6. Unfortunately, Bartlett (1932) gives no references to underpin his exam ple. As an exercisein the history of "constructivist" approaches it would be interesting to track down the basis ofhis remarks.

ReferencesBarnes, B. 1983. Social life as bootstrapped induction. Sociology 4524-45.. 1988. The nature of power. Cam bridge, UK: Polity.Barnes, B., D. Bloor, and J. Henry. 1996.Scienh~c know1ed ge:A ociological analysis.Chicago:Chicago University Press.Bartlett, E C. 1932.Remembering: A study in experimental and socialpsychology. Cambridge,UK: Cambridge University Press.Bloor, D. 1991. Knowledge and social imagery. 2d ed. Chicago: Chicago University mess .. 1982. Durkheim and Mauss revisited: Classification and the sociology of know ledge .

Studies in History and Philosophy of Science 13:267-97.Broadbent, D. E. 1973. In defence of empirical psychology. London: Methuen.Cole, S . 1996. Voodoo sociology: Recent developments in the sociology of science. In Theflightfrom science and reason, edited by P. R. Gross, N. Levitt, and M . W. Lewis, 274-87. NewYork: New York Academ y of S ciences.Haugeland, J. 1990. The intentionality all-stars. In Philosophical perspectives: Action theoryandphilosophy of mind, edited by J. E. Tomberlin, 383-427 . Atascadero, CA: Ridgeview.Latour, B. 1992. One more turn after the social turn. In The social dimensions of science, editedby E . McMullin, 262-94. Notre Dame, IN: University of Notre Dame Press.Popper, K. R. 1963. Conjectures and refutations: The growth of scientific know ledge . London:Routledge and Kegan.Rouse, J. 1996. Engaging science . Ithaca, N Y Cornell University Press.Shapin,S. 1980. Social uses of science. In The ferment of knowledge, edited by G . S. Rousseauand R. Porter, 93-139. Cambridge, UK: C ambridge University Press.. 1994. A social history of mcth: Civility and science in seventeenth-century England.Chicago : Chicago University Press.Shapin, S., and S. Schaffer. 1985. Leviarhan and the air pump: Hobbes, Boyle and theexperimental life. Princeton, NJ: Princeton University Press .Wanvick, A. 1992. Cambridge mathematicsandCavendish physics: Cunningham, Campbell andEinstein's relativity 1905-1911: Part I: The uses of theory. Studies in the History andPhilosophy of Science 23:625-56.


14/14

Bloor l Remember the Strong Program? 385

. 1993. Cambridge mathematics and Cavendish physics: Cunningham, Campbell andEinstein's relativity 1905-1911 : Part 11:Comparing traditions in Cambridge physics . Studiesin the History and Philosophy of Science 24:l-25.

David Bloor Remember the Strong Program

Documents