How to end the science wars: a review of Harry Collins and Jay Labinger, The One Culture? A Conversation About Science, part I/II

The science wars were a series of skirmishes that took place between scientists and sociologists (loosely speaking) in the 1990s. Sociologists of science were accused of using bad arguments and shoddy scholarship to undermine science; scientists were accused of misunderstanding the sociologists, idealising science, and conspiring to shut down legitimate debate. In 1997 some of the protagonists met at a ‘Science Peace Workshop’ in the hope of finding common ground and clarifying the issues at stake. The result was The One Culture? A Conversation About Science (2001), edited by the sociologist Harry Collins and the chemist Jay Labinger. It has been said, not without justice, that the book spelled the end of the science wars. But the book has its flaws, including several irritants and two serious omissions. This post and the next one are a guide to the 'science peace process.' These remarks are cobbled together from insights I found in the book and from my own reflections on such things as the symmetry principle and the internal/external distinction.

Be charitable in debate. Of the 35 chapters in The One Culture, the one I found most enlightening was by N. David Mermin, a theoretical physicist who took on a whole squadron of sociologists in reviews and letters published in the magazine Physics Today in the mid-1990s. His targets were The Golem (1993), by Harry Collins and Trevor Pinch, and Scientific Knowledge: a Sociological Analysis (1996), by Barry Barnes, David Bloor and John Henry. In his contribution to The One Culture, Mermin describes the slow process by which he came to understand what his interlocutors were trying to say, and he concludes with three lessons that are as simple as they are indispensable:

Focus on the substance of what is being said and not on alleged motives for saying it.

Do not expect people from remote disciplines to speak clearly in or understand the nuances of your own disciplinary language.

Do not assume that it is as easy as it may appear to penetrate the disciplinary language of others (97-8).

Mermin says that sociologists, not just scientists, fell foul of these rules. This is worth stressing, since the ‘Introduction’ and ‘Conclusion’ of the book mention the scientists’ habit of dismissing sociologists as science-haters but do not mention the sociologists’ habit of dismissing scientists as science-worshippers (see esp. 5-6, 297). Mermin was a victim of this habit, to judge from his account of his exchanges with sociologists:

...the response was not to what I was saying, but to why I might have been saying it. Both times I was taken to be responding to a perceived violation of something I held to be sacred [ie. science], and my actual criticism--they were paying insufficient attention to the broad coherence of an extensive body of knowledge--was read as a charge that they were personally biased (against relativity, for astrology). In both cases the imagined charge was both denied and dismissed as irrelevant, and the substance of my criticism was not addressed (94).

Acknowledge past relativism. Now for one of the serious omissions. Arguably, the main reason for the progress of the ‘peace talks’ between scientists and sociologists is that the latter no longer make radical claims about the limits of scientific knowledge. None of the contributors to The One Culture seem to agree with Barry Barnes and David Bloor that ‘there is no sense attached to the idea that some standards or beliefs are really rational as distinct from merely locally accepted as such.’* Nor do they deploy the arguments that were routinely invoked to support such claims before the year 2000, such as the Quine-Duhem thesis and the underdetermination of theory by the evidence. Some contributors mention these arguments, but only to attribute them to other writers (eg. 31-3, 196). If the science wars are over, this is largely because the more radical sociologists have beaten a hasty retreat.

Remarkably, this retreat is scarcely mentioned in The One Culture. There is a glimmer of recognition in a footnote by Harry Collins in which he says that he changed his mind about relativism some time around 1980, abandoning ordinary relativism in favour of the methodological version (of which more below). This is an isolated case, however, and it does not sit well with the claim, repeated throughout the book, that the sociology of science has never had any consequences for our evaluation of scientific knowledge, and that those who charge sociologists with ‘undermining’ science are being precious or malicious or both. In the 1970s, Barnes and Collins both implied that there is no objective reason—and could never be an objective reason—for preferring the theories of present-day Western scientists over any other theories.** If that does not count as undermining science, what does?

Be methodological relativists about reasons not truth. The full-blooded relativism of the 1970s has been replaced by the dictum that historians and sociologists should not refer to the truth or falsity of the beliefs they try to explain. This dictum is known as ‘methodological relativism’ or ‘the symmetry principle’, and The One Culture takes it very seriously indeed. The editors say, rightly, that it is the most common source of disagreement between the book's contributors (297). It is also the topic with the most entries in the book’s index, outrunning ‘relativism’, ‘Thomas Kuhn’, and even ‘science studies.’ Some contributors argue that methodological relativism is a precondition for a mature sociology of science, others that it makes for aimless and ill-formed history, and others that it leads back to the epistemological relativism that it was supposed to replace.

I believe that this entire debate is red herring. The problem is that The One Culture frames methodological relativism as the thesis that the truth of a belief has no place in a sociological explanation of that belief. The scientists dispute this thesis, pointing out that many beliefs (eg. ‘it is raining today’) are caused, in part, by their truth (eg. the fact that it is raining today). The sociologists seem to concede that the truth of a belief can (partly) explain the belief, but they deny that this kind of cause is a proper subject for a sociologist.

There is a genuine disagreement here, but it has little to do with the science wars, since it is possible to explain a theory in a manner that is triumphalist and rationalistic in the extreme without ever referring to its truth. All you need to do is idealise the experiments and arguments that the scientists gave for the theory, ignore the interests and prejudices of those scientists, and play up the interests and prejudices of their rivals. If the only novelty of the sociology of scientific knowledge (SSK) was that it made no reference to the truth of theories, it would be scarcely distinguishable from the explanations that historians, scientists, and sociologists have traditionally given of scientists’ beliefs.

In my view, what was really new about SSK, and what raised the ire of scientists, was that it tried to reverse the usual pattern of explanation. It drew attention to the arguments in favour of false theories and to the interests and prejudices that lay behind true theories. In other words, it explained true beliefs in terms of bad reasons and false beliefs in terms of good reasons. Its methodological relativism was directed at reasons, not at truth.

The contributors to The One Culture are unable to see this because they do not distinguish clearly between reason and truth. Even the scientists are guilty of this conflation. Consider Jean Bricmont and Alan Sokal. They use the ‘it is raining’ example to show that the truth of a belief can explain that belief. But later they accuse sociologists of eliminating evidence, not truth, from their explanations: ‘Could one conceivably explain scientists’ beliefs about the Earth’s climate without making any reference to the currently available evidence concerning the Earth's climate?’ (245, original emphasis).

Insofar as Bricmont and Sokal do distinguish between truth and evidence, they seem to assume that they are very strongly correlated. They assume, that is, that every true theory has always had better evidence in its favour than its false rivals. The One Culture would have been more fruitful if the authors had confronted this assumption head-on instead of wrangling over the question of whether the truth of a belief can be part of the explanation of the belief.

Not only is the latter question irrelevant to the science wars—it is also more easily resolved than one might infer from the intensity with which it is debated in The One Culture. Harry Collins is the pioneer of methodological relativism and one of its staunchest supporters. Yet even he concedes that the ban on truth only applies to some kinds of history and not to others (192). It seems that Collins has no in-principle objection to the idea that (for example) the moons of Jupiter are part of the explanation of Galileo’s belief in the moons of Jupiter. If this is what methodological relativism amounts to, it is hard to see what all the fuss is about.



*Barnes, Barry, and David Bloor, ‘Relativism, Rationalism, and the Sociology of Knowledge’, in Rationality and Relativism, ed. Martin Hollis (Cambridge: MIT Press, 1982), pp. 21–47, on 27.

**Barnes, Barry, Scientific Knowledge and Sociological Theory (London: Routledge & Kegan Paul, 1974), 154; Collins, Harry, and Graham Cox, ‘Recovering Relativity: Did Prophecy Fail?’, Social Studies of Science, 6 (1976), 423–44. Expand post.

Malheureusement je ne traduis plus tous les posts que j'écrit sur ce blog. Si vous voudriez lire les posts qui ont été traduits, veuillez cliquer ici.

Agrandir ce message.