Harry Collins, Methodological Relativism, and Sociological Explanation, Pt. 2 August 20, 2011Posted by Will Thomas in Uncategorized.
Tags: Bruno Latour, David Edgerton, Harry Collins, Imre Lakatos, Jerry Ravetz, John Ziman, Joseph Agassi, Karin Knorr Cetina, Karl Popper, Mary Hesse, Paul Feyerabend, Peter Winch, Simon Schaffer, Stephen Toulmin, Steven Shapin, Thomas Kuhn, W. V. O. Quine
In my previous post on Harry Collins’ ideas about “methodological relativism”, I discussed how in the early 1980s Collins began explicitly using relativism as a “natural attitude” that could be used to produce “sociological explanations” of scientists’ behavior. Methodological relativism was premised on a clear delineation of tasks, which makes it appropriate for the sociologist, but not for scientists.
However, this delineation of tasks remained incomplete: in particular, the relationship between sociology, philosophy, and history of science remained confusingly unresolved. Further, it was unclear what sociological fruits would actually be obtained via methodological relativism. Finally, it left unclear what the relationship was supposed to be between the sociology of scientific knowledge and the more general sociology of knowledge, upon which STS appears to be based.
A key point to keep in mind here is that the apparent gains of new approaches could be easily amplified as a potential resolution to the frictions at that time between philosophy and history of science. (See also this blog’s series on history’s “Great Escape” from philosophy of science.) While philosophy traditionally addressed questions bearing on how knowledge could be considered valid, historians demanded more realistic portrayals of how knowledge came to be validated.
In his paper, “Stages in the Empirical Programme of Relativism” (1981), Collins explicitly referenced the crucial role of temporal processes in the validation of scientific work: “Theories are now seen as linked to each other, and to observations, not by fixed bonds of logic and correspondence, but by a network, each link of which takes time to be established as a consensus emerges and each link of which is potentially revisable — given time.”
Ostensibly, philosophy was already well on the way to allowing the necessity of time with Karl Popper’s falsification, W. V. O. Quine’s underdetermination of theories, and Thomas Kuhn’s paradigms. Collins also cited the work of Stephen Toulmin, Mary Hesse, Imre Lakatos, Paul Feyerabend, Jerry Ravetz, and physicist John Ziman.
While Collins conceded that “it is sometimes rational to disagree,” he saw it as important to apply sociological rather than rational explanations within periods in scientific work when it was impossible to arrive at rational agreements. Indeed, Collins’ empirical work on gravity wave detection has thrived on that field’s slow march to robustness. Still, one should remember that Collins insists that it is possible to achieve purely sociological descriptions of all stages of scientific work, including the ones where rational interpretations are made.
It is, though, unclear to me that methodological relativism ever proved very fruitful in generating sociological knowledge. As near as I can tell, only two major points emerged from the relativist program. First, there is Collins’ own sociology of calibration (dating to the mid-’70s), which holds that disputes over the validity of observational results require some basic level of trust in observers and their equipment. Second, “boundary negotiation” (dating to sociologist Thomas Gieryn’s work in this same period) could be used to include or exclude figures where closure is sought but agreement cannot be achieved whether rationally or by a confluence of interests.
The possibility of programmatic stagnation was detectable at the time as well. As soon as Collins articulated methodological relativism, the ethnologist Karin Knorr Cetina expressed skepticism about its potential. In her reply to Collins’ “Stages” paper*, she suggested that it might lead only to the pronouncement of the simple fact that there exist “processes of negotiation which surround experimental ‘observation’.” Beyond this point, it was likely that there were few generalizable sociological points — only historical particulars:
…it was after all to be hoped that their privileged access to real life data would enable sociologists of science to arrive at (social) mechanisms of consensus formation which escape, or are of no concern to, the historiography of science.
If these mechanisms, as Feyerabend said again and again in regard to science, are constituted afresh within each research tradition, what general patterns could we arrive at? To paraphrase [philosopher of social science Peter] Winch, if these mechanisms exist only in and through the ideas which impose themselves in a particular social context, does it not follow that they must be an unsuitable subject for generalizations?
It is my impression that proponents of a relativist programme, whenever they did move to establish the substantial characteristics of a field of study (as in anthropology and linguistics), quickly abandoned their original relativist preoccupations in order to immerse themselves in the study of the intricate structures and processes of their domains.
If methodological relativism was supposed to achieve sociological explanation, but it only led to an affirmation of historical (or ethnographical) particularity, then along what lines should work proceed? This important question was never addressed, I would argue because the division between generalized sociology and particularized historiography, which Knorr Cetina deemed important, collapsed amid a new alignment between the fields that was developing at just this time.
This alignment was legitimized, I would argue, primarily because proponents of the alignment were able to link the sociology of scientific knowledge with the already-accepted importance of anti-Whiggism. In his 1981 TRASP paper, Collins himself observed the “Radical Programme” would be useful for historians who wished not to be “wise before the event”. However, acceptance of the program was by no means necessary to achieve this aim. In fact, in using the “wise before the event” phrase, Collins was quoting a 1963 paper by Joseph Agassi†, who was certainly no relativist.
Beyond this vaguely salubrious anti-Whiggism, sociological relativism also gained currency in historiography because — in the form of a social constructionist realization — it was touted not just as a useful tool for the historiography of science, but as a path to the total reformation of what was often claimed to be a fundamentally compromised historiography. In 1982 Steven Shapin wrote‡ of the “sociological reconstructions” of the history of science, inverting Lakatos’ ahistorical, philosophically structured “rational” reconstructions.
It remains unclear what exactly it means to sociologically reconstruct the historiography, and whether sociologically reconstructed history is meant to connote proper history. In his 1991 criticism of Bruno Latour’s Pasteurization of France, Simon Schaffer would appear to have argued for something like a regressive search for the origins of the various strategies scientific figures have found plausible or compelling. This accords with Collins’ own observations a decade earlier that the clearest place for sociological work is in the formative stages of scientific ideas.
Alas, the alignment of sociology and historiography has not produced its promised reformation of historiography, using a regressive methodology or otherwise. Mainly it has resulted in a much-remarked-upon surfeit of studies of various rhetoric-laden disputes taking place prior to the “closure” of ideas, or the successful appropriation of ideas (technologies, etc.) within new contexts.
Such studies are deemed important, apparently because closure is supposed to equate with an ascendancy to social authority described by the general sociology of knowledge. Per Gieryn’s work, closure within the ambit of “science” is supposed to lend a particular authority (though I believe this point has always been more assumed than demonstrated). Meanwhile, Latour’s “networks” approach emphasizes the more heterogeneous sources of authority available to program-builders, and the necessity of continually building new networks to achieve new spheres of authority. Ultimately, though, all approaches in this tradition focus on closure or reinterpretation in appropriation, because the subsequent career of ideas that have achieved authority (in the form of technologies, policies, pedagogy, etc.) is apparently regarded as a trivial extension of some initial task of closure or appropriation.
David Edgerton’s focus on “old” technologies attempts to question this view of what is important in history. He suggests that by focusing on instances of technological novelty and appropriation, historians have not recognized how little they know about the patterns of technology use over time. The same could be said of scientific ideas, instruments, policies, and so forth.
I would argue that the history of methodological relativism allows us to partially understand why it has become nearly impossible to focus on patterns (as Edgerton would recommend) rather than on isolated instances of novelty or appropriation.
Basically, I believe that Knorr Cetina was quite correct in her assessment that the intellectual prospects of methodological relativism were slim. I would go on to argue that, in fact, no robust SSK program ever materialized. We can reconcile this claim with the revolutionary effects often attributed to SSK by supposing that the one or two gains that could be attributed to methodological relativism were touted as crucial advances over a traditional history, philosophy, and sociology.
Thus methodological relativism enjoyed a second career in STS as a kind of crucial critical (rather than methodological) insight, where others’ failure to apprehend it is habitually used to explain various discords in the science-society relationship. Dissatisfied with what he views as the lack of intellectual productivity in STS’s critical program, Collins has himself moved on to what he views as the more methodologically productive sociology of expertise and experience.
Meanwhile, by virtue of being presented as the wellspring of all good historiographical things, methodological relativism’s second career also licensed the growth of a fragmented historiography, which remains preoccupied by sociological/cultural themes surrounding the origins or appropriation of allegedly authoritative ideas. While this has helped spur many useful studies in the culture of and surrounding science, it has not resulted in a substantially revised synthesis of the history of the sciences.
Full references for Collins’ pieces are given in Pt. 1.
*Karin Knorr Cetina, “Relativism — What Now,” Social Studies of Science 12 (1982): 133-136.
†Joseph Agassi, “Towards an Historiography of Science,” History and Theory, Beiheft 2 (1963): 1-23.
‡Steven Shapin, “History of Science and Its Sociological Reconstructions” History of Science 20 (1982): 157-211.