Harry Collins, Methodological Relativism, and Sociological Explanation, Pt. 1

When I was at SEESHOP5 in Cardiff last month, I had an opportunity to talk a little with Harry Collins about the history of his work, its relationship to the history of science, relativism, radicalism, and STS.

People involved in Collins’ “Sociology of Expertise and Experience” (SEE) project would like their work to inform future STS scholarship.  However, by their estimate, STS has been reluctant to take up SEE.  This has led the SEE crowd to chart their own course, distinguishing their work as committed to a constructive deliberation about the nature and social operation of expertise, which they would contrast to an argumentation-averse, and ultimately nonconstructive critical orthodoxy prevailing in STS.

Now, STS distinguishes itself by a sort of ambivalently* radical relativist intellectual position, descending from the sociology of scientific knowledge (SSK) project of the late-1970s and 1980s.  By attempting to define the bounds of expert authority, the SEE project is often taken to be a retreat from STS-brand radicalism to a more traditional set of ideas about expertise.  It has sometimes been paired with Bruno Latour’s own apparent retreat (pdf) around the same time as the SEE project got started, in the early 2000s.

Collins denies that SEE represents any shift in his critical position: for him it is just a shift to a different methodology and a different sort of problem.  His pre-SEE work was simply grounded in the idea that radical relativism was a heuristic posture necessary for gaining a purely sociological knowledge about how scientific communities come to agreement and manage disagreement.

I remarked to Collins that it was odd, given this oft-stated stricture on his relativism, that he was ever perceived by STS insiders as having taken a radical position.  At which point Collins told me that, in fact, he did take a radical position up until 1981, when he felt he could no longer make a strong case for a radical relativism.  He pointed me to his paper:

(1) “What is TRASP?: The Radical Programme as a Methodological Imperative,” Philosophy of the Social Sciences 11 (1981): 215-224.

This pair of posts looks at that paper, as well as his introduction to a special issue of Social Studies of Science, and his reply to criticisms and comments on that issue:

(2) “Stages in the Empricial Programme of Relativism,” SSS 11 (1981): 3-10.

(3) “Special Relativism — The Natural Attitude” SSS 12 (1982): 139-143.

If Collins did have a change of mind in the early 1980s, he did not declare it in the papers.  However, the papers do indeed function as an explicit mission statement for a specifically methodological (aka, “special”) relativism.

In (1), Collins explicitly distances himself from SSK as epistemology: “I do not want to defend relativism.  I do not want to talk about what exists in the natural world nor how we ground our knowledge of it.  Ontology and epistemology are not the subjects of this paper, the subject is methodology of social science.”  Instead, he opts to delineate tasks, reserving “special” relativism for social scientists seeking to explain social phenomena, leaving realism to the natural scientists.  In (3) he follows the vocabulary of epistemological phenomenologists, referring to this as a “natural attitude”:

The sociologists’ métier is the social world.  While the good sociologist of science will understand and, from time to time, participate in his (or her) subjects’ natural attitude to natural reality, his job is also to suspend that attitude in order to make other kinds of observations.  The social scientist can be helped to accomplish this by attending to those occasions when the natural scientist — as a result of some disturbance in taken-for-granted-reality — experiences the natural world as socially constructed.  A relativistic attitude to the natural world, on the part of the social scientist, will encourage attentiveness on these occasions and help him to observe the way that an apparently fixed reality is built up in the community of the natural scientist.

To attain purely sociological explanations, in (1) Collins advised sociologists to adhere to the “Radical Programme”, which was defined as explanations that eschew TRASP — the “true, rational, successful or progressive”.

Collins distinguished his Radical Programme from David Bloor’s “Strong Programme”.  Of the four components of the Strong Programme — (a) causality, (b) impartiality, (c) symmetry, and (d) reflexivity — Collins’ Radical Programme found (a) and (d) dispensible.

Strong Programme SSK was seen as universally applicable to all forms of knowledge (actually, it was an explicit attempt to extend Karl Mannheim’s sociology of knowledge to include the scientific).  So, for the sake of consistency, it held that relativism applied reflexively to SSK as well.  For critics like Larry Laudan, SSK’s claims to the universal applicability of relativisitic symmetry were undermined by reflexivity.  Why should SSK be universally applicable if all knowledge was relative?  Collins’ delineation of tasks side-stepped this problem by simply disavowing “currently fashionable” reflexivity.  Such matters would only be the concern of a prospective “sociologist of sociology of scientific knowledge”.

As to causality, where Bloor sometimes would use the word “cause”, Collins preferred a more delimited “explanation”.  In abandoning the universality of the Strong Programme, Collins’ Radical Programme produced what he called “sociological explanations” appropriate for sociologists.  Collins elaborated on these explanations a little in footnote 5 of (1):

A more general methodological prescription that would save a great deal of confusion is ‘avoid all reference to what is inside actors’ heads in all explanations of scientific knowledge’.  The crucial point for the investigator is not, ‘Why did various actors act as they did?’, but, ‘What effect did these acts have on the reception of views about nature?’  Attempts at the attribution of degree of rationality, motive, personality make-up and so forth all lead to a psychologistic regress.

The question (for me, anyway) is what “sociological explanations” could accomplish, if they did not purport to account fully for the history of scientific thought and work.†  I will address this issue in Pt. 2.

—

*I call STS radicalism ambivalent, because STS does assert (I think rather speciously) that its ideas about the social authority of science are markedly different from others’ appreciation of that authority.  At the same time, STSers habitually deny that they hold an especially radical view about the ability of scientists to produce valid knowledge.

†Note that Latour’s later Actor-Network Theory would actually attempt to make a full account of history while skirting the inscrutability of causation and psychology by referring to the persuasive power of networks and indivisible natural-social “hybrids”.  In 1992 Collins (with Steven Yearley) would deride this tactic as playing “epistemological chicken”.