Read and Woolley on Wray on Kuhn

Thu 17 Oct 2013 01:13 PM

In a recent issue of BJPS, Rupert Read and Jessica Woolley review Brad Wray's book Kuhn's Evolutionary Social Epistemology.* I've been thinking about Kuhn recently, too, because I'm teaching Scientific Revolutions this term.

Read and Woolley survey Wray's argument against critics of Kuhn who would say that there is no sharp different between normal science and revolution. Wray appeals to the fact, mentioned in Structure and developed by Kuhn in more detail later, that revolutions involve a substantial change in taxonomy. The world seen after the revolution is divided along different lines than it was before. Shared standards and data can be (perhaps must be) insufficient to determine the choice between rival taxonomies. So these transitions are importantly different than normal science.

This line of reasoning relies on underdetermination at the object-level of scientific theories and taxonomies. Read and Woolley note that it applies just as well at the meta-level, to the theories and taxonomies used in science studies. Yet the difference between normal science and revolution is just part of the taxonomy of Kuhn's account of science. Contra Wray, they conclude that we are not entitled to the conclusion that there really are revolutions.

They put the point this way:

Wray defends Kuhn’s account in the way that he does, aiming to show that there 'really are' Kuhnian revolutions in science (p. 34). If overall taxonomies are underdetermined even in disciplines such as physics where there is ordinarily widespread agreement, it seems unclear what standards Wray could appeal to in order to justify the socio-historical claim that there 'really' are Kuhnian revolutions.

That much seems right to me, and it is anticipated in Structure. In Section VIII, Kuhn treats Hypothetico-Deductive philosophy of science as if it were a paradigm. It does not fit the history of science very well, so it faces anomalies. According to the the Kuhnian picture, adherents of a paradigm faced with such phenomena will make ad hoc revisions to accommodate them, but the puzzling phenomena may be tautological for a later paradigm. And (he suggests) this is just what obtains for the HD account and his proposed alternative. This does not prove that Kuhn's approach is true, however, any more than the anomaly proved that HD was false. So Kuhn himself makes the reflexive move. Just as he denies that (e.g.) a paradigm in chemistry gets to what's really real, he'd deny that his own historiographic paradigm does.

The concluding sentences of the review go wrong, however. Read and Woolley write:

'Paradigm' is not a scientific phenomenon but a term of art, a tool for enabling the historian to bring an order to the deeply challenging task of understanding defunct science without 'Whiggish' preconceptions. The idea of being a realist or an anti-realist about paradigms involves a misplaced concreteness. How, after all, could one meaningfully be a 'relativist' (or indeed a 'realist') about something that is only a term of art?

First, this seems confuse the word ('paradigm') with the things (paradigms). It's a word that's a term of art, but the question of realism arises for the things. It is coherent to ask whether there are really paradigms, just as it is coherent to ask whether there are anglerfish or pions.

Second, this seems to infer from the fact that the term is a tool crafted for a specific purpose to the conclusion that the category cannot be part of the world. This is what I call amphibolic pragmatism, and I rail against it at some length in SENK.

Third, this asks too much of underdetermination. There is no sharp boundary in uncertainty between our knowledge of commonplace stuff (which might be hidden or far away), unobservable entities (which we can't see or sense in the way that I can see or sense a burrito), or natural kinds (which are structural features of the world rather than isolated particulars). Natural kinds, which may be picked out by terms of art, have as much claim to being real as commonplace stuff does. This is what I call equity realism.

Equity realism contrasts with deep realism, the project of fundamental metaphysics. (Again, this distinction is in the book.)

If Kuhnian considerations underwrite an argument against realism, it's against deep realism. Grant that ontology should be relativized to a way of seeing the world. Even within a paradigm, we can distinguish between the substantive commitments and the parts which are self-consciously arbitrary. So we can ask whether revolutions are posited to be features of the world or merely used as wrenches to dismantle Whiggish history. We can ask whether they are real, on par with the way trees and the plague are real, even if we can't say whether they are really real in real reality.

* DOI: 10.1093/bjps/axs032 [direct link]