Realist Methodology in Contemporary Genetics

Abstract

The present paper is intended to support three main claims. All three, I believe, represent minority views among historians and philosophers of science; they are, therefore, probably of general interest. Boldly stated, the claims to be defended are:

• 1. Neighboring subdisciplines within a single science need not be methodologically uniform, at least not with respect to realist vs. non-realist methodology.

• 2. There is an important distinction between realist methodology on the one hand and ontological (semantic) realism and epistemic realism on the other. One piece of evidence for this claim is the fact that, among working scientists, there is no simple correlation between the use or rejection of realist methodology and the acceptance or non-acceptance of ontological or epistemic realism with respect to the leading entities or theories of concern to them in their (sub)discipline.

• 3. In some very weak sense, methodological realism (which is allied to what Laudan has called ‘axiological realism’¹) is prior to semantic or epistemic realism. The reason is that it is not possible to ground either epistemic or semantic realism adequately without working through the concrete ways in which putative theoretical (or deep sub-structural) entities might enter into concrete experimental and natural situations. Thus one’s way of working (methodology) must presuppose the existence of the very entities about whose existence one is uncertain if one is to accumulate and assimilate the evidence that might establish their existence.

Work on this paper was supported in part by an NSF Scholar’s Award for calendar 1984. This support is gratefully acknowledged. Preliminary versions of the paper were read in April, 1984, at a Conference on Realist Methodologies at the Center for the Study of Science in Society, Virginia Polytechnic Institute and State University and in April, 1985, at the Pacific Division Meetings of the American Philosophical Association. I have been greatly helped by critical discussion on both occasions, by the comments of John Dupre, and for the suggestions of numerous colleagues, especially M. Grene, J. Griesemer, L. Laudan, R. Richardson, P. Siegel, and B. Wallace.