Reply to Sober

Abstract

The paper (Shimony, 1989) which Elliott Sober (1989) has commented upon grew out of my efforts to understand the structure of the theory of evolution and to explain it to students. The absence of anything approaching a broadly accepted axiomatization of the theory of evolution makes it difficult to articulate this structure. But one should not complain about this fact. It may very well be that the theory of evolution resists axiomatization or would be little served by it. At a minimum, a good informal articulation of the structure of the theory is a precondition for an adequate formalization. Let me try, then -- before attempting to answer Sober's criticisms -- to lay down some desiderata for an informal articulation of any theory. (a) The domain of phenomena to which the theory applies should be indicated. (There are different ways in which this step may be performed. A theory applying to a very broad domain may be presupposed, and the scope of the theory of interest may be indicated by specialization; or the domain may be described phenomenologically, in a language which is fairly characterized by practitioners as an "observation language"; or some combination of these procedures may be used.) (b) The fundamental concepts of the theory should be presented explicitly. (These will be somehow connected to experience, but in this note I can reasonably ask to be excused from entering into the intricate problem of how this linkage is accomplished. In any case, what makes a concept fundamental is that it is not explicitly defined in terms of other concepts within the theory; definability in terms of concepts in a more basic or more general theory is not precluded, but that is another matter, which can be set aside in the formulation of the theory under consideration.) (c) A set of propositions, expressed in terms of the fundamental concepts mentioned in (b) and possibly also of phenomenological concepts and those of other scientific theories, should be designated as underived within the theory, and these are its principles. (Steps (b) and (c) should not be rigidified. We have plentiful examples from geometry and physics of equivalent formulations of the same theory which differ in the choice of principles and fundamental concepts. In other words, even