Abstract: While the topic of natural kinds has long been a focus of work in the philosophy of science, as well as in other areas of philosophy, a generally accepted account of natural kinds is still lacking. Moreover, there even is no general agreement about the kind of account that is being searched for, or about the criteria that a good account of natural kinds should meet. In response to these problems, in the philosophy of science (though not in other areas) there is a trend to move away from the metaphysics of kinds and classification and to turn to epistemological issues and questions regarding the use of kinds and classifications in various contexts. This trend fits well with the renewed interest in doing naturalistic philosophy of science and the increasing calls for bringing philosophy of science closer to scientific practice.
However, abandoning the search for a metaphysics of kinds and classifications is too quick. A metaphysical account of kinds is a crucial element of the explanation why some kinds and classifications are used in the sciences with more success than others, and some ways of grouping things turn out not to be useful at all. After all, barring cases of epistemic luck the reason for the epistemic and practical success of kinds and classifications must be that they adequately represent some aspect of the world or other. But problem for naturalistic philosophers wanting to elucidate the metaphysics of scientific kinds and classifications is whether naturalistic and practice-oriented philosophy of science can at all come up with an account of the metaphysics of kinds and classifications, rather than staying with epistemological and practical issues. Metaphysics cannot be read off from either epistemology or practice: simply examining scientific kinds and classifications and the ways in which investigators in the various areas of science employ them will not reveal their underpinnings. Thus, some a priori considerations need to enter into the picture – but a priori metaphysics is suspect from a naturalistic viewpoint.
The challenge for a naturalistic and practice-oriented metaphysics of kinds and classification, then, is to bring a priori considerations into play without rendering the account insufficiently naturalistic. In this talk I addresses this challenge and explore what a thoroughly naturalistic metaphysics of kinds and classifications should look like. I will do this by examining two core notions in the debate on kinds and classification, namely the notions of naturalness and, in particular, normativity.