Abstract

Millikan seeks to provide “new foundations for realism”, and I argue that she gets half-way to this goal. For she provides a robustly realist account of nature′s kinds: what holds the members or samples of a kind together are properties that recur together for a common causal reason. But full-strength realism requires realism not just about kind-membership, but about persistence across time. I argue that Millikan does not, but could and should, embrace the traditional idea that the conditions on kind-membership double as persistence conditions for members of the kind. Thus could she generate realism about persistence from realism about kinds