Abstract

Biological species are often taken as counterexamples to essentialist accounts of natural kinds. Essentialists like Ellis (2001) agree with nominalists that because biological kinds evolve, any distinctions between kinds of biological kind must ultimately be arbitrary. The resulting vagueness in the extension of natural kind predicates in the case of species has led to the claim that species ought to be construed as individuals rather than kinds (Ghiselin 1974, 1987; Hull 1976, 1978). I examine the possibility that causal features extrinsic to the properties of natural kinds are responsible for establishing the unity of the properties of a natural kind. I reject the intuitive idea that laws of nature might act as such an external mechanism because this would entail an account of ceteris paribus biological laws, where there are no plausible truthmakers in terms of kinds or properties. I suggest instead that symbiosis is a plausible external causal mechanism, which explains the evolution of homeostasis in natural kind clusters. This involves the acceptance of an expanded account of evolutionary development as cooperative symbiosis