Natural Kinds of Substance

This paper presents an extension of Putnam's account of how substance terms such as ‘water’ and ‘gold’ function and of how a posteriori necessary truths concerning the underlying microstructures of such kinds may be derived. The paper has three aims. (1) I aim to refute a familiar criticism of Putnam's account: that it presupposes what Salmon calls an ‘irredeemably metaphysical, and philosophically controversial, theory of essentialism’. I show how all of the details of Putnam's account—including those that Salmon believes smuggle in such essentialist commitments—can be squared with a rejection of any such essentialist metaphysics. (2) I aim to reveal why Steward is wrong to suppose that, by helping himself to the claim that ‘H2O’ is a rigid designator of a substance, Kripke, too, presupposes something controversially ‘metaphysical’. (3) I aim to show how my proposed account also sidesteps a variety of objections raised by Needham and others who argue that Kripke's and Putnam's accounts of how ‘water’ and ‘gold’ function founder upon the sheer microstructural complexity of the phenomena in question.