Are There Semantic Natural Kinds of Words?

<jats:p>Gareth Evans proposes that there are semantic natural kinds of words. In his development of this theory, he argues for two constraints on the identification of these kinds. I argue that neither of these constraints are justified. Furthermore, my argument against Evans’ second constraint constitutes a direct argument for the existence of semantic natural kinds, something Evans himself never offers. I conclude by sketching some positive details of a more plausible theory of semantic natural kinds.</jats:p>