OUP accepted manuscript

I analyze Quine’s later writings on analyticity from a linguistic point of view. In Word and Object Quine made room for a “strictly vegetarian” notion of analyticity. In later years, he developed this notion into two more precise notions, which I have coined “stimulus analyticity” and “behaviorist analyticity.” The latter characterization is in many respects similar to Carnap’s characterization of analyticity based on semantic rules and can be seamlessly incorporated in a Carnapian project of explication. I explain why Quine failed to see the importance of analyticity thus defined. I argue that Quine’s views on language are deeply rooted in the American structuralist tradition in linguistics and that the resulting strict descriptivism and the strong emphasis on language change bring down the usefulness of analyticity.