In this paper, I examine and defend McCawley’s position that a notion of ‘prototype’ is relevant to the representation of syntactic categories in grammatical competence. I first discuss some background on prototype theories of categorization, including McCawley’s (1982, 1987, 1998) theory of syntactic categories, in Sections 2–3 below. Then, in Section 4, I present linguistic evidence in support ... [Show full abstract] of a prototype theory of syntactic categories. The evidence is based on McCawley’s (1987) analysis of English adjectival nouns, and is further supported by data from Culicover and Jackendoff (1997) involving subordinating uses of the English conjunction and. Both cases involve unacceptable sentences that are neither syntactically nor semantically ill-formed, but that deviate from the prototypical syntax-semantics correspondence. Based on this evidence, I argue that deviations from the prototypical syntax-semantics correspondence for a category can have consequences beyond those predictable from independent principles of syntax and semantics. The notion of ‘prototype’ is therefore significant for grammatical competence because it constitutes an EXPECTATION which, when violated in certain ways, has direct consequences for the grammar.