A classic puzzle about CPs is that they distribute differently than nominal arguments. This fact is reflected, among other things, by the order of complements in English (Stowell 1981) and the right-peripheral position of CPs in many OV languages (Hindi, Farsi, German). This distribution has traditionally been seen as a reflex of grammatical function, most famously encoded by Stowell’s case resistance principle (or modern variants, Pesetsky and Torrego 2004).

In this talk I argue that the syntactic distribution of clausal arguments has its source in the semantic type of CPs. I begin by establishing a puzzle: that clause-taking predicates only form non-event nominals. I argue that the explanation for this puzzle requires that CPs are never able to saturate an argument position (cf. Grimshaw 1990). This also prevents CPs from combining as the internal argument of verbs. The only solution, I claim, is to turn the vP into something with the meaning of a one-place, non-verbal predicate, with which the ‘complement CP’ can combine. We then show how this motivates a vP raising analysis of the rightward position of CPs (Larson 1988a,b) as opposed to an anti-symmetry analysis (Zwart 1993, and following)

Keir Moulton (McGill University) will give this job talk as a candidate for the Linguistics department’s Syntax faculty position.