Colloquium: Peter Svenonius (Tromsø), "When can a single word by biclausal? Word structure and the limits of monoclausality" Monday, September 16, 3-4:30pm in 370 Dwinelle (with reception to follow). Abstract:
"Various diagnostics have been identified to determine whether a complex predicate, such as a causative structure, involves a single clause or two. The diagnostics pick out clausal properties of subparts of the predicate, for example an agent or a distinctly modifiable event description is sometimes understood as a sufficient condition for clausehood, and if the subordinate part of a complex predicate has one of these things independently of the superordinate part, the complex predicate is deemed to be biclausal. This sometimes holds even when the predicate is considered to consist of a single word (as with Japanese indirect causatives, as argued by Shibatani and Kuno in the 1970s; see Matsumoto 1996 CSLI for discussion and references).
However, there are striking limits on biclausal single words, which reveal constraints on the architecture of grammar and the interface between grammar and the lexicon. First, restrictions on the conceptual content associated with listed lexical roots mean that certain kinds of information can only be expressed in syntactically complex forms. Second, functional structure is constrained --- I argue that the constraints follow from a certain conception of extended projections, which are responsible for a sharp difference between what can be expressed with and without embedding of one extended projection in another. This leads to motivated definitions of mono- and biclausality."
