Abstract for Heejeong Ko's talk

Scrambling and linearization
Heejeong Ko (Stony Brook University)
November 29, 2005 (Tuesday)
6:30 PM - 8:00 PM; Room 7102, The CUNY Graduate Center

The question of how successive cyclic movement is derived in the grammar has been at the center of intensive research. A large body of work has argued that particular syntactic domains are impenetrable domains, with the necessary consequence that syntactic escape hatches need to be postulated (Chomsky 2000, 2001, Uriagereka 1999, Nissenbaum 2000). Recent work by Fox and Pesetsky (2003, 2004), however, presents an alternative approach, which eliminates the notion of escape hatch and derives successive cyclicity effect from the interface conditions on the mapping between syntax and phonology (Cyclic Linearization). In this talk, I provide novel evidence for the latter approach. In particular, I show that linear order in constructions with scrambling is constrained by cyclic linearization of syntactic structure at the syntax-phonology interface. I further show that the interactions of cyclic linearization and locality conditions on movement lead us to understand hitherto unexplained behavior of an argument externally-merged on the edge of spell-out domains. I draw evidence for my arguments from a variety of types of asymmetries in scrambling, with special attention to floating quantifier and possessor raising constructions in Korean and Japanese . I conclude the paper by showing that my arguments may provide a diagnostic for resolving a long-standing controversy concerning underlying constituency in NP-split constructions and the phasehood of vP and VP.