|
|
MORE PROBLEMS FOR THE 'LITTLE
v'—AND A LEXICALIST ALTERNATIVE
Steve Wechsler University of Texas, Austin Tuesday, October 11, 5:30 PM MJH Rm 126 Sponsored by the Stanford Humanities Center/Mellon Foundation Graduate Research Program According to the "Little v Hypothesis" (LvH), external thematic roles such as agent roles are not assigned directly by their verb but rather by a silent "light verb" acting as a secondary predicate. This talk will re-examine two key arguments for the LvH: the argument from idiom asymmetries and the argument from nominalizations (Kratzer, Marantz, i.a.). Both arguments are problematic. For example, on the LvH analysis of event nominals the agent ("the soldier") in the ing-of mixed nominal "the soldier's looting of their home" receives its thematic role from v, while other (non-ing) event nominals such as "the soldier's destruction of their home" lack v. But nouns of the two types are freely coordinated and can take shared dependents, as in this attested example: 1. With nothing left after the soldier's destruction and looting of their home, they re-boarded their coach and set out for the port of Calais. Sentences like (1) can be shown to involve N-zero coordination ("destruction and looting") rather than right node raising, based on quantifier scope (Partee and Rooth 1983) and other properties. Under the little v hypothesis such examples are problematic because of the necessary assumption that "looting" requires an extra vP shell that "destruction" lacks. Finally, we consider what might be right about the LvH: the insight that agency can be a "supralexical" semantic feature, that is, one that is not tied to particular words. We consider how that insight might be captured while avoiding the problems of the LvH. 
|
