| Case is generally understood to be semantically empty. That is the reason that a Case feature is assumed to be an LF-uninterpretable feature in Minimalism. It is perplexing that an LF-uninterpretable Case feature should be present in the computational system.; I present evidence that (structural) Case is not always completely semantically inert. The stacked Case in Korean behaves like a focus marker. This focus-like Case is checked by a syncretic head: T or nu with a matching focus feature.; Moreover, I suggest that Case should not be a reflex of &phis;-feature agreement. Rather, Korean Raising supports the separation of Case valuation from &phis;-feature agreement, instead of the unification of the two.; Finally, we provide an account of why the English Resultative Construction differs from the Korean one in terms of movement. The former shows the selectional restriction of resultative predicates, while the latter does not. The presence of the selectional restriction in English forces the small clause to merge as a complement of the matrix verb. Conversely, the absence of the selectional restriction in Korean enables the small clause to behave like an adjunct. Thus, the small clause can have various adjunction sites in the verbal domain. Unlike the limited availability of the Object Resultative in English, in Korean the Subject, Object, and Goal Resultative are all possible in compliance with minimality. |
