허사 there의 병합과 이동

There in [Spec, TP] : via Merge or Move?

Critically reviewing Chomsky’s (2001, 2004~2005) [Spec, TP] approach to the expletive there, this paper argues for Kim and Lee’s (2004) proposal that expletive there externally merges in [Spec, υP], which is supported by the complementarity of Object Shift and Transitive Expletive Construction, as observed in Richards and Biberauer (2005). [Spec, υP] analysis is shown to have advantages over [Spec, TP] analysis with respect to the following: There, merged in [Spec, υP], raises to [Spec, TP] like any other formal subject. Since [Spec, υP] is in T’s c-command domain, there can be probed and valued by T like any other nominal. And, like other nominals, it is rendered active by a Case feature. There is, therefore, no need to assume that its defective ø-set is unvalued: its [Person]-feature has default, third-person value, hence it is not entitled to be a probe, contra Chomsky (2005). If there merges inside T’s Agree domain, then the EPP problem in Chomsky’s analysis is also naturally overcome, since it satisfies T’s EPP via Move in the same way that nominal arguments do. In the current analysis, [Person]-feature of there being interpretable, there is no need to assume, contra Kim and Lee (2004), that there enters Agree with its associate. The Case features of the expletive and its associate are valued via multiple Agree with T, which implies that the two share the same Case. As for the locus of Merge-there, it is argued that there is allowed to merge in [Spec, υP], provided that this position is not occupied by a thematic element (an external argument or causing event). On the observation that only a subset of un-accusative verbs allows there-construction, we specify this to be the specifiers of non-inchoative unaccusatives.