| This dissertation aims to explore issues of null/empty arguments in Chinese. Earlier study on empty arguments in East Asian Languages, such as Chinese and Japanese (Huang 1984, 1987, Xu 1986) has argued that Chinese/Japanese empty arguments have their own peculiar properties that are different from big PRO and small pro in the standard theory based on European languages, such as English and Italian. For example, Huang (1984, 1987) proposes the Generalized Control Rule (GCR) for the empty subjects in Chinese in order to account for the interpretative asymmetry between empty subjects and empty objects. Xu (1986), on the other hand, remarks that such subject-object asymmetry does not exist in Chinese, and the interpretative asymmetry is by large determined by the discourse and linguistic contexts. The latter, however, may turn out to be too unrestricted since empty arguments do display grammatical restrictions that can hardly be fully accounted for by pragmatics alone. In view of this problem, and given evidence from the Null Object Construction in Chinese, Li (2005 et seq.) has clearly shown that there indeed is a grammaticalized subject-object asymmetry with respect to the LF properties of empty arguments in Chinese, and such asymmetry can be explained if we postulate a true empty category, whose existence is nonetheless subject to a fundamental grammatical property, namely, subcategorization. Li argues that the subcategorized object can remain empty in Chinese during syntactic derivation, and their semantic contents are `recovered' at LF via LF-copying or from the discourse context. Li's proposal therefore provides an elegant solution out of the existing dilemma.;The dissertation finds solid evidence for Li's proposal of True Empty Category. It is argued that the theory of True Empty Category be applied to the following three constructions, all of which involve theoretical complications that are difficult to be explained in standard theories: the Null Object Construction, the NP-ellipsis, and the Right Node Raising Construction. One of the general properties across the three constructions is that they all contain missing elements (gaps) in sentences. It has been demonstrated that the missing elements in these constructions cannot be derived via PF-deletion or syntactic movement. In view of the fact that the missing elements in these constructions are all subcategorized/licensed by their corresponding heads, I show that True Empty Category, when applied to these three constructions, yields satisfying results. Semantically, in contrast to the PF-deletion approach, the interpretation of the missing element (True Empty Category) can be obtained via LF-copying (or the discourse context/convention) without resorting to the presence of linguistic antecedent. It is further proposed that although True Empty Category may enjoy a wider range of interpretative possibilities, it is constrained, nevertheless, by syntactic and semantic well-formedness conditions. To spell out these conditions, True Empty Category must be syntactically subcategorized by a head that selects the [N] feature. Semantically, LF-copying of True Empty Category must observe other language-specific LF constraints, such as the subset rule and the identification rule in two different topicalization constructions in Chinese. I therefore show that the theory of True Empty Category represents a more explanatorily adequate improvement on both theoretical and empirical grounds. |
