Derivationa Nd Representation Of Relative Constructions

The major focus of this study is to propose a unified theory of derivation and representation of relative constructions (RCs) within the Minimalist framework. We propose that relativizations are motivated by the EPP-feature carried by the functional head C, which is called the EPP-feature orientated analysis (EOA).It is argued that the EPP-feature is the property of the head C in both wh-questions and RCs, but not in declarative sentences and that-clauses. According to the EOA, a relative internal nominal element raises to SpecCP triggered by the uninterpretable EPP-feature carried by the functional head C in order to check the uninterpretable feature. This operation takes place in a phase C and never beyond such a phase. So it can be regarded as an autonomous reform of the phase C. The result as a syntactic object is selected by an external determiner in the English-like languages. When all the uninterpretable features have been checked and eliminated, the whole structure is a convergent restrictive RC In the course of the presentation, the label has been argued to have nothing to do with the categorial feature, and not the label but the categorial feature plays the essential role. We argue that there is no such a categorial feature [D] but [N], and the relative head is an NP in nature although it may be represented as a DP superficially. Based on the property of the phonological component, this thesis claims that a relative pronoun is a resumptive pronoun at PF, and this resumptive strategy can extend to many other similar cases cross-linguistically. The application of the EOA to the derivation yields the representation of a complementation structure which is compatible with the economy principle of universal grammar.Concerning the Chinese RCs, de is an obligatory element but can never be regarded as the only factor for a relativization. With the EOA, the relativization depends on the cooperation of two Cs. One is de which has been construed to be a functional head C and its EPP-feature must be checked by a phonologically overt nominal element; the other is a phonologically covert C # which has also been proved to be true. The occurrence of a de in a relativization is only for the sake of scope, similar to the occurrence of the external determiner in the other languages. It is the property of de that makes the Chinese RCs show a rather marked word order. Regarding the stacking and coordination, we further develop Chomsky's definition of chain-formation and claim that two elements share the same features may form a chain. This proposal may well explain the coordination structure.When the EOA extends to the derivations of adjunct RCs and possessive RCs, thedistinction between primary adjunct RCs and secondary adjunct RCs proposed by Ning undergoes a revision. The primary one has been argued to be pseudo-RC, and the secondary one is the real adjunct RC. In contrast with the possessive RCs in other languages, the Chinese counterpart shows the same characters and hence is identified. Throughout this study, the EOA can successfully derive the RCs and all the discussions prove that the EOA is an ideal relativization theory.