The Locative Zai Phrase In Chinese: A Predication Approach

This thesis examines the semantic functions and the syntactic representations of the locative zai phrase in Chinese and aims to provide a unified analysis of the derivation of all the surface structures of the zai construction. The empirical evidence is mainly taken from the standard Chinese and partly from the Wu dialects.Our research starts with the scope of our study and the brief introduction to Bowers' (2002) predication approach, the framework within which our study is carried out (chapter 1). We then identify three semantic functions the locative zai phrase can perform, and assume that the zai-PP with a separate function is base generated in a different syntactic position (chapter 2). Following Bowers' approach, we argue that those denoting the time when an actor is located at some place are T'-adjoined PPs, those denoting the location of an action or a state of affairs are Pr'-adjoined PPs, and those denoting the location of concrete objects as a result of an action are base generated in the [Spec, VP] position. Thus it can be seen that the relationship between the syntactic positions and the semantic functions is clear cut. Chapter 3 and chapter 4 focus on the derivation of the sentences with the third kind of zai-PPs. On the basis of the previous studies, we propose a V-movement-with-PP-movement approach, under which all the surface structures are correctly derived due to the different syntactic operations triggered by different uninterpretable features.The basic assumptions of the latest minimalist program are adopted unanimously in our study, in the hope of our research mking some contributions to the study of the nature of human languages.