The Study Of Classifier Reduplication In Mandarin Chinese In The Framework Of Generative Grammar

This thesis, within the framework of Generative Grammar, explores the syntactic and semantic features of Classifier Reduplication (CC construction) in Mandarin Chinese.In Chinese, classifiers (Cl) have an individualizing function:it has the ability to pick out a single instance of whatever is described by NP. Classifier reduplication is a Multiply-Individualizing (MI) process, which is conducted at the lexical level. The whole set discribed by NP could be determinated by multiply-individualizing all the members in it and therefore we can get the universal-quantification interpretation. Plurality meaning is achieved by using the special grammatical form, namely, reduplicating the classifier, so a CC construction hosts the feature [+P]([+Plural-Number]),[+Q]([+Quanlified]) and [+D]([+Definite]). We test all the classifiers in Chinese to figure out which sub-category of classifier could be reduplicated. The result shows that most of one-syllable classifiers (about90%) can be reduplicated and the reason why the rest cannot is feature conflicts.Conducted within the Minimalist Program outlined in Chomsky (1995) and based on the Multiple Feature Checking theory, our study focuses on the syntactic generating process of NPs and sentences which contain CC construction.Both CC and single classifier are lexical items in lexicon where the MI operation takes place. CC’s projection is the head of CIP, but [CC N] is a DP structure. At LF level, the move of feature [+D] is driven by feature checking between [D] which is generated on an empty D position and [+D] hosted in CC. Besides, we compare the syntactic and semantic differences between "CC N" and "-CC N".After searching CCL corpus, we find that CC constructions usually appear in Universally-Quantified Subject Sentences and Topic Sentences, where "Dou"(都)plays an important role in sentence construction. After studying semantic and syntactic functions of Dou, we draw up the original structure of Dou sentence:[Doup Spec [Dou-都VP]]. Dou hosts strong EPP feature OCC, feature [+P] and strong feature [+Q]. Dou’s multiply-quantifying function allows Dou to have two spec positions. Argument/arguments in VP is/are driven to move, checking strong features with Dou in computational system. In topic sentences, the topic is base-generated at spec position of TopP. The rest generation process of topic sentences is similar to that of universally-quantified subject sentences. Moreover, we also analyze ambiguous sentences and explain why one sentence could have two or three meanings.In the last part, we compare CC and other universal quantification phrases. From the perspective of logic semantics, CC should be considered as a sum operator. It sums up NPs instead of events represented by the predicate. We explain why Dou could be taken out in some circumstances. CC is a sum operator, while Mei(每)is a distributive operator and suoyou(所有) is a universal quantifier.