A DR-Theoretic Account Of Chinese Donkey Anaphora

This thesis addresses two issues, one is that there is only one type of donkey anaphora in Mandarin Chinese, and the other is how a unified account of Chinese donkey anaphora can be obtained by appealing to the Discourse Representative Theory (DRT).Chapter 1 schematically introduces the charateristics of donkey phenomenon and two popular theories designed for the donkey puzzles in the literature, i.e., DRT and the E-type pronoun approach (EPA). Chapter 2 is devoted to reviewing previous accounts of Chinese donkey anaphora, in which, except Pan and Jiang(1997), all accounts hold the view that two types of donkey anaphora are available in Mandarin Chinese, i.e., bare conditionals and ruguo/dou-conditionals. Chapter 3 is dedicated to the claim that there is only one type of Chinese donkey anaphora by examining those unreliable syntactic and semantic diagnostic tests. Chapter 4 focuses on the discussion how a unified account of Chinese donkey anaphora is obtained by resorting to a DR-theoretical analysis. As a precondition, the implication of the other competitive candidate-EPA is proved invalid in Chinese by taking issues of the lack of independent evidence and the inefficiency of those motivations in previous accounts. DRT, however, can give a satisfactory explanation on the basis of the following refinements over previous accounts: (i) both bare conditionals and ruguo/dou-conditionals are derivatives built out of the canonical tripartite structure of conditionals, in which, the quantificational domain is over minimal situations; (ii) there is a division of labor between licensing and binding of indefinite wh-words. Specifically, the conditional construal of donkey sentences per se is responsible for the former, whereas (implicit) adverbs of quantification the latter; (iii) the interplay between semantics and pragmatics determines readings / readings, and (a)symmetric readings in various contexts, enhancing the role qf pragmatics in the interpretation of donkey anaphora and hinting the empirical flavor of donkey anaphora; (iv) dou-conditionals can alsoget a straightforward explanation under the DRT framework on the basis of the consideration that the adverb dou 'all' is used only if there is an element that can give us a plural individual or sum individual.If I am on the right track, my proposal may contribute to the study of Chinese donkey anaphora the following respects:(i) there is only one type of donkey anaphora in MandarinChinese; (ii) a unified semantics of Chinese donkey anaphora is obtained,i.e., a DR-theoretical analysis portrays all cases of Chinesedonkey sentences; (iii) w/z-words, interrogative or noninterrogative, also get a unifiedsemantics, i.e., all of them are restrictive free variables; (iv) given the universality of the (a)symmetric readings of donkeyanaphora, the unselective binding spirit of classical DRT isdiscarded, pace Chierchia( 1992,1995); (v) the uniqueness requirement is waived and obviated in anatural way.