Based On Back To The Heart Of The Uighur Discourse Of Research

Centering theory (Grosz, Joshi and Weinstein,1986,1995; Walker et al,1998) is a theory of local discourse coherence and pronominalization. The theory concerns relationships among focus of attention, choice of referring expression, and perceived coherence of utterances within a discourse segment. It is simple, easy to operate and verify, therefore it has been widely used in research discourse anaphora (Kameyama,1985,1986,1988,1998;Walker,Iida and Cote,1990,1994;Di Eugenio,1990; Rambow,1993; Prince,1994; Turan,1996,1998; Ryu, 2001; Mitsuko et al.2001; Prasad,2003; Yeh and Chen,2001,2003;王德亮,2004；许余龙et al.2008)。Anaphora is one of the most frequently encountered phenomena in natural languages, therefore its research is important to discourse understanding and production. So far, many researchers have investigated anaphora phenomena of various languages from different angles. However, very few researchers have studied anaphora in Uyghur discourse. The only work on discourse anaphora we have found in Uyghur linguistic literature is the author (2007). Although the author contrasted third-person pronominal anaphora in Uyghur discourse with Chinese in the light of their distribution in discourse, she did not do any further study on null anaphora and full Noun phrase anaphora in Uyghur discourse. That's the reason why we will attempt to make a comprehensive study on the application of centering theory to the analysis of anaphora in Uyghur discourse.This dissertation takes centering theory as a theoretical basis. Three questions are addressed concerning discourse anaphora:1. Which noun-phrases can evoke discourse entities in Uyghur discourse and which are not?2. Which factors affect forward-looking centers in Uyghur discourse?3. What are the functions of null anaphora, pronoun anaphora and full noun phrase anaphora in Uyghur discourse? An exploration regarding question 1 indicate that, the noun phrases which can evoke discourse entities in Uyghur include:(1) Possessive marked NPs(2) NPs which are modified by demonstratives;(3) Accusative case-marked NPs;(4) Accusative case-marked NPs in a generic sentence.The noun phrases which can not evoke discourse entities in Uyghur include:(1) Predicative noun phrases;(2) Indefinite nonspecific NPs within the scope of a negative operator;(3) Indefinite nonspecific NPs within the scope of yes/no questions;(4) Some bare objects (objects which are not accusative case-marked) During the investigation of question 2, we found the factors that affect the forward-looking center ranking. These factors include grammatical functions, experiencer objects, psychological verbs and so on.In order to answer question 3 we conducted an empirical study. The data in this study include selected discourse samples from published narratives. A total of 1010 tokens are analyzed, of which 83 tokens are excluded because they are expressed from the subjective point of view. The results show that null anaphora, pronoun anaphora and full NP anaphora in Uyghur discourse can encode a continue transition. Among them, null anaphora realize the previous preferred center and encode a continue or shift-subject transition. If the backward-looking center is a stressed entity, proun anaphora will be used in continue transitions. The functions of full NP anaphora which are used in a Continue transition are to add new information or to signal a new discourse segment.Furthermore, full NP subjects can also encode a smooth or rough-shift transition, because either a new entity is introduced into the discourse, or because other non-subject entity in the forward-looking center of a previous utterance is promoted into the center of attention in the subject position in the current utterance.