Zero Anaphora In Chinese Narratives: From A Cognitive Perspective

熊学亮 (1998) successfully explains the underlining cognitive mechanism of the first person zero anaphora in Chinese from the perspective of Cognitive Grammar. However, the "self-effacement" ideology cannot explain other zero anaphora phenomena such as the third person zero anaphora, which is more prevalent in written narratives. Within the framework of Cognitive Grammar, we intend to extend 熊's research to account for the third person zero anaphora in Chinese written narratives from van Hoek's reference point model.In van Hoek (1995, 1997, 2003), she develops Langacker's reference point/ dominion configuration into a Reference Point Model to account for constraints on pronominal anaphora. The reference point model provides a good interpretation for the intra-sentential pronominal anaphora without resorting to syntactic devices like c-command. However, van Hoek does not attempt to analyze the discourse anaphora.To cope with zero anaphora in discourse, we employ quantitative measurements to assess the occurrence and distribution of zero anaphora in Chinese written narratives, using first 40 pages of each of three novels: (1) 《芙蓉镇》(古华 1981), (2) 《洗澡》(杨绛982), and (3) 《二马》 (老舍 1984), about 76,800 words, in which we find 851 instances of zero anaphors. After the preliminary analyses, we propose two criteria for the delimitation of dominion at the discourse level:(1) a dominion is ended when switch of reference occurs; or(2) a dominion is ended when the same referent which remains the reference point in both the preceding clause and the current clause is referred to by an NP, showing the low accessibility of the referent.Based on the study of the distribution of zero anaphora and the antecedent in Chinese, we propose and define two notions, i.e. the Prominence Hierarchy and the Macro-Connectivity. We believe that they are crucial to account for the zero anaphora resolution in Chinese narratives. The Prominence Hierarchy denotes the prominence of the discourse referents in different positions:Subject-Subject > Object-Subject > Subject-Object / Modifier of the Subject-Subject > Topic- Subject/Object-Object > Nominal in the existential/ presentative sentence-Subject.We note that the positions of the antecedent and the zero anaphor in discourse are related to the prominence of the referent. The Macro-Connectivity refers to the intra-dominion or inter-dominion macro-connectivity. The notion of intra-dominion macro-connectivity refers to the connectivity between the embedded clause and the projected clauses in the embedded dominion. The notion of inter-dominion macro-connectivity refers to the connectivity among the clauses in different dominions. The inter-dominion macro-connectivity includes two subgroups. The connectivity between dominions in discourse not only influences the choice of anaphoric forms, but also helps the reader resolve zero anaphora.In terms of the Prominence Hierarchy and the Macro-Connectivity, we come out with three Recovery Principles. We test the principles with natural language data, i.e. first 40 pages of each of eight novels in Chapter 6. Of the 2234 instances of zero anaphors found in the data, 2231 instances are solved, accounting for 99.87%.Thus our research not only extends 熊学亮 (1998, 1999)'s research on first person zero anaphora to third person zero anaphora, explicating its cognitive mechanism, but also provides feasible method to resolve zero anaphora in Chinese discourse.