A Corpus-based Study Of Object-referring Behaviour In Chinese

Referring to object is a common activity type in daily communication. This dissertation makes a contrastive study of the object-referring behaviour in Chinese between adults and children by adopting a corpus-based approach.Referential expressions are usually employed in referential behaviour. Utilizing the tripartite approach to analyzing the referential behaviour to human entities in Ma(2005) as a source of reference, we establish a tripartite model to examine the referential behaviour to objects in naturally-occurring Chinese conversation. This model proposes a three-line analysis scheme—the object-ontology analysis, the cognitive analysis and the linguistic realization analysis. Along the line of object-ontology analysis, different types of objects are distinguished in three worlds, namely the physical, mental and discourse world. Along the line of cognitive analysis, the object-referring behaivor is interpreted as a process in which the speaker construes the referent in the current discourse space and factors upon the construal operations are also identified. These factors include the construal perspectives, the contextual factors and the ontological features of objects. Along the line of linguistic realization, a system of referential expressions for objects is constructed, which is further divided into the subsystem of the present-object referential expressions, that of the non-present-object referential expressions, that of the mixed referential expressions and of the kind-denoting expressions. In addition to the system of object-referring expressions, the line of linguistic realization analysis also includes the discussions of the following phenomena related to the referential expressions:the semantic-connection types between the introducing referential expressions and the maintaining ones, the switch of categorization names, the metonymic referential expressions for objects, the usages of demonstratives "zhe(this)" and "na(that)"; the referential repair and the distribution pattern of referential expressions encoding non-present objects.Besides referential expressions, the object-referring behaviour also involves some non-verbal actions such as hand pointings and iconic gestures. We define these non-verbal gestures as gesture modes and propose an analytical framework for the gesture modes in the object-referring behaviour. In this framework, we distinguish diverse types of the gesture modes for three types of objects, namely the present, the non-present and the generic. The gesture modes include the pointing, the iconic and the placing, among which the pointing is further divided into the big pointing and the small pointing, and the iconic into the model, the action and the description. Moreover, we proceed to analyze three patterns of interaction between the gesture modes and the referential expressions:the former in a subordinate position to the latter, the latter in a subordinate position to the former and both modes in an equal position.Based on the tripartite model and the analytical framework of the gesture modes, we design the annotation scheme for the audio data as well as the video data. The former is collected for the contrastive analysis of the object-referring expressions between adults and children while the latter of the gesture modes between these two groups of people.In this dissertation, the contrastive analysis has been conducted from eleven aspects and the major findings are listed as follows:(1) Concerning the use of the referential expressions, the strategies adopted by both groups are largely identical except that:(i) adults exploit more forms of referential expressions than children do;(ii) when introducing the non-present objects into the discourse world, adults prefer definite forms while chidren prefer indefinite forms;(iii) with regard to the concrete forms of referential expressions, it is more frequent to find the na expression, the de expression, the double-referring expression and the zero form in adults’referential behaviour while children incline to use the quantitative noun phrase, the reference-point noun phrase, the impersonal pronoun "ta (it)" and the referential expression in quotation;(iv) in terms of the switch of categorization names, adults prefer the strategy of switching from the subordinate names to the superordinate names while children between the names at the same level;(v) regarding the usages of demonstratives, adults use more frequently the recognitional "na" while children use the maintaining "zhe" and "na" with a higher frequency. (2) Concerning the use of the gesture modes, similarities and differences coexist between adults and children. The similarities include:in terms of the pointing, most of the big ones are used to identify the present objects while most of the small ones to identify the non-present and the generic; when referring to the last two types, both groups rely more on the referential expressions than on the pointing; as for the iconic, most of them are employed to represent the non-present and the dominating interaction pattern between the iconic and the referential expression is that the former in a subordinate position to the latter. As for differences, when referring to the present, children resort more to the pointing while adults the referential expression. When employing the iconic to represent the non-present as well as the generic, adults prefer the model while children the description.To sum up, this dissertation finds that the overall referential ability of children aged between4and9is still developing and there exists some differences between the referential ability of children and that of adults.