A Study Of Linguistic Ambiguity And Its Resolution

Ambiguity is recognized as a common phenomenon of human language, prevailing across different linguistic levels, such as lexical, syntactic and pragmatic levels. However, many empirical studies conducted at home and abroad of language users (particularly those who have a particular language as their mother tongue ) indicate their usual insensitiveness or unawareness to ambiguity of various types of the language they are using in communication. Certain ambiguity turns sharper to their attention only when there is a lack of the necessary conditions under which such ambiguity is normally resolved. To the questions of"what is ambiguity"and"what are the necessary conditions for the resolution of ambiguity", there exist various answers from different linguistic schools. This dissertation has attempted a comparison and contrast between some of these different ways in which ambiguity might be identified and resolved, during the process of which, the mechanical and inflexible features are presented of the traditional ways in which ambiguity is identified and resolved. This dissertation thus holds that the ambiguity- identifying and resolving process correlates to a cognitive one, introducing the cognitive accounting mechanism (which consists of conceptualization, categorization, proto-categorization, schematization, figure/ground, frame, perspective choosing) into the process during which ambiguities on different levels are identified and resolved. In so doing, it is assumed that human beings'competence of identifying and resolving ambiguity consists of experience and knowledge acquired through cognitive activities, and stored, activated, and retrieved in the form of modules, which are represented by the linguistic forms. The said cognitive ambiguity-identifying and resolving mechanism has made available both a physical and mental basis for the account of automaticity and instancy in ambiguity identification and resolution. In this regard, the dissertation has applied the'Frame'theory by Fillmore and others(F.Ungerer and H.J.Schmid:1996:205-249) and made reference to"Event-domain Cognitive Model"(ECM)(Wang Ying:2005:17-26) in the cognitive account of ambiguity and its resolution on all the three linguistic levels within one cognitive frame. With a degree of application of both the"Connectionism"and"Module theory"in data processing, information activation and processing ,are not done as they traditionally were exclusively level by level up and in an all closed way before the output , but are two-way operations, with each level open to other levels. The result is a reduction of burden the working memory bears during the linguistic information processing. This is all attributed to the introduction of the cognitive mechanism which consists of prototypical categorization and schematization to construal-context analysis. In EMC model, this cognitive mechanism is used to account for physical, mental as well as cognitive basis for the human being's insensitiveness to ambiguity and ability of instant resolution. The dissertation thus proposes the application of the model to related scientific research fields, such as artificial intelligence field, and the transference of the human competence of instant and automatic resolution of linguistic ambiguity in artificial intelligence with a view to reducing its mechanicality and inflexibility in linguistic information processing, and making its linguistic competence closer to that of human beings.