Conceptual Integration Approach To Semantic Analysis Of Chinese Noun-Noun Compounds

As society and science are developing unceasingly in China, new things emerge continually. The existent words cannot satisfy the requirement of nominating new things and expressing abstract concepts. Under such a circumstance, Chinese Noun-Noun Compounds (CNNCs hereafter) emerge as times requires. CNNCs are composed of two monosyllabic nouns on the surface, but they imply complex semantic relations. Therefore, the semantic construction of CNNCs has been the research focus of domestic linguists for a long time. However, few people have made systematical research and overall demonstration on the semantic construction of CNNCs. And also, in the previous studies, we do not find a cognitive mechanism that can analyze the semantics of CNNCs powerfully and systematically. In consideration of this situation, this thesis aims to find such a cognitive mechanism and tries to apply it to the semantic analysis of CNNCs. With reference of many domestic and foreign literatures in this regard, we find that Conceptual Integration Theory (hereafter CIT) of Fauconnier (1998(22):133-187) can serve to adequately account for the semantics of CNNCs. The theory includes a set of optimality principles and a representation of Conceptual Integration Networks, which are classified into four types: simplex network, mirror network, single-scope network and double-scope network. As the means of conceptual integration, these four types of networks can present the semantic construction process of CNNCs approximately. And also, this thesis, combined with the ideas of Liu Zhengguang and Liu Runqing (2004), makes a new classification of CNNCs: relational type, paralleling type, feature-mapping type and mixed type. These four types of CNNCs correspond to the four networks exactly, which thus ensures a successful semantic analysis of CNNCs.CNNCs have two main prominent features: metaphoric quality and semantic indirectness. Simply speaking, metaphoric quality means that the components constituting a compound take on a metaphorical feature in the course of meaning construction, for example,"银发","雨丝","瓜葛", etc.; while semantic indirectness means that if the meaning of a compound cannot be obtained from the literal meanings of the two components constituting the compound or from the construction meaning based on these two components, we can say this compound owns semantic indirectness, for example,"马车","矛盾","党棍", etc.. And furthermore, we find that CNNCs which own metaphoric quality necessarily possesses semantic indirectness, but conversely, CNNCs having semantic indirectness do not necessarily own metaphoric quality. Taking"马车"as an example, this compound has semantic indirectness but the two nouns constituting this compound do not take metaphoric quality. This thesis takes these two prominent features as the point of departure to make the analysis of meaning construction of CNNCs. Besides, the notion of continuum of Conceptual Integration Networks is also applied to the dynamical analysis of meaning construction, which is also one of the innovations in this thesis.