【作者】 张雪珠；

【导师】 刘宇红；

【作者基本信息】 湘潭大学 ， 英语语言文学， 2006， 硕士

【摘要】 由于多义词(多义聚合体)的分析和理解对于准确阅读、语言习得、语言教学、机器翻译、辞典编撰、计算机语言学、语义学和语用学都至关重要，因此近年来成为语言学研究的焦点。传统语言学对多义的研究过多地关注其外部原因如历史、文化和社会发展等方面的因素，以及多义词与同形异义词或同形异音词，多义与单义的区分，而没有把握多义的本质与全部。 20世纪80年代认知语言学的兴起，多义成为认知语言学研究的焦点之一。学者们(Geeraerts，1993；Lyons，1995；Taylor，1995；Tuggy，1995；Croft，1998；Tyler & Evans，2001)对多义的程度，多义的表征，多义词义项之间的关系以及多义的定义进行过论述，但是却没有在这些问题上达成一致。有些学者(Croft，1998；Briscoe &Copestake，1995)关注多义词体系的划分，有些(Lakoff，1987；Tyler & Evans，2001)着重研究个别词的意义延伸，特别是介词多义性的研究。这些研究都有利于提高人们对多义的理解和认识。但以认知模型来研究多义词内部语义关系的还不多见。 本文主要以认知语言学的基本理论为基础，构建分析和解释多义网络的理论框架。并在这一理论框架下，对多义的本质和特点进行讨论分析，旨在探讨多义的认知特征和多义网络建构的可能性。 文章分析了认知语言学的范畴观—范畴化是人类认识世界的基本过程。在此过程中，人们以原形理论为基础，按照家族相似性对世界进行范畴化，从而形成不同的范畴和概念。语言结构也是认知范畴，具有原型特征。因此多义范畴就有原型意义和边缘意义之分，即原型意义与扩展意义。扩展意义都是由原型意义通过转喻、隐喻和意象图式派生而来。这些意义又以连锁或放射的形式构成一个相互联结的复杂结构。由于意义之间至少有一个节点相连，因此多义词的意义就构成了一张语义网络。 到目前为止，学者们(Lakoff，Brugman，Taylor)主要侧重对介词的分析。大多数的实义词还需进一步系统研究。Lakoff曾对bachelor，lie和mother的ICM及其认知模型进行了详尽分析，但未详细讨论它们的引申义。本文试图以ICM为理论框架，对汉语的“打”进行个案研究，分析揭示多义网络扩展的机制。由于ICM包括意象图式，命题结构，隐喻，和转喻4个认知模型，它不仅可以储存信息，而且可以对输入的信息进行重组。因而意义更广，覆盖面更大，更适于解释多义网络。汉语“打”的主要认知模型有4个，由这些认知模型衍生出来的子模型大致可以分为25个。每个认更多还原

【Abstract】 Polysemy, "the association of two or more related senses with a single linguistic form" (Taylor, 1995: 99) has received much attention in recent linguistic semantics, Since the analysis and understanding of polysemy and polysemization processes are significant for accurate reading, language acquisition, language teaching, machine translation, dictionary compiling, computational linguistics, semantics and pragmatics. Traditional approaches to polysemy overemphasized external causes such as historical, cultural and social factors and the distinctions such as polysemy and homonymy, polysemy and monosemy without grasping the nature of polysemy.With the emergence of cognitive linguistics in the 1980s, polysemy has been one of its foci. Scholars have done research into polysemy, but disagreed with each other on the frequency, representation, behaviours and definition of polysemy. Some of them (Croft, 1998; Briscoe & Copestake, 1995) have paid close attention to the division of the semantic system of polysemous words while some of them (Lakoff, 1987; Tyler & Evans, 2001) are mainly concerned with the semantic extension of some individual words, especially the research into polysemy of prepositions. These studies are helpful for us to understand polysemy. However, the research which systematically studies inner semantic relations of polysemous words is rare in terms of cognitive models.This thesis sets up the theoretical framework of analyzing and interpreting polysemy network based on fundamental theories of cognitive linguistics. The aim of it is to discuss the nature and features of polysemy and to explore the possibility of the construction of polysemy network.From a cognitve view, categorization is a fundamental cognitive process. In the course, people conceptualize the world around us based on prototype theory and on the principle of family resemblance, and then form cognitve categories and concepts. And linguistic structure is a kind of cognitive category which bears prototypicality.Polysemy category, as a kind of linguistic structure, is organized with respect to a primary sense from which the extended meanings are derived via metonymy, metaphor and image schemas. These meanings form an interrelated complex structure by radiation and concatenation. Since all the meanings are related at least one node, and thus they form a polysemy network.Lakoff has made a detailed analysis of ICM and its cognitive models of Bachelor, Lie更多还原