【作者】 李知恩；

【导师】 郭锐；

【作者基本信息】 北京大学 ， 汉语言文字学， 2011， 博士

【摘要】 本文在类型学的背景下,对量词进行了跨语言的研究。绪论部分介绍了本文的研究方法、目标、语料来源和相关术语等问题。首先,本文对国内外分类词语言的研究成果进行了综述,并讨论了数词分类词语言的特点及其功能。西方类型学家认为汉语的量词是用于数量结构并对名词进行分类的数词分类词(numeral classifiers)。而本文的观点是：汉语个体量词的基本功能是表示“计量单位”,而不是对名词进行分类或个体化。从这个角度,我们在前人研究的基础上,继续对世界数词分类词语言进行了考察,并讨论了数词分类词语言的共性和特点。此外,据Aikhenvald (2000)的研究,汉语普通话、泰语、越南语和广东话等语言有多种分类词系统。以普通话为例,既有数词分类词,也有指示分类词,属于多重分类词语言(multiple classifier languages)。如“这本书”的“本”是依附于指示词,并且对名词进行分类的指示分类词。本文不赞同这一观点,认为这是量词的扩张功能,而不是对名词进行分类。从这一点出发,本文借鉴语义地图模型(Semantic Map Model)讨论了量词在各语言中的扩张功能。考察发现,量词有多种扩张功能,我们所确定的功能有14种：计量、数量转指、非定指转指、量名无定、类指、指量名、指量转指、定指、定指性转指、无核关系化、有核关系化、无核领属性定语、有核领属性定语、名词类别词。我们基于对98种语言的量词功能的考察,勾画出量词功能的概念空间(conceptual space),并尝试对这些概念之间的语义关联及其扩张机制作出了初步的解释,认为量词的无定标记和定指标记功能分别来源于数词“一”和指示词的脱落。我们的考察结果还显示,量词的定语标记用法(关系化标记和领属定语标记)不是来源于定指标记,而是来源于指量转指。此外,传统的观点认为,语言的量词和复数标记会构成互补关系,即有量词的语言没有复数标记,而有复数标记的语言没有量词。但是汉语和韩语既有量词,又有复数标记,汉语的是“们”,韩语的是“deul”。我们采用Corbett (2005)的观点,认为汉语和韩语的名词是泛数(general number),而不是类指。在此基础上,我们进一步借鉴他所分的数量范畴：联系性复数(associatives plural)、分配性复数(distributives plural)和集合性复数(collectives plural),对汉语的“们”和韩语的"deul"进行了对比。最后总结前面讨论的内容,并指出本项研究的成果及不足之处。更多还原

【Abstract】 On the background of linguistic typology, the dissertation is designed to present the hitherto unraised question pertaining to Chinese numeral classifiers through the researches made within the broader frame of world languages.The introduction concerns general issues such as achievements accumulated, methodologies, the corpus of the field, and relevant terminology, etc.Firstly, the dissertation reviews Chinese and international achievements on classifier languages with special emphasis on features and functions of numeral classifiers:Western typologists argue that the individual classifier in Chinese is used in the numerical structure and its main function is to classify nouns on the basis of the attributes such as shapes and functions. Meanwhile, this dissertation maintains that Chinese individual classifiers function as the measuring unit rather than as the assumed function, i.e. classification or individualization of nouns. From this perspective, the dissertation is devoted to investigating the various strata of numeral classifier languages in the world and to probing common and unique characteristics of the languages. In the same vein, Aikhenvald (2000) claiming that Mandarin, Thai, Vietnamese, Cantonese and other languages have multiple classifier systems holds attention. According to her, the Mandarin has both numeral- and deictic classifiers, belonging to the group of multiple classifier languages. To illustrate, the ’ben’ in the ’zhe ben shu’(literally ’this book’) is a deictic classifier attached to the demonstrative which classifies the noun. However, the author does not agree with her view, maintaining that it is an expanded function of the classifiers rather than the classification of nouns.Following the line of thought, the dissertation refers to the semantic map model for the purpose of highlighting the expanded function of classifiers in many languages. According to the author’s observation, the numeral classifiers has expanded functions which are classified as fourteen:measure unit, Num-Cl transferred referent, indefinite transferred referent, indefinite marker, generic reference maker, Dem-Cl-N, Dem-Cl transferred referent, definite marker, definite transferred referent, relative clause marker without head, relative clause marker with head and possessive modifier marker without head, possessive modifier marker with head, noun classifier. And, further researches on the classifiers’functions in ninety eight languages show the conceptual space of classifiers functions which is a convenient diagram to allow for the comparison of language-particular categories in some functional domain and an give initial explanation on semantic relationship between the concept and the expanded mechanism, maintaining that the definite- and indefinite marking functions of the classifiers stem from the omission of numeral ’one’ and directives respectively. It then follows that the attributive marker of the classifier (relative clause markers and possessive attributive markers) is not derived from the definite maker but from the Dem-Cl transferred referent.Further researches reveal that the traditionally upheld view that the numeral classifiers and the plural markers constitute a complementary relationship in the sense that the language with classifiers has no plural markers is mistaken. Rather, Chinese and Korean have both the numeral classifier and plural marker,’men’ and ’deul’ respectively. Siding with Corbett’s (2005) viewpoint that nouns in Chinese and in Korean is general number rather than generic reference, the author further learns from his category of quantification:associatives plural, distributives plural and collectives plural. Thereby, the author draws a contrast between the Chinese ’men’ and the Korean ’deul.’In conclusion, the summary is given and the author’s contributions together with the possible drawbacks are taken into account.更多还原