【作者】 孙翠兰；

【导师】 盛玉麒；

【作者基本信息】 山东大学 ， 语言学及应用语言学， 2014， 博士

【摘要】 中动结构是一种既常见又特殊的具有跨语言特性的语言现象。中动结构在句法上为主动式而在语义上又隐含着被动的意义。该结构特殊的语法形式、语义属性以及复杂的二元对立特征①,已成为国内外语言学界研究的热点和争议的话题。以往对该结构研究的焦点主要集中在中动结构的生成机制、副词的使用以及时体的限制等方面。然而,在中动结构的界定上仍没有统一的标准,学界从不同的视角衍生出不同的制约标准,如副词效应标准、隐含施事标准、情态语义标准、动词体貌制约标准、主语有责条件、动词影响标准以及论元敏感制约标准等诸多纷繁复杂的制约标准,而且每一种制约标准都能找出违反该标准的反例。为了去繁就简,突破藩篱,以加深对中动结构本质的认识,本文从语言类型学的跨语言视角,在总结和借鉴了国内外对中动结构研究成果的基础上,着眼于对汉英两种语言结构的语义特征、施事论元的隐含动因、中动主语的内在属性等进行整体的全方位的分析和研究。为了避免在跨语言研究中比附英语、生搬硬套、止于直觉的表观和片面,本文以语料库语言事实的统计为依据,以汉语为本位以英语为参照,排斥主观臆断,以“他山之石,可以攻玉”的理念,参照外位,来充实本位,本文对英汉语料库真实语料的调查、统计进行了定量、定性对比分析,对前人的假说进行了求证,证实了“中动具有跨语言特性”的论点。解决了英汉语中动结构统一的语义特征②以及其定义特征和本质特征的界定问题,明确了研究对象和范围；进而论证了和确立了汉语中[NP+V-起来+AP]和[NP+好-V]构式的中动地位。通过对比研究,发现了英汉两种语言的中动结构方面所具有的共性和差异：都是介于主动结构与被动结构之间的独立语言构式；都具有“主语有责性”特征③和非事件性特征④；都有凸显受事隐含施事等特征。从语料库的分布情况和使用范围看,英、汉语中动结构都属于口头语言、中动动词多数是日常用语中单音节常用动词；在实际语料中,不只副词／形容词能用于中动词后,介词短语也能用于中动词后来表述主语名词的属性或特征,而且副词／形容词或介词短语的语义指向都不指向隐含的施事等。不同之处：英语动词后多用副词修饰而汉语中动词后多用形容词修饰；语料中发现英语中动结构存在大量光杆动词结构而汉语中没有；在构句条件允准程度方面,汉语比英语宽容度强、能产性高。论文不但从组合关系入手,在句法结构上将其分为典型的和非典型的两种类型,研究其句法结构和语义表达。文章还从认知的视角,对中动结构进行动态研究,从聚合关系入手,提出了在同一句法位置上的对立归一系统中存在“类指”与“定指”两种类型的中动结构。不但论证了英、汉两种语言在中动结构的类指与定指上具有的共性,考察了与之对应的语义变化、语用差异以及在隐含施事方面的深层机理,还对国内外普遍认为的所有中动结构都具有类属性和普遍性特征的论点提出了自己的不同观点。通过研究发现“定指”中动结构表述在特定语境下,具象的(实体的)客体主语在动作过程中的属性或状态。指出了只有“类指”中动结构才有类属性、规律性和普遍性的特征。论证了两种语言在“定指”和“类指”中动结构对比分析中有着如下相同的差异：隐含施事认知动因的差异、名词主语语义属性的差异以及语义等级层级的差异。用类指和定指中动结构以及其物性结构,对困扰学界的一些违反界定中动结构标准的反例进行了统一解释。全文共分六个部分。第一章作为绪论首先是问题的提出,介绍了研究的对象和内容、研究的目的及意义。该章同时作为文献综述部分,回顾并介绍了国内外不同流派对英、汉语中动结构的相关研究；评述了英、汉语中动结构的研究以及比较研究中所取得的成就及存在的问题。然后提出了指导理论和研究方法,简介了自建双语语料库、美国当代英语语料库以及北京大学现代汉语语料库的情况,最后介绍了语料的来源和结构安排。第二章英、汉中动结构的语料考察和研究。首先,介绍了自建双语语料库的建设情况。其次,在通用大型语料库和自建英、汉双语平行语料库的支持下,本文作者对语料库的真实语料进行定量分析,给出定性结论。然后,文章统计分析了中动结构在不同语料库中的分布情况,以查证该结构的使用范围和频次。对英语中动结构所对应的汉语结构进行了统计分析和研究,对满足中动结构构句条件的动词以及修饰语进行语料考察、统计和分析,以查证该结构的汉语结构有哪些?最后,采用定性与定量分析相结合、描写与分析相结合的研究方法,论证语料中哪些动词和副词能用于英语中动结构也能/不能与汉语对应?以及它们使用频次的异同。分别对汉英语中动结构的约束条件和中动结构修饰语的类别以及选择制约进行了对比研究。本章的目的是通过呈现在真实语料方面英汉语中动结构的一些独特特征,分别为第三章的中动结构的界定和第五章的汉语中动结构研究提供依据。第三章中动结构的界定。本章以第二章的语料统计分析的结果为依据,以论元结构理论和认知语言学理论为指导,在前贤们研究成果的基础上,对中动结构进行明确界定,有利于把握该结构的全貌。同时,根据英语中动结构的界定标准,结合在语料库中与英语对应的汉语结构特征,论证和确立了汉语中[NP+V-起来+AP]和[NP+好-V]格式的中动地位。第四章对中动结构进行了句法和语义分类。本章从组合关系上,将中动结构分成典型的和非典型的进行研究。还从聚合关系上提出了中动结构存在“类指”与“定指”两种类型。论证了英汉两种语言在中动结构的类指与定指上具有的共性和差异,考察了与之对应的语义变化、语用差异以及在隐含施事方面的深层机理。第五章是对汉语中动结构及其语义特征的研究。结合第二章的语料分析,本章对汉语结构“V-起来”与“好-V”的句法和语义特征进行了研究。借鉴英语中动研究中已经达成共识的中动句法标准和中动语义特征,检验汉语非施事主语V-起来句[NP+V-起来+AP](简称S1)和非施事主语“好-V”句[NP+好-V](简称S2),发现这两种句法结构都体现了英语中动结构的所有中动语义特征和非常相似的句法表征。据此,本文在第三章中确立过的这两种汉语中动结构进行构句成分分析和结构层次分析,并从句法特征和语义特征上发现了它们的异同。对句首名词NP的语义属性、谓语动词V的选择限制、以及修饰语AP的语义指向进行全面系统的分析。第六部分是全文的结语。首先概括了本文研究对比的结果和理论思考,然后阐明本研究的贡献和应用启示,最后说明有待努力之处。本论文基本的研究方法是基于自建英汉语双语平行语料库和通用语料库的真实语料,进行定量统计与定性分析相结合、描写与分析相结合的方法。“定量”包括英语和中动结构在语料库中存在形式的分布情况、出现频次的定量统计和分析,以及与其对应的汉语中动结构在语料库中存在形式的分布情况、出现频次的定量统计和分析。“定性”包括英汉语中动结构的对应情况的共性和差异的分析,英汉语中动结构语义特征的共性与差异的分析。同时“描写”是定量的基础,“分析”是定性的必由之路,通过内省的方法来弥补语料库统计中的不足。本研究的创新点主要体现在以下几个方面：1)以语料库的真实语言统计为依据,以语言类型学理论为指导,解决了中动结构的定义特征和本质特征的界定问题,明确了研究对象和范围；根据通用语料库和自建双语语料库的调查、统计,论证了和确立了汉语中[NP+V-起来+AP]和[NP+好-V]格式的“中动”地位。2)在大量语料的支持下,对英、汉语中动结构的特征进行定量、定性分析,对前人的假说进行求证,证实了“中动具有跨语言特性”的论点,以期深化汉语言所具有的共性和差异的本体研究。3)从聚合关系入手,提出了中动结构的“类指”与“定指”两种类型,不但论证了英、汉两种语言在中动结构的类指与定指上具有的共性和差异,还分析了与之对应的语义变化、语用差异以及在隐含施事方面的深层机理。就国内外普遍认为的所有“中动结构都具有类属性和普遍性特征”的论点,本文作者以语料库中检测到的真实语料为依据,从主语名词在不同类型中的语义属性和认知动因上论证了该“类属性”论点的偏颇。指出了只有“类指”中动结构才有类属性和普遍性的特征,“定指”的中动结构并不表述这两种遍性特征。论证了两种语言在“定指”和“类指”中动结构对比分析中有着如下相同的差异：隐含施事认知动因的差异、名词主语语义属性的差异以及语义等级层级的差异。4)对困惑语言学界的复杂的二元对立特征和许多违反界定中动结构标准的反例进行了统一解释①,了解这些特征会有助于更加深刻地探讨中动结构的本质。不足之处在于,在语料的统计方面,由于在删减不符合构句条件的句子时需要人工操作,可能出现部分数据的偏差；自建英汉双语语料库规模小,达不到海量检索,也可能会出现数据偏差。根据上述不足我们将多次核对、穷尽检索,尽可能地减少偏差；对于自建双语语料库,今后将联合能联合的力量尽可能扩大其规模更多还原

【Abstract】 Middle Construction (MC), also known as the Middles is a kind of common and special language phenomenon with cross-language properties. It is active in syntactic form and passive in semantic meaning implicitly. The qualities of its special syntactic form, unique semantic properties and complex binary opposition has made the study of MC a wide and controversial reaseach topic in the linguistic circles both at home and abroad since the1980s. The previous studies on MC were mainly concentrated around the mechanism of the structure, the use of adverbs and the selectional restrictions of tenses and aspects. But there is still no normal standard for its proper definiting, resulting in great differences standard from different perspectives in linguistic world, so much complicated constraints, as adverb effect restriction, implicit agent restriction, modal semantic standard. verb effect restriction, non-double-objects restriction, action effects and argument sensitive standard. Each kind of standard, however, can find counterexamples in violation. In order to break through the barriers the complicated constraints and to deepen our understanding the true nature of MCs, this thesis, with the Cross-linguistic perspective of the Linguistic Typology, on the base of the research results in this field achieved at home and abroard, makes the intrinsic properties of the overall comprehensive analysis and research, focusing on the semantic characteristics、implicit-agent in cognitive motivation、the intrinsic properties of both Chinese and English MCs.In order to avoid cross-language study just applying English MC mechanically, corpus-based quantitative and qualitative analyses on English and Chinese MCs with the concept of" the jade may be refined from stones coming from other hills", that is, theories abroad can be used as reference to enrich the Chinese language study. The cross-language study, from the perspective of typology, is advantageous to the theory of ontology research on Chinese and helpful for the application of Chinese language international promotion. By way of the reasearches on the previous hypothesis on these studies, the argument that "the MCs are of cross-language characteristics" have been confirmed and the characteristics of universality and heterogeneity have been found in Chinese MC. And then according to the investigation and demonstration of the statistics from the corpora, the quantitative and qualitative analyses have been carried out on the corpora characteristics of MCs both in English and Chinese. Thus the status of the Chinese format [NP+V-qilai+AP] and [NP+hao-V] has been set as the Chinese MCs. Through comparative study, the commonalities and differences have be found. The commonalities are that both English and Chinese MCs are of independent constructions between active and passive structure, of "the subject duty and non events " features, and of highlighted-patient implied-agent characteristics. In the distribution of corpus and range of using, both English and Chinese MCs belong to oral language, and the verbs in MCs are most commonly used in single syllable routinely used verbs; not only adverbs/adjectives can be used after verbs in MCs, but prepositional phrases, the semantic orientation of which pointing to the implicit agents, can also be found to describe attributes or characteristics of the serface subject NP. The differences:adverbs are used to modify after the English verbs while adjectives after Chinese verbs; and in real corpus a lot of bare verb English MCs, that is no modifies after verbs, while no bare verb Chises MCs can be found in corpus; and the degree of consent in the condition of sentences is that Chinese MCs are more tolerant and productive than English MCs.In this paper, MCs have been classified into typical and atypical models in syntactic structure by the syntagmatic-relation approach. And researches have been made on the syntactic stuctures and semantic expressions of the two models. In addition, the distinction between the two types of MCs, namely,"the generic reference" type and "the definite reference" type, have been put forward by the paradigmatic-relation approach from the perspective of cognition. With an argument on the similarities between the two types of MCs both in English and Chinese, an examination has been conducted on their corresponding semantic changes, pragmatic features and the deep cognitive motivation in the implicit-agents. This method of classification has overturned the beliefs held by most of domestic scholars that all the MCs are characteristic of "genericity" and "universality", which, in fact, is only true of the "the generic reference" middles. In the comparative studies on the two types of MCs, the similar distinctions have been found between English and Chinese: differences between the implicit-agents in cognitive motivation, differences in semantic attributes of the subject noun and the differences of the semantic hierarchy levels. Whereas "the definite reference" functions the contextal representation of object properties or state in the process of the action. And unified interpretations have been made to complex binary opposition and many counter-examples of linguistics and to the features of binary opposition, which have long confused the linguistic field.The full text of the thesis is divided into six parts. Chapter One, as the introduction to the whole thesis, is an explanation to the reason of the establishment of the reasearch program as well as an illustration of the research object, content, purpose and significance. The chapter, as a part of literature review, is designed as an introduction to the relative studies on MCs held by the different genres of English and Chinese linguistists both at home and abroad. It is a general review of the studies on English and Chinese MCs as well as a comparative study on the achievements obtained and the problems still existed. Finally, the guiding theory and research methods has been put forward with an interpretation of the source of corpus and structure arrangement.Chapter Two is an investigation into and an analysis on the corpuses of English and Chinese MCs. With the support of large-scale corpus and the self-built English-Chinese parallel corpus, a research has been carried out on the distribution of and the correspondence with English and Chinese MCs with an investigation into the verbs and the modifiers in the MCs as well in the corpus. First of all, an examination and statistics have been conducted on the corresponding MC structures in Chinese with those in English in bilingual corpus. Secondly, with an investigation into the distribution of the MCs in different corpus, the range and frequency in use of the structures have been verified. Finally, a discussion has been made on how many of the verbs and the adverbs which can be found in English MCs could or couldn’t be correspond with those in Chinese MCs in the corpus. The other object of the discussion is the difference of its frequency in use of those verbs and adverbs. A contrastive analysis on English and Chinese sentence conditions of the MCs in corpus is another task of this chapter---comparative studies being made respectively on the constraint conditions, the category of the modifiers and the restricts of choice. By way of description and illustration, a conclusion has been made with qualitative and quantitative analyses in this chapter that the MCs are also a general language phenomenon in Chinese.Chapter Three, while combing the standards and biased errors in the previous studies on defining MC has presented the significance in strictly definiting MC. And then the contrast researches have been made on the MC structures easy to be confused, and the syntactic semantic standards in defining MC have been put forward with the grammatical category of MC structures being analysized.Chapter Four is the classification of the MC structures. In this chapter, the classification has been conducted syntactically and semantically. The general MCs are divided into typical and atypical models with the syntagmatic-relation approach so that investigations could be made in Chinese and English MCs. And the distinction between the "the generic reference" and the definite reference types of MCs has been made with paradigmatic-relation approach. As for the the generic reference" and definite reference types of MC, a fruitful discussion have been held on the commonness and differences between English and Chinese MCs.Chapter Five is a study on the Chinese MCs and their semantic characteristics. A research has been made on the syntactic and semantic features of the Chinese sentence structures "V-qilai起来" and "hao好-V". With the commonly acknowleged syntactic standards and semantic features of MCs obtained in English MC studies as the reference, it is found that the very similar semantic features and syntactic representations in English MCs can be found in the Chinese syntactic structures of "V-qilai"[NP+V-qialai+AP](SI) and the "hao-V"[NP+hao-V](S2). Thus the status of the Chinese format [NP+V-qilai+AP] and [NP+hao-V] has been set as the Chinese MCs. And the similarities and differences between them are easily distinguished from their syntactic and semantic features. However, a research has been also made on the sentence structures and the hierarchy structures of the patterns. Moreover a sytematic anlysis has been carried out on the semantic properties of NP, the choice restrictions of V, the semantic orientation of AP and the semantic features.The sixth part of the thesis is the conclusion. It is first a summary of the results of the comparative studies and the theoretical thinking of the author, and then a clarification of the research contributions and application revelation. And the weakpoints of the studies have been revealed at the end of the part.In this paper, quantitative statistics with qualitative analysis and description with analysis are adopted as the basic research methods by which studies are conducted on the basis of the self-built English-Chinese parallel corpus and the large-scale corpus, such as COCA and the Beijing University Chinese Corpus. The studies in terms of "Quantitative" include the quantitative statistics and analysis on the distribution and frequency of MCs in English corpus as well as those of the corresponding MC in Chinese corpus. And the studies in terms of "Qualitative" include the analysis on the commonness and difference between English and Chinese MCs, especially on their semantic characteristics. Meanwhile "Description" is the basis of quantitative analysis and "Analysis" is the essence of qualitative approach. And the lack of corpus statistics can be compensated by the introspective method.Innovation points of this research are mainly in the following several aspects:1) From the perspective of linguistic typology, the problems have been solved in the defining and essential characteristics of MCs, and the research object and scope made clear about. And according to the investigation and the statistics made on the large corpora and the self-built bilingual corpus, the Chinese [NP+V-qilai+AP] and [NP+hao-V] format have been demonstrated and their status in MC established.2) With the support of a large corpus, the quantitative and qualitative analyseson the features of MCs in English and Chinese are helpful to prove the hypothesis of previous patch, offering a confirmation on the argument that "the MCs are of cross-language characteristics," in order to deepen the researches on the universality and heterogeneous ontology of the Chinese language.3) From the perspective of the syntagmatic and the paradigmatic relations, two types of MCs, namely,"the generic reference" and "the definite reference," have been put forward in this paper. With an argument on the similarities between the two types of MCs both in English and Chinese, an examination has been conducted on their corresponding semantic changes, pragmatic features and the deep cognitive motivation in the implicit-agents.. The beliefs held by most of domestic scholars have been overturned that all the MCs are characteristic of "genericity" and "universality," which, in fact, is only true of the "the generic reference" middles". In the comparative studies on the two types of MCs, the similar distinctions have been found between English and Chinese:differences between the implied agencies in cognitive motivation, differences in semantic attributes of the subject noun and the differences of the semantic hierarchy levels.4) In oder to help understand the nature of MC more deeply, unified interpretations have been made to complex binary opposition and many counter-examples of linguistics and to the features of binary opposition, which have long confused the linguistic field.Its insufficients are first in corpus statistics because of some deviation of the data in cutting the sentences not in conformity with the conditions by way of manual operation. And secondly data deviation may occur due to the small capacity of the self-built English-Chinese bilingual corpus which offers no massive amounts of retrieval. According to the shortcomings, the author will try her best to reduce the deviation and to enrich the self-built English-Chinese bilingual corpus in the future.更多还原