【作者】 王雅刚；

【导师】 刘正光；

【作者基本信息】 湖南大学 ， 外国语言文学， 2017， 博士

【副题名】从精致描写走向精致解释

【摘要】 本文采用基于用法的构式语法理论重新解释语义韵研究的核心概念,并以个案形式来展示新研究路径的理论与方法论收益。语义韵研究历经二十多年的发展,为语料库语言学提供了具有洞察力的词汇语义描写工具,拓展了弗斯学派的经典搭配研究。由于语义韵研究依托语料库,采用概率统计方法,因而丰富了意义研究的手段,避免了单纯依赖研究者内省的不足。语义韵研究的成果为语言教学、辞书编撰和语言对比开辟了新的探索领域。虽然语义韵研究已成为语料库语言学中的一支显学,但学界对于语义韵的概念界定、归属范畴、极性描写、调查手段,以及语义韵与隐含义、语义韵与语义偏好的关系等问题争议不断。这些问题的长期纷争既显出一批前沿学者的洞见,又揭示了该研究领域的隐忧。本文重新审视诸问题之根由,发现主流的社会-文化量化研究路径无法对语义韵现象做出整体一致的解释,颇多龃龉难以调和。语义韵术语虽具有描写便利,但尚属经验术语范畴,于语言学理论构建贡献有限。所以本文跳出以词为中心的传统研究思路,从构式语法视角重新检视语义韵及意义单位和语义偏好等相关概念,以拓宽其研究领域,提升理论价值。有鉴于此,本文分两步走以改进现有的主流语义韵研究。前三章为第一步。主流研究路径在一些影响研究全局的根本性问题上有前后不一致之处。我们以这些不自洽问题为研究起点,提出了构式语法的解决思路。我们认为:1)语义韵本质属构式语义;2)作为意义单位的“节点词+搭配词”型式即构式;3)语义韵非语义偏好的抽象,语义偏好为(构素)词汇义,语义韵为构式义;4)语义韵依附构式,对语境和文本类型敏感。从构式语法角度重新理解语义韵及相关概念有诸多好处。比如,能清晰界定语义韵属性和从属单位,将态度评价义描写融入构式语义分析,道破语义韵与语义偏好的根本区别,避免对语境大小和文本类型作任意性划分,贯通语义韵的共时与历时研究。通过对语义韵、意义单位等核心概念作构式语法重释来为语义韵实证研究寻找适切的理论依托。这样做非但将传统研究路径下该领域存在的老大难问题一一化解,形成一个合理自洽的理论模型,更能为实证研究获取的描写数据寻找新的理论增长点。第二步是按照新的研究思路,以案例分析的形式来检测构式语法研究路径的可行性,展示其优势。本文选取了三个有代表性的新兴评价义构式,分别为“这是X的节奏”、新被字式和“不要太X”,从语义韵的生成、发展和扭转三个方面进行构式语法分析。案例研究始于语义韵调查。沿用基于用法的研究惯例,调查语料均采自真实发生的自然语料。研究结果证实了以构式为单位的语义韵判断的确与现有文献中的某些结论有较大出入,我们对此做出了详细分析。第四章从新节奏构式[这是X的节奏]演化来讨论构式化中的识解干预。该构式在真实语境中确实表现出较明显的消极评价倾向。消极韵主要由X槽位的搭配词贡献。除了有语篇特点、语用策略、语境因素和种系发生等方面的影响外,该消极评价倾向还有认知心理动因。新节奏构式发展经历了“这是VP的节奏”-进程>[这是VP的节奏]-情态>[这是X的节奏]-情态三阶段,体现了原型范畴的拓展。包含「规律」、「顿变」和「渐进」认知拓扑特征的“节奏”意象图式为范畴原型,而对三者的识解变换则引发构式微变,最终导致构式化。第五章以新被字式[被X]的生成为例说明“雪克隆”生成的构式压制动因。语料库调查显示新被字式含消极语义韵的观点证据不足。[被X]所含的反讽语义是承继自原初构语的一种识解视角。常规被字句与新被字式的搭配动词的语义区别体现为指称性与陈述性的不同。[被X]雪克隆形成前,“被X”构语的产出与解读依靠与常规被字句的在线类比,句法语义的系统性反差为形义图式化的动因。雪克隆形成后,其形义特性来自于原初构语对[被X]雪克隆的构式压制。第六章从“不要太X”的评价义扭转来说明语义韵的范畴化是以构式为基础的。“不要太X”的消极和积极语义韵分属[不要太X]-遏止与[不要太X]-极限。两个构式中X槽位的语义类型差别显著。[不要太X]-极限中的量度类搭配词种类明显少于[不要太X]-遏止,且比重远不如点评类搭配词。情感类搭配词在[不要太X]-遏止 中主要表达负面情绪,而在[不要太X]-极限 中则表达愉悦感受。[不要太X]遏止的四类搭配词都表现出「+意志」、「+主观可控」特点,但这些不是[不要太X]极限的主要特征。态度评价指向的扭转有其构式化动因,由溯因推理引发。由于[不要太X]-遏止 中的语义预设凝固为常规会话含义,[不要太X]-极限 才得以析出。以上个案从构式义的生成、拓展和识解上为语义韵及构式发展提供了认知功能分析。可见,将语义韵研究纳入构式语法研究范畴不但理论上可行,而且从实际效果来看也能让这两个传统领域取长补短,相得益彰,比单一的社会-文化量化路径更为可取。构式语法观照下的语义韵研究使得数据挖掘能与理论问题对接,同时也为增强构式语法的解释效力提供了实证数据支持,实现了数据与理论、描写与解释的双赢。更多还原

【Abstract】 This dissertation presents a usage-based constructionist account of key concepts in semantic prosody(hereinafter SP)research.Three detailed case studies are then employed to demonstrate what theoretical and methodological gains can be expected from this new approach.Over the last two decades,SP research has made major advances in the field of corpus linguistics.It has provided an insightful tool for profiling lexicosemantic behavior and contributed to the expansion of classic Firthian studies on collocation.Moreover,the incorporation of corpus-based statistics,a valuable quantitative supplement to introspective data,has also enriched the way how we approach meaning.Discussions in SP research have opened new avenues for language teaching,lexicography,and contrastive studies.Although SP constitutes an important focus in corpus linguistics,disputes continue unabated over its definition,provenance,polarity description,investigation strategies,as well as its differences from connotation and semantic preference,etc.These recurring problems not only reflect the insights of the leading scholars but also call attention to the hidden dangers in this research field.By revisiting these problems,we find that the mainstream sociocultural quantitative approach to SP offered no coherent account,and that some disparities and contradictions were not amenable to solutions.Despite its descriptive value,the notion of SP is nothing more than a heuristic shorthand.As such,it has failed to make substantive contributions to contemporary linguistic theorizing.As a result,this dissertation departs from the traditional word-centered approach,by reconceptualizing SP,unit of meaning,and semantic preference in constructionist terms.It is to be shown that such an attempt can consolidate and extend SP research and enhance its theoretical validity.To revamp the current mainstream approach,we take two steps.Step 1 comprises the first three chapters,in which a selective critique is included,detailing inconsistencies in mainstream research findings.Taking the incoherent account of some global issues as a departure point,we propose a constructionist reworking of these SP-related postulates.It is concluded that 1)SP is an aspect of constructional meaning;2)units of meaning in the form of "node+collocates" are essentially constructions in nature;3)SP differs from semantic preference insofar as the former is constructional whereas the latter(intra-constructionally)lexical;4)SP is tied to specific constructions and is therefore as context-sensitive as is constructional meaning.The constructionist approach has added benefits in that it 1)settles the disputes over the nature and locus of SP,2)incorporates attitudinal evaluation into the wholesale analysis of constructional meaning,3)pinpoints intrinsic differences between SP and semantic preference,4)avoids arbitrariness in determining the context size and text types,5)redresses the balance between diachrony and synchrony in SP research.Moreover,reinterpreting key notions in this field within the framework of Construction Grammar lends more theoretical import into SP studies.In doing so,the new account not only addresses vagaries and controversies in the old one so that a coherent model can be offered,it also uncovers new theoretical insights from empirical data.We then follow up Step 1 with case studies to evaluate the feasibility and strengths of this new approach,whereby the discussion of SP is couched within the framework of Construction Grammar.Three constructions with evaluative functions,i.e.[zheshi X-de jiezou],[bei X]and[buyaotai X],are analyzed,taking into consideration the emergence,changes and shifts of SP.All case studies start with a thorough investigation of their respective SP features,based on naturally-occurring corpus data.Findings show that construction-based SP readings do differ from what was assumed in the literature,for which we offer a detailed account in each case.Chapter 4 examines the role of construal shifts in constructionalization,as evidenced by the emerging epistemic construction[zheshi X-de jiezou].Corpus findings show that this construction is markedly negative in evaluation in common language use.Its negativity is largely due to the lexical meaning of the collocates in the X slot.In addition to factors such as text-types,pragmatic strategies,contextual situations,and phylogenic tendencies,the negative SP of the construction has its cognitive-psychological motivations.The development of this construction undergoes three stages:"zheshi VP-de jiezou"-PROCESS>[zheshi VP-de jiezou]-EPISTEMIC>[zheshi X-de jiezou]-EPISTEMIC.The process is claimed to represent a prototype-based construction extension,in which the prototype is an invariant JIEZOU image-schema while constructionalization proceeds through micro-steps,driven by construal shifts of its topological structures,namely REGULARITY,ALTERNATION and PROGRESSION.Chapter 5 addresses the new[bei X]construction,with a focus on the role of constructional coercion in snowcloning.Corpus findings belie the commonly held view that this new construction carries a negative SP.In fact,ironic effects characteristic of[bei X]reflect the perspective inherited from the cognitive construal of its source.The semantic difference between the X-slot items in the canonical passive bei-construction and the new[bei X]construction lies in their descriptive and referential functions respectively.The production and understanding of "bei X" constructs prior to the emergence of[bei X]snowclones are reliant on their analogy with canonical bei-passives.Syntactic and semantic contrasts between the two bei-construction types motivate the schematization of the[bei X]snowclone.However,with the snowclone in place,the form-meaning idiosyncrasies of the snowclone result from the constructional coercion of the source expression on[bei X]instead.Chapter 6 discusses the SP polarity shift of the "buyaotai X" pattern.Corpus findings show that this pattern instantiates two distinct constructions,[buyaotai X]-RESTRAIN and[buyaotai X]-MAXIMIZE,which are characterized by negative and positive SPs respectively.The proportion of MEASUREMENT collocates of[buyaotai X]-MAXIMIZE is less than that of[buyaotai X]-RESTRAIN,and also less than the proportion of its COMMENT collocates.Whereas the SENTIMENT collocates of[buyaotai X]-RESTRAIN mainly express negative emotions,the SENTIMENT collocates of[buyaotai X]-MAXIMIZE express favorable feelings.Semantic features of[+VOLITION]and[+CONTROLLABILITY]are visible in all the four collocate types of[buyaotai X]-RESTRAIN but not necessarily in the collocates of[buyaotai X]-MAXIMIZE.The shift of the evaluative meaning of "buyaotai" is motivated by the abduction-driven constructionalization.The construction of[buyaotai X]-MAXIMIZE emerges out of conventionalization of the conversational implicature of the semantic presupposition of[buyaotai X]-RESTRAIN.These case studies provide cognitive-functional explanations for the emergence of SP and construction development through an analysis of the emergence,extension and construal of their constructional meanings.It is therefore argued that reinterpreting SP-related concepts constructionally is both convincing in theory and rewarding in practice,hence an attractive alternative to the monolithic sociocultural quantitative approach.Looking into SP from the Construction Grammar perspective brings the two established disciplines closer for further cross-fertilization,as it enables data-mining methodology to speak to theoretical questions.Data thus derived also feed back into the refinement of the Construction Grammar model and,consequently,lead to a win-win situation for both theoretically-and empirically-minded researchers who seek a balance between data and theory.更多还原