竞争模型的语义线索：论元特征满足度

论文摘要

本文在竞争模型(Bates & MacWhinney,1982,1989)的框架下探索词汇语义在句子理解中的作用。通过对竞争模型文献的回顾我们发现,该模型没有足够的线索代表词汇语义信息。首先,仅有的语义线索是生命性,但是,竞争模型研究采用的都是要求有生命名词作为主语的简单动词。这种选择性用词所带来的问题是,生命性效应是单纯由生命性引起的还是由于使用了简单动词而引起的?其次,虽然竞争模型承认词汇语义是一类线索,但除了生命性之外,名词的其它语义特征在该模型中并没有被提及。再次,竞争模型中没有代表动词语义的线索。但是,越来越多的研究表明,词汇语义是句子理解中的重要信息。任何句子理解的理论如果不承认词汇语义的作用都很难成立。 为了解释词汇语义在句子理解中的作用,本文提出了一条代表动词和名词语义的线索:论元特征满足度。我们假设,名词—论元投射是句子理解的重要部分,这是一种多对多的投射。在句子理解中,名词为了投射到功能层上的论元而相互竞争。论元带有语义特征,如果一个名词比另一个名词满足一个论元更多的语义特征,那么该名词在对那个论元的特征满足度就比较高,更有可能被投射到该论元上。 为了检验该假设,我们做了三个实验。实验采用常用的施事辨认作业,受试被要求辨认出及物动词简单句中的施事。三十个汉语单语受试和三十个英语单语受试参加了实验。实验一的目的是检测生命性效应是否是使用简单动词的产物。实验使用了三类动词:简单动词、刺激物—感知者动词(简写为SE动词)和非词动词。实验二的目的是探索论元满足度在句子理解中的作用。实验区分了三种论元特征满足度:第一种包括动词倾向选择第一个名词作为主语的句子,第二种包括动词倾向选择第二个名词作为主语的句子,在第三种情况里,动词对两个名词都没有特定的倾向。实验三有两个目的,一是探索生命性之外的其它名词特征在句子理解中的作用,二是检验名词—论元多对多投射的假设。在实验中,生命

论文目录

• Abstract
• Acknowledgement
• Abbreviation
• List of Figures
• List of Tables
• CHAPTER 1 INTRODUCTION
• 1.1 Research orientation
• 1.2 Sentence processing: a Competition Model perspective
• 1.3 Research rationale
• 1.4 Summary
• CHAPTER 2 THE COMPETITION MODEL: THEORETICAL FRAMEWORK AND EMPIRICAL STUDIES
• 2.1 Overview
• 2.2 Sentence processing
• 2.2.1 Approaches to sentence processing
• 2.2.1.1 The principle-based approach
• 2.2.1.2 The constraint-based approach
• 2.3 The Competition Model
• 2.3.1 Form-function mapping
• 2.3.2 Coalition in form-function mapping
• 2.3.3 Competition of cues in form-function mapping
• 2.3.4 Cue strength
• 2.3.4.1 Task frequency
• 2.3.4.2 Cue validity
• 2.3.4.3 Cue cost
• 2.3.5 Research paradigm
• 2.3.5.1 Design
• 2.3.5.2 Materials
• 2.4 Empirical studies in support of the Competition Model
• 2.4.1 Monolingual sentence processing
• 2.4.2 Cross-linguistic comparisons of sentence processing
• 2.4.3 Second language sentence processing
• 2.5 Comparisons of Chinese and English sentence processing strategy
• 2.6 Summary
• CHAPTER 3 A BROADER PICTURE OF CUES IN THE COMPETITION MODEL
• 3.1 Overview
• 3.2 Cues in Form-Function Mapping
• 3.2.1 Definition of Cues
• 3.2.2 Cues in the prosodic level
• 3.2.3 Cues in the morphological level
• 3.2.4 Cues in the syntactic level
• 3.2.5 Cues in the lexical semantic level
• 3.3 Lack of cues in representing lexical semantics
• 3.4 The role of lexical semantics in sentence processing
• 3.5 Potential difficulties for the Competition Model
• 3.5.1 Reanalysis of the animacy effect
• 3.5.2 The role of verb semantics
• 3.5.3 The role of nominal semantics
• 3.6 Summary
• CHAPTER 4 A POSSIBLE CUE TO BE TESTED: ARGUMENT SPECIFICATION SATISFACTION
• 4.1 Overview
• 4.2 Research questions
• 4.3 a possible cue for lexical semantics
• 4.3.1 Argument Specification
• 4.3.2 Noun-argument mapping
• 4.3.3 Argument specification satisfaction
• 4.3.4 Sentence processing as many-to-many noun-argument mapping
• 4.3.5 ASS as a source of the animacy effect
• 4.3.6 Animacy per se effect
• 4.4 Hypotheses for the research questions
• 4.5 Overview of the empirical study
• 4.5.1 Introduction
• 4.5.2 Rationale of the experiments
• 4.5.3 The experiments
• 4.5.3.1 Participants
• 4.5.3.2 Materials
• 4.5.3.3 Arrangement of the experiments
• 4.5.3.4 Experimental task and procedures
• 4.6 Summary
• CHAPTER 5 DECOMPOSING THE ANIMACY EFFECT: VERB SEMANTICS AND ANIMACY PER SE
• 5.1 Introduction
• 5.2 Experiment 1
• 5.2.1 Hypotheses
• 5.2.2 Design and stimuli
• 5.2.3 Subjects and procedures
• 5.2.4 Predictions
• 5.2.5 Results
• 5.2.5.1 Overview of the results
• 5.2.5.2 Processing of real verb sentences
• 5.2.5.3 ASS effect: reanalysis of the data
• 5.2.5.4 Animacy per se effect: processing of nonword verb sentences
• 5.2.5.5 A general picture of the animacy effect
• 5.2.6 Discussion
• CHAPTER 6 ARGUMENT SPECIFICATION SATISFACTION AS A SEMANTIC CUE
• 6.1 Introduction
• 6.2 Experiment 2
• 6.2.1 Hypotheses
• 6.2.2 Design and materials
• 6.2.3 Subjects and procedures
• 6.2.4 Predictions
• 6.2.5 Results
• 6.2.5.1 Processing of Chinese sentences
• 6.2.5.2 Processing of English sentences
• 6.2.5.3 Summary of the results
• 6.2.6 Discussion
• CHAPTER 7 COMPETING FOR BOTH ARGUMENTS: THE ROLE OF NOMINAL semantics
• 7.1 Introduction
• 7.2 Experiment 3
• 7.2.1 Hypotheses
• 7.2.2 Design and stimuli
• 7.2.3 Subjects and procedure
• 7.2.4 Predictions
• 7.2.5 Results
• 7.2.5.1 Processing of Chinese sentences
• 7.2.5.2 Processing of English sentences
• 7.2.5.3 Summary of the results
• 7.2.6 Discussion
• CHAPTER 8 GENERAL DISCUSSION AND CONCLUSION
• 8.1 Overview
• 8.2 major findings for the research questions
• 8.3 Representing lexical semantics in the Competition Model
• 8.3.1 ASS as a semantic cue
• 8.3.2 The animacy per se effect
• 8.3.3 The animacy effect in simple verb sentences
• 8.3.4 The animacy effect in SE verb sentences
• 8.4 Implications ofthe present study
• 8.4.1 Implications for the Competition Model
• 8.4.2 Implications for sentence processing in general
• 8.5 Limitations of the present study
• 8.6 Recommendations for further studies
• 8.7 Summary
• APPENDICES
• Appendix 1: Languages that have been investigated under the Competition Model
• Appendix 2: Materials for Experiment 1
• Appendix 3: Materials for Experiment 2
• Appendix 4: Materials for Experiment 3
• Appendix 5: Experiment 1 Chinese processing data: first-noun choice percentage
• Appendix 6: Experiment 1 English processing data: first-noun choice percentage
• Appendix 7: Experiment 2 Chinese processing data: first-noun choice percentage
• Appendix 8: Experiment 2 English processing data: first-noun choice percentage
• Appendix 9: Experiment 3 Chinese processing data: object-position noun choice percentage
• Appendix 10: Experiment 3 English processing data: object-position noun choice percentage
• REFERENCES: