汉语“驴句”的儿童语言习得研究

发布时间：2016-05-05 10:35

Chapterl Introduction

The earliest research on donkey sentences was dated back toChrysippos-sentences by Stoic philosopher Chrysippos in the fifth century, with arenaissance in the early 1980s when linguists found out that the formal semanticscannot offer a reasonable explanation towards this language phenomenon. Ever since,there have been mass discussions on the two types of donkey sentences^ namely,conditional sentences and relative clauses. The intensive issue with regard to theirforms particularly concerns indefinite NPs and pronouns with which they are relatedto in sentences like (1) and (2), while the debate embracing the nature has to do withanaphor, quantification, scope and binding (Geach 1962; Lewis 1975;-Evans 1977,1980; Kamp 1981; Heim 1982,1990; Kadmon 1987; Kratzer 1995; Chierchia 1992,2000).(1).......If a fanner owns a donkey, he beats it.(2).......Eveiy fanner who owns a donkey beats it.In Mandarin Chinese, since Cheng & Huang(1996) identified certain types ofconditionals as Chinese donkey sentences,several other accounts are offered(Linl996; Pan and Jiang 1997; Wen 1996,1997; Chierchia 2000;Huang 2003; Wen2006),which naturally fall into four classifications according to their divisions ofChinese donkey sentences, namely, two types of donkey conditionals, one type ofdonkey conditionals, three types of donkey sentences, and a wh-...wh- correlativestructure.

1.1 The Purposes and Significance of the Study
The main purposes of the present study are to investigate the acquisition of thedonkey sentences in Mandarin-speaking children of different ages and to discuss thetypes of donkey sentences in Mandarin Chinese from the child language acquisitionperspective.A number of studies have put emphasis on the classification and interpretationof donkey sentences in Chinese. However, there is no specific empirical investigationon child language acquisition. Using a truth value judgment task,Grain & Thornton(2000) tested children's understanding of both relative-clause and conditional donkeysentences. Even though there are abundant of theoretical studies on the donkeysentences in Mandarin Chinese, there is no consistent view on the types of donkeysentences yet. Knowledge of language cannot be divorced from language acquisition.The principles that constitute knowledge of language must therefore relate toacquisition (Cook & Newson 2000). The present study therefore attempts to obtainthe empirical evidence and explore child's first language acquisition of donkeysentences in Mandarin Chinese.
...........

1.2 Organization of the thesis
The thesis consists of six chapters.Chapter One is the introduction part in which purposes and significance of thestudy are briefly introduced.Chapter Two briefly reviews the previous literature on the donkey sentences inEnglish and Chinese, including theoretical backgrounds and empirical studies.Chapter Three involves the detailed description of the research design andmethodology of the present study, including the research questions, the subjects, themethods and the procedure.Chapter Four reports in details the experimental results by analyzing thecollected data with SPSS Statistics 19.0.Chapter Five gives a brief discussion on the results.Chapter Six is the conclusion part, in which major findings are displayed. Thenthe implications and limitations of the current study as well as some suggestions forfurther study are put forward.
.........

Chapter 2 Literature Review

Donkey sentences in Mandarin Chinese, which is an important languagephenomenon that cannot be ignored, has aroused great interest in linguisticresearches since 1996. A great number of studies devoted to donkey sentences havebeen conducted from the aspect of quantification, scope and anaphora. However, theresearch on the child language acquisition of donkey sentences in Mandarin Chineseis a new field, and little specific research has been done on this topic. In this chapter,tiie author will briefly review the theoretical and empirical studies of donkeysentences.

2.1 Theoretical Studies on Donkey Sentences
In English there are two types of donkey sentences: one is conditional donkeysentences as (1), and the other is relative clause donkey sentences as (2). Heim (1982)has given an explicit definition of them:Donkey sentences are sentences that contain an indefinite NP which isinside an if-clause or relative clause,and a pronoun which is outside thatnif-clause or relative clause,but is related anaphorically to the indefinite NP(Heim 1982: 44).According to Heim's definition, a critical characteristic of donkey sentences canbe derived: there exists an anaphoric relation between a pronoun and its indefiniteNP antecedent,though the pronoun is beyond the syntactic scope of that indefinite.The other defining chmacteristie of donkey sentences is the universal reading ofindefinites.Beginning with Heim (1982) and Kamp (1981),and continuing through currentversion of Discourse Representation Theory (DRT) (e.g. Kamp and Reyle 1993),linguistic analyses of donkey sentences have had two main goals: one goal is toprovide a semantics that assigns the same truth conditions to both relative-clause andconditional donkey sentences?,tiie other is to ensure that the truth conditions forsentences of both types correspond to the strong reading,according to which everyfarmer feeds every donkey that he owns (Groenendijk and Stokhof 1991). (On theweak reading, each farmer must feed at least one of the donkeys he owns,he mayfeed them all, but this is not required for the sentence to be true.) Heim achievesthese goals by treating the universal quantifier as an "unselective binder,,. In therelative-clause donkey sentences like (2), the universal quantifier is taken tounselectively bind nominal, namely the farmer and donkey. As a result,therelative-clause donkey sentences can get the same readings as conditional donkeysentences like (1), i.e. they both get the strong readings.
.........

2.2 Child Acquisition Studies on Donkey Sentences
If there are distinctions to be drawn between relative-clause donkey sentenceand conditional donkey sentences, as Chierchia suggested, then we csm expectchildren to distinguish them more clearly than adults,for the following reasons.Adults' judgments about the alternative interpretation of ambiguous sentences maybe easily influenced by general world knowledge. Acquiring such knowledge requires experience. Because children's experience is more limited,their judgmentsshould not be subject to the influence of general world knowledge to the same extentas adults' judgments. If so,children's judgments may directly reflect basic principlesof the language apparatus.Grain and Thorthon (2000) conducted an experiment designed to test children'sunderstanding of both conditional donkey sentences like (1) and relative-clausedonkey sentences like (2),using a variant of the Truth Value Judgment Task. Theexperiment showed that children did not treat the two types of donkey sentences alike.With relative-clause targets,all children demonstrated that the weak reading wasavailable in their grammars; no child required the strong interpretation for allrelative-clause. By contrast, with conditional targets, the responses by individualchildren were extremely consistent: the child either assigned the weak reading on alltrials or assigned the strong reading on all trials.
........

Chapter 3 Experiment.......21
3.1Research Questions......21
3.2Research Method.......23
3.3Participants.......25
3.4Stimuli.......26
3.5Procedure.............28
3.5.1The Pilot Study.......30
3.5.2The Formal Study.......30
Chapter 4 Results.............32
4.1Results of the Adult Group...........33
4.2Results of the Filler Sentences...........35
4.3Results of the Experiment Groups............38
4.3.1 Group Results..........40

Chapter 5 Discussion

From Chapter 4,we can see that children's performance differs from that of theadults in the existential context. In this chapter,we will discuss the results from thefollowing aspects: the types of donkey sentences,the two readings of donkeysentences and the distribution of wA-words.

5.1 Discussion on the Types of Donkey Sentences

According to the previous study,Cheng & Huang (1996) think that there aretwo types of donkey sentences: bare conditionals and ruguo-conditionals. WhileHuang (2003) believe that there is only one type of donkey conditionals. In Cheng &Huang's theory,we can predict that adults should treat the two types different. And inHuang's theory, we can predict that adults will not distinguish the two types ofdonkey sentences. In this point, from the prediction of adults' interpretation, we alsocan infer that in Cheng & Huang's theory, children will distinguish the two types ofdonkey sentences,while in Huang's theory,children will treat the two types ofdonkey sentences alike.In our experiment, the statistical analysis between bare conditionals andruguo-conditionals showed that there was no significant difference in theinterpretations of the two types of donkey sentences. The result reflects that bothchildren and adults treated bare conditionals and ruguo-conditionals alike. Thenon-distinctive interpretation of two donkey sentences in our experiment may be areflection of Huang's (2003) theory that there is only one type of donkey sentence inChinese. She gives a unified account that Chinese donkey anaphora can be obtainedby appealing to the Discourse Representative Theory. In a word, our experimentalresults indicated that the two types of donkey sentences can be interpreted in asimilar way.

........

Conclusion

The present study is a tentative research on the child acquisition of donkeysentences in Mandarin Chinese. In this final chapter, major findings of the presentstudy are summarized. Then implications and limitations of the present study andsuggestions for future research are put forward.The present study is an attempt to explore the acquisition of donkey sentences inMandarin Chinese. The major findings are that: (1) both the children and the adultsdid not distinguish the bare conditionals and ruguo-conditionals; (2) children did nottreat the two readings of donkey sentences like adults: adults only accepted theuniversal reading, while both the universal and existential readings were available tochildren. Children preferred the universal reading in universal context and preferredexistential reading in existential context; (3) the different positions of wh-words hadno effect on children's understanding; (4) after 5 years old,children performed moreadult-like in interpretation of the donkey sentences.From the major findings,we can draw the following implications.Firstly, the present study provides an empirical evidence for the argument of thedifferent classification about the donkey sentences in Chinese. Since there is nosignificant difference between ruguo-conditionals and bare conditionals, we cananalyze the two types of donkey sentences in the same way since they convey thesame meaning.
.........
The reference (omitted)

，

本文编号：42133