| English relative clauses(RCs) are a great difficulty for English EFL learners because of its unique syntactic structure and difference from Chinese attributive clauses. In the filed of SLA,the factors affecting the acquisition of RCs(including three language universals which are accessibility, embeddedness, animacy, and also L1 transfer) have attracted great attention from scholars both at home and abroad. Although many empirical studies have been conducted on the effects of these factors, they still left many gaps to fill in. First, previous studies on the effects of accessibility, embeddedness and animacy only involve three or four types of RCs and the data collected through sentence combination task or translation task can’t reflect learners’ natural use of RCs. Second, L1 transfer was determined by the previous studies merely based on subjunctive judgments, which is not rigid enough. Besides, few studies have probed into the relationship between L2 proficiency level and language transfer. Thus, through a picture-cued written task,this thesis aims to explore the effects of accessibility, embeddedness and animacy on Chinese undergraduates’ use of RCs on the one hand, and adopts the comparison-induction methodological framework proposed by Cai(2015) to investigate the dynamic effect of L1 transfer on the other hand. The paper addresses the following questions:1) How do accessibility, embeddedness and animacy affect Chinese undergraduates’ use of RCs in English narratives over the first two years?2) At individual level, how is Chinese undergraduates’ use of English RCs influenced by Chinese over the first two years?3) At group level, how is Chinese undergraduates’ use of English RCs influenced by Chinese over the first two years?This thesis is a part of the National Social Science Project(No.11CYY021) and the data were collected collaboratively by the project members. The participants in the study were 79 English majors chosen from Qufu Normal University, North China University of Water Resources, Luoyang Normal University, Jiangnan University, and Xi’an Institute of Post and Telecommunications. The data were collected at five times ─ the beginning of the first semester, the beginning and the end of the second semester, the beginning and the end of the fourth semester. The data were 395 English narratives together with their Chinese versions produced according to five series of pictures. Then, the data were coded along five dimensions:the type of RC, the animacy of the antecedent, the correctness of use, the congruity between English and Chinese expressions, and the type of transfer.Through data collection and data analysis, conclusions are drawn as follows.1. In terms of accessibility, the frequency orders of different types of RCs are SU > DO >GEN & OPREP > IO for Test 1, SU > DO > OPREP > IO > GEN for Test 2, SU > DO >OPREP > IO & GEN for Test 3, SU > DO > IO > OPREP & GEN for Test 4, SU > DO > IO >OPREP & GEN for Test 5, and SU > DO > OPREP > IO > GEN for the total result. Thus, it’s clear to see that the results of the five tests all show partial support for the Noun Phrase Accessibility hierarchy hypothesis. Specifically, GEN RCs are produced more than IO RCs in Test 1, and have the same number of RCs with IO RCs in Test 3 and with OPREP RCs in Tests 1,4 and 5. Moreover, OPREP RCs are produced more than IO RCs in Tests 1, 2, 3 and in total.However, in general, the more accessible RCs are produced more frequently and less accessible ones are less produced. In terms of embeddedness, the results of the five tests all lend full support to the Perceptual Difficulty Hypothesis, and suggest that learners tend to produce more matrix object RCs than matrix subject RCs. In addition, although the effects of accessibility and embeddedness have been proven by the study, the results lend little support to the Subject Object Hierarchy Hypothesis, which suggests that accessibility and embeddedness can’t be simply mixed together. In terms of animacy, when the antecedents are animate, there is no preference for learners to use matrix object RCs or matrix subject RCs; however, when the antecedents are inanimate, learners prefer to use matrix object RCs instead of matrix subject RCs.2. At individual level, since the English RCs and Chinese expressions(mainly referring to attributive clauses) are incongruent in form, positive transfer mainly refers to positive semantic transfer. Negative transfer occurs for two reasons. One is due to the presence of certain features in Chinese but not in English(e.g., resumptive pronouns) and this kind of transfer is intrusive transfer. The other is due to the absence of certain features in Chinese(e.g., relative pronouns,the morphological changes of verbs) and this type of transfer is inhibitive transfer. Moreover, the comparison made between learners shows that the distributions of the different types of transfer vary from person to person both synchronically and diachronically.3. At group level, the effect of L1 can be judged as having happened in 75% of the total RCs. Positive transfer takes up 39% of the total RCs in Test 1, 62% of the RCs in Test 2, 58% in Test 3, 64% in Test 4 and 60% in Test 5 while negative transfer takes up 22%, 19%, 17%, 16%,10% respectively. Within negative transfer, inhibitive transfer occupies 18%, 15%, 11%, 16%,10% of the RCs in each test and intrusive transfer accounts for 4%, 4%, 6%, 0%, 0% respectively.In general, positive transfer takes up 59% and negative transfer takes up 16%, in which inhibitive transfer accounts for 14% and intrusive transfer accounts for 2%. The special Z-tests show that there are significant differences between positive transfer and negative transfer in Tests 2, 3, 4, 5and between inhibitive transfer and intrusive transfer in all the five tests. Besides, the results of Change-point Analyzer show that neither of these different types of transfer increase or decreaselinearly but continually fluctuate as learners’ English proficiency level rise up.The present study not only enriches the studies on the acquisition of English RCs but also is of great importance to RC teaching and learning. On the one hand, the sequence of teaching different types of RCs can be arranged in a reasonable way according to the Chinese learners’ RC acquisition order. On the other hand, based on the knowledge of the positive and negative effects of Chinese on English RC acquisition, English teachers are able to design specific teaching methods to improve the teaching effect on English RCs and students will be able to increase the quantity and improve the quality of RCs used. |
PDF Full Text Request