Reading-time Studies of the Resolutionof Main Verb versus Reduced Relative Clause Ambiguities by English Learners in Taiwan

• Main Authors: Pi-lan Yang, 楊碧藍
• Other Authors: Su-chin Shih
• Format: Others
• Language: en_US
• Published: 2009
• Online Access: http://ndltd.ncl.edu.tw/handle/repyp7

博士 === 國立高雄師範大學 === 英語學系 === 97 === ABSTRACT This dissertation aimed at investigating (a) whether English learners in Taiwan immediately resolve main verb versus reduced relative clause (MV/RRC) ambiguities such as those boldfaced ones in (1) and (2) on-line in a similar way as native English speakers, and (b) whether English learners in Taiwan at various English proficiency levels show diverse profiles of the MV/RRC ambiguity resolution. (1) a. The boy kissed by the girl was cute. b. The boy who was kissed by the girl was cute. c. The apple kissed by the girl was cute. d. The apple that was kissed by the girl was cute. (2) a. A boy and a dog were playing with a girl in the playground. The girl kissed the boy but not the dog. The boy kissed by the girl was cute. b. A boy and a dog were playing with a girl in the playground. The girl kissed the boy but not the dog. The boy who was kissed by the girl was cute. c. Two boys were playing with a girl in the playground. The girl kissed one of the boys but not the other. The boy kissed by the girl was cute. d. Two boys were playing with a girl in the playground. The girl kissed one of the boys but not the other. The boy who was kissed by the girl was cute. More specifically, it compared behaviors of English learners in the MV/RRC ambiguity resolution with those of native English speakers in terms of sensitivity to use of reduction, use of noun animacy cue, and use of referential noun phrase information in a discourse context. These three sensitivities are defined as follows. First, the sensitivity to the use of reduction is that the minimal attachment strategy (i.e., postulating no potentially unnecessary nodes) is immediately used to form an incorrect MV structure for the reduced relative clause (1a). This leads to a revision. The revision got the reading times for “kissed by” in (1a) to be significantly longer than those in (1b). Next, the sensitivity to the use of noun animacy cue is that the inanimate subject noun phrase “The apple” in (1c) can immediately help construct a correct RRC reading, but the animate subject noun phrases “The boy” in (1a) cannot. Such an immediate help shortens reading times for “kissed by” in (1c). These shorter reading times result in a significant difference in reading times for “kissed by” between (1c) and (1a). Finally, the sensitivity to the use of referential information is that the 2-NP-Referent discourse context “Two boys” in (2c) can immediately bias towards a correct RRC interpretation. This immediate bias reduces reading times for “kissed by” in (2c), and further attenuates a significant difference in reading times for “kissed by” between (2c) and (2d). Three groups of English learners in Taiwan at various L2 English proficiency levels (i.e., English learners who passed the elementary level of General English Proficiency Test (GEPT), English learners who passed the intermediate level of GEPT, and English-major Ph.D. students) and a comparison group of native English speakers participated in two on-line self-paced reading time experiments. Each group consisted of 20 subjects. The first experiment detected the sensitivity to the use of reduction and the use of noun animacy cue in the resolution of MV/RRC ambiguities such as those in (1). The second experiment explored the sensitivity to the use of reduction and the use of referential noun phrase information contained in a discourse context such as those in (2). Major findings of the dissertation are summarized as follows. First of all, the reduction results showed that, as the comparison group, the native English speakers were not immediately sensitive to the use of reduction in Experiment 1, but they were in Experiment 2. The elementary English learners and the intermediate English learners were immediately sensitive to the use of reduction in both Experiments 1 and 2. Like the native speakers, the English-major Ph.D. students were not immediately sensitive to the use of reduction in Experiment 1, but they were in Experiment 2. Second, the noun animacy cue results indicated that the native speakers were immediately sensitive to the use of noun animacy cue. Like the native speakers, both of the elementary English learners and the intermediate English learners were immediately sensitive to the use of noun animacy cue. However, the English-major Ph.D. students were not immediately sensitive to the use of noun animacy cue. Third, the referential information results displayed that all native speakers and English learners were not immediately sensitive to the use of referential information contained in a discourse context. In combination of the results mentioned above, it was found that the three groups of English learners immediately resolved the MV/RRC ambiguities on-line in a similar way as the native English speakers to varying extent. That is, the English-major Ph.D. students behaved most comparably to the native English speakers, followed by the intermediate English learners, and finally by the elementary English learners. Finally, it was found that the profiles of the MV/RRC ambiguity resolution by the elementary English learners, by the intermediate English learners, and by the English-major Ph.D. students were different from each other in Experiments 1 and 2 respectively. And the profiles by each learner group also differed in Experiments 1 and 2. By using experimental psycholinguistic techniques, this dissertation clearly makes two contributions. Firstly, it clarifies that L2 English learners are able to make immediate use of lexical, semantic, and syntactic information relevant to the MV/RRC ambiguity resolution simultaneously. This clarification disagrees with the hypothesis that L2 learners cannot use phrase structure-based parsing strategies of the sort that have been attested in L1 on-line sentence processing. Moreover, it proves that L2 English learners are likely to reach native-like competence of the MV/RRC ambiguities and to develop such native-like on-line processing strategies. Key words: main verb, reduced relative clause, MV/RRC ambiguities, sensitivity, reduction, noun animacy cue, referential noun phrase information, reading times