| Summary: | 博士 === 國立高雄師範大學 === 英語學系 === 94 === This thesis aimed to examine the English vowel production and perception by Taiwanese EFL learners and to unveil the possible interaction in between. Also, an OT(Optimality Theory)-based discussion was provided to interpret the main results in the acoustic experiments and to elaborate the re-ranking processing in learnability.
Sixty Taiwanese EFL (English as a foreign language) students (TS) and twenty native speakers of American English (NS) were recruited for the current research. These Taiwanese EFL learners were composed of two subgroups: thirty college English-majors (HEFL) and thirty senior high school students (LEFL). Both groups took part in the questionnaire survey, the general English proficiency test, the production experiment and the perception experiment. As for the native speakers of English, they merely did the English vowel production task and the English vowel perception task.
According to the statistical results and OT-based discussion, the following are major findings related to the four research questions. Firstly, greater “similarity effect” was identified in TS English vowel production; the acoustic deviance among similar vowels was greater than that among new vowels. Under greater L1 interference, LEFL learners’ English pronunciation approximated closer to their L1 counterpart pronunciation. HEFL, on the contrary, produced English vowels in a more native-like way. Additionally, vowel duration offered an important cue for TS to differentiate English tense-lax vowel contrasts. Secondly, less “similarity effect” was found in TS English vowel perception; the acoustic deviance among similar vowels was less than that among new vowels. LEFL experienced greater L1 influence and diverged to a greater extent from NS English perception than HEFL. Thirdly, with regard to within-group production-perception asymmetry, TS were less consistent in these two speech tasks than NS, indicating that TS showed greater asymmetry in speech tasks. Concerning the TS cross-group asymmetry between vowel production and perception, both HEFL and LEFL were less dispersed in the perception task than in the production task. Fourthly, Optimality Theory could adequately interpret the acoustic findings and the issue of learnability. Second language acquisition was closely associated with such linguistic factors as L1 filter, universal constraints, markedness constraints and constraint re-ranking.
Based on the research findings and discussion, there were several primary pedagogical implications. To start with, if language learners could be directed to the significant differences between their native language and the target language, for example, in tense-lax vowel distinction and in the dissmilatory constraints of backness, it would help lower the possible L1 interference and overcome the pronunciation problems. Secondly, besides the duration cue, other phonetic properties of English vowels should be introduced to Taiwanese EFL learners, for instance, formants, muscle tension, tongue height and backness, all of which might be beneficial in English vowel discrimination. Thirdly, among three kinds of phonetic symbols, the Daniel Jones system (DJ), the Kenyon-knott system (KK) or the Trager Smith system, the Trager Smith system truly reflected the phonetic nature of tense vowels with the final gliding; therefore, the Trager Smith system would be of greater help in teaching English tense-lax contrasts. Fourthly, concerning the future pronunciation courses, computer-assisted acoustic software could be adopted to offer speaking and listening practice and to focus on the confusing vowel contrasts with minimal pairs. Finally, from the perspective of OT, since many production or perception difficulties for Taiwanese EFL learners resulted from the L1 phonological transfer and universal as well as markedness constraints, learning a language involved an ongoing process, during which the set of interlanguage constraints should be adequately re-ranked to achieve the ultimate goal of native-like speech performance.
|
|---|
