Self-repairs in Taiwanese EFL Learners'' English Oral Production

• Main Authors: I-Lin Chen, 陳怡霖
• Other Authors: Agnes Huang
• Format: Others
• Language: en_US
• Published: 2012
• Online Access: http://ndltd.ncl.edu.tw/handle/08616059701296551749

碩士 === 國立高雄第一科技大學 === 應用英語研究所 === 100 === Second language learners in their oral production frequently create reformulations to correct their speech errors, which is known as self-repairs. Though self-repairs may influence the quality of learners’ speech, L2 self-repairs are unavoidable and could serve special functions in speech. To examine the role of self-repairs in L2 speech, this study investigated Taiwanese EFL learners’ self-repair patterns in their English monologues under more and less pressured conditions. Three research questions are addressed: (a) What categories of self-repairs do Taiwanese EFL learners create in their English monologues under two pressured conditions (more pressured vs. less pressured) as shown in a self-compiled learner corpus? (b) Quantitatively, how does the amount of self-repairs differ in the two pressured conditions? (c) Qualitatively, what are the discourse functions of each self-repair category in the two pressured conditions? The learner data in this study were collected from 56 English-major juniors in three different classes of a university of science and technology in southern Taiwan. Four monologic tasks were chosen for this study: impromptu speech, collaborative story telling, narration of personal information, as well as question and answer. The former two tasks were classified into the more pressured condition whereas latter two were classified into the less pressured condition. All the data were audio-recorded and then transcribed. Self-repairs were coded into six categories: (a) lexical error repairs, (b) grammatical error repairs, (c) appropriate level of information repairs, (d) inappropriate information repairs, (e) message abandonment, and (f) same information repairs. Quantitatively, the raw frequency and percentage of each self-repair category were counted and compared across the two pressured conditions by using Wilcoxon matched-pairs signed-ranks tests. Next, qualitative analyses were carried out by identifying the functions of each self-repair category in each pressured condition. The quantitative results revealed that learners created the most same information repairs, followed by appropriate level of information repairs in both pressured conditions. The former repairs mainly functioned to buy time for speech planning whereas the latter aimed to enrich the content of speech. Of all the six self-repair categories, only three categories reached significant differences between the two pressured conditions: appropriate level of information repairs, inappropriate information repairs, and message abandonment. The quantities of the first two categories were significantly greater in the less pressured condition, indicating that learners had more resources available in working memory to monitor speech content, but familiar information tended to be less organized and require more inappropriate information repairs. On the other hand, the amount of message abandonment was significantly larger in the more pressured condition possibly because higher anxiety caused disruption in the speech planning process. In terms of functions, lexical error repairs mainly served to change expressions to match speakers’ intentions and grammatical error repairs primarily functioned to adjust verb forms for tense and aspect. Appropriate level of information repairs were created mainly to clarify the content of speech while the major function of inappropriate information repairs was to modify the meaning of expressions. Finally, message abandonment was mostly used to move to the next speech plan whereas the main purpose of same information repairs was to buy time for learners’ online planning. As for pedagogical implications, as learners tend to prioritize speech content, teachers should design tasks that aim for communicative purposes to assist students in the organization of speech content. Teacher should also encourage learners to carefully plan familiar information and use transition words strategically to switch topics without overusing message abandonment. To reduce same information repairs, teachers can help students analyze their own speaking patterns and teach them useful strategies to cope with anxiety in L2 oral production.