Recasts in the EFL classroom : a comparison of native and nonnative teachers

Recasting (i.e., corrective feedback in which the teacher paraphrases a learner's incorrect utterance without explicitly labeling it as erroneous) is a frequent phenomenon in classroom discourse. Despite its frequency and naturalness, educators continue to debate its efficacy. At issue is whether learners notice such implicit feedback in order to make use of it. This on-going debate centers on the following question: What makes a teacher's recast noticeable to a student? While most of the studies in the recast literature have emphasized student factors such as working memory and/or developmental readiness (e.g., Havranek & Cesnik, 2001), few studies have explored how teacher factors affect learner perceptions of and receptivity to recasts. This study fills this gap by employing qualitative methods to investigate student perceptions of their teacher recasts. Six classes in Applied English Departments at three different institutes of technology in mid-southern Taiwan participated in this study. Different methods were employed to gather student and teacher data in order to arrive at a more complete understanding of classroom recasts: classroom observations, individual student interviews, group stimulated recall interviews, and teacher interviews. According to student interview data, seven teacher categories (e.g., nativeness, teacher language use, teacher affect, etc.) were found to have a significant impact on students’ noticing of and receptivity to recasts. In particular, the findings indicate that EFL student perceptions of recasts are profoundly affected by teachers’ language use (e.g. phonetic and syntactic features) and teachers' nativeness (e.g., native vs. nonnative). Other factors such as teacher-student rapport also mediate students’ attention to and understanding of feedback. In addition to the student self-report data, classroom observation data of teachers' behavior indicated striking differences as well--native teachers tended to correct more grammatical errors while nonnative teachers corrected more phonological errors. In light of these findings, suggestions for improving student awareness of corrective recasts are given to both native and nonnative teachers. It is hoped that the qualitative categories uncovered in this study will lead to more rigorous, testable hypotheses for future quantitative analysis.