Disambiguating recasts with prosodic and extra-linguistic cues in task-based interactions among young learners

Abstract

This study investigates the relative effects of enhanced and unenhanced recasts on young learners’ question development in L2 English, examining whether enhanced recasts can contribute to disambiguation. Forty-six Secondary One students in Hong Kong were evenly divided into an enhanced recast group and an unenhanced recast group. Each participant participated in a pretest, followed by three treatment sessions in three consecutive weeks. Participants in the enhanced recast group received enhanced recasts with prosodic (tonal stress) and extra-linguistic (gestures and facial expressions) cues from the teacher, whereas those in the unenhanced recast group received normal recasts when they made mistakes when asking questions. After that, an immediate post-test and a two-week delayed post-test were carried out to examine participants’ progress in L2 question formation. The results of the immediate post-test showed that both types of recasts contributed to L2 advancement, but the effects of enhanced recasts were relatively strong as compared with unenhanced recasts. Both groups performed significantly better in the delayed post-test than in the pretest, indicating the robust effects of recasts (enhanced or unenhanced) on learning L2 question formation. However, the differences between the two groups were significantly reduced in the delayed post-test. In addition, the enhanced group scored significantly lower in the delayed post-test than in the immediate post-test, suggesting a greater degenerating effect of enhanced recasts than unenhanced recasts.