earticle

영어

This study explores why different effects are obtained through recasts, drawing attention to the nature of the target construction of recasts. During last decades, the negotiated interaction model have got much popularity in the study of language learning including EFL learning. One of the issues in this model is how errors are corrected through interaction, and researchers have got interested in cognitive processes that can be identified through conversational interactions. In particular, it has been found that recasts serve as a tool for providing negative evidence and constitute contingent responses to L2 learners’ ungrammaticality. It has been tested in several studies whether or not and how much recasts are effective in error correction (Mackey & Philp, 1998; Long, Inagaki, & Ortega, 1998; Phip, 2003). The findings reveal that the effects of recasts are quite different among target grammatical constructions. I argue in this study that the observed differences in the effects of recasts may root in the different grammatical properties of the target constructions among other factors and L2 learners’ sensitivity to those properties. For pedagogical implications, I suggest that recasts should be taken to be a tool not for direct error correction but for building implicit knowledge of certain grammatical items.