| The present study addressed the following under-explored issues: (1) learners' perceptions about recasts (i.e., as the interlocutor's response to content, negative evidence, and/or positive evidence), (2) relationship between learners' perceptions about recasts and different features of recasts, and (3) relationship between learners' perceptions about recasts and their subsequent L2 learning.; While previous research has supported claims about the facilitative effects of recasts in SLA, there is considerable controversy regarding the precise developmental benefits of recasts. Researchers generally agree on the need for positive evidence; however, the role of negative evidence has been controversial. Given the multiple discourse functions of recasts, they are often perceived as interlocutors' reactions to content rather than corrective feedback. Researchers have suggested that such an interpretation of recasts may limit their developmental benefits. Under the premise that learners' perceptions about recasts may constrain their utility, the present study investigated a possible link between learners' perceptions about recasts and L2 learning to examine the developmental benefits of recasts.; Forty-nine high-beginning to intermediate learners of Japanese carried out communicative tasks with a native speaker, during which they received recasts of their non-targetlike productions of targeted items (the te-form verbs and numeral classifiers). Short-term and more sustained learning was assessed by a combination of standard and customized immediate and delayed posttests. The learners' perceptions about recasts were measured by immediate reports (Egi, in press) or stimulated recall (Gass & Mackey, 2000).; The study provided a finely grained analysis of learners' various perceptions about recasts. A significant relationship was found between their perceptions about recasts and (1) the length of recasts and (2) the number of changes made to the learner's original utterance. Furthermore, the study empirically supported the claim that recasts are ineffective when learners perceive them as responses to content. A (non-significant) trend suggested that there may be greater benefits for recasts when learners perceive them as positive evidence rather than as negative evidence. In summary, this study suggests that the utility of recasts may in part be constrained by learners' perceptions about recasts. |
