| This study investigated the connection between the process of second language acquisition (SLA) and negative feedback, i.e. feedback to learners that informs them of their lack of success in L2 production. Previous studies have pointed to explicitness as key factor in the impact that negative feedback can have on the learner. Also important is the types of linguistic information supplied by negative feedback to learners, e.g., models or hint for more accurate production, or requests for clarification. To explore these features in relation to their potential contributions to SLA, this study focused on three feedback types: (1) implicit negative feedback that provided correct forms (e.g., recasts), (2) implicit negative feedback that encouraged modifications focusing on meaning (e.g., clarification requests), and (3) explicit negative feedback that provided the metalinguistic information utilized for modifications (e.g., metalinguistic cues). Their relative contributions to SLA were determined through a comparison of the extent to which learners modified their production toward greater accuracy after receiving each type of feedback. Additional factors were the grammatical focus of the feedback and the passage of time. Thirty-eight beginning-level learners of Japanese as a foreign language were randomly divided into three experimental groups: (1) the recast group (N = 10), (2) the clarification request group (N = 9), (3) the metalinguistic cue group (N=9), and one control group (N=10). Learners engaged in picture-describing tasks with the researcher that solicited their production of an existential structure. During the tasks, the researcher provided designated types of negative feedback to errors in learners' production. The major findings were: (1) negative feedback showed the greatest facilitation on L2 production when encoded as metalinguistic cues followed by recasts, however, (2) clarification requests showed no significant facilitation. These findings underscored the importance of explicitness making negative feedback effective for SLA, and pointed to its encoding through models and metalinguistic information as crucial to this process. |
