The Effects of Explicit and Implicit Recasts on the Acquisition of Two Grammatical Structures and the Mediating Role of Working Memory

Abstract

The efficacy of recasts – a strategy for providing corrective feedback – is a matter of controversy in the field of second language acquisition (SLA). According to one position recasts are effective because they provide learners with both negative and positive evidence. However, according to another position they are ineffective because they frequently fail to push learners to correct their own errors. SLA researchers have also disputed the value of recasts on the grounds that they constitute an implicit type of corrective feedback that learners fail to attend to. However, as other researchers have pointed out, recasts are not invariably implicit; in many cases their corrective force is transparent. This thesis compared the effects of corrective recasts (i.e. repetition plus recasting with stress) that are relatively explicit and implicit recasts (i.e. simple recasting) on the acquisition of two English grammatical structures (3rd person -s and embedded questions). The effects of these two types of recasts were examined in terms of both learners’ acquisition of implicit knowledge (measured by means of elicited imitation and oral production) and their acquisition of explicit knowledge (measured by means of written production and an untimed grammaticality judgment test). It also investigated the role of uptake-with-repair and working memory in recasts-driven L2 learning. This was a classroom-based quasi-experimental study, with four groups of EFL learners (n = 109) in a Chinese university (i.e. a corrective recast group, an implicit recast group, a task control group, and a test control group). All groups completed a pretest, an immediate posttest, and a delayed posttest. The treatment for the experimental groups consisted of corrective feedback (i.e. corrective recasts and implicit recasts) provided online while the learners performed focused tasks designed to provide contexts for the use of the target structure. The task-based lessons were audio-recorded and transcribed, enabling all occasions where uptake occurred following corrective feedback to be noted. All groups also completed working memory tests. Language knowledge scores were compared by means of parametric/non-parametric ANOVAs or ANCOVAs to examine the effect of feedback. Correlational analyses were conducted to investigate the relationships between repair rate, knowledge gains and working memory capacity. Simply performing the tasks without any feedback benefited L2 acquisition, as shown by the fact that the task control group outperformed the test control group in gains in both implicit and explicit knowledge with a large effect size. Corrective recasts facilitated the development of implicit and explicit knowledge but implicit recasts only led to significant gains in explicit knowledge. Corrective recasts were more effective than implicit recasts on the implicit knowledge of embedded questions and the explicit knowledge of 3rd person -s. These results suggest that explicitness in recasts makes a difference to their efficacy, and the superior effect of corrective recasts over implicit recasts on the development of implicit or explicit knowledge depends on the target structure. This relative effect on explicit knowledge was evident for those learners who had low prior knowledge of the target structure, pointing to the mediation of learner proficiency. The repair rate was also higher after corrective recasts than implicit recasts. However, whether repair did or did not occur was found to be generally irrelevant to L2 learning through recasts. Neither phonological short-term memory (tapped by a non-word span test) nor complex working memory (tapped by a listening span test) was, by and large, a decisive factor in recasts-driven L2 learning in this classroom-based study. The thesis concludes with a discussion of the theoretical and pedagogical implications of the study. I propose that in the context of Chinese university classroom L2 acquisition can be best promoted by means of focused tasks and corrective feedback consisting of explicit recasts, for low proficiency learners in particular.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 32 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.