(2006), there are very few studies that have investigated the relationship between different types of recasts and acquisition. “Loewen and Philp (in press) investigated the relationship between Recast characteristics and scores on tailor-made posttests. The characteristics that they examined were linguistic focus, length of recast, segmentation (i.e. whether the recast repeated all or just part of the learner’s utterance), number of changes, and complexity (i.e. whether the corrective episodes were simple or complex, involving several turns).
Of these, only the feature concerning the number of changes predicted test scores: if the Recast made only a single change, the learners were more likely to get the right answer when tested” (p. 593).
In short, while there is clear evidence that Recasts can facilitate acquisition, there is still no clear picture of when they will do so. Learner factors, the nature of the targeted features, and the characteristics of the recasts help to determine, in complex ways, when Recasts work for acquisition and when they do not. If Recasts are intensive, focused, and individualized (as has been the case in laboratory studies), they are likely to be effective, yet so are other form-focused techniques (Ellis & Sheen, 2006).
2.2.4. Claimed Implicitness of Recast
The second problem is concerned with the claimed implicitness of Recast. The prevailing view is that recasts constitute an implicit form of negative feedback. Long (2007: 76) asserted that “implicit negative feedback in the form of corrective Recasts seems particularly promising”. In Ellis et al. (2006), the implicit corrective feedback in their study takes the form of Recast. So is the case with Ammar and Spada (2006), Long et al. (1998), and so on.
However, as pointed out recently by Ellis and Sheen (2006), Recasts are not always as implicit as Long (1996, 2007) claimed. For example, it might be argued that the Recasts used in Doughty and Varela’s (1998) study [see (2)] contain clear signals ,such as repetition and stress, which made their corrective force quite explicit.
Therefore, Recasts should not be viewed as necessarily implicit, but, depending on the linguistic signals that encode them, they should be taken as being more or less implicit or explicit. In Ellis and Sheen’s (2006: 583) words, “Recasts can lie at various points on a continuum of linguistic implicitness-explicitness”. The terms “explicit Recast” and characteristics of Recast “implicit recast” are only introduced by Sheen (2006: 388) after his study of the characteristics of Recast.
2.2.5. Research Rationale
To a large extent, the solutions to the afore-mentioned problems converge. Specifically, to solve the definitional discrepancy we should analyze the characteristics of Recasts and know more of different types of Recast. To abandon our incorrect assumption that Recasts are implicit in nature, we should acknowledge that there actually are implicit Recast and explicit Recast. It is in this way that the study of explicit Recast and implicit Recast gains importance.
The descriptive studies reviewed previously reveal that learners experience difficulty in interpreting Recast as being corrective. However, when Recast has features that make their corrective force explicit, learners are more likely to interpret Recasts as being corrective. So it is possible that Recasts which are combined with explicit features are more effective in leading to L2 development than those which are not. Moreover, the experimental studies reviewed show that Recasts are less effective than explicit types of negative feedback. Their inferiority is due to their ambiguous corrective force and the overt nature of explicit negative feedback. It can also be inferred that if Recasts are made more explicit to overtly indicate their corrective purpose, it may be more effective. However, so far there has been no such study conducted to testify to this. The purpose of the present study is to fill the research gap to compare the acquisitional benefits of explicit and implicit Recasts.
2.2.6. Recast in second and foreign language acquisition research
As Nicholas, Lightbown, and Spada (2001) rightly observed, “The research findings on Recast in L2 acquisition come from observational studies in classroom contexts and in NS dyadic situations as well as from experimental studies in laboratory contexts. Explanation for some contradictory interpretations of research findings appear to lie partly in the difference between classroom contexts and dyadic interaction in laboratories, but the evidence in both types of setting seems to point to greater effectiveness of Recasts in situations in which learners are given additional cues that help them recognize the Recast as feedback on error” (p.735).
For example, Lyster and Ranta (1997) studied the feedback behavior of four teachers in French immersion classes where teachers taught science, social studies, mathematics, and language arts lessons to young students. In these classes, the instructional focus was on subject matter content, not language form. The researchers analyzed transcripts based on over 18 hours of audio-recorded classroom interaction, with detailed examination of teachers’ feedback to more than 1100 student utterances containing errors. This enabled them to distinguish six categories of teacher feedback mentioned earlier: explicit correction, recasts, clarification request, Metalinguistic information, elicitation, and repetition.
Lyster and Ranta (1997) report that for three of the four teachers, 60% or more of the feedback was in the form of a recast. The fourth teacher, teaching a more advanced class, had a lower percentage (39%) of Recasts. Lyster and Ranta concluded that this is because teachers of more advanced students can make use of a greater variety of options to challenge the students. Overall, Recasts were the single most frequent feedback type (55%).
Lyster and Ranta (1997) further classified the learners’ “uptake” in response to the teachers’ feedback. They defined uptake as “a learner’s utterance that immediately follows the teacher’s feedback and that constitutes a reaction in some ways to the teacher’s intention to draw attention to some aspect of the student’s initial utterance” (p.49). Uptake covered a range of learner responses, from a simple “yes” admitting that the student had heard the teacher’s utterance to a repetition of the teacher’s feedback utterance and “self-repair,” in which the student produced a more accurate utterance even though the teacher’s correction did not provide the necessary information.
Lyster and Ranta found that the most frequent corrective feedback type, i.e. Recasts resulted in least uptake “because the teacher often continued with his/her turn after Recasting the student utterance, not waiting for a student response and not appearing to expect the student to provide a reaction to the feedback.” (Nicholas et al., 2001, p.739)
Lyster (1998) examined the relationship among error types, feedback types, and immediate Learner repair in 4 French immersion classrooms at the elementary level. He coded the 921 learner errors initiating each sequence as grammatical, lexical, or phonological, or as unsolicited uses on L1 (English) and corrective feedback moves as negotiation of form, recasts, or explicit correction. Findings indicate that lexical errors favored the negotiation of form; grammatical and phonological errors invited Recasts, but with differential effects in terms of learner repair.
Overall, the negotiation of form proved more effective at leading to immediate repair than did Recasts or explicit correction, particularly for lexical and grammatical errors, but not for phonological errors. Phonological repairs resulted primarily from Recasts. Ellis, Loewen, and Erlam (2006), cited in McDonough and Mackey (2006), compared the effects of different types of interactional feedback on learners’ performance on tests of explicit and implicit knowledge of regular past tense in English. Thirty-four ESL learners in three classes carried out two story narration tasks and several tests over a three week period. One class received interactional feedback in the form of Metalinguistic information and opportunity to respond, while the second class received Recasts and the third class did not receive any interactional feedback. The results of the explicit knowledge test (a grammatical judgment task) showed that the Metalinguistic information group outperformed the no-feedback group and the recast group only on the delayed post-test. The researchers proposed that “interactional feedback in the form of Metalinguistic information with learner response might have been more effective than Recasts because learners might be more likely to perceive it as overtly corrective” (McDonough & Mackey, 2006, p. 696).
In a dyadic laboratory study, Long, Inagaki, and Ortega (1998) provided different kinds of feedback and input to learners of Japanese and Spanish on different linguistic features. In interaction with native speakers and with taped-recorded cues, learners received either Recasts or modeling. The researchers report that recasts were more effective than models in creating short term changes in the learners’ interlanguage. As in the L1 literature, however, the effect of Recasts on different features was not uniform. Moreover, in some cases, there were no differences between the performance of learners in either of the two conditions, i.e. modeling versus Recast.
Carroll and Swain (1993) investigated the effects of different feedback strategies on adult ESL learners’ ability to recognize verbs which do or do not alternate in dative sentences. They measured the effect of different treatment types using the same items as in the treatment in the post-treatment recall test, and found that the group that received recast feedback “performed second best in the short-term recall session next to the group who received Metalinguistic explanation feedback, but only the latter maintained a long-term advantage” (Nabei & Swain, 2002, p. 46).
Likewise, through pre- and post- treatment production, Mackey and Philp (1998) indicated that intensive Recast treatment had a positive delayed effect on learners’ use of target question forms. They compared learners’ accurate and complex production of questions before and after the treatments. The advanced learners who enjoyed intensive Recast feedback during the treatment period produced developmentally higher question forms 78% of the time in the post-test while the other advanced learners who benefited from interactive negotiation treatment produced these only 17% of the time. Meanwhile, no differences were reported for the less advanced groups.
Perdomo (2008) assessed the effectiveness of oral Recasts in the EFL classroom. Thirty-eight college students and a female teacher participated in the study in a western state in Venezuela. Learners were