B.B Providing corrective feedback 020292

What is feedback?

In education, feedback is viewed as crucial for both encouraging and consolidating learning and this significance has also been recognized in the area of second language writing. Indeed, feedback is a key component of second language writing programs around the world, with product, process and genre approaches all employing it as a central part of their instructional repertoires (Hyland &Hyland, 2006b: 15) $$Insert formula here$$

Feedback 

Interactional feedback is an important source of information for learners. Most generally, it provides them with information about the success (or, more likely, lack of success) of their utterances and gives additional opportunities to focus on production or comprehension. There are numerous ways of providing feedback to learners from the explicit (stating that there is a problem) to the implicit (feedback during the course of an interaction). In this and the subsequent sections, we address the role of feedback and suggest ways that different types of feedback may impact learning. Figure 10.1 illustrates this concept with the mediating factor of attention.

Through interaction, learners’ attention is drawn to some element(s) of language with the possible consequence that that element/those elements will be incorporated into a learner’s developing system.

Discussed the role of negative evidence (information that a particular utterance is deviant vis-à-vis target language norms), it was pointed out that, at least with regard to children, it cannot be a necessary condition for acquisition. What, then, about second language learning? It is undoubtedly the case that adults (at least those in formal learning situations) do receive more correction than children, and it may further be the case that adults must have negative evidence

(i.e., that it is a necessary condition) in order to accomplish the goal of learning a second language (Birdsong, 1989; Bley-Vroman, 1989; Gass, 1988a; Schachter, 1988). While this research has been based primarily on theoretical arguments, there is some empirical evidence that negative evidence is in some instances necessary for second language acquisition.

White (1991) considered the development of adverb placement by French children learning English. She was interested in the question of how learners learn not to do something in the L2 that is present in the native language. In particular, French learners of English have to learn that English allows subject–adverb–verb (SAV) order (He always runs) and that it does not allow subject–verb–adverb–object (SVAO) order (*He drinks always coffee). White’s study consisted of five classes of French NSs learning English as a second language (two classes at grade 5 and three classes at grade 6) and one control group of monolingual NSs of English.

One of the grade 5 groups and two of the grade 6 groups were given explicit instruction on adverb placement as well as exercises and correction on adverb placement; the other groups were given instruction on questions using the same type of exercises but no explicit instruction on adverbs. The classroom treatment lasted two weeks. All children were given pretests, posttests immediately following the treatment sessions, a second posttest five weeks later, and a follow-up test a year later. The tests consisted of grammaticality judgment tasks (with correction), preference tasks, and a sentence-manipulation task. By comparing the groups’ performance,

White was able to show that negative evidence did indeed promote the learning of adverb placement. However, the effects of the treatment were not as long-lived as anticipated, as the two groups did not differ on their performance one year following the treatment negotiation

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">Negotiation serves as a catalyst for change because of its focus on incorrect forms. By providing learners with information about incorrect forms, negotiation enables learners to search for additional confirmatory or non-confirmatory evidence. If we accept that negotiation as a form of negative evidence and as a way of providing feedback serves the function of initiating change, we need to ask what factors determine whether the initiated change results in permanent restructuring of linguistic knowledge.

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">As with any type of learning, there needs to be reinforcement of what is being learned. This is schematized in Figure 10.2. If additional input is not available, learners do not have the opportunity to obtain confirmatory/non-confirmatory evidence. This, in fact, may explain the results of White’s study. Without additional focused evidence, it is not surprising that the learners did not retain knowledge of English adverb placement. In other words, acquisition appears to be gradual and, to state the matter simplistically, takes time and often requires numerous “doses” of evidence. That is, there is an incubation period extending from the

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">time of the initial input (negative or positive) to the final stage of restructuring and output.

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">Although White’s study is important in showing that negative evidence may be necessary t trigger a permanent change in a learner’s grammar, it does not show that positive evidence (i.e., input) alone is insufficient. (In fact, the question group of White’s study received little information about adverbs from the naturalistic classroom data to which they were exposed.

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">Trahey and White (1993) conducted a follow-up study to determine the effect of positive evidence. Their study consisted of two grade 5 classes of French students learning English. Both classes were given an input flood of English adverbs (positive evidence only) over a two-week period.

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">The same timetable as that used in the earlier White studies was used with the exception of three-week rather than five-week follow-up testing and no testing one year later. What they found was that input was sufficient for learners to notice that SAV order is possible in English, but that it was not sufficient to detect the ungrammaticality of SVAO sentences. Thus, these two experiments showed that positive evidence can reveal to learners the presence of information in the second language that is different from their native language, but that negative evidence is necessary to show what is not possible in the second language when it is possible in the native language. Trahey (1996) showed that an abundance of positive evidence a year after exposure yielded knowledge of grammatical sentences, but did not succeed in eradicating the ungrammatical sentences. Thus, positive evidence alone is not sufficient.

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">Other studies of feedback have also suggested that feedback obtained through negotiation serves a corrective function (Gass and Varonis, 1989; Pica, Holliday, Lewis, and Morgenthaler, 1989). The latter study is interesting in that the authors provided the first systematic evidence that learners respond differentially to different types of feedback. In their study one important focus was on different types of NS signals to NNS errors. They found that the greatest amount of modification comes in response to clarification requests, as in the following example (Nobuyoshi

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">and R. Ellis, 1993, p. 204): (10-27)

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">  NNS: He pass his house.

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">NS: Sorry?

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">NNS: He passed, he passed, ah, his sign. as opposed to seeking confirmation through modeling. What this suggests is that the fact that the NNS is “forced” to make the actual

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">correction, as opposed to hearing and perhaps thinking about the correct form, is in itself a facilitator to acquisition. But again, we are left with the unknown factor of longer-term retention.

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">One study that suggested longer-term retention after focused attention is that of Nobuyoshi and R. Ellis (1993). Learners had to describe a series of pictures that depicted events that had happened the previous weekend and the previous day. The experimental group received feedback through clarification requests that focused on past tense forms. The control group did not receive such focused feedback. The results can only be considered suggestive given the very small sample size. However, in the experimental group, two of the three subjects were able to reformulate the correct forms after feedback and were able to maintain the correct forms at a subsequent administration one week later. In the control group, none of the subjects showed an accuracy gain.

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">Similarly, Lightbown (1992) compared corrective feedback provided by teachers immediately after the occurrence of an error in a communicative activity versus feedback on audiolingual drills or pure practice activities.

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">She found that in both cases learners were able to self-correct, but only in the first case was the self-correction incorporated into their second language systems, as evidenced by use of the targeted form outside of the classroom.

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">An early study on the effect of corrective feedback on grammatical reorganization was carried out by Carroll, Roberge, and Swain (1992). The comparison was between groups with corrective feedback and groups with no corrective feedback. The linguistic focus was on regular noun formation in French. After receiving training on the relationship between verbs and nouns (e.g., attelé-attelage, “harnessed”/“harnessing”), learners were given new words to manipulate. Some participants were corrected and others were not. The results showed that corrective feedback was important in the learning of individual items, but that it had little effect on a learner’s ability to generalize this information to new items.

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">Takashima (1995), in a study of Japanese learners of English, investigated the effects of feedback that was focused on particular morphological form (past tense and plural)5 versus feedback that was communication-oriented. The focused feedback was in the form of clarification requests (Sorry?, What did you say?). Groups of students had to work together to make up a story based on a sequence of pictures, of which each student in the group had only one. One student was then nominated to tell the story to the class. This was the actual feedback session as the teacher provided either focused morphological feedback or content feedback. The accuracy rate for past tense increased at a faster rate during the time of the study (11 weeks) in the focused morphological correction group as opposed to the content correction group. Further, the magnitude of the difference increased as a function of time. Improved accuracy was noted for the particular student who was corrected (in front of the class) as well as for those students who were in the class observing the interaction. Interestingly, when considering the actual reformulations by individual students, there was no correlation between the reformulated utterances and improvement on the use of the structure on tests. This further suggests that the actual interaction does not constitute change itself, but is only a catalyst for later change. Illustrative of this is the following excerpt from Takashima (1995, p. 77), in which the first clarification request appears to fall short of the mark in that the student makes no change, but as the storytelling continues, the student seems to be more sensitive to the past tense forms, even self-correcting in the last turn.

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">(10-28) S = student; T = teacher

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">S: One day, the fairy, sting the magic wand to Cinderalla.

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">T: Sorry?

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">S: One day, the fairy sting the magic wand to Cinderalla.

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">T: OK.

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">S: Cinde, ah, Cinderaella changed into, the beautiful girl.

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">(Laugh) Ah, and, the, Cin, Cinderella wen Cinderella

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">went to the palace by coach. The, the prince fall in

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">love at a first glance.

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">T: Sorry?

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">S: Ah, the prince fall in, falled falled in love Cinderella at

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">a first glance. And they dance, they danced. . . Ah,

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">Cin, Cinderella have, Cinderella have to go home.

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">Here, the input has been enhanced through clarification requests and

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">the output has similarly been enhanced (Takashima’s term), apparently as

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">a function of the input enhancement.

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"> 12.0pt;font-family:"Times New Roman","serif"">  <span style="font-size:12.0pt;font-family:"Times New Roman","serif"">Recasts ''' <p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">Recasts are another form of feedback, though they are less direct and more subtle than other forms of feedback. A recast is a reformulation of an incorrect utterance that maintains the original meaning of the utterance, as in 10-29, where the NS reformulates the NNS’s incorrect question (Philp, 1999).
 * '''<span style="font-size:

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">(10-29) NNS: Why he want this house?

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">→NS: Why does he want this house?

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">Recasts are complex. For example, is it a partial recast? A full recast? A response to a single error or to multiple errors (how many changes are made)? We present two examples that illustrate forms that recasts can take. In 10-30, a recast with rising intonation, the auxiliary is added and the verbal morphology is corrected (Philp, 1999, p. 92). In 10-31 the verb form is corrected (from future to subjunctive, required after avant que) without rising intonation (Lyster, 1998, p. 58).

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">(10-30) NNS: What doctor say?

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">NS: What is the doctor saying?

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">(10-31) S = student; T = teacher

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">S: Avant que quelqu’un le prendra. before someone it will take

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">“Before someone will take it.”

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">T: Avant que quelqu’un le prenne. before someone it takes

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">“Before someone takes it.”

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">There have been a number of recent reviews of recasts in the second language literature, focusing on experimental as well as theoretical concerns (Nicholas, Lightbown, and Spada, 2001; R. Ellis and Sheen, 2006; Long, 2007; Mackey and Goo, 2007). Because recasts are an indirect form of correction, it is not clear to what extent they are relevant to acquisition. There have been a number of empirical studies focused specifically on the effectiveness of recasts. The results from these studies are mixed.

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">Lyster and Ranta (1997) collected data from children in Grades 4–6 enrolled in French immersion programs. Their research considered recasts by teachers following errors and, importantly, the reaction by the student (“uptake,” in their terminology) in the subsequent turn. They argued that uptake “reveals what the student attempts to do with the teacher’s feedback” (p. 49). Even though there were numerous instances of recasts found in the data, they did not appear to be particularly effective.

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">Rather, students were more prone to repair utterances following other types of feedback.

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">Unfortunately, an immediate response may not be revealing, in that learners may be “mimicking or repeating without true understanding” (Gass, 2003, p. 236). This makes recasts a somewhat elusive concept to deal with and research often produces mixed results. For example,

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">Mackey and Philp (1998) found that an immediate response by a learner was not necessarily related to development, whereas Nabei and Swain (2002) and Lyster (2004) found the reverse. As noted elsewhere in this chapter, it is not always possible to judge what the effects of learning are by immediate reactions.

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">Lyster (1998), using the same database as reported on in the Lyster and Ranta (1997) study, divided recasts into four types depending on two features: (a) declarative; (b) interrogative; (c) confirmation of the original utterance; or (d) additional information. Lyster found that there was some confusion between the corrective and approval functions of recasts. He argued that recasts may not be particularly useful in terms of corrective feedback, but they allow teachers to move a lesson forward by focusing attention on lesson content rather than on language form. Lyster (2004), in a study that took place in immersion classrooms, compared the benefits of recasts and prompts. By prompts, he includes

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">the following four types:

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">(10-32) Clarification requests

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">Student: Et le coccinelle. . . “And the (M) ladybug. . .”

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">Teacher: Pardon? “Sorry?”

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">Student: La coccinelle. . . “The (F) ladybug. . .”

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">(10-33) Repetitions

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">Student: La chocolat. “(F) Chocolate.”

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">Teacher: La chocolat? “(F) Chocolate.”

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">Student: Le chocolat. “(M) Chocolate.”

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">(10-34) Metalinguistic clues

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">Student: Parce qu’elle cherche, euh, son, son carte.

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">“Because she’s looking for, um, her, her (M)

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">card.”

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">Teacher: Pas son carte. “Not her (M) card.”

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">Student: Euh, sa carte? “Um, her (F) card?”

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">(10-35) Elicitation

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">Teacher: Il vit où un animal domestique? Où est-ce que ça

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">vit? “Where does a pet live? Where does it live?”

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">Student: Dans un maison. “In a (M) house.”

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">Teacher: Dans. . .? Attention. “In. . .? Careful.”

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">Student: Dans une maison. “In a (F) house.”

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">Data were collected within the context of a fifth-grade-content French immersion classroom. Teachers either provided no feedback, recasts, or prompts. The focus was on French grammatical gender. Determination of learning was made through both oral and written tasks following the five-week treatment sessions. He found that form-focused instruction with prompts was more successful than with recasts, based on the written measures. There was not a significant difference on the oral assessment measures. This study was conducted in a content-based classroom where there are numerous nonlinguistic demands made on the learner, possibly making it difficult to focus on the subtle corrective function of recasts.

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">The results of a study by Ammar and Spada (2006) are similar to those of Lyster (2004). Their study took place in intensive English classes (L1 French) in Montreal with Grade 6 pupils. The target grammatical area was possessive determiners (his/her), a structure notably difficult for French learners of English. Prompts turned out to be more effective than recasts. The effectiveness of recasts depended, in part, on proficiency level, with more advanced learners receiving more benefit than learners of lower proficiency.

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">Ellis, Loewen, and Erlam (2006) looked at metalinguistic explanation (explicit feedback) and recasts (implicit feedback), finding that on tests of both explicit and implicit knowledge the metalinguistic explanation group outperformed the recast group, most likely due to recognition of the overtly corrective nature of metalinguistic feedback. Explicit feedback benefited both implicit and explicit knowledge.

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">In general, a number of studies have suggested that there is a positive effect for recasts on later learning (see Nicholas, Lightbown, and Spada 2001 and Mackey and Goo, 2007 for reviews). Leeman (2003) looked at noun–adjective agreement in Spanish in attempting to determine the benefits of recasts, particularly because they serve to provide positive evidence in a salient way. She had three experimental groups: (1) recasts, which she proposed provided both negative evidence as well as enhanced salience of positive evidence; (2) negative evidence; and (3) enhanced salience of positive evidence. She found that the first and third groups (recast group and enhanced salience of positive evidence group) showed post-treatment benefits. In this way she was able to separate out the various parts of recasts (positive and negative evidence). Thus, it appears that recasts are useful due to the enhanced salience provided in recasts rather than negative evidence. Han (2002) investigated consistency of use of past tense morphology. She found that recasts were beneficial, but proposed four conditions for their usefulness: individualized attention, consistent focus, developmental readiness, and intensity.

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">McDonough (2007), in a study of the acquisition of past tense in an interactive context, compared clarification requests and recasts, finding that both positively influenced the acquisition of past tense. However, in a study on the acquisition of the comparative and past tense, R. Ellis (2007) considered the effect of recasts and metalinguistic feedback, not finding a positive effect for recasts. However, the treatment time in his study was much shorter than in other studies investigating the impact of recasts on the development of English past tense morphology. Ishida (2004) considered Japanese morphology in her study of recasts. where there are numerous nonlinguistic demands made on the learner, possibly making it difficult to focus on the subtle corrective function of recasts.

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">The results of a study by Ammar and Spada (2006) are similar to those of Lyster (2004). Their study took place in intensive English classes (L1 French) in Montreal with Grade 6 pupils. The target grammatical area was possessive determiners (his/her), a structure notably difficult for French learners of English. Prompts turned out to be more effective than recasts. The effectiveness of recasts depended, in part, on proficiency level, with more advanced learners receiving more benefit than learners of lower proficiency.

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">Ellis, Loewen, and Erlam (2006) looked at metalinguistic explanation (explicit feedback) and recasts (implicit feedback), finding that on tests of both explicit and implicit knowledge the metalinguistic explanation group outperformed the recast group, most likely due to recognition of the overtly corrective nature of metalinguistic feedback. Explicit feedback

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">benefited both implicit and explicit knowledge.

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">In general, a number of studies have suggested that there is a positive effect for recasts on later learning (see Nicholas, Lightbown, and Spada 2001 and Mackey and Goo, 2007 for reviews). Leeman (2003) looked at noun–adjective agreement in Spanish in attempting to determine the benefits of recasts, particularly because they serve to provide positive evidence in a salient way. She had three experimental groups: (1) recasts, which she proposed provided both negative evidence as well as enhanced salience of positive evidence; (2) negative evidence; and (3) enhanced salience of positive evidence. She found that the first and third groups (recast group and enhanced salience of positive evidence group) showed post-treatment benefits. In this way she was able to separate out the various parts of recasts (positive and negative evidence). Thus, it appears that recasts are useful due to the enhanced salience provided in recasts rather than negative evidence. Han (2002) investigated consistency of use of past tense morphology. She found that recasts were beneficial, but proposed four conditions for their usefulness: individualized attention, consistent focus, developmental readiness, and intensity.

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">McDonough (2007), in a study of the acquisition of past tense in an interactive context, compared clarification requests and recasts, finding that both positively influenced the acquisition of past tense. However, in a study on the acquisition of the comparative and past tense, R. Ellis (2007) considered the effect of recasts and metalinguistic feedback, not finding a positive effect for recasts. However, the treatment time in his study was much shorter than in other studies investigating the impact of recasts on the development of English past tense morphology.

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">Ishida (2004) considered Japanese morphology in her study of recasts. Their study included three posttests and development was operationalizedas two questions with unique lexical items in different tasks.

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">Both recasts and primed production were predictive of ESL question development. What was particularly interesting is that mere repetition of the recasted form (uptake in Lyster and Ranta’s framework) was not correlated with development.

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">Other studies that show a positive effect for recasts point to two main problems with recast studies: the concept of uptake and the data to be included in analysis. Mackey and Philp (1998) pointed out that uptake (as defined by Lyster and Ranta, 1997) may be the wrong measure to use in determining effectiveness. Their data represented an attempt to go beyond the turn immediately following a recast. They make the point (cf. Gass, 1997; Gass and Varonis, 1994; Lightbown, 1998) that, if one is to consider effectiveness (i.e., development/acquisition), then one should more appropriately measure delayed effects. In particular, Mackey and

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">Philp considered the effects of interaction with and without recasts on learners’ knowledge of English questions. Their results showed that, for more advanced learners, recasts plus negotiation were more beneficial than negotiation alone. This was the case even though there was not always evidence for a reaction by the learner in the subsequent turn. A study by Long, Inagaki, and Ortega (1998) also attempted to determine the role of recasts (in this case as opposed to models). They investigated (a) the acquisition of ordering of adjectives and a locative construction by English learners of Japanese, and (b) the acquisition of topicalization and adverb placement by English learners of Spanish.

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">Their results were mixed inasmuch as only one of the learner groups (Spanish) showed greater learning following recasts as opposed to models. Furthermore, these findings were true for adverb placement only. A problem having to do with the data used for analysis was noted by Oliver (1995). Frequently, after a recast, there is no opportunity for the original speaker to make a comment. This may be due to a topic shift, as in 10-40 (Oliver, 1995, p. 472), or the inappropriateness of making a comment because the recast had been in the form of a yes/no question and the appropriate response would not be a repetition, but a yes/no response.

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">(10-40) From Oliver (1995, p. 472)

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">NNS: A [c]lower tree.

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">NS: A flower tree. How tall is the trunk?

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">When the lack of opportunity/appropriacy is included, the percentage of “incorporated” recasts greatly increases. Lyster (1998) argued that the context of language use in these studies (child–child dyadic interactions in Oliver’s research and teacher–student interactions in his own research) is different and that, in fact, in classrooms the teacher often keeps the floor, thereby (as mentioned earlier) drawing attention to content and not to language form. In his 2004 study, Lyster compares recasts with prompts (see examples 10-32 to 10-35 above) finding the superiority of prompts to recasts given the opportunity for some form of uptake.

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">There is one final issue to address before concluding this section on feedback. What do learners perceive? In a study by Mackey, Gass, and McDonough (2000), data were collected from 10 learners of English as a second language and 7 learners of Italian as a foreign language. The study explored learners’ perceptions about feedback provided to them through task-based dyadic interaction. In the interactions, learners received feedback focused on a range of morphosyntactic, lexical, and phonological forms. After completing the tasks, learners watched videotapes of their previous interactions and were asked to introspect about their thoughts at

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">the time the original interactions were in progress. Examples of the interactions and the recall comments of the learners follow.

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">(10-41) Morphosyntactic feedback (perceived as lexical feedback)

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">NNS: C’è due tazzi.

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">“There is two cups (m. pl.).”

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">INT: Due tazz-come?

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">“Two cup—what?”

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">NNS: Tazzi, dove si puó mettere té, come se dice questo?

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">“Cups (m. pl.), where one can put tea, how do you

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">say this?”

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">INT: tazze?

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">“Cups (f. pl.)?”

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">NNS: ok, tazze.

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">“Ok, cups (f. pl).”

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">Recall: I wasn’t sure if I learned the proper word at the

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">beginning.

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">(10-42) Phonological feedback correctly perceived

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">NNS: Vincino la tavolo è.

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">“Near the table is (the correct form is vicino).”

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">INT: Vicino?

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">“Near?”

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">NNS: La, lu tavolo.

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">“The? table.”

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">Recall: I was thinking. . . when she said vicino I was

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">thinking, OK, did I pronounce that right there?

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">(10-43) Lexical feedback correctly perceived

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">NNS: There is a library.

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">NS: A what?

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">NNS: A place where you put books.

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">NS: A bookshelf?

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">NNS: Bok?

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">NS: Shelf.

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">NNS: Bookshelf.

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">Recall: That’s not a good word she was thinking about

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">library like we have here on campus, yeah.

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif""> The results showed that learners were relatively accurate in their perceptions about lexical, semantic, and phonological feedback. However, morphosyntactic feedback was generally not perceived as such.

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">Consequently, it is not always clear that learners perceive feedback in the way it was intended (see also, Hawkins, 1985). Thus, there may be a differential role for feedback in different linguistic areas, as suggested by Pica (1994). For example, perhaps morphosyntactic feedback is not noticed because, as is typical in a conversational context, individuals are focused on meaning, not on language form. Phonological and lexical errors can interfere with basic meaning and hence need to be attended to on the spot if shared meaning is to result; the morphosyntactic examples in the Mackey, Gass, and McDonough (2000) study generally dealt with

<p class="MsoNormal" style="margin-bottom:0in;margin-bottom:.0001pt;line-height: normal;mso-layout-grid-align:none;text-autospace:none"><span style="font-size: 12.0pt;font-family:"Times New Roman","serif"">low-level nonmeaning-bearing elements.

<p class="MsoNormal"><span style="font-size:12.0pt;line-height:115%;font-family: "Times New Roman","serif"">