In this section, two new forms of recasts will be introduced and the psycholinguistic basis of the two forms of recasts will be addressed. But first we will look at some of the deficiencies of standard recasts.
One of the major criticisms of standard recasts is that they have a tendency to go unnoticed by language learners (Egi, 2010; Lasagabaster & Sierra, 2005). This is because recasts are often confused with confirmation checks, which naturally occur in conversational discourse. In the following example, the teacher uses a standard recast to correct the student‘s past-tense error. The student appears to not notice the correction, perhaps because his focus was on meaning and the recast was confused with a confirmation check. The student‘s error is in italics and the teacher‘s recast is in bold.
Teacher: There‘s a…did you know there‘s a Halloween party tonight?
Student 1: I know.
Student 1: They even try to invite me.
Teacher: They tried to invite you.
Allan: They try to invite me every year. (Everyone laughs.) Teacher: Have you ever gone to the Halloween party?
Student 1: No. Never…because they have to make up. (laughs) I don‘t know how to make up.
Teacher: Oh…you don‘t know how to dress up…
Student 1: Oh, dress up…O.K…dress up…
(Example taken from a transcript of a conversation between the author and a student)
Because the tendency for students to not notice recasts results from the similarity between recasts and clarification checks (see the underlined sentence in the example above), teachers might be able to increase student noticing of recasts by changing the way that recasts are presented to students. One possible alternative to standard recasts
would be to provide students with a paraphrased recast. A paraphrased recast would reformulate the student‘s sentence and provide positive evidence much like a standard recast, but instead of simply repeating the student‘s sentence minus the error, it would
use a slightly different sentence structure. The idea is that the change in sentence structure would be different enough to make the student stop and consider it more carefully. According to the following example, the teacher uses a paraphrased recast to decrease the chance that the student might misconstrue it as a confirmation check (see the underlined statement below). A paraphrased recast is a reformulation of the student‘s utterance, plus the correct form of the student‘s incorrect feature(s). It changes the sentence structure slightly but maintains the student‘s original meaning.
Teacher: There‘s a…did you know there‘s a Halloween party tonight?
Student 1: I know.
Student 1: They even try to invite me.
Teacher: They tried to get you to go to the party.*
Allan: They try to invite me every year. (Everyone laughs.) Teacher: Have you ever gone to the Halloween party?
Student 1: No. Never…because they have to make up. (laughs) I don‘t know how to make up.
Teacher: Oh…you don‘t know how to dress up…
Student 1: Oh, dress up…O.K…dress up…
A second alternative might be an elaborated recast. An elaborated recast consists of two utterances. The first utterance provides additional relevant, contextual information that aims to clarify the semantic content of the sentence in which the learners‘ to-be-learned structure or vocabulary is embedded. (Meaning is negotiated and elaborated before form is presented). The additional information would serve the dual purpose of (1) getting the student to notice the recast as a result of having to mentally reorient himself to process the novel input, and (2) providing additional context for the forthcoming recast. The second utterance would be a standard recast as defined above. The underlined statement in the example below demonstrates how the teacher could implement an elaborated recast.
Teacher: There‘s a…did you know there‘s a Halloween party tonight?
Student 1: I know.
Student 2: Yes.
Student 1: They even try to invite me.
Teacher: They tried to get all the teachers and students to go. They even tried to invite you.*
Allan: They try to invite me every year. (Everyone laughs.) Teacher: Have you ever gone to the Halloween party?
Student 1: No. Never…because they have to make up. (laughs) I don‘t know how to make up.
Teacher: Oh…you don‘t know how to dress up…
Student 1: Oh, dress up…O.K…dress up…
In summary, recasts can be operationalized in at least three manners: a standard recast, a paraphrased recast, and an elaborated recast. Paraphrased and
elaborated recasts are premised on two possible scenarios for how second language learners process form and meaning in communicative input. Specifically, as mentioned above, there is some debate as to whether language learners are capable of simultaneously processing input for form and meaning (Hulstijn, 1989; Han &
Peverly, 2007), or whether language learners process form and meaning sequentially (VanPatten, 1990; Wong, 2001). If the latter sequential view is correct, learners will not be able to perceive the recast as corrective feedback before meaning has been clarified. In this case, to ensure and maximize that the recast can effectively serve as a pedagogical device, teachers need to first engage the learners in ―negotiation of meaning‖ before the recast is provided. This is what reduces the possibility that the recast might be merely perceived as a confirmation check and is what also allows the learner to understand the content of the recast, thereby freeing up attentional resources for him to understand the form correction. An elaborated recast first presents the language learner with a sentence to assist in comprehending the meaning, and secondly with a sentence containing the form correction. Elaborated recasts therefore allow language learners to sequentially process meaning and form.
On the other hand, if the simultaneous view is correct, it means that learners are capable of processing grammatical form in the input (or form-focused corrective feedback like recasts) while simultaneously making sense of the meaning of the input.
The simultaneous view suggests that no prior negotiation or clarification of meaning is required in order for learners to perceive recasts as corrective feedback.
Paraphrased recasts are designed in consideration of the simultaneous processing view, and present the language learner with slightly different forms that still encode the same semantic content in the same sentence. This requires the language learner to simultaneously process both form and meaning in the input. Depending on whether language learners process meaning and form simultaneously or sequentially, one type of the proposed recasts should result in more learning than the other. In order to better understand which of the proposed recasts will prove more effective, the next section will review the empirical research on simultaneous and sequential processing of input.
2.4 Sequential Processing of Meaning and Form: Insights from Studies of
Intermediate L2 Learners
Although empirical evidence that language learners may simultaneously process form and meaning in written input does exist (Han & Peverly, 2007; Hulstijn, 1989),
there seems to be sufficient reason to argue that aural input is processed sequentially by intermediate language learners (Li & Pu, 2010; VanPatten, 1990; Wong, 2001).
For example, VanPatten (1990) conducted an experiment to investigate the ability of Spanish learners to process aural input in the L2 for both form and meaning at the same time. Research participants consisted of two hundred university Spanish language learners. Participants consisted of three levels of L2 proficiency: Level-one were first semester Spanish students; level-two were fourth semester Spanish students;
and level-three were third year Spanish conversation students. The research participants were divided into four groups with each group containing an equal number of students from all three proficiency levels. All research participants were instructed to listen to a tape-recorded passage about inflation in Latin America.
VanPatten‘s experiment then subjected each of the four groups of participants to one of four different conditions: 1) one group of participants were instructed to listen only
for content, 2) one group were instructed to listen for content while listening for a key content word ‗inflacion‘, 3) one group were instructed to listen to content while
attending to a past-tense verb morpheme ‗n‘, and 4) one group were instructed to listen to content while listening for the Spanish article ‗la‘. According to the results,
language learners under conditions (1) and (2) significantly outperformed those under conditions (3) and (4) on listening comprehension. Attending to key vocabulary did
not have a negative impact on the comprehension of aural input, whereas attending to
form features did have a negative impact on the comprehension of aural input. This confirmed VanPatten‘s hypothesis that attending to form and meaning at the same
time poses difficulty for language learners, especially language learners of lower proficiency.
Replication studies by Li and Pu (2010) and Wong (2001) arrived at similar results to those of VanPatten (1990), supporting the idea that language learners sequentially process meaning and form in communicative input. However, Wong‘s (2001) study, conducted in both the written and aural mode, only confirmed the results of VanPatten‘s (1990) study in the aural mode. In the written mode, Wong (2001) found there to be no significant differences between processing input for meaning only and processing input for form and meaning simultaneously. This led Wong (2001) to conclude that language learners may process written input differently than aural input—a factor that is relevant when reviewing Hulstijn‘s (1989) research.
In Hulstijn‘s (1989) first experiment, one hundred and forty-five adult intermediate level students of Dutch were asked to read a series of nine opinion questions presented in Dutch. The participants were divided into four groups. The first group was given a task to keep them focused on form in the input. The second and third groups were given a task to keep them focused on the meaning in the input. The
fourth group was instructed to pay attention to both the form and the meaning of the sentences, and was not given any task to direct their attention. According to the results, the meaning-focused groups outperformed the form-focused group in retrieving the content of the opinion questions, and the form-focused group outperformed the meaning-focused groups in reconstructing the syntactic forms of the questions.
Interestingly, however, the results also revealed that the form and meaning-focused group (group four), which had been given no specific task to focus their attention, outperformed both the form-focused group on measures of form retrieval and the meaning-focused group on measures of content retrieval. In the sentence copying tests, however, there were no significant differences between groups. The results of Hulstijn‘s (1989) second experiment in which the researcher used an artificial language to control for participants‘ previous knowledge, were similar to the first, except that the form and meaning-focused group only outperformed the form-focused group on reconstructing the forms in the input. In the sentence copying tests, all groups significantly outperformed the control group; the form group significantly outperformed the meaning group; but the form-meaning group showed no significant differences between both the form group and the meaning group.
The results of Hulstijn‘s (1989) experiments seemed to both confirm and contradict those of VanPatten (1990). On the one hand, Hulstijn‘s (1989) experiments
suggested that language learners who attend to form are more adept at retrieving form, and language learners who attend to meaning are more adept at retrieving content.
However, Hulstijn‘s (1989) experiments also showed that on measures of form and content retrieval, the form-and-meaning-focused group performed as well as or better than both the meaning-only-focused group and the form-only-focused group. This suggests that when language learners‘ attention is simultaneously directed toward form and meaning, language learners are capable of simultaneously processing form and meaning (content) in a relatively efficient manner. In this sense, Hulstijn‘s (1989) study was in direct contrast to VanPatten‘s (1990) study, which claimed that language learners process form and meaning sequentially. However, VanPatten‘s (1990) was conducted in the aural mode, whereas Hulstijn‘s experiment was conducted in the written mode. As Wong (2001) pointed out, it is entirely possible that modality may be a factor in language learners‘ ability to simultaneously process form and meaning.
Simultaneous processing of form and meaning in aural input may be more challenging than simultaneous processing of form and meaning in written input. One reason for this may be that written input is constantly available for reprocessing, whereas aural input requires listeners to simultaneously process form and meaning within a limited time frame (Danks, 1980). The assertion that aural input is more difficult to process than written input has been supported by some empirical studies into the processing
ability of second language learners (Lund, 1991; Murphy, 1997 as cited in Wong, 2001). Considering that VanPatten (1990), Li and Pu (2010), and Wong (2001) all provided evidence that meaning and form are sequentially processed in the aural mode, Hulstijn‘s study, conducted in the written mode, does not necessarily conflict with their results.
In addition to modality being a factor in input processing, the level of the language learner may also impact how input is processed. Han and Peverly (2007) investigated input processing strategies in language learners, who were absolute beginners and had no previous knowledge of the target language. The participants were divided into two groups: a group placed under a simultaneous processing of form and meaning condition, and a group placed under a sequential processing of form and meaning condition. Each group was tested on two measures: A written free recall task, aimed at assessing content comprehension, and a fill-in-the-blank test, aimed at assessing the ability to recall forms in the input. Although results showed there to be no statistically significant differences between groups, there did appear to be a pattern. The sequential group consistently scored better on the form-focused tasks; whereas the simultaneous group consistently scored better on the meaning-focused tasks. Han and Peverly (2007) concluded that absolute beginners have no choice but to process the input for form, and supported this conclusion by mentioning
that the participants‘ retrospective protocols referred almost exclusively to
form-focused strategies in attempting to break the code. The authors argued that VanPatten‘s (1990, 2004) principles of processing input sequentially only apply to
language learners with some knowledge of the L2, and although sequential processing of meaning and form may be the optimal processing strategy for intermediate learners,
absolute beginners process meaning and form simultaneously. In any case, Han and Peverly (2007) only demonstrated that VanPatten‘s (1990, 2004) input processing
principles may not entirely apply to absolute beginners. According to evidence from VanPatten (1990), Li and Pu (2010), and Wong (2001), intermediate-level language learners seem to process meaning and form sequentially when presented with input in the aural mode.
Taken together, the findings of VanPatten (1990) and other researchers (Li & Pu, 2010; Wong, 2001) provide strong evidence to support the hypothesis that elaborated recasts, which sequentially present meaning first and form second in aural input, may be more easily processed by intermediate-level, second language learners and thus more effective for intermediate-level, second language learning.