Languagehttp://ltr.sagepub.

com/ Teaching Research

Long-term effects of noticing on oral output

Paul Mennim Language Teaching Research 2007 11: 265 DOI: 10.1177/1362168807077551 The online version of this article can be found at: http://ltr.sagepub.com/content/11/3/265

Published by:
http://www.sagepublications.com

Additional services and information for Language Teaching Research can be found at: Email Alerts: http://ltr.sagepub.com/cgi/alerts Subscriptions: http://ltr.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav Permissions: http://www.sagepub.com/journalsPermissions.nav Citations: http://ltr.sagepub.com/content/11/3/265.refs.html

>> Version of Record - Jul 19, 2007 What is This?

Downloaded from ltr.sagepub.com at Sookmyung Women's University on October 22, 2013

Language Teaching Research 11,3 (2007); pp. 265280

Long-term effects of noticing on oral output

Paul Mennim

Aoyama Gakuin University, Japan

This paper reports on the effects of classroom exercises that encourage noticing and conscious attention to form, which were part of a university EFL oral presentation course in Japan. The students on the course were given a set of exercises that encouraged them to notice and to reect on L2 forms of their own choosing throughout one academic year. Records of their noticing were tracked throughout the year and recordings of their oral output made over the same period were analysed to determine whether there was any development in the use of the forms that the students had noticed. The paper describes an initial analysis of the tracking of two students noticing and subsequent use of a non-count noun, which presented them with difculties at the start of the year. Nine months later their accuracy in the use of this word was much improved. The paper considers how the students noticing of the word might have related to this improvement.

I Introduction
This study considers the outcome of a series of noticing exercises in a university presentation course in Japan. In this course, students, working in groups, selected one topic to research for a whole academic year and gave three oral presentations of their ndings over that period. As they conducted their research, they came into contact with English in published material or through accessing the Internet, and in this way they were responsible for their own selection of second language (L2) input. Although this was not a language course per se (as I will describe below), I anticipated that the process of both researching and presenting on a research topic in English would help develop their L2 prociency. I wanted to try to focus the students attention on language forms as I had a good idea that they would want to know that their English was as accurate as it could be before they made their public performance. I therefore set noticing exercises to increase the likelihood that they would attend to linguistic form in both the input and in their own L2 output. I hoped that this attention might then feed into the language learning process. In order to understand how this might come about, it will be helpful, before I describe the course in more detail, to consider the ideas of noticing and intake in second language acquisition theory and in classroom studies.
Address for correspondence: Paul Mennim, Aoyama Gakuin University, 4-4-25 Shibuya, Shibuyaku, Tokyo 150-8366, Japan; email: mennim@hotmail.com 2007 SAGE Publications 10.1177/1362168807077551

266

Long-term effects of noticing on oral output

1 Noticing and intake

Noticing and intake describe processes that mediate between L2 input and a developing interlanguage system. Schmidt and Frota describe how learners might notice input in the normal sense of the word, that is, consciously (1986: 311). Schmidt has rened his concept of noticing to account for interlanguage development that is not necessarily conscious but does require attention. He describes (1990), for example, how most learners of French are unaware that there are phonological clues as to the gender of nouns, yet they seem to make use of these clues when guessing noun gender. This ability may be gained through all of their exposure to French nouns. The general point here is that focal attention is necessary for learning but conscious attention is not. For Schmidt, intake is simply that part of the input that the learner notices (1990: 139). Much of his early data came from his own efforts at learning Portuguese (Schmidt and Frota, 1986). He spent ve months in Brazil following a language course and at the same time he kept a journal describing what progress he felt he was making in the language. This included new target structures that he heard or tried to use. In addition he was interviewed once a month on tape and a Portuguese speaker examined the recordings to compare what he thought he had learnt with what he had actually learnt. He found a strong similarity between what he noticed in the input to what he was actually using himself. The structures he noted in his diary as noticed structures began to appear in his speech. The input included his lessons and what he heard in daily life. The study therefore provides some evidence that forms have to be noticed before they are used actively. Nevertheless, noticing was not sufcient as some forms appeared in his journal but did not emerge on the tapes as part of his output. Even so, Schmidt makes strong claims for noticing (1990: 144): Those who notice most learn most, and it may be that those who notice most are those who pay attention most, as a general disposition or on particular occasions. Noticing and intake are dened somewhat differently by Gass (1997), who presents them as separate processes in a ve-stage model of second language acquisition, beginning with the apperception of input. Apperception is similar to Schmidts idea of noticing. It represents some part of the input getting through to the learner, perhaps due to recognition that the input represents a gap in his or her interlanguage. Prior knowledge is one factor that might lead to apperception of a language form; others include frequency or saliency in the input. The stage between apperception and intake is comprehended input, which accounts for instances when a learner can comprehend the general meaning of the input for the immediate purpose of communication but cannot process it any further, perhaps due to lack of time. In Gasss model, intake, the third stage, represents a deeper analysis of L2 than Schmidts intake; it is the stage at which learners analyse input, perhaps gaining an understanding of its syntactic structure. It is also here that generalizations and hypotheses

Paul Mennim

267

about the L2 are formed and tested. If further input conrms a hypothesis, which, as a consequence, becomes more securely integrated into the learners interlanguage, then the fourth stage, integration, has been reached. The nal stage is output, the point at which newly acquired language is produced. Although, ostensibly, this stage concludes the model, output, too, can feed into the learning process, as we shall consider below.

II Noticing in the language classroom

This section looks at ways in which ideas about noticing and intake have been investigated or utilized in the classroom.

1 Noticing oral output

Swain (2000) has argued that limited attention can prevent students from noticing structures when they are speaking in the L2. Her students, quite reasonably, tended to focus on meaning during conversation and did not always succeed in developing accuracy in the L2, in this particular case French. Returning to the model outlined above, we might say that although Swains students were getting comprehended input, they were not always continuing to the next stage of intake, the crucial stage for acquisition. Swains output hypothesis (1995) proposes that, for the purposes of the development of syntax and morphology, it is more benecial for learners to use the L2 actively rather than to listen to it. She argues for three functions of output. First is the question of noticing. When actively constructing L2 utterances, learners may be more likely to notice gaps in their interlanguage, as, when producing output, they are pushed into syntactic processing to a greater extent than is the case when they are attending to input. Noticing the gap was alluded to above as a possible trigger of apperception. It is a term used to describe learners reection on the difference between what they themselves can or have said (or even what they know they cannot say) and what it is more competent speakers of the target language say instead to convey the same intention under the same social conditions (Doughty, 1999: 21). The second function Swain attributes to output, analogous to Gasss idea that output feeds back into the acquisition process at the level of intake, is the chance it gives learners to test their hypotheses about the L2. Finally, output provides opportunities for conscious reection about the L2, characterized by metatalk: discussion about language form that may or may not include explicit metalinguistic terminology. This third function is somewhat different from the rst two, which could arise in natural speech. Metalinguistic reference to oral output is more likely to arise in classroom situations where it is required by a task. Swain (1998) encouraged noticing in the classroom through the use of Wajnrybs dictogloss (1991), which provides learners with the opportunity to discover gaps in their L2 knowledge and to address them verbally. In the

268

Long-term effects of noticing on oral output

dictogloss, the teacher reads out a short passage once to the class who must then, working in groups, reconstruct it. In the process, the students are meant to identify and discuss language problems and pool their linguistic resources to solve them. Swain calls this discussion negotiating about form, which is manifested by language related episodes (LREs), dened as any part of a dialogue in which students talk about the language they are producing, question their language use, or other- or self-correct (1998: 70). A crucial aspect of LREs is that they are a product of the students attention directed towards their own L2 output. This line of research has produced some evidence that the noticing of L2 forms can result in observable gains in the L2. Swain (1998) used the dictogloss task to encourage the production of LREs. In a French immersion programme, 13-year-old students were given a series of three dictogloss tasks to complete and for the third one, their negotiation about form was recorded and analysed. In order to evaluate the effect of this task on the students learning, a post-test was produced for each group based on the language forms discussed in the LREs. For example, two students were unsure about the gender of the word rve in French and discussed this problem. They were therefore asked the gender of this noun a week later in the post-test. Swain described a strong tendency (1998: 78) for students to respond correctly in the posttests to questions about problem forms that they had correctly solved during the dictogloss task. A related effect was observed in that students who had solved language problems in a non-target-like direction were likely to give the same non-target-like answer in the post-test. This suggested that conscious reection about language form might be a source of language learning.

2 Noticing and learner autonomy

The effect of classroom discourse initiation on students noticing of language forms was considered by Slimani (1989). She investigated uptake from what learners claimed to have learnt from a series of grammar lessons. Uptake in this study was what students noticed during a lesson in terms of grammar, vocabulary and pronunciation, recalled and reported in written questionnaires three hours after the lesson. Thus, it is similar to Gasss idea of apperception: part of the input that is noticed but may or may not be further analysed as intake. Slimani found that a great deal of the topicality of the lesson (the grammar structures, items of vocabulary or any other language forms covered in the class) was rarely recalled as uptake from the evidence of forms and questionnaires that were completed by the students after the class. Of the total number of topics focused on during the lessons as much as 36% seemed to go unnoticed. There were, however, more positive ndings when she analysed the uptake of the topics initiated by the students themselves. Such occasions were much less common than that teacher-initiated content, yet 74% of language forms initiated by students during the lesson was conrmed

Paul Mennim

269

as uptake in the questionnaires compared to just 49% of the forms introduced by the teacher.
Though the discourse initiation appears to be predominantly in the hands of the teacher, it looks as if, given the chance, the informants benet more from topics initiated by the learners. (Slimani, 1989: 227)

As to why student-initiated topicalization had a better chance of being noticed, Slimani suggests that the student contributions may have been pitched at a more appropriate level to the learners interlanguage compared to those of the teacher and may have been more relevant to their learning needs. Insofar as this study reveals a decidedly limited role for the teacher in drawing attention to language and a strong role for student autonomy in the potential uptake of L2 forms, albeit in one particular setting, it suggests that allowing students more freedom in attending to forms of their own choosing might better result in apperception or intake. This idea of student autonomy is relevant to natural order hypotheses in SLA (Pienemann, 1989), and a learners noticing of a form can be interpreted as a readiness to acquire it (e.g. Williams and Evans, 1998; Williams, 1999). Despite this potential, teachers might be interested to know, before they begin to relinquish control of input in the classroom, to what extent learners by themselves are able to notice and analyse the L2. While Schmidt was adept at noticing language forms as he learned Portuguese, he is aware that, as he is a trained linguist, it might be objected that his metalinguistic awareness would be far superior to that of non-specialist learners, who may have little aptitude for noticing (Schmidt and Frota, 1986). A study relevant to these concerns is that of Williams (1999), which investigated learner-generated attention to form, in this case LREs produced spontaneously (which means without the direction of the teacher) by eight students aged between 18 and 25 in an intensive English programme, who wore clip-on microphones as they completed a selection of classroom tasks and exercises such as role plays, correcting homework in pairs, grammar activities and free conversation. The resulting 65 hours of taped dialogue yielded 268 LREs but Williams expresses disappointment with the amount and range of students noticing, and questions whether the granting of autonomy to students to initiate their own attention to form is worthwhile. Williams judged that the students in her 1999 study did not spontaneously or consistently attend to form (they averaged between three and seven LREs in a 45minute session). In addition, the forms they did attend to were overwhelmingly lexical (around 80% of LREs were lexically oriented). Williams concludes that teachers cannot expect learners to consistently ferret out and notice morphosyntactic features (1999, p. 620). Despite the low number of LREs reported by Williams, it may be possible to develop classroom tasks and exercises that result in a greater amount of

270

Long-term effects of noticing on oral output

noticing and of a wider range of forms. Thornbury (1997), acknowledging that noticing is not guaranteed to take place in the language classroom, suggests pedagogical interventions that might encourage it. His examples include scrutinizing the language of recorded conversations, having students proof read texts or list notice the gap incidents in a journal. My study was motivated by the research cited above in the following ways. First, it seemed that, taken together, the studies by Slimani and Williams presented language teachers with a possible dilemma that although learnerinitiated noticing may be benecial to learning, it may not often occur in the classroom. Because the course I was teaching gave students freedom to work with their own choice of L2 input, I was keen to consider the efcacy of such an approach. Second, I felt that noticing exercises that produced LREs might be an effective way to give students feedback on their own output, as they can result in a deep level of L2 analysis. Instead of input from a dictogloss, the learners in this study scrutinized transcripts of their own presentations, while I helped to correct any errors that appeared in them, as I describe below.

III The study

The study focuses on an academic presentation course for rst-year students at a private Japanese university. The students were aged between 18 and 20, and, as this course was designated upper level, they had TOEFL scores of 500 or above, determined by the in house Institutional Testing Program version of that examination. The students were all enrolled in the Faculty of Policy Studies and were not majoring in English; they followed courses in subjects such as law, economics, Asian studies and comparative culture. The academic presentation course was one of four compulsory English classes per week. Students on this course researched a topic they were interested in, read about this topic in English, discussed it with their peers and gave oral presentations about their research. The duration of the course was nine months (approximately 25 classes of 90 minutes each), during which time students gave three presentations in groups of two or three. The class size was relatively small with 17 students. It is relevant to this study that the course objectives were not of a linguistic nature; rather, they were drawn up in terms of appropriate preparation of content, quality of research and organization. Class work included the consideration of information sources and presentation structure. Because the ostensible focus of this approach was successful research, it was not always obvious to the students to what extent and in what ways their English improved during this nine-month process. Even classes concentrating specifically on oral skills do not always give students the chance to scrutinize their output as would be the case with the written drafts of a composition class. I therefore incorporated various noticing exercises, which aimed to focus the students attention on L2 form.

Paul Mennim

271

The noticing exercises encouraged students to notice both L2 input from information sources and their own L2 output (recordings of their presentations). The four core components of the course were as follows.

1 The presentation course

The students prepared three different presentations: one in Month 3 (10 minutes), one in Month 4 (20 minutes) and a nal presentation in Month 9 (30 minutes). This last presentation was preceded by a rehearsal in Month 8. The Month 3 presentation required the students to explain to the rest of the class why they had chosen their topic, what they already knew and what they wanted to nd out about it. The subsequent presentations would allow them to report on their actual research. These presentations were made without the use of scripts, though the students were allowed to use small cue cards as an aide-mmoire. The students completed several different exercises throughout the year that encouraged noticing.

a Language development awareness (LDA) sheets: The students lled these out every week. They were asked to write down any new language that they had noticed over the previous week. Categories helped to focus their attention on different types of language form, such as grammar, vocabulary and pronunciation. b Post-presentation questionnaires: These helped focus the students
attention on their own L2 output immediately after their presentations and after the Month 8 rehearsal. The students were asked to report on the language they used in the presentation, including whether they realized that they had made an error, whether they had managed to correct an error as they spoke, or whether they used any new or recently learned language. Again, the questionnaire was intended to focus their attention on different types of language form.

c Transcription exercises: This was completed three times during the

year. The presentations and the rehearsal were recorded and the students were asked to take their recording and transcribe approximately ve minutes of their own speech. For most students, this resulted in a double-spaced script covering one side of A4 paper. They were encouraged to transcribe their speech warts and all, even if they happened to notice errors while they made the transcription. This would provide me with a clearer presentation of the errors that they would spot later on, as the corrections were to be made in red pen over the printed transcript. For the Month 3 and Month 7 transcription exercises they went on to correct these transcripts together in their groups. The ndings of a pilot study reported elsewhere (Mennim, 2003), supplied encouraging evidence that a similar set of students were able to successfully

272

Long-term effects of noticing on oral output

correct errors working on their own. The students interactions as they discussed language problems were recorded. The last stage was to pass their completed work on to me. I added any corrections that they had missed and returned the scripts to them. Attention to form was therefore largely unprompted by the instructor and allowed students to focus on whatever forms they wanted to.

2 Selection and tracking of forms

Because the course was process oriented and did not present any set list of forms for the students to learn, it was necessary to cast a wide net in the search for L2 gains. Progress could only be identied for those forms that came up in the course of the students presentations and also with sufcient regularity, so the study relied on a large quantity of recorded speech (approximately 100 minutes for each group of students). Although this methodology required extensive analysis of recorded student output, I considered it worthwhile insofar as it might be more likely to uncover the re-emergence and development of language forms over the nine-month period. However, my starting point in tracking was not to trawl the presentation transcripts to nd L2 development, as I would be unable to speculate about the source of every improvement found there. Rather, because my interest lay in what students had noticed, I began my search in the noticing data. Only if an L2 form was noted or discussed there did I go on to trawl the output data to nd subsequent re-emergences of that form. Of course, one of the inevitable restrictions associated with this procedure was that there was no guarantee that forms noticed would ever re-emerge as output during the presentations. I identied only seven examples where a noticed form appeared over 10 times in the presentation recordings (Mennim, 2005). I considered this number of re-emergences high enough to make some determination as to whether or not a students use of the form became more target-like. This paper focuses on one of these seven examples: the noticing and output of two students appearing here under the pseudonyms Toru and Katsu, who were researching the effects of garbage disposal on the environment. Although their noticing of language forms was considerable and diverse (they produced 36 LREs during the Month 7 transcription exercise alone), for the purposes of this paper, I trace their noticing throughout the year of one language form: the non-count noun garbage. The data shows that Toru noticed the form at least three times and Katsu noticed it at least ve times over the year, as I will go on to illustrate. Because the idea of garbage disposal remained at the centre of this pairs research throughout the year, the word reemerged fairly often in the recorded data. Toru used it 13 times in the years three presentations and one rehearsal, while Katsu used it 25 times over the same period. Although there are many aspects involved in vocabulary acquisition (semantic properties, pronunciation, spelling and so forth), because my

Paul Mennim

273

Table 1 Target-like and non-target-like occurrences of garbage in Month 3 and Month 4 presentations Student /occurrence of form Garbage Garbages Toru 3 3 Katsu 5 7

interest lies in the effect of the students own noticing, in what follows I track only that aspect of the word that they noticed and discussed themselves, namely, whether or not it was grammatically countable. The word is of interest as both of the students had a problem with it at the start of the academic year. In the Month 3 and Month 4 presentations, three out of Torus six uses of garbage were non-standard as they related to countability, while seven out of Katsus 12 uses of garbage over the same two presentations were nonstandard in this way. (See Table 1.) In what follows, I track Toru and Katsus noticing throughout the year as represented by the data from the classroom exercises, looking for evidence of conscious reection about this word. I then track re-emergences in their output to see whether any improvement can be observed.

IV The tracking process 1 Initial noticing

It seems that the word was unknown to Toru until the start of the academic year. In the rst post-presentation questionnaire in Month 3, he cited garbage in the section: Did you use any new (recently learned) vocabulary? However, because he did not include the word in any of his weekly language development awareness forms, it was uncertain how recently he had learnt it. As for Katsu, in a LDA entry in Month 2, he noted:
I noticed garbage rubbish US UK

It seemed that Katsu may have been aware of one or both of these words before, but, unlike Toru, it is not possible to tell from his remarks whether garbage in particular had been recently learnt. What is more, although Katsu has noticed something about the geographical spread of the words usage (a further aspect of vocabulary acquisition), these rst comments said nothing about the grammatical behaviour of garbage, and, as mentioned above, their use of the word during the subsequent presentation in Month 3 provided no evidence that they were aware that it is non-count, although the above entry may suggest such awareness. In terms of Gasss model of second language acquisition, we might say that these rst notes are evidence of either apperception or of comprehended input.

274

Long-term effects of noticing on oral output

2 Grammatical reection
As they made corrections to Katsus rst presentation transcript in Month 3, the two students reected for the rst time on the grammatical behaviour of garbage. The text in block capitals indicates where the students are reading verbatim from their presentation transcript. The question mark in turn 5 indicates one short inaudible section. The underlined text in turn 8 denotes stress.
Transcript 1: ONE DAY, HE VISITED TO JAPAN AND HE SAID, OH THERE ARE MANY GARBAGES IN JAPAN 1 Katsu: ONE DAY, HE VISITED TO JAPAN AND HE SAID, OH THERE ARE MANY GARBAGES garbages? 2 Toru: IN JAPAN you dont need to say that. Japan because he is 3 Katsu: Ah okay he is in Japan. garbages or garbage? 4 Toru: garbage? 5 Katsu: I looked into the dictionary [?] garbages 6 Toru: you cant say that 7 Katsu: yeah you cant say that [garbage 8 Toru: [garbage like informations its like information not informations 9 Katsu: yeah yeah 10 Toru: garbage.

Katsu rst questions his use of garbages in turn 1. Torus attention is taken up with another point in turn 2, but Katsu repeats his query in turn 3. Torus answer seems unsure as it is given with a questioning, high-rising terminal intonation. In turn 5 Katsu mentions looking up this word in the dictionary. This particular recording is of poor quality and his whole statement is inaudible. It is likely that the lost comment is to do with the countability of garbage as he immediately agrees with his partner when he judges garbages unacceptable in turn 6. It is interesting that Toru seems to lack the grammatical terminology to describe this point, and instead uses the analogy with information in order to make the point clear. This is nevertheless a clear example of metatalk. It seems from the above negotiation that the students had some awareness that garbage is not a countable noun. It is not possible from my data to determine where this insight came from. As mentioned above, I did not offer any instruction or advice about this word. It is conceivable that, during their reading about the topic in the L2 (as the course required them to do), they noticed the form garbage when they would have expected to read garbages, and this might have provided the negative evidence behind the doubts voiced in this LRE. In the way that Katsu has used his dictionary to try to conrm or reject his hunch about the way garbage is used, this episode suggests that he is rening his hypotheses about the way the word behaves and that the LRE is helping him advance from the intake stage to the integration stage, according to Gasss model.

Paul Mennim 275

3 Subsequent noticing
The next reference the students made to this word was in Month 4. In the post-presentation questionnaire after his second presentation, Katsu made the following short note in the section: Did you use any new (recently learned) grammar?
garbages garbage.

In fact, all three of Katsus uses of garbage in the presentation were standard. A possible interpretation of this note is that he realized that he had managed to overcome this error in the Month 4 presentation. This is interesting as it suggests that he had been conscious even of his target-like use of garbage. Whereas noticing is often illustrated in terms of negative evidence and noticing the gap, Katsu, consciously reecting on his target-like performance, seemed to demonstrate a more positive form of noticing; that is, that he had gained new knowledge, lling a previous gap in his L2. Katsus note therefore gives a further indication that the countability of the form has been further integrated into his interlanguage system. The students next noticed the word during the second transcription exercise in Month 7 when they were correcting a transcript based on the second presentation in Month 4. In the following segment, they are discussing Torus script.
Transcript 2: MANY GARBAGES THAT REPLACE OUR LIVING SPACE AND ENDANGER OUR LIFE 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Katsu: garbage. Not s Toru: ah no. garbage Katsu: yeah Toru: that replace our living space Katsu: many garbage? Toru: amount of amount of amounts of Katsu: a lot of Toru: hmm Katsu: hmm because Toru: its not uh Katsu: We cannot we cannot say many? Many is not good in this sentence Toru: amounts of maybe Katsu: mm hmm Toru: amounts of garbage that replace

This LRE is markedly different from the corresponding one in Month 3. It seems clear that both students are now aware of the countability of garbage. No discussion or justication of Katsus correction seems to be necessary; he simply states that no s is needed and Toru immediately agrees. It is conrmed that they realize they are dealing with the same grammatical form from turn 4, when they go on to change their choice of determiner to one that co-occurs with non-count nouns.

276

Long-term effects of noticing on oral output

The word attracted attention for the last time in the data in Month 8 after the rehearsal of their nal presentation, which was to take place in Month 9. The rehearsals were recorded so that the students could perform a nal transcription exercise around their oral output. Due to a lack of time, the students did not discuss their rehearsal output in groups; they simply made their transcript and corrected it alone before handing it on to the teacher, who made a further search for errors. Katsu made one alteration in his script from garbages to garbage: the only occurrence of the error during the Month 8 rehearsal. Meanwhile, there was no evidence that Toru noticed or attended to this form during the same transcription exercise. His use of the word during the rehearsal was always target-like. Neither student made reference to garbage in the post-presentation questionnaire after the nal presentation in Month 9.

V Noticing and language development

We now consider the students developing use of this form over the academic year from Month 3 to Month 9. Table 2 charts the students performance in terms of the number of times garbage is used in a standard or non-standard way. For Toru, the gures do not suggest any progress at all over Month 3 and Month 4. In both months, he is as likely to use garbage incorrectly as correctly. However, his transcription exercise negotiation showed there was progress in terms of his declarative knowledge of the grammatical behaviour of the word. In Month 3, we saw from his use of an analogy with information that he was aware that garbage was a non-count noun. In Month 4, he conrmed Katsus correction of one of his own non-target-like utterances of garbage. Looking at the gures for the end of the course, with his attention having been drawn to non-target-like occurrences of garbage twice in the two tape transcription exercises, his subsequent use of this word is always target-like.

Table 2 Toru

Occurrences of garbage from Month 3 to Month 9 Month 3 1 1 Month 4 2 2 Month 8 2 0 Month 9 5 0

Target-like Non-target-like

Katsu Target-like Non-target-like

Month 3 2 7

Month 4 3 0

Month 8 4 1

Month 9 6 2

Paul Mennim

277

For Katsu, progress seems to have been less consistent. Although he was aware that he produced only target-like occurrences of the word during the Month 4 presentation, the next two recordings showed three examples of the non-target-like garbages. This is interesting as Katsu was more inclined to notice this word during the transcription exercises than Toru; Katsu initiated both of the LREs about garbage in Month 3 and Month 4. He also positively reported his own progress with this word in the post-presentation questionnaire in Month 4. It might therefore have been thought more likely that Katsu would have eradicated the error. However, a non-linear pattern in grammatical acquisition has long been observed in second language acquisition studies, so it does not follow that increased declarative knowledge of an L2 form will guarantee target-like use during real time communication. But in any case, Katsu clearly made progress in his target-like use of garbage over the year. Table 3 shows more clearly how the mostly non-target-like occurrences at the beginning of the course became mostly target-like towards the end. This applies to both of the students.

VI Conclusions
There was evidence that although the students had some knowledge of the meaning of the word garbage at the start of their presentation course, they were unable to use it in a grammatically target-like way as a non-count noun. However, by the end of the year, their increased accuracy in its use in this way suggests a long-term gain in language learning. The long-term nature of this study is important to the evaluation of the noticing exercises employed in the course. This 9-month time period is in marked contrast to the typically short-term cross-sectional nature of other studies in this area. For instance, in my pilot study, cited above, I observed the effects of noticing over just two weeks. Similarly, Swains 1998 study, also cited above, involved a post-test one week after the nal dictogloss. This study suggests that it has taken time for the students progress to show itself. Had the analysis of the students use of garbage ended in Month 4, much of Torus eventual improvement would have gone unobserved. Although it was already clear in Month 3 that Toru possessed the declarative knowledge necessary to solve the problem of the countability of garbage, consistent proceduralization of this knowledge was only observable towards the end of the year. His use of the form was target-like just 50% of the time in the Month 4 presentation, yet the Month 8 rehearsal and Month 9 presentation show a 100% accuracy rate. This was despite the fact that the only evidence of his attention returning to that form was a very brief correction during the Month 7 transcription exercise; this correction may have been a sufcient reminder of his use of garbage to get him back on the right track. On the other hand, the fact that Katsu reverted to garbages twice in the Month 9 presentation suggests that language problems can take a considerable time to be fully eradicated.

278

Long-term effects of noticing on oral output

Occurrences and accuracy of garbage throughout year Accuracy high consciousness about garbage problems people threw their garbages away if there are many garbages that replace there are too much garbages that wait for the garbage weve already made throw it into the garbage can! environmental distress caused by garbage and other Japanese consciousness about garbage is low through the garbage and economy research. the future for garbage problem. we are interested in garbage problem space for garbage disposal to throw garbage away as much as we want X X X Month 3 3 4 4 4 4 8 8 9 9 9 9 9

Table 3 Toru 1 2 3 4 5 6 7 8 9 10 11 12 13

Katsu 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 reduce the quantities of garbages we think hardly as much as garbage in Japan. There are many garbages in Japan. his country has garbage garbages but environmental problem caused by garbages. thoughts about garbages equal to our visions Why is full of garbages in Japan? Where does it come in the world forever uh garbage will be produced especially eh about garbages. And so our topic this mans vision and thoughts about garbage resemble oh, there is a lot of garbage in Japan. And then, Why is it full of garbage in Japan? Eh where does it come he was so amazed at the quantity of garbage. Then environmental problem caused by garbages. in which we throw away garbage. Uh because of er this mans vision and thoughts about garbage equal to Why is Japan full of garbage? Where does it come from? he was so amazed at the quantity of garbage. Then he environmental problems caused by garbage. declamation in which we throw away garbage. this mans vision and thoughts about garbage equal our vision damage caused by garbages and other poisonous waste the technology of garbage disposal developed and the when we throw away these garbages ah we must pay why is Japan full of garbage? Eh-where does it come from?

Accuracy X X X X X X X X X X

Month 3 3 3 3 3 3 3 3 3 4 4 4 8 8 8 8 8 9 9 9 9 9 9 9 9

Paul Mennim

279

It is worth restating that the students made their own decisions about which forms to attend to during the exercises used in this course. With regard to the transcription exercises, given the teachers partial relinquishing of classroom control, there was always the risk that the students would be unable to scrutinize their transcripts effectively, perhaps due to a lack of condence or to face-saving issues associated with criticizing other students output. But the time they spent in their groups seems to have been well spent. Toru and Katsu discussed 36 LREs over 20 minutes during the Month 7 transcription exercise, while the initial analysis of just some of those showed that, for these adult learners at least, it provided opportunities to notice the gap and to discuss and verbalize hypotheses about the L2. Of course, whether this would be observed in different student populations remains uncertain. The nature of the data collection carried certain risks, as there was no guarantee that any form would emerge with sufcient regularity to provide a clear picture of L2 gains. In fact, though, the number of occurrences of garbage (13 by Toru and 25 by Katsu) made it possible to make reasonable claims about how that form developed over the year. However, the fact that the students were required to stay with the same research topic over the year helped ensure some repetition of lexical items at least. But one further factor in Toru and Katsus presentations resulted in more recycling of language forms compared to those of the other groups; at the start of each of their talks they gave a brief review of their previous presentations. In hindsight, this element would have represented a potentially useful requirement for all of the groups. The repetition of the review may have resulted in repeated noticing of forms and perhaps in greater accuracy in the use of those forms, as was the case in studies of task repetition reported by Bygate (1996), Gass et al. (1999) and Lynch and Maclean (2001). Finally, the sequence of noticing that is followed here is just one of several appearing in my data, which record the noticing of 17 students over the course of the academic year. A more extensive study of the classroom data sampled here (Mennim, 2005) offers a deeper and more extensive analysis of the variety of forms that attracted the students attention and reports further tracking of re-emergences of those forms, showing more evidence of longterm L2 gains, attributable to noticing.

VII References
Bygate, M. 1996: Effects of task repetition: Appraising the developing language of learners. In Willis, J. and Willis, D., editors, Challenge and change in language teaching. Oxford: Heinemann, 13646. Doughty, C. 1999: Cognitive underpinnings of focus on form. University of Hawaii Working Papers in ESL 18(1): 169. Gass, S. 1997: Input, interaction, and the second language learner. Mahwah, NJ: Lawrence Erlbaum.

280

Long-term effects of noticing on oral output

Gass, S., Mackey, A., Alvarez-Torres, M. and Fernandez-Garcia, M. 1999: The effects of task repetition on linguistic output. Language Learning 49(4): 59481. Lynch, T. and Maclean, J. 2001: A case of exercising: effects of immediate task repetition on learners performance. In Bygate, M., Skehan, P. and Swain, M., editors, Researching pedagogic tasks: Second language learning, teaching and testing. Harlow, Essex: Addison Wesley Longman, 14162. Mennim, P. 2003: Rehearsed oral L2 output and reactive focus on form. ELT Journal 57(2): 13038. 2005: Noticing tasks in a university EFL presentation course in Japan: Their effect on oral output. Unpublished PhD thesis. Edinburgh University, UK. Pienemann, M. 1989: Is language teachable? Psycholinguistic experiments and hypotheses. Applied Linguistics 10(1): 5279. Schmidt, R. 1990: The role of consciousness in second language learning. Applied Linguistics 11(2): 12958. Schmidt, R. and Frota, S. 1986: Developing basic conversational ability in a second language: A case study of an adult learner of Portuguese. In Day, R., editor, Talking to learn: conversation in second language acquisition. Rowley, MA: Newbury House, 237326. Slimani, A. 1989: The role of topicalization in classroom language learning. System 17(2): 22334. Swain, M. 1995: Three functions of output in second language learning. In Cook, G. and Seidlhofer, B., editors, Principle and practice in applied linguistics. Studies in honour of HG Widdowson. Oxford: Oxford University Press, 12544. 1998: Focus on form through conscious reection. In Doughty, C. and Williams, J., editors, Focus on form in classroom second language acquisition. Cambridge: Cambridge University Press, 6481. 2000: French immersion research in Canada: Recent contributions to SLA and applied linguistics. Annual Review of Applied Linguistics 20: 199212. Thornbury, S. 1997: Reformulation and reconstruction: Tasks that promote noticing. ELT Journal 51(4): 32635. Wajnryb, R. 1991: Grammar dictation. Oxford: Oxford University Press. Williams, J. 1999: Learner-generated attention to form. Language Learning 49(4): 583625. Williams, J. and Evans, J. 1998: What kind of focus and on which forms? In Doughty, C. and Williams, J., editors, Focus on form in classroom second language acquisition. Cambridge: Cambridge University Press, 13955.