Vous êtes sur la page 1sur 27

Journal of Second Language Writing

10 (2001) 277 303

Exploring the role of noticing in a three-stage


second language writing task$
Donald S. Qi*, Sharon Lapkin
Modern Language Centre, Ontario Institute for Studies in Education, University of Toronto,
252 Bloor Street West, Toronto, ON, Canada M5S 1V6
Received 31 August 2001

Abstract
The importance of noticing as a cognitive process in second language (L2) acquisition
has been increasingly recognized by applied linguistics researchers. However, issues
concerning how noticing is related to composing and subsequent feedback processing, and
what impact such noticing has on L2 writing improvement, need to be addressed. We
conducted a case study to investigate these issues with two Mandarin background adult
English-as-a-second language (ESL) learners. The study documents the relationship of
noticing, both in the composing stage (Stage 1) and the reformulation stage (Stage 2,
where learners compare their own text to a reformulated version of it), to the improvement
of the written product in the posttest (Stage 3) of a three-stage writing task. The findings
suggest that while composing and reformulation promote noticing, the quality of noticing,
which relates directly to L2 writing improvement, is different for learners with different
levels of L2 proficiency. We argue that while promoting noticing is important, promoting
the quality of that noticing is a more important issue to be addressed in L2 writing
pedagogy. D 2001 Elsevier Science Inc. All rights reserved.

The research reported here was funded by Grant No. 410-99-0269 from the Social Sciences and
Humanities Research Council of Canada (SSHRC) to Merrill Swain and Sharon Lapkin, and by a
graduate assistantship from the Ontario Institute for Studies in Education, University of Toronto to the
first author. We wish to thank Lindsay Brooks, Alister Cumming, Monika Smith, Merrill Swain, and
Agustina Tocalli-Beller and the anonymous reviewers for comments on an earlier draft of this article.
* Corresponding author. Tel.: +1-416-923-6641x2645; fax: +1-416-926-4769.
E-mail address: sqi@oise.utoronto.ca (D.S. Qi).
1060-3743/01/$ see front matter D 2001 Elsevier Science Inc. All rights reserved.
PII: S 1 0 6 0 - 3 7 4 3 ( 0 1 ) 0 0 0 4 6 - 7

278

D.S. Qi, S. Lapkin / Journal of Second Language Writing 10 (2001) 277303

Introduction
Noticing may be of crucial importance to human development. As Bennett
(1976) has asserted: Unless we notice, we cannot be in a position to choose or
act for ourselves. It is a transition from one state of existence to another (p. i). In
the past decade or two, noticing has been receiving an increasing amount of
attention from applied linguistics researchers (e.g., Ellis, 1993, 1995; Robinson,
1995; Schmidt, 1990, 1994; Schmidt & Frota, 1986; Sharwood Smith, 1981,
1991, 1993; Swain, 1985, 1995; Swain & Lapkin, 1995). Its significance for
second language (L2) acquisition can be seen and understood from such claims as
those who notice most, learn most (Schmidt & Frota, 1986, p. 313) and no
noticing, no acquisition (Ellis, 1995, p. 89).1
Noticing is also an important cognitive process in L2 composing. Issues such
as how noticing is related to L2 composing and what impact it has on L2 writing
improvement still need to be further addressed and investigated, especially
through empirical research studies. Indeed, in order to help learners improve
their L2 writing skills and assist them to achieve native-like writing proficiency,
teachers need a better understanding of the L2 writer, the native-speaker reader,
the L2 text, and the context, as well as their interactions (Silva, 1990). Due to the
complex nature of writing itself and the unique characteristics inherent in L2
writing (Silva, 1993), comprehension of these interrelations requires research
studies at various analytic levels and from both sociolinguistic and psycholinguistic perspectives (Cumming, 1998). This article examines some of the
interactions that take place as L2 writers engage in composing an L2 text and
then compares the text to a reformulated version of it. In particular, this article
attempts to investigate, using think-aloud protocols produced by L2 learners,
what learners notice when they are composing by themselves and when they are
comparing their own written text to a reformulated version of it, and what effects
such noticing has on their resulting L2 text.
Noticing and writing
Research on noticing in L2 acquisition has largely focused on input. Based on
Schmidt (1990), Batstone (1996) defines noticing as the intake of grammar
as a result of learners paying attention to the input where intake refers to
input which becomes part of the learning process (p. 273). In this definition,
noticing is equated with intake that is derived from the source of input. It can be
argued, however, that in the broader context of language acquisition, it is not just
intake that stems from an input source that can become part of the learning
1 Although specific functions of noticing with respect to L2 acquisition are not without
controversy (see, e.g., Truscott, 1998, for a critical review), the researchers investigating the concept
of noticing, including Truscott, all seem to agree that noticing could play a helpful role in the process
of L2 learning.

D.S. Qi, S. Lapkin / Journal of Second Language Writing 10 (2001) 277303

279

process. There are types of intake that do not stem from the input source but are
generated in output (Swain, 1985) that may also be significant for L2 acquisition.
For the purpose of this article, we would like to define noticing broadly as
awareness of a stimulus via short-term memory (see Robinson, 1995, for a
comprehensive review and discussion of various dimensions and constructs of
noticing). We refer to stimulus as anything that rouses ones attention, in
particular, for our purposes, with respect to language (input or output). Our stance
in this article is that while noticing of input is exceedingly important, noticing as
a result of producing the target language (TL), as in the context of L2 composing,
also has important roles to play in L2 development.
Swain (1985) has proposed the output hypothesis and argues that comprehensible output is a necessary mechanism of acquisition independent of the role
of comprehensible input. She points out that producing the TL may be the trigger
that forces the learner to pay attention to the means of expression needed in order
to successfully convey his or her own intended meaning. This will move the
learner from a purely semantic analysis of the language to a syntactic analysis of
it. Swain (1995, 1998) further argues that the noticing/triggering function of
output can prompt L2 learners to recognize consciously some of their linguistic
problems. It may make them aware of something they need to find out about their
L2. L2 writing studies that employ think-aloud research techniques support the
claim that output stimulates noticing of problems that prompts learners to engage
in some kind of analysis of their existing linguistic resources in order to resolve
these problems (e.g., Cumming, 1990; Qi, 1998; Swain & Lapkin, 1995).
Swain and Lapkins (1995) study examines directly the noticing function of
output in the context of L2 writing, asking if learners own output can lead them
to a conscious awareness of language problems they are experiencing, if
cognitive processes are triggered in response to the problems they are aware
of, and if learners engage in grammatical analysis in the processes. Eighteen
Grade 8 French-immersion students (average age 13) participated in that study.
The participants were asked to think aloud while composing in response to a
writing task. Language-related episodes (LREs, the unit of analysis developed by
the researchers) were identified from the think-aloud protocol data and then
classified into descriptive categories according to the cognitive processes the
researchers thought were reflected in the changes the participants made to their
output. The study demonstrated that the young French-immersion L2 writers did
indeed notice gaps in their linguistic knowledge while producing their L2 in the
composing process. The researchers also found that when they encountered
difficulties in producing the TL, they did engage in certain thought processes that
may play a role in L2 learning, including grammatical analysis that was
considered essential to accurate production.
This study also yielded other findings. One of the results is that for these
learners, the substance of their thoughts was sometimes faulty, leading to
incorrect hypotheses and inappropriate generalizations. This suggests that relevant feedback can play a central role in improving L2 writing development.

280

D.S. Qi, S. Lapkin / Journal of Second Language Writing 10 (2001) 277303

Another finding was that the most proficient participants appeared to pay more
attention to grammar and relied more on applying grammatical rules than the least
proficient participants, suggesting that L2 proficiency may play a role in
linguistic awareness. These results pointed to some important issues that need
further consideration in future research. These include what feedback teachers
should provide to promote the L2 writers noticing and learning, and how L2
proficiency is related to the L2 learners noticing in composing and his/her
reactions to feedback. Our study attempts to address these issues. Before we turn
to our study, we will first briefly review some relevant literature about feedback
and its relation to noticing.
Feedback and noticing
When L2 learners incorrect hypotheses and inappropriate generalizations lead
to errors in their written texts, appropriate feedback from the teacher is needed in
order to help learners correct these errors. Zhangs (1995) research indicates that
L2 learners genuinely and overwhelmingly welcome feedback, especially from
the teacher. Error correction is also what most learners want. For example, Lekis
(1991) research found that 70% of the 100 learners surveyed expected all their
errors to be corrected. In the context of L2 acquisition, error is referred to as the
discrepancy between the learners interlanguage (IL) and the native speakers
version, i.e., the TL (James, 1998, p. 63). However, research results regarding
what is considered to be appropriate and effective feedback on students errors in
writing have been inconclusive, sometimes contradictory, and in L2 writing,
sparse (Leki, 1990, p. 66).2
Some research has tried to identify factors that may influence the effectiveness
of written feedback. One such factor is that the teachers feedback may be
unclear, inaccurate, and may lack balance among form, content, and style (Cohen
& Cavalcanti, 1990, p. 155). A second factor may relate to a possible lack of
sensitivity of teachers to different contexts as well as to varying levels of need,
ability, and other individual differences of students in providing feedback (e.g.,
Conrad & Goldstein, 1999; Ferris, Pezone, Tade, & Tinti, 1997; Hyland, 1998).
Another critical factor, as we view it, is that the type of feedback the teacher
offers to the learner does not provide optimal conditions to help learners notice
their errors, i.e., the gap between their IL and the TL when they receive and

2 Issues discussed in the literature on feedback concern whether the focus of feedback should be
on form or content (see Fathman & Whalley, 1990; Hyland, 1998), if written commentary is useful
(see Leki, 1990, pp. 60 63), if editing instruction and grammar correction are effective (see Polio,
Fleck, & Leder, 1998), or if teachers feedback is what learners need (Cohen & Cavalcanti, 1990). A
review of research in the literature suggests that there seems to be a debate over issues such as form/
meaning focus (e.g., Horowitz, 1986; Hyland, 1998; Paulus, 1999; Zamel, 1988) and effectiveness of
teachers comments or error correction (e.g., Ferris, 1999; Polio et al., 1998; Semke, 1984; Truscott,
1996; Zamel, 1985).

D.S. Qi, S. Lapkin / Journal of Second Language Writing 10 (2001) 277303

281

process the feedback. James (1998) compares noticing the gap as similar to
learners doing their own error analysis (EA), as both entail making a comparison
between IL and TL (p. 258). Ellis (1995) notion of cognitive comparison
roughly refers to this same idea. However, James has identified a difference
between Ellis cognitive comparison and EA. Cognitive comparison refers to
the learners noticing of a linguistic entity in TL input before comparing it with
his/her own IL version of it. However, EA proceeds in the converse direction:
The learner first notices a problematic formulation in his/her own production
before comparing it with a TL version of it (p. 258). According to James, it is not
just preferable but necessary that the forms learners notice and cognitive
comparisons they make are based on their own recent learning experience,
particularly where that experience is negative (p. 258).3 This EA sense of
cognitive comparison is consistent with the output hypothesis in that both
emphasize the important role of noticing derived from output in L2 learning.
Reformulation and noticing
The idea of reformulation was first proposed by Levenston (1978) to challenge
Corders (1971) notion of reconstruction.4 Reformulation refers to a native
speakers rewriting of an L2 learners composition such that the content the
learner provides in the original draft is maintained, but its awkwardness,
rhetorical inadequacy, ambiguity, logical confusion, style, and so on as well as
lexical inadequacy and grammatical errors are tidied up (Levenston, 1978). The
rewritten text provides a TL model so that the learner can make a comparison of
his/her own draft with a native writers version of it.
In the early 1980s, Cohen (1982, 1983a, 1983b) conducted several research
studies to investigate the effectiveness of reformulation and reported that learners
at intermediate levels and above seem to benefit from using this technique
(Cohen, 1983b, p. 5). His research results show that the L2 writers benefited from
reformulation in such aspects as vocabulary, syntax, and paragraphing, as well as
cohesion. Through reformulation, the learners gained insights about problems of
cohesion, grammatical rules, precision in the use of vocabulary, and differences
in levels of formality in the TL. Cohen (1989) concludes that despite some
potential problems as with any pedagogical technique, reformulation has much
3 It needs to be pointed out, however, that James (1998) use of the term error analysis here is
different from the orthodox use of the term error analysis that has been established and recognized in
the field as a research approach rather than something a learner does.
4 Corders (1971) notion of reconstruction refers to replacement of a learners idiosyncratic
sentence that carries grammatical errors with a well-formed sentence with the same meaning. The
focus of this notion of reconstruction is on correction of surface grammatical errors at the sentence
level only. Reformulation, as defined by Levenston (1978), addresses more than what reconstruction
does. A reformulator of a learners writing needs to take into consideration such issues as lexical
choice, syntactic blend, conceptual clearness, rhetorical adequacy, stylistics, and so on as well as
spelling and grammatical correctness.

282

D.S. Qi, S. Lapkin / Journal of Second Language Writing 10 (2001) 277303

promise. It provides an opportunity for the learner to obtain deeper feedback


than in the simple correction of surface errors, which is often what learners
receive as feedback on their essays (p. 9).5
Reformulation has advantages that other types of feedback may lack. For one
thing, it provides relevant TL structures to allow the learner to appropriate from
them according to his/her own needs and interests and to the context they
themselves provided for a particular writing task. This is important because
preferences for the teachers feedback may vary from student to student (Cohen
& Cavalcanti, 1990, p. 165) and the teachers response should respect the
students right to their own expression and their intention in a given piece of
writing (Leki, 1990, p. 64). Secondly, a text reformulated by a competent TL
writer provides appropriate TL forms for the given context. Without this, L2
learners are left to figure out the solutions themselves since they have a smaller
backlog of experience with English grammatical or rhetorical structure to fall
back on, not having had the same exposure to those structures as native speakers
have had (Leki, 1990, p. 58). Thirdly, this type of feedback provides a good
balance between focus on form and focus on meaning since it exploits both the
meaning-driven and form-focused potential: By the manipulation of task design
or the choice of text, teachers can harness the technique for either meaningdriven or form-focused needs (Thornbury, 1997, p. 334). It needs to be noted,
nevertheless, that reformulation may give rise to problems (e.g., text appropriation); every effort should be made by the reformulators to ensure maximal
respect for the content of the original text (e.g., through conferencing with the
writer before or after reformulation).
Sanaoui (1984) studied the use of reformulation in a classroom setting. The
results indicate that although there are differences between better writers and
poorer writers, all her French-as-second-language students benefited from the
use of the reformulation approach in such areas as selection of vocabulary,
syntactic structures, markers of cohesion, discourse functions, overall organization, and stance towards the reader. She reports that the results obtained
over a short period of time surpass results I have achieved with other students
through any type of explicit or direct instruction in teaching creative writing
(p. 145). Similar findings with respect to the effectiveness of reformulation
were made in Mantellos (1996) research exploring the use of reformulation in
a classroom setting. Allwright, Woodley, and Allwright (1988) conducted a

5 The idea of reformulation has also been largely employed and researched in the context of
promoting learners TL accuracy at the oral level through the use of corrective feedback called
recasts in the course of conversational interaction (e.g., Farrar, 1992; Long, 1996; Mackey & Philp,
1998; Wilberg, 1987). In addition to potential usage differences between speaking and writing in terms
of genre and style, recasts generally address a speakers lexical and grammatical problems at a
sentence or a specific local level, given the nature of the context in which the technique is used,
whereas reformulation, as written feedback, provides the writer with a model for analysis at the textual
as well as sentence levels.

D.S. Qi, S. Lapkin / Journal of Second Language Writing 10 (2001) 277303

283

study investigating nonnative writers texts resulting from reformulation followed by a class discussion. They found that while the reformulation strategy
may be a viable way of helping nonnative writers move nearer to native-writer
norms, the class discussion that follows the provision of the reformulated
version may be more influential than the reformulation itself. As the study did
not directly examine the relationship of the students reaction to reformulation
to the subsequent class discussion, it is hard to evaluate what direct and
indirect effects the reformulation technique itself had on the students final
written texts.
A key function of reformulation is its provision of opportunity for noticing. If
cognitive comparison is essential to learning as discussed earlier, we should select
feedback types on the basis of their capacity to promote noticing and EA. Some
criteria for appropriate types of feedback should be their potential to encourage
learners to pay attention to form and, moreover, provide learners with TL data so
that they can make comparisons between their IL and a TL model of it.
Thornbury (1997) proposes that reformulation is one feedback type that meets
these criteria. Johnson (1988, cited in Thornbury, 1997) argues that exposing
learners to the target behavior after the event rather than providing a model
beforehand has greater psychological validity, in that the learners are predisposed to look out for (and notice) those features of the modeled behavior that
they themselves had found problematic in the initial trial run (or first draft) (see
Thornbury, 1997, p. 328.). This idea is also argued for in output theory. That is,
the problems learners encounter in output could trigger an analysis of incoming
data, that is, a syntactic analysis of input (Swain & Lapkin, 1995, p. 375). The
purpose of the study reported below is to investigate the processes underlying the
learners initial output and the relationship of these processes to those occurring
as they confront the reformulated version. We also examine the impact of these
processes on the resulting text in a three-stage L2 writing task. Including two
participants with two different L2 proficiency levels permits us to explore the
relationship of L2 proficiency level to the qualities of noticing in a particular
writing task.
The study
This research is the result of one of the pilot studies conducted for a large-scale
research project. One of its objectives is to explore the roles played by output
(speaking and writing) through collaborative dialogue in L2 learning. Preliminary
to the main research, this particular case study was to investigate the roles of
output in a three-stage L2 writing task performed on an individual basis by two
learners with two different levels of L2 proficiency.
The writing task consisted of the following three stages (see detailed
description under Procedure below). First, a participant wrote an L2 text in
response to a picture prompt (Stage 1). Then he/she was asked to engage in a
comparison of his/her written draft with a reformulated version of it followed by

284

D.S. Qi, S. Lapkin / Journal of Second Language Writing 10 (2001) 277303

an immediate retrospective interview intended to clarify what he/she had noticed


(Stage 2). Finally, we gave the participant his/her original text and asked him/her
to revise it (Stage 3). By examining each of the three stages, we intended to
investigate the effects of noticing in output: (a) as each individual participant
composed an L2 text (Stage 1); (b) as he/she subsequently compared his/her L2
text with a reformulated version of it (Stage 2); and (c) as he/she revised the L2
text (Stage 3). Thus, our research questions in this study were as follows:
1. What aspects of language do L2 learners notice in/during an output-only
writing condition (Stage 1 of a three-stage writing task)?
2. What do L2 learners notice as they compare their text to a reformulated
version of it while thinking aloud (Stage 2 of a three-stage writing task)?
3. How is such noticing related to changes in the written product from Stage 1
to Stage 3 (posttest) of the L2 writing task?

Methodology
Participants
The participants in this study were two adult Mandarin-speaking English-as-asecond language (ESL) learners. The first participant (hereafter Wu), male, aged
27, had a bachelors degree in textile engineering from China, had been in
Canada for 3 years, and had never been to any other English-speaking countries
before. He had studied English as a foreign language for over 10 years before
immigrating to Canada. One year after he arrived in Canada, he enrolled in a
2-year computer network and technical support diploma program at a Toronto
community college. At the time of our data collection for this study, he was in his
last semester of the program. An analysis of his oral sample based on the ACTFL
(American Council on the Teaching of Foreign Languages) proficiency guidelines (1986) as well as an analysis of his written text based on global scales for
writing (Hamp-Lyons, 1991) revealed that his oral English was at an advanced
level on the ACTFL scale and his written English was intermediate-high on the
global scale for writing. The second participant (hereafter Su), female, aged 32,
had a diploma in fine arts from a college in Taiwan, had been in Canada for 5
months, and had never been to any other English-speaking countries before. She
had studied English as a foreign language for 1 year before immigrating to
Canada. One month after her arrival in Canada, she enrolled in an ESL class at
the Toronto District School Board and was placed in a high-basic level class. At
the time of the data collection, she was in her fourth month of an ESL class. An
analysis of her English oral sample and her written text indicated that she had a
low-intermediate level of English proficiency on these scales. In short, Wu had a
relatively high level of ESL proficiency and Su had a relatively low level of ESL
proficiency. Both participants were interested in being involved in our research

D.S. Qi, S. Lapkin / Journal of Second Language Writing 10 (2001) 277303

285

after they were informed of the general purpose of the study, i.e., to document
what L2 learners think while doing a writing task in order to help teachers
understand their learners needs.
Procedure
Each participant wrote a narrative in response to a picture prompt of The
Scene of the Crime (see Appendix A). The picture prompt was chosen for two
reasons.6 First, the picture did not provide any verbal data so the participant
would have to produce the TL in an output-only writing condition. Second, it was
open-ended so the participants could compose the story in whatever way they
liked, thus giving them control over the content. Before the data collection, each
participant was trained to produce think-aloud protocols; that is, each was given
two multiplication problems and was asked to think aloud while solving these
problems. Each was asked to think out loud in the language of his/her choice
(following procedures in Cumming, 1989, p. 89). The picture prompt was not
given to the participant until he or she practiced the think-aloud method several
times and felt comfortable enough to use the method. Each participant was given
30 minutes to do the composing (Stage 1), and each did the writing task at about
the same time of day. We used audiotape to record think-aloud verbalizations and
videotape to capture the participants writing of texts. The audio-track of the
videotape also recorded the think-aloud and provided a backup copy of the
verbalizations (following procedures in Bracewell & Breauleux, 1994, p. 58).
One of the researchers was present to operate the video camera and to remind the
participant to keep thinking aloud if he/she stopped talking for very long. To
reduce possible psychological reactivity effects caused by the use of video
camera and audio recorder as well as the presence of a researcher, we consulted
both participants in advance. Both expressed the opinion that this would not
affect their performance so long as it was for research purposes only. When asked
about the same questions concerning the reactivity effects after they finished their
tasks, Su confirmed in Chinese that her mind was too busy with the task itself to
be affected, and Wu said that I really dont think they were affecting my
thinking and writing at all.
Each participant was asked to proofread his/her writing in the end before
submitting it to us. The purpose was to make sure that the draft represented the
participants best possible version and that the errors were not slips of the
tongue or pen that could be self-corrected by the author (James, 1998; Poulisse,
1999). As in Swain and Lapkins (1995) study, the participants in this study were
also advised that they could not have access to a dictionary or any other aid and
that the researcher would not be able to help either. We intended to find out what

6 This type of picture prompt has drawbacks in that its open-ended nature does not provide any
particular focus for the task.

286

D.S. Qi, S. Lapkin / Journal of Second Language Writing 10 (2001) 277303

students would do without such resources and whether they would try to work
out solutions on their own.7
After the task was completed, the researchers collected the draft and
reformulated it from beginning to end to produce a native-like model. In the
reformulated version, the participants original ideas were maintained. We
corrected all the syntactic and morphological errors, and removed any problems
in stylistics and logical sequencing at the discourse level. Four days later, we
gave each participant his/her own original draft along with the reformulated
version of it and asked him/her to think aloud while comparing his/her own
draft with the reformulated version of it and noticing the differences between
the two versions (Stage 2). Again both the audiotape and videotape were used
to record the comparison process. Immediately following the comparison stage,
a researcher rewound the videotape and showed the participant the video of the
comparison and noticing process that had just been taped. The researcher
stopped the tape wherever noticing took place before asking the participant to
clarify what he/she was noticing specifically at that particular moment (immediate retrospective interview). The evidence on which we based our judgement
of the occurrence of a noticing episode includes their pointing to a specific line
in a text, their exclamatory utterances such as oh, yeah, right, etc., their
hesitation, and their thinking pauses. The researcher who could speak both
English and Mandarin conducted the interviews primarily in English,8 and the
learners used the language of their choice to respond.
Each participant did a posttest 1 week after Stage 2. The participant was not
informed of the posttest in advance. For the posttest, each participant received his/
her own original written draft, which was typed in a triple-spaced format, and was
asked to revise his/her draft right on the triple-spaced typed version (Stage 3) based
on what he/she had learnt throughout this entire task process. In sum, each participant produced two protocols, one from Stage 1 and the other from Stage 2. The four
think-aloud protocols, produced by the two participants both at Stage 1 and Stage 2,
were transcribed for analysis. Formal permission to use the think-aloud protocols
and produced texts was given by the participants (see Appendix B for the texts
produced by the participants as well as the reformulated versions of the draft texts).
Data analysis
Researchers have proposed the use of concurrent think-aloud protocols as a
useful source of information about cognitive processes in L2 research (e.g.,
Cohen, 2000; Leow, 1997). Schmidt (1990) also mentions that noticing can be
operationally defined as availability for verbal report, and it can sometimes even
7 Because this was a pilot study in the context of a Swain/Lapkin project on Extending the
Output Hypothesis, we followed the same method here as that planned for the main study.
8 Conducting the interviews in English in this pilot study made the data more accessible to the
project team.

D.S. Qi, S. Lapkin / Journal of Second Language Writing 10 (2001) 277303

287

take place when it cannot be easily verbalized (p. 132). In other words, what is
verbalized is normally what has been attended to. Moreover, it has been argued
that the very act of producing a written form itself may allow a writer to bring
aspects of speech, such as sentences, words, and phonemes, into consciousness
(Olson, 1994). In our study, we focused our analysis on LREs from the four
think-aloud protocols produced by the two participants.
Swain and Lapkin (1995) defined LRE as any segment of the protocol in
which a learner either spoke about a language problem he/she encountered while
writing and solved it either correctly or incorrectly, or simply solved it (again,
either correctly or incorrectly) without having explicitly identified it as a problem
(p. 378). As our study involved both writing and comparing (a draft to a
reformulated text of it), we extended the definition so it could be applied to
the comparison stage as well. Thus, in addition to what is defined in Swain and
Lapkins research, an LRE in our study refers to a segment of the protocol in
which a learner noticed a language-related problem he/she encountered while
comparing his/her text to a reformulation and addressed it either by accepting the
reformulation and providing a reason, or only noticing the difference without
giving a reason.
In this study, we did not count as LREs verbalizations of segments of the
text as they were being produced, as we had no evidence that they represented
the cognitive process of solving language-related problems; with Cumming
(1990, p. 504) we considered these verbalizations to be an automatic act of
writing behavior that may present L2 learners little opportunity to develop
linguistic knowledge. However, we counted language-related noticing which
was not verbalized (e.g., self-correction, see Leow, 1997) but captured in the
videotape and verified in the interview as LREs. For example, in Stage 1, while
proofreading her initial draft, Su crossed out the word discussed and wrote
disscused above it without verbalizing the process.
The first author of the present article identified the LREs in the writing (Stage 1)
and noticing (Stage 2) transcripts, and discussed any problematic cases with the
second author. Each of the authors then coded the LREs into three broad categories
as follows: (1) lexical includes adjective, adverb, noun, preposition, pronoun,
and verb; (2) form includes comparative (adjective), superlative (adjective),
plural, possessive marker, possessive pronoun, punctuation, sentence structure,
spelling, subject verb agreement, verb form, and verb tense; (3) discourse
includes achieving logical sequencing (cohesion, coherence), achieving intersentential clarity, and stylistics (see Swain & Lapkin, 1995). The interrater reliability
for identification and coding of LREs was 91% (based on simple percentage
agreement) before the two researchers conferenced until an agreement on the
coding for each LRE was reached. The categorization of the lexical and form
LREs depended on the linguistic focus of the LREs (e.g., searching for a lexical
item, applying a grammatical rule, noticing a morphological difference in Stage 2,
etc.), and the categorization of the discourse LREs depended on the way in which
ideational unity across the text was achieved through the appropriate use of

288

D.S. Qi, S. Lapkin / Journal of Second Language Writing 10 (2001) 277303

language. The following three examples serve to illustrate briefly the way in which
we coded the LREs into the three categories. In all the examples, we use words in
boldface italics to represent a participants L1 (Mandarin) use. Words that are
underlined represent written text.
1. Lexical (verb resolving the problem incorrectly) Wu (Stage 1):
Console Mikes mother. Console console how to say it? What word should
I use? The maid is helping them console his mother. It will be fine enough by
using help his mother.

Wu was searching for the word console in his first language. But he failed
to translate it successfully into English and then decided that use of help would
suffice. In this example, Wu first talked about a lexical problem he encountered.
Then he conducted the lexical search in memory by using his L1. He came up
with the word help. Since console and help were not synonyms, Wus
solution as displayed in this LRE was counted as incorrect.
2. Form (verb tense resolving the problem incorrectly) Su (Stage 1): So
they asked the second son but he didnt, no should be now, doesnt want to take
it out.

The focus of this LRE was on the tense. Su first used the verb form didnt.
Then she felt that she might have used the wrong verb tense as it should be now.
So she changed it into doesnt, which happened to be an incorrect solution.
3. Discourse (logical sequencing accepting reformulation with a reason) Su
(Stage 2): Oh yeah I forgot to turn on the light.

Earlier in her draft text, Su mentioned that the light suddenly went out and the
room became dark. Later in the text, she described the scene in such a way as it
could only have been observed after the light was turned on. However, she did not
mention when the light was turned on. This gap was resolved in the reformulated
text. While comparing the two texts, she noticed the flaw in her original. This
example indicates that in Stage 2 the participant noticed a language-related
problem she encountered while comparing her own draft text with a reformulated
text and then addressed it by accepting the reformulation with a good reason. We
refer to reason as a participants verbalizing the nature of a difference in a point
being compared between the learners draft text and its reformulated text.
Although in our instructions we did not ask the participants to provide a reason
while they were comparing, we asked them to verbalize their thoughts in their
comparing processes. Their reasons were clarified and verbalized in the
retrospective interview and are reflected and incorporated in the coding results.
Findings
Noticing in the composing stage (Research Question 1)
It took each participant about the same length of time (30 minutes each) to
complete the composing task. Table 1 presents the number of LREs generated in

D.S. Qi, S. Lapkin / Journal of Second Language Writing 10 (2001) 277303

289

Table 1
Number of LREs for the two participants in Stage 1
Wu

Su

LRE
categories

Correctly
resolved

Incorrectly
resolved

Total

Correctly
resolved

Incorrectly
resolved

Total

Lexical
Form
Discourse
Total n
Percent

8
6
2
16
64

5
4

9
36

13
10
2
25
100

1
3

4
25

3
9

12
75

4
12

16
100

Stage 1 in each broad category (lexical, form, and discourse) and the number
resolved correctly or incorrectly. The table indicates that while almost half of the
total 41 language-related problems encountered in the composing stage were
resolved correctly (16 for Wu and 4 for Su, a total of 49%), just over half of the
problems were identified but not resolved correctly (9 for Wu and 12 for Su,
accounting for 51% of the LREs).
We assume that the failure to reach a satisfactory solution to a problem with
existing linguistic knowledge may result in a sense of uncertainty or lack of
fulfillment on the part of a learner. For example, Wu recalled a noticing
experience during the interview when he was referring to the difference between
his choice of talk about in the composing stage and the reformulators choice
of the word chat in the reformulated text: . . . I tried to use a good word for
talk about. I knew thats not a good use here talk about. But I couldnt find a
better word. The teacher used chat. Thats a good word. But I couldnt find that
at the moment when I wrote the article. It is perhaps this sense of lack of
fulfillment that may push a learner to look out for any future relevant information
available that he/she believes might help solve the problems in a better way.
An important characteristic of the LREs in Stage 1 is the identification of
problems followed by the modification of a linguistic form in the meaning form
mapping process; that is, Step 1, the writer verbalized an idea, and Step 2, the
writer modified the linguistic realization of the idea before mapping the idea onto
what he/she thought was an appropriate form of the target language. For example:
4. Form (spelling resolving the problem incorrectly) Wu (Stage 1): From
where he received the phone call? From um um the famous bullfighter . . . bull
fighter Johns home.

Wu first verbalized the idea of bullfighter in Chinese. After writing down


bull-fighter Johns home, he crossed out the word bull-fighter (compound
word) and wrote bull fighter (two words) above it without verbalizing the
process. The more appropriate form should be one word.
Although the LRE indicates that the modification turned out to be an incorrect
one, Wus output had triggered his noticing of the form of this particular word.
Thus, this type of noticing may promote an L2 writers search for an appropriate

290

D.S. Qi, S. Lapkin / Journal of Second Language Writing 10 (2001) 277303

match between ideational content generated in thought and a linguistically


appropriate form available in memory, a process that is essential to improving
L2 writing proficiency or even L2 proficiency in general.
As we know, Wu had a relatively high level of English (L2) proficiency, and
Su had a relatively low level of English (L2) proficiency. The results shown in
Table 1 reveal differences between the two participants both in the quantity and
quality of their Stage 1 LREs. As shown in Table 1, 25 LREs were identified for
Wu while the corresponding figure for Su was 16. This suggests that L2 writers
with a relatively high level of L2 proficiency might engage more in languagerelated noticing than L2 writers with a relatively low level of L2 proficiency.
Table 1 shows that for Wu 64% of the language-related problems noticed in Stage
1 were resolved, and 36% were not resolved, whereas for Su 25% were resolved
and 75% not resolved. In this case study, the L2 writer with a higher level of L2
proficiency seems more capable of solving language-related problems on his own
than the lower-proficiency L2 writer.
Noticing in the reformulation task (Research Question 2)
It took Wu about 9 1/ 2 minutes to complete the comparing task and Su about
12 minutes. By viewing the videotapes, we found that both participants focused
more on reading the reformulated text than reading their own draft while
comparing. This seems to indicate that they were more interested in looking
out for anything that was different from what they were familiar with, or anything
that was new to them. Most of the episodes we identified in this stage included
such exclamatory utterances as Oh!, Yeah!, Ha!, etc. as well as such
comments as I forgot this, this one is better, I was supposed to use . . .,
Oh, its much clearer, etc. to express realization of a difference he/she noticed.
This demonstrates that the noticed features of the modeled TL behavior were
being constantly compared to the learners own written text and that the
participants own recent experience of output in Stage 1 was an important factor
in influencing what he/she noticed in Stage 2. Example 5 provides part of the
transcript from Stage 2 that builds directly on example 1 (from Stage 1):
5. Lexical (verb accepting reformulation with a reason). Wu (Stage 2): Yes,
console is good. Console Mikes mother. Yes. It is much better than this one
[referring to the incorrect solution help he came up with in Stage 1].

In Stage 2, Wu immediately and readily accepted the reformulation by


commenting that console in the reformulated text represented much better
the meaning he had struggled to express in Stage 1.
Table 2 summarizes the LREs identified for the two participants in Stage 2.
More importantly, the table reveals that the participants were able to provide
reasons for accepting the reformulation in 28 cases (21 for Wu and 7 for Su for a
total of 47% of the LREs). In over half of the LREs produced by the two
participants in Stage 2, however, they merely noticed the differences between the
reformulation and their own text or accepted the reformulation without stating a

D.S. Qi, S. Lapkin / Journal of Second Language Writing 10 (2001) 277303

291

Table 2
LREs by the two participants in Stage 2
Wu

Su

LRE categories

Notice only

Accepting
reformulation
with a reason

Lexical
Form
Discourse
Total n
Percent

8
0
0
8
28

9
8
4
21
72

Total
17
8
4
29
100

Notice only

Accepting
reformulation
with a reason

Total

10
10
4
24
77

1
5
1
7
23

11
15
5
31
100

reason. Examples 6, 7, and 8 illustrate these three ways in which language-related


noticing occurred.
6. Lexical (noun accepting reformulation with a reason) Wu S2: Gunshot,
Its gunshot not gun fighting. Heard the gunshot. Simply use a noun.

Wu accepted the reformulated noun and verbalized a correct explanation for


doing so.
7. Form (subject verb agreement noticing only) Su S2: [Original: Some
parents they doesnt want to share theirs wealth/Reformulation: Some parents do
not want to share their wealth]. Here is different: dont and doesnt.

As Su does not make any judgmental comment or indicate any sign of


accepting the reformulation, this LRE was coded into noticing only.
8. Form (superlative adjective accepting reformulation without stating a
reason) Su S2: [Original: The older is short guy/Reformulation: She has three
sons. Her eldest son is the short guy]. Yeah. Eldest.

Su seemed to have accepted the reformulation but did not provide a reason or
rationale. So the LRE was coded into noticing only.
These examples suggest that while reformulation may generally prompt
learners to notice differences between two texts, the quality of that noticing is
variable. For the purpose of this study, we refer to the quality of noticing as the
extent to which an instance of noticing is perfunctory (i.e., noticing only and
without giving reasons) or substantive (i.e., noticing and providing reasons). As
shown below in our study, the quality of the learners Stage 2 noticing generally
has a direct impact on the final product (see our discussion under Research
Question 3 below).
Table 2 illustrates the quality of noticing identified from the LREs for each
participant in the comparison stage (Stage 2). It shows that although Wu (with a
higher level of L2 proficiency) and Su (with a lower level of L2 proficiency)
produced a similar number of LREs (29 for Wu and 31 for Su), Wu provided a
reason for accepting the reformulation more frequently than Su (72% vs. 23%).
On the other hand, 77% of Sus LREs involved noticing the differences without
making any comments, while for Wu, the noticing-only episodes accounted for a

292

D.S. Qi, S. Lapkin / Journal of Second Language Writing 10 (2001) 277303

relatively small proportion of his LREs (28% of his own total). These results
indicate that a learner with a lower level of L2 proficiency may not be so capable
as a learner with a higher level of L2 proficiency of understanding the nature of
noticed gaps despite the availability of a model text. This seems to support the
finding from Mantellos (1996) study that lower-level learners would not benefit
as much from reformulation as higher-level learners. In short, the results of our
study suggest that the quality of noticing in the reformulation task may be directly
related to level of L2 proficiency.
The results from both Tables 1 and 2 demonstrate that there were substantially
more LREs in Stage 2 than in Stage 1 (especially for Su), suggesting that
learners engaging in a comparison of their own text with a reformulated version
of it could greatly promote noticing. In Stage 2, the participants memory was
shown to be fresh from their earlier experience (Stage 1) of problem solving,
when they did not always reach a satisfactory solution. An item-to-item
comparison between the two stages of noticing revealed that of the nine problems
that Wu noticed and failed to resolve correctly in Stage 1, seven (78%) were
noticed in Stage 2, and that of the 12 problems that Su noticed and did not resolve
correctly in Stage 1, nine (75%) were noticed in Stage 2. The fact that most of the
problems that were noticed and not resolved correctly in Stage 1 were noticed in
Stage 2 confirms our earlier assumption that the participants failure to reach a
satisfactory solution to a problem with existing linguistic knowledge may lead to
a sense of lack of fulfillment. They hoped to check out their earlier solutions
against their reformulated texts. This suggests that the output process that takes
place in an early stage of a writing task may play an important role in influencing
noticing of the related language data at a later stage.
Relationship of noticing to changes in the revision stage
(Research Question 3)
Table 3 illustrates the effects of the LREs in Stage 2 on the changes in the
posttest (Stage 3). The table shows that in the case of Wu, of his 21 withreason LREs, 15 (71%) contributed to an improvement in written product, 1
Table 3
Relationship between the LREs in Stage 2 (S2) and the changes in Stage 3 (S3, posttest)

Wu

Su

Relationship to
changes in S3

No relationship to
changes in S3

Quality of
LREs in S2

Total number of
LREs in S2a

Better

Percent

Same

noticing only
accepting with
a reason
noticing only
accepting with
a reason

8
21

3
15

38
71

5
5

62
24

24
7

4
5

17
71

29

13
2

54
29

a The figures correspond to those in Table 2.

Percent

Percent

Example

S1
written product

Wu: Detective Jack received a phone call


from the famous bull fighter Johns home.
[bull fighter incorrect spelling]

10

Su: So they asked the second son, but he


doesnt want to take it out.
[incorrect tense]

11

Wu: Johns niece Mike is dead


[niece incorrect word choice according
to the picture prompt]
Su: . . . father who was die many years ago.
[incorrect verb form]

12

13

Su: The older is short guy . . . [incorrect


superlative adjective]

LRE in S1

LRE in S2

Change in S3

From where he received the phone


call? from um um the famous
bullfighter . . . the famous bull
fighter Johns home [wrote down
bull-fighter Johns home, and
crossed it and rewrote bull fighter
above it. [incorrect solution]
So they asked the second son, but
he didnt, no um should be now um,
doesnt want to take it out.
[incorrect solution]
Johns niece, I dont know
[referring to the word niece],
is dead. [incorrect solution]
[verbalization of the text while it is
being produced No LRE
was identified]
The older um is short short one
who . . . [No LRE was identified]

Oh, bullfighter is one word.


It should be one word! I thought
they were two words
[accepts with a reason]

Change bull fighter into


bullfighter [better]

Here [referring to didnt] the


grammar changed.
[accepts with a reason]

Changed doesnt into


didnt [better]

Oh, its a spelling mistakea


[referring to niece]. It should be
nephew. [accepts with a reason]
Who was die changed to
who died [noticing only]

Changed niece into


nephew. [better]

Eldest. Its good. [noticing only]

Underlined older with


no correction [same]

a Though Wu called it a spelling mistake, it was actually coded as a lexical LRE due to the nature of the problem.

who was die changed to


he was died [same]

D.S. Qi, S. Lapkin / Journal of Second Language Writing 10 (2001) 277303

Table 4
Examples of written product in Stages 1 and 3 and the underlying noticing processes

293

294

D.S. Qi, S. Lapkin / Journal of Second Language Writing 10 (2001) 277303

LRE (5%) resulted in a change but not an improvement, and 5 LREs (24%) were
not linked to any changes in Stage 3. There were similar findings for Su. That is,
of her total of seven with-reason LREs in Stage 2, five (71%) resulted in
improvements to her written story and two (29%) were not linked to any changes
in Stage 3. These results suggest that while some of the LREs where participants
accept the reformulation and provide a reason may not effect a change in the final
product, most of them play a positive role and contribute directly to L2 writing
improvement. Table 3 also shows that while some of the noticing-only LREs may
lead to improvement of the final written product (38% for Wu and 17% for Su),
most of the noticing-only LREs led either to no changes (62% for Wu and 54%
for Su) or to changes that did not represent an improvement in the written product
(e.g., 29% for Su).
Table 4 provides some examples of the changes in Wu/Sus written product
from Stage 1 to Stage 3 (posttest) along with relevant LREs from his/her thinkaloud protocol. Examples 9, 10, and 11 trace the relationship of Stage 1 LREs to
Stage 2 (noticing) LREs and show the effect of accepting the reformulation with a
reason on the written product in Stage 3 (the posttest). They indicate how LREs
in Stage 1 triggered noticing of relevant data in Stage 2 and then led to
improvement in Stage 3. Examples 12 and 13 indicate that noticing only without
providing a reason or demonstrating an understanding of the nature of the gap
between the learners IL and the TL may not lead to improvement. All these
results demonstrate that noticing without understanding or noticing for no
articulated clear reason does not have the same impact on learning in L2 writing
performance as does noticing with understanding. In turn, this suggests that the
quality of noticing in the reformulation task has direct implications for the written
product in a three-stage L2 writing task.

Discussion
Given that only two participants were involved in this case study, we consider
that the above results constitute tentative rather than definitive answers
(Allwright et al., 1988, p. 250) to our three research questions. Overall, the
results of the study have shown that in a three-stage L2 writing task, languagerelated noticing may contribute to the improvement of L2 writing. Furthermore,
the results indicate that language-related noticing in an output-only writing
condition, along with production in the process, may not only promote subsequent problem-solving performance during the solo composing activity but
also trigger noticing of relevant information from the modeled TL data when
available in the later stage of a writing task. The results of the study also
demonstrate that quality of noticing in Stage 2 has direct implications for the final
written product in Stage 3. This finding suggests that while promoting noticing in
a reformulation task may be important, improving the quality of noticing may be
even more important.

D.S. Qi, S. Lapkin / Journal of Second Language Writing 10 (2001) 277303

295

As shown in this study, due to the availability of a relevant text model (based
on their own prior written expression), learners were often able to notice the gap.
However, the quantity of language-related noticing in the output-only composing
stage, and the quality of such noticing in both composing and reformulation
comparison stages of the writing, may be different for learners with different
levels of L2 proficiency. The results demonstrate that learners with a higher level
of L2 proficiency may be able to conduct more language-related noticing while
composing alone than learners with a lower level of L2 proficiency. Further,
learners with a higher level of L2 proficiency may be more capable of solving the
noticed problems by themselves while composing alone than learners with a
lower level of L2 proficiency. When comparing their own written draft with a
reformulated version of it, learners with a higher level of L2 proficiency may
accept more reformulated items or structures, verbalizing the reasons for doing
so, than learners with a lower level of L2 proficiency. This suggests that learners
with a lower level of L2 proficiency may have more difficulty identifying the
nature of the gap between their IL and the TL even though a TL model is
provided to them. Thus, quality of noticing may be related to level of L2
proficiency: i.e., the higher the L2 proficiency, the better the Stage 2 noticing may
be in a three-stage L2 writing task. This finding is in line with Cohens (1983b)
claim that reformulation may benefit learners at intermediate levels and above
and may have its greatest impact among advanced students (p. 5, also cf.
Mantello, 1996). One explanation for the influence of the level of L2 proficiency
on the quality of noticing may be that learners with a low level of L2 proficiency
have limited linguistic knowledge of the TL, which may affect their judgement
about what is right or wrong and why (Swain & Lapkin, 2000).
While the results of the study generally point to some positive effects of
noticing conducted on an individual basis, they also reveal that some problems or
differences remain unnoticed in Stage 1 or Stage 2. As many of the problems that
were noticed and not resolved in the composing stage were noticed and addressed
correctly in Stage 2, and the correct form was maintained in Stage 3, promoting
noticing in composing might be important. It might also help to enhance the
quality of noticing in Stage 2.
The fact that some LREs where the reformulation was accepted for the right
reason were not utilized in the revision stage (Stage 3) suggests that even noticing
with comprehension may need some reinforced rehearsal in memory (Robinson,
1995). Moreover, despite the availability of the native speakers model text, there
was still a substantial percentage of differences (28% for Wu and 77% for Su)
that were noticed without the participants demonstrating evidence of understanding, particularly for the learner with the lower level of L2 proficiency.
Based on the above findings, we derive several implications for L2 writing
pedagogy. Like Cohen (1989) and Thornbury (1997), we suggest that reformulation is a valid pedagogical tool: The positive modeling of native-like writing
may be more helpful to the learner than error correction (negative feedback). It
appears important for teachers to promote language-related noticing in L2

296

D.S. Qi, S. Lapkin / Journal of Second Language Writing 10 (2001) 277303

composing. Teachers of writing routinely assess individual learners and identify


which L2 problems each learner needs to work on before completing a writing
task. This pilot study suggests the usefulness of directing learners to pay
particular attention to certain areas (e.g., lexical, grammatical, and discourse)
according to their needs. The research of Polio et al. (1998) demonstrates that if
L2 writers are given more time to reflect on their drafts, they can be given an
additional chance to focus on the accuracy of their language, thus promoting their
noticing in composing. Teachers could also teach the learner some languagerelated noticing strategies in composing by demonstrating them using the
thinking-aloud method. Verbalization itself may be an effective strategy. For
example, asking learners about what they are thinking and the rationale for their
grammatical decisions can promote metacognitive processing and lead to
effective problem solving (Dominowski, 1998). Our study indicates that even
though a noticed problem is not resolved successfully, such experience may lead
to noticing of relevant information in incoming input data.
Second, the teacher may need to train learners, especially those with a lower
level of L2 proficiency, how to notice the gap between their own draft text and
the reformulated text. This may mean that the teacher may need to organize some
awareness-raising activities in reformulation tasks. A list of such proposed
activities can be found in Thornbury (1997, p. 333).
Third, since quality of noticing in reformulation tasks may be crucial for
improvement of L2 writing, learners should be encouraged to work collaboratively
with the teacher or with both peers and native speakers to increase opportunities
for noticing as well as to improve the quality of noticing (Swain & Lapkin, 2001).
This may be indispensable for learners with a lower level of L2 proficiency.
In conclusion, the results of the study suggest that language-related noticing
does have a direct impact on students written products, and that output can
promote opportunities for such noticing both in an output-only writing condition,
and via the feedback provided by a reformulation of learners written texts.
Further, the quality of the noticing may differ according to the learners level of
proficiency and may affect their capacity to benefit from the reformulation. In
short, how to improve noticing quality especially for learners with a lower level of
L2 proficiency may be a crucial issue to be addressed in L2 writing pedagogy.

References
Allwright, R. L., Woodley, M. P., & Allwright, J. M. (1988). Investigating reformulation as a practical
strategy for the teaching of academic writing. Applied Linguistics, 9, 236 256.
Batstone, R. (1996). Noticing. ELT Journal, 50, 273.
Bennett, J. G. (1976). Noticing. Sherborne, England: Coombe Springs Press.
Bracewell, R. J., & Breauleux, A. (1994). Substance and romance in analysing think-aloud protocols. In P. Smagorinsky (Ed.), Speaking about writing: Reflections on research methodology
( pp. 55 88). Thousand Oaks, CA: SAGE Publications.
Cohen, A. D. (1982). Writing like a native: The process of reformulation (ERIC ED 224 338).

D.S. Qi, S. Lapkin / Journal of Second Language Writing 10 (2001) 277303

297

Cohen, A. D. (1983a). Reformulating second-language compositions: A potential source of input for


the learner (ERIC ED 228 866).
Cohen, A. D. (1983b). Reformulating compositions. TESOL Newsletter, 17 (6), 1 5.
Cohen, A. D. (1989). Reformulation: A technique for providing advanced feedback in writing.
Guidelines: A Periodical for Classroom Language Teachers, 11 (2), 1 9.
Cohen, A. D. (2000). Exploring strategies in test-taking: Fine-tuning verbal reports from respondents. In G. Ekbatani, & H. Pierson (Eds.), Learner-directed assessment in ESL ( pp. 127 150).
Mahwah, NJ: Erlbaum.
Cohen, A. D., & Cavalcanti, M. C. (1990). Feedback on compositions: Teacher and student verbal
reports. In B. Kroll (Ed.), Second language writing: Research insights for the classroom ( pp.
155 177). Cambridge: Cambridge University Press.
Conrad, S. M., & Goldstein, L. M. (1999). ESL student revision after teacher-written comments: Text,
contexts and individuals. Journal of Second Language Writing, 8, 147 179.
Corder, S. P. (1971). Idiosyncratic dialects and error analysis. International Review of Applied Linguistics, 9, 147 160.
Cumming, A. (1989). Writing expertise and second language proficiency. Language Learning, 39,
81 141.
Cumming, A. (1990). Metalinguistic and ideational thinking in second language composing. Written
Communication, 7, 482 511.
Cumming, A. (1998). Theoretical perspectives on writing. Annual Review of Applied Linguistics, 18,
61 67.
Dominowski, R. L. (1998). Verbalization and problem solving. In D. J. Hacker, J. Dunlosky, & A.
C. Graesser (Eds.), Metacognition in educational theory and practice ( pp. 25 45). Mahwah,
NJ: Erlbaum.
Ellis, R. (1993). Interpretation-based grammar teaching. System, 21, 69 78.
Ellis, R. (1995). Interpretation tasks for grammar teaching. TESOL Quarterly, 29, 87 105.
Farrar, M. J. (1992). Negative evidence and grammatical morpheme acquisition. Developmental
Psychology, 28, 90 98.
Fathman, A. K., & Whalley, E. (1990). Teacher response to students writing: Focus on from versus
content. In B. Kroll (Ed.), Second language writing: Research insights for the classroom ( pp.
178 190). Cambridge: Cambridge University Press.
Ferris, D. (1999). The case for grammar correction in L2 writing classes: A response to Truscott
(1996). Journal of Second Language Writing, 8, 1 11.
Ferris, D. R., Pezone, S., Tade, C. R., & Tinti, S. (1997). Teacher commentary on student writing:
Descriptions and implications. Journal of Second Language Writing, 6, 155 182.
Hamp-Lyons, L. (1991). Reconstructing academic writing proficiency. In L. Hamp-Lyons (Ed.),
Assessing second language writing in academic contexts ( pp. 127 153). Norwood, NJ: Ablex.
Horowitz, D. (1986). Process, not product: Less than meets the eye. TESOL Quarterly, 20, 141 144.
Hyland, F. (1998). The impact of teacher written feedback on individual writers. Journal of Second
Language Writing, 7, 255 286.
James, C. (1998). Errors in language learning and use. Essex, UK: Addison Wesley Longman.
Johnson, K. (1988). Mistake correction. ELT Journal, 42, 89 96.
Leki, I. (1990). Coaching from the margins: Issues in written response. In B. Kroll (Ed.), Second
language writing: Research insights for the classroom ( pp. 57 68). Cambridge: Cambridge
University Press.
Leki, I. (1991). The preferences of ESL students for error correction in college-level writing classes.
Foreign Language Annals, 24, 203 318.
Leow, R. P. (1997). Attention, awareness, and foreign language behaviour. Language Learning, 47,
467 505.
Levenston, E. A. (1978). Error analysis of free comparison: The theory and the practice. Indian
Journal of Applied Linguistics, 4 (1), 1 11.
Long, M. H. (1996). The role of linguistic environment in second language acquisition. In W. C.

298

D.S. Qi, S. Lapkin / Journal of Second Language Writing 10 (2001) 277303

Ritchie, & T. K. Bhatia (Eds.), Handbook of language acquisition: Vol. 2. Second language
acquisition ( pp. 413 468). New York: Academic.
Mackey, A., & Philp, J. (1998). Conversational Interaction and second language development: Recasts, responses and red herrings? Modern Language Journal, 82, 338 356.
Mantello, M. A. (1996). Selective error correction in intermediate extended French writing programs:
A comparative study of reformulation and coded feedback. Unpublished masters thesis, Ontario
Institute for Studies in Education, University of Toronto, Toronto, ON.
Olson, D. R. (1994). The world on paper: The conceptual and cognitive implications of writing and
reading. Cambridge: Cambridge University Press.
Paulus, T. M. (1999). The effect of peer and teacher feedback on student writing. Journal of Second
Language Writing, 8, 265 289.
Polio, C., Fleck, C., & Leder, N. (1998). If I only had more time: ESL learners changes in
linguistic accuracy on essay revisions. Journal of Second Language Writing, 7, 43 68.
Poulisse, N. (1999). Slips of the tongue: Speech errors in first and second language production.
Amsterdam: John Benjamins.
Qi, D. S. (1998). An inquiry into language-switching in second language composing. Canadian
Modern Language Review, 54, 413 435.
Robinson, P. (1995). Review article: Attention, memory, and the noticing hypothesis. Language
Learning, 45, 283 331.
Sanaoui, R. (1984). The use of reformulation in teaching writing to FSL students. In L. Young (Ed.),
Carleton papers in applied language studies ( pp. 139 147). Ottawa, ON: Carleton University.
Schmidt, R. (1990). The role of consciousness in second language learning. Applied Linguistics, 11,
129 158.
Schmidt, R. (1994). Deconstructing consciousness in search of useful definitions for applied linguistics. AILA Review, 11, 11 26.
Schmidt, R., & Frota, S. (1986). Developing basic conversational ability in a second language: A case
study of an adult learner of Portuguese. In R. Day (Ed.), Talking to learn: Conversation in second
language acquisition ( pp. 237 326). Rowley, MA: Newbury House.
Semke, H. D. (1984). The effects of the red pen. Foreign Language Annals, 17, 195 202.
Sharwood Smith, M. (1981). Consciousness raising and the second language learner. Applied Linguistics, 2, 159 168.
Sharwood Smith, M. (1991). Speaking to many minds: On the relevance of different types of language
information for the L2 learner. Second Language Research, 7, 118 132.
Sharwood Smith, M. (1993). Input enhancement in instructed SLA: Theoretical bases. Studies in
Second Language Acquisition, 15, 165 179.
Silva, T. (1990). Second language composition instruction: Developments, issues, and directions in
ESL. In B. Kroll (Ed.), Second language writing ( pp. 11 23). Cambridge: Cambridge University Press.
Silva, T. (1993). Toward an understanding of the distinct nature of L2 writing: The ESL research and
its implications. TESOL Quarterly, 27, 657 677.
Swain, M. (1985). Communicative competence: Some roles of comprehensible input and comprehensible output in its development. In S. M. C. Gass, & C. G. Madden (Eds.), Input in second
language acquisition ( pp. 235 253). Rowley, MA: Newbury House.
Swain, M. (1995). Three functions of output in second language learning. In G. Cook, & B.
Seidhofer (Eds.), Principles and practice in applied linguistics ( pp. 125 144). Oxford: Oxford
University Press.
Swain, M. (1998). Focus of form through conscious reflection. In C. Doughty, & J. Williams (Eds.),
Focus on form in classroom second language acquisition ( pp. 64 81). Cambridge: Cambridge
University Press.
Swain, M., & Lapkin, S. (1995). Problems in output and the cognitive processes they generate: A step
towards second language learning. Applied Linguistics, 16, 371 391.
Swain, M., & Lapkin, S. (2000). What two learners notice in their reformulated writing, what they

D.S. Qi, S. Lapkin / Journal of Second Language Writing 10 (2001) 277303

299

learn from it, and their insights into the process. Paper presented at the annual conference of the
Canadian Association of Second Language Teachers (mimeo).
Swain, M., & Lapkin, S. (2001). Focus on form through collaborative dialogue: Exploring task effects.
In M. Bygate, P. Skehan, & M. Swain (Eds.), Researching pedagogic tasks: Second language
learning, teaching and testing ( pp. 99 118). London: Longman.
Thornbury, S. (1997). Reformulation and reconstruction: Tasks that promote noticing. ELT Journal,
51, 326 335.
Truscott, J. (1996). The case against grammar journal correction in L2 writing classes. Language
Learning, 46, 327 369.
Truscott, J. (1998). Noticing in second language acquisition: A critical review. Second Language
Research, 14, 103 135.
Wilberg, T. (1987). One-to-one. Hove: Language Teaching Publications.
Zamel, V. (1985). Responding to student writing. TESOL Quarterly, 19, 195 202.
Zamel, V. (1988). The author responds to comments on Vivian Zamels Recent research on writing
pedagogy. TESOL Quarterly, 22, 520 524.
Zhang, S. (1995). Reexamining the affective advantage of peer feedback in the ESL writing class.
Journal of Second Language Writing, 4, 209 222.

Appendix A. The writing prompt

Balas, R., & Rice, D. Quest-ce que se passe, Second Edition. Copyright 1984 by Houghton Mifflin
Company. Used with permission.

300

D.S. Qi, S. Lapkin / Journal of Second Language Writing 10 (2001) 277303

Appendix B
B.1. Wus draft text
Its ten oclock at the weekend. Twenty minutes ago, detective jack received a
phone call from the famous bull fighter Johns home. He said that his niece has
stolen his money from his house and was shoot by him by accident.
When the detective Jack and his colleague come to his house, Johns niece
Mike is dead and laying on the floor with the money in his hand in the dining
room. There is a gun around him and a box is sitting beside him. His parents are
so sad and his uncle and aunty in law are so angry. The table clock was falling
down the floor. The servant is helping Mikes mother.
Bull fighter John who has earned a lot of money tell the detectives that his
brother family come to visit him today. His brother is old and retire and his niece
is twenty years old and doesnt work at the moment. They had a great dinner at
8 oclock after that they went to the meeting room and talk about the family.
About 9:20 Mike get out the meeting room and go to the washroom. About ten
minute later, they heard some sound from bedroom. and then John left to take a
look. He found Mike get out of his bed room and carrying a box. Also there is
some cash in his hand. He realize that hes stealing. Then he try to stop him doing
that. But Mike is so crazy and fight his uncle and try to escape. John followed
him to the dinning room and get a gun on the wall and ask him not to move. Mike
try to get the gun. In the mess, the gun fight and Mike is shoot by accident.
When all the people heard the gun fighting and get out from meeting room to the
dining room. Mikes parents are so sorry and sad about what happened. Then John
let the servant to call the detective. Mike is dead. The things like that happened
every day. So I think the money is the source of crime. Even between the relative.
B.2. Reformulated version of Wus written draft
It is ten oclock at the weekend. Twenty minutes ago, Detective Jack received
a phone call from Johns home. John is a famous bullfighter. He told the detective
that he had shot his nephew by accident when his nephew was stealing his money
in his house.
When detective Jack and his colleague arrived at the scene, Johns nephew
Mike was already dead. Now the body lies on the floor of the dining room with
some money in the left hand. Near the right hand is a suitcase. A gun is sitting on
the floor, just beside the left side of the body. A table clock has been knocked
over and has fallen to the floor. Mikes parents look miserable and his uncle and
aunt look very angry. A housemaid is trying to console Mikes mother.
Bullfighter John who has earned a lot of money tells the detectives that his
brothers family came over to visit him today. His brother is old and retired and his
nephew is twenty years old and does not have a job at the moment. They had a
great dinner at 8 oclock this evening. After that, they went to the living room and

D.S. Qi, S. Lapkin / Journal of Second Language Writing 10 (2001) 277303

301

had a good chat. At around 9:20, Mike left the living room for the washroom.
About ten minutes later, they heard some noise from the bedroom. John decided to
go there and take a look. When John got there, he saw Mike coming out of his
bedroom, carrying a suitcase in his right hand and some cash in his left hand. He
realized that Mike was stealing from him. So John tried to stop him. But Mike was
so crazy at that time. He fought his uncle and at the same time tried to escape. John
chased Mike to the dining room. When John found that he was unable to stop his
nephew, he took a gun off the wall and asked him not to move. Mike tried to grab
the gun. In the midst of the fight, the gun was fired and Mike was shot by accident.
When all the people heard the gunshot, they rushed to the dining room. John
let his maid call the police. Now Mikes parents are so sad about what has
happened to their son. Mike is dead. Things like this happen almost every day. I
believe that money is a source of crime, even among relatives.
B.3. Wus revised text
Its ten oclock at the weekend. Twenty minutes ago, detective Jack received a
phone call from the famous bullfighter Johns home. He said that his nephew has
stolen his money from his house and was shot by him by accident.
When the detective Jack and his colleague come to his house, Johns nephew
Mike is dead and laying on the floor with the money in his left hand in the dining
room. There is a gun around him and a suitcase is sitting beside him. His parents
are so sad and his uncle and aunty-in-law are so angry. The table clock was
knocked down to the floor. The servant is helping console Mikes mother.
Bullfighter John who has earned a lot of money tells the detectives that his
brother family came to visit him that day. His brother was old and retired and his
nephew is twenty years old and doesnt work at the moment. They had a great
dinner at 8 oclock. Then they went to the meeting room and had a chat. About
9:20 Mike went to the washroom. About ten minutes later, they heard some noise
from Johns bedroom. John left to take a look. He found Mike coming out from
his bedroom and carrying a suitcase. Also there was some cash in his left hand.
He realize that Mikes stealing from him. He try to stop Mike. But Mike is so
crazy that he fought his uncle and tried to escape. John followed him to the
dinning room, grab a gun on the wall and ask him not to move. Mike try to get
the gun. In the mess, the gun was shot and Mike was shot by accident.
When all the people heard the gunshot, they ran out from the meeting room to
the dining room. Mikes parents are so sorry and sad about what happened. Then
John let the servant to call the detective. Mike is dead. Things like that happened
every day. So I believe the money is the source of crime, even among relatives.
B.4. Sus draft text
There is a rich family. The old woman who she has three sons. The older is
short guy He was married. And hes wife is a chill weman. She felt angry about

302

D.S. Qi, S. Lapkin / Journal of Second Language Writing 10 (2001) 277303

hers husband get a little money. The dead guy is the old women second son. He
staied at home no job lives with his mother. The third one who sit at the table he
drank. He didnt care about anything.
One day, they have a dinner and meeting because their mother want to share
hers wealth after dinner the old mother think they need share all of the family
wealth. Because of the old mother is too old. So after dinner. They discussed
about how to share theirs money. The sisters-in-law was very angry because
she though they just got a little bit. So they asked their mother about a big bag
where is it? No boday can find it. it is belong theirs father who was die many
years ago. So they asked the second son, but he doesnt want to take it out.
Finelly he took it out. and suddenly the light off. There is dark. Then they heard
a gun bomn.
They found the second son lie down. nobody found who kill him. So the
witness called the police. In a short time police came in. They told about the
situation then the smoking police saied: The third son killed then second son. So
theirs mother was very sad. I think that is a sad family story. Right now, some
parents hey doesnt want to share theirs wealth to give their children. They think
about if they keep their money, they can keep theirs children. If they shared their
money, they loose their children.
B.5. Reformulated version of Sus written draft
There is a rich family. The mother of the family is an old woman. She has
three sons. Her eldest son is the short guy in the picture. He is married. His wife is
a greedy woman. She feels unhappy because she thinks her mother-in-law has not
given her husband enough money. The dead guy is the old womans second son.
He did not have a job and lived with his mother. The third son is the one who sits
at the table. He is drunk. He does not seem to care about anything.
That day, the family had dinner together at the mothers home. After dinner,
the mother wanted to have a family meeting to discuss how to divide her money
among her three sons. She wanted to share all her money with her sons because
she felt she was too old. So after dinner, the family discussed how to divide the
money. The eldest sons wife was very angry because she thought she and her
husband only got a little bit. So she and her husband asked the mother where the
suitcase was. They were unable to find it. The suitcase belonged to their father
who died many years ago. So they asked the second son, but he did not want to
get it. Finally he brought it out. Suddenly the light went out. The room turned
dark. Then a gunshot was heard.
After the light was turned on, the second son was found lying dead on the
floor. It was not known who killed him. Someone called the police. A little while
later, two detectives arrived. The family reported what had happened before the
gunshot. Then the detective who smoked a pipe said: The third son killed the
second son. The mother was very sad. I think this is a sad family story.
Nowadays, some parents do not want to share their wealth with their children.

D.S. Qi, S. Lapkin / Journal of Second Language Writing 10 (2001) 277303

303

They believe if they keep their money, they can keep their children. If they share
their money with their children, they can lose their children.
B.6. Sus revised text
There is a rich family. The old woman who she has three sons. The older is
short guy He is married. And hes wife is a greedy weman. She felt angry about
hers husband get a little money. The dead guy is the old mothers second son. He
staied at home no job lives with his mother. The third one who is sitting at the
table he drunk. He didnt care anything.
One day, they have a family meeting because their mother wants to share hers
wealth. After dinner, they discussed their family wealth. Because of the old
mother is too old. So after dinner. They discussed how to share theirs money.
The sister-in-law was very angry because she just got a little bit. So they asked
their mother about a big bag where it is? It is belong theirs father who was died
many years ago. So they asked the second son, but he didnt want to take it out.
Finelly he took it out. Suddenly the light off. There is a dark. Then they heard a
gun sound.
When the light turned on, they found the second son lay down. Nobody found
who killed him. Someone called the police. In a short time two police came in.
They told him about the situation. The policeman with a pipe saied: The third
son killed the second son. So the old mother was very sad. I think that this a sad
family story. These days, some parents they dont want to share theirs wealth with
their children. They think if they keep their money, they can keep theirs children.
If they shared their money. They lost their children.