Professional Documents
Culture Documents
This article reports a case study investigating the relationship between three
teachers' stated beliefs about and practices of focus on form in intermediate
level ESL communicative lessons. Focus on form was de®ned and studied in
terms of incidental time-outs taken by students and teachers to deal with issues
of linguistic form during communicative lessons. The teachers' statements of
belief about focus on form were compared to their management of focus on
form during lessons in which all the teachers used the same communicative
task. Results showed some inconsistencies in the teachers' stated beliefs, in
particular in relation to when it is legitimate to take time out from a
communicative activity to focus on issues of form, and preferred error
correction technique. While some statistically signi®cant dierences in the
teachers' practices were re¯ected in dierences in their stated beliefs, others
were not. These results indicated a somewhat tenuous relationship between the
teachers' practices and stated beliefs regarding focus on form. It is argued that
future investigations of teachers' beliefs, especially of unplanned elements of
teaching such as focus on form, need to be based on both stated beliefs and
observed behaviours.
INTRODUCTION
The study reported in this article takes up Borg's (2003) call for investigations
that address speci®c aspects of language teaching in relation to teacher
cognition. The speci®c aspect of teaching that will be considered is focus on
form instruction. The aspect of cognition to be examined is teachers' beliefs.
We brie¯y examine these below.
Focus on form
Focus on form is a feature of communicative language teaching (CLT). CLT
can be broadly de®ned, at least in its `strong version' (Howatt 1984), as
teaching oriented primarily towards exchanging meaning or messages. It
typically involves the use of communicative tasks, de®ned by Skehan (1998:
268) as activities where: meaning is primary; there is a goal that needs to be
worked towards; the activity is outcome-evaluated; and there is a relationship
between the task and real life. Although CLT, so de®ned, is meaning-centred,
244 INCIDENTAL FOCUS ON FORM AND CLASSROOM PRACTICES
it need not preclude attention to form if this takes place in the context of
performing a communicative task. Long (1991) uses the term `focus on form'
to refer to such attention to form and has argued that it promotes language
acquisition because it enables learners to `notice' linguistic elements
(grammar, vocabulary, pronunciation, and discourse features) that might
otherwise have been missed, and because it is compatible with the learners'
internal syllabus (that is learners are free to attend to forms that are
`learnable').
Ellis (2001) has pointed out that focus on form can be either planned or
incidental. In planned focus on form, the teacher/researcher preselects a form
for attention and designs a focused communicative task that will provide
opportunities for its use. This type of focus on form has served as the subject of
a number of experimental studies (for example Doughty and Varela 1998). In
incidental focus on form, the teacher/researcher makes no attempt to
predetermine which form or forms will be attended to; rather, the focus on
form arises naturally out of the performance of a communicative task with no
pre-targeted language forms. Our concern in this paper is with incidental
focus on form.
Recent research (Lyster and Ranta 1997; Lyster 1998; Ellis et al. 2001a, b)
has documented the fact that incidental focus on form occurs frequently in
CLT. It can consist of responses to errors made by students (`reactive focus on
form') or queries raised by either the teacher or the students about a linguistic
item even though no error has occurred (`preemptive focus on form'). Our
previous research (Ellis et al. 2001a, b) has shown that incidental focus on
form episodes (FFEs) (that is, episodes in which the discourse is related to a
speci®c linguistic item) varied across classes in a number of ways, such as in
the complexity of episodes, in their linguistic focus, and in the teachers'
responses within them.
Teachers' beliefs
In recent years, there has been a shift away from a view of teachers as people
who master a set of general principles and theories developed by experts
towards a view of teaching as a thinking activity and teachers as people who
construct their own personal and workable theories of teaching (Fang 1996;
Borg 2003; Richards 1998). Concomitant with this conceptualization of
teaching as a thinking activity has been an increase in research into teachers'
beliefs (Borg 2003).
Although the concept of belief has attracted considerable research interest
in education in recent years, there is a lack of consensus about what the term
denotes (S. Borg 2003, M. Borg 2001; Woods 1996). In the present study, the
term beliefs is de®ned as statements teachers made about their ideas,
thoughts, and knowledge that are expressed as evaluations of what `should
be done', `should be the case', and `is preferable'. This de®nition is in line with
HELEN BASTURKMEN, SHAWN LOEWEN, and ROD ELLIS 245
Linde's (1980: 2) de®nition of beliefs systems as `values, that is, about what
ought to be the case' as reported in Woods (1996: 70).
It is generally acknowledged that teachers' stated beliefs play an important
role in relation to instructional practice. Burns (1992: 58) describes the `beliefs
that motivate' instructional practices in the classroom. Johnson (1992) reports
research in the teaching of literacy supporting the notion that `beliefs tend to
shape' teachers' instructional practices. M. Borg (2001) proposes that beliefs
`guide' teachers' thought and behaviour. Some writers emphasize the
mutually informing nature of the relationship (Borg 2003; Fang 1996; Pajares
1992). Pajares (1992: 307) states: `Few would argue [dispute the fact] that the
beliefs teachers hold in¯uence their perceptions and judgements, which, in
turn, aect their behaviour in the classroom.'
Researchers have made use of both observational and self-report data to try
to understand how teachers make sense of their work. Burns (1992)
investigating the beliefs of six teachers about the use of written language in
teaching beginner-level students, encouraged the teachers to re¯ect on their
lessons and was able to identify a number of themes in the teachers'
`personalised theories'. Borg (1999) attempted to get teachers to identify their
implicit theories by examining their own instructional decision-making, and
to make what was implicit in practice, explicit as theory. Breen et al. (2001)
investigated how eighteen teachers in one particular teaching situation
described their classroom practices and explained the pedagogic principles
they based these practices on. They investigated whether a particular shared
disposition among a group of teachers would be materialized through similar
or diverse practices. They found that in some cases shared practices appeared
to be driven by dierent stated principles and that in other instances the
teachers associated a shared principle with diverse practices.
Some studies have used self-report and observational data to compare the
extent to which teachers' stated beliefs are correlated with their practices. For
a review of such research in general education, see Fang (1996). Johnson
(1992) studied the ®t between the `theoretical orientations' (in terms of
methodological preferences) towards teaching ESL literacy of three teachers
and their instructional practices over a number of lessons. Some studies have
found consistency between stated beliefs and practices. For example, Johnson
found that ESL teachers holding clearly de®ned theoretical beliefs provided
instruction consistent with their theoretical orientations. Hsiao-Ching (2000)
identi®ed the stated beliefs of a seventh-grade biology teacher about gender-
related dierences in learning style and classroom participation, and found
these beliefs re¯ected in the teacher's classroom practices. However, a number
of studies have found inconsistencies between stated beliefs and practices. A
study by Borko and Niles (1982) found that teachers' stated educational
beliefs were unrelated to how they grouped students for instruction. ESL
studies of self-styled communicative teachers have shown that these teachers
may no more create conditions for `genuine communication' in their
classrooms than other traditional teachers (Nunan 1987; Kumaravadivelu
246 INCIDENTAL FOCUS ON FORM AND CLASSROOM PRACTICES
1993). Fang (1996) surveys research on this topic and concludes that studies
based on multiple sources of data, including classroom observation, have
tended to show inconsistencies between stated beliefs and practices.
A key issue, then, is why teachers' practices sometimes contravene their
stated beliefs. A considerable body of literature now exists documenting the
role of context, and particularly constraints, that can hinder teachers from
implementing their stated beliefs (Borg 2003; Fang 1996). For example, a
study by Graden (1996) investigated the stated beliefs and practices of
teachers of reading instruction. The study found that the teachers had a
number of clearly articulated beliefs about reading instruction (for example,
the bene®ts of using the target language in reading practice) that generally
guided their practice. However, mismatches between their stated beliefs and
their practices were evident in observations of their lessons. The teachers
explained these in terms of constraints, such as time and lack of appropriate
materials. Woods (1996) reports a case study investigation of the thinking and
practices of experienced ESL teachers in university settings in Canada. The
study revealed that the teachers' interpreted classroom events through their
evolving networks of beliefs, assumptions, and knowledge (BAK). Woods'
study related `hotspots' in the data (elements which seemed incoherent) to
the dynamic and evolving nature of the teachers' BAK over time (1996: 248)
and showed that these hotspots came to be resolved through developments in
teaching experience and expertise.
An apparent mismatch can also be explained in terms of con¯icting belief
systems. The teachers in Graden's (1996) study reported that, in the main, it
was their perceptions of poor student performance and lack of motivation
that prevented them from providing instruction consistent with their stated
beliefs. Graden concluded that the teachers subordinated their beliefs about
reading instruction to their beliefs about the motivational needs of their
students. We can draw similar conclusions from studies that have speci®cally
investigated the reasons for teachers departing from their lesson plans (see
review in Borg 2003). Studies by Richards (1998) and Bailey (1996) found
that student factors most often prompted teachers' to make modi®cations.
For example, the teachers in Bailey's study were prompted to make changes
by their perceptions of the need to promote students' involvement and to
encourage more equal distribution of participation in the discourse in the
classroom.
It may be the case that teachers draw on dierent sources of knowledge
when planning for teaching, as opposed to when they make on-line
decisions during teaching. Eraut (1994) and Ellis (1997) distinguish between
teachers' technical and practical knowledge. Technical knowledge denotes
the body of explicit ideas derived by a profession from deep re¯ection or
empirical investigation. These are general ideas that can be applied to many
cases. Johnson's (1992) study, mentioned above, found consistency
between teachers' stated beliefs and their planned practices, such as the
types of activities the teachers included in their lessons. Practical knowledge
HELEN BASTURKMEN, SHAWN LOEWEN, and ROD ELLIS 247
RESEARCH QUESTIONS
Our research questions were as follows:
1 How do teachers practise incidental focus on form?
2 What beliefs do teachers hold about incidental focus on form? To what
extent are these beliefs internally consistent?
3 To what extent are teachers' beliefs about incidental focus on form
congruent with their observed practices?
METHOD
Teaching context and participants
This study focuses on the practices of three teachers. It is part of a larger study
(Loewen 2002) investigating focus on form in intact classes with dierent
teachers all from the same private language school in Auckland, New Zealand.
The full data set consisted of 48 lessons from 12 teachers, from 40 to 65
minutes in length. In recognition of the fact that individuals have their own
interpretations of what constitutes communicative teaching, the researcher
simply asked to observe lessons that the individual teachers themselves
considered to be communicative. No eort was made by the researcher to
guide the teachers in their choice of lesson plans. Following the observations,
a second researcher interviewed the teachers to elicit their stated beliefs of
focus on form in communicative teaching.
In the school there is a set of `communicative tasks' prepared in-house that
248 INCIDENTAL FOCUS ON FORM AND CLASSROOM PRACTICES
teachers can elect to use. During the original observations, the researcher
noticed that three teachers used the same communicative task (the Prisoner
Task as shown in Appendix A). This provides the opportunity to investigate
the stated beliefs and practices of these three teachers in relation to their use
of the same instructional material. That is, it allows us to compare teachers'
practices while keeping the task variable constant. The observational data for
this study, then, consists of one lesson for each of the three teachers, using the
same task, and in the same school.
The three teachers were all male, native speakers of English. Table 1 shows
their experience in teaching ESL, the length of time they had been at the
school, as well as their ESL quali®cations and the class they were teaching.
The teachers are listed in order of teaching experience.
Data collection
The study involved a combination of observational and self-report data. The
observational component involved observation of the teachers' lessons. The
self-report component comprised statements of beliefs about focus on form
elicited from teachers through the use of in-depth interviews, cued response
scenarios and stimulated recall. These instruments will now be described in
further detail.
HELEN BASTURKMEN, SHAWN LOEWEN, and ROD ELLIS 249
Observations of lessons
One researcher was present as a non-participant observer in the naturally
occurring communicative lessons of the teachers at the school described
above. Each lesson was audio-recorded using a wireless clip-on microphone
that was attached to the teacher. This arrangement allowed the researcher to
record all teacher±student interaction during the class, including talk between
the teacher and individuals, pairs, small groups as well as the whole class. The
lessons were transcribed in their entirety.
Our previous work on focus on form served as a basis for selecting the
speci®c aspects of focus on form to address using these introspective
techniques These aspects included the type of episode (reactive and
student-initiated or teacher-initiated pre-emptive episodes), the linguistic
focus of episodes (grammar, vocabulary, spelling, or pronunciation), the
reason for focusing on forms (because of problems in understanding the
message or because of concerns about language forms per se, that is accuracy),
and how the episodes are managed once they are in progress (the length of
episodes and the type of response given by the teacher).
(a) In-depth interview. The ®rst part of the interview was based roughly
around a semi-structured interview protocol. This involved some closed item
lead-in prompts that aimed to focus the respondents' attention on the subject
at hand. The main items were open-ended. Open-ended items have the added
advantage of eliciting ideas expressed in the respondents' own words
(Oppenheimer 1992). The principle that guided the design of the interview
was to avoid direct questioning in favour of indirect items. For example, we
did not ask the teachers what a good communicative lesson should be like, but
rather asked them to recall a successful communicative lesson they had
recently taught. See Appendix B for sample questions from the in-depth
interviews.
(b) Cued response scenarios. One problem in the design of some studies has
been that an attempt to assess beliefs has been made without instantiating the
context. To reduce this problem, the teachers were presented with a set of
scenarios of typical classroom situations and asked to comment on them.
These represented vignettes of critical incidents related to focus on form that
teachers might encounter in the classroom. The scenarios were derived from
our previous studies of focus on form in the communicative classroom (Ellis et
al. 2001a, b). They were based on typical focus on form episodes we had
observed in the CLT of a number of teachers at similar institutions. For
example, one scenario showed a student-initiated episode while another
showed a reactive (error correction) episode. One scenario was concerned
with pronunciation while others were concerned with grammar or
vocabulary. An example is shown in Appendix C. This scenario was devised
to represent a student-initiated enquiry about a vocabulary item. It re¯ected a
common type of focus on form episode, as demonstrated by our previous
research. However, it was not feasible to specify in the scenarios all the
individual and contextual factors which may aect teachers' responses, such
as a teacher's knowledge of the particular students involved and the mood of
the class.
The teachers were asked to comment on what they felt they should do in
these situations, because we wanted to ®nd out what they felt as desirable
behaviour in accordance with our de®nition of belief. We did not ask them
what they could do, as that would have invited them to list a number of
possibilities. All three teachers were shown the same episodes and this
technique enabled us to compare their responses to the same events.
HELEN BASTURKMEN, SHAWN LOEWEN, and ROD ELLIS 251
Data analysis
Two of the researchers ®rst analysed the observational and self-report data on
their own and then compared their analyses. The teachers themselves were
not involved in analysing or commenting on the analysis of the data.
Although presenting teachers with an analysis and eliciting their comments
can provide a further source of validation for the analysis (see Woods 1996), it
was not possible to incorporate this into the design of the study.
(a) Identifying focus on form episodes (FFE). The recordings of the lessons were
transcribed and focus on form episodes (FFEs) were identi®ed. In line with
our wider research study, an FFE was de®ned as `the discourse from the point
where the attention to linguistic form starts to the point where it ends, due to
a change in topic back to message or sometimes another focus on form' (Ellis
et al. 2001a: 294). In order to check the reliability of the identi®cation of FFEs,
the data were coded by a second rater and an agreement rate of 89.9 per cent
was reached.
(b) Coding of characteristics. The FFEs were then coded, based on Ellis,
Loewen, and Basturkmen (1999) by the researchers for the characteristics
shown in Table 2. To illustrate this scheme, a number of focus on form
episodes will now be discussed.
Example 1
* The teacher's initiation of the FFE parallels the options of provide or elicit described
in teacher response.
254 INCIDENTAL FOCUS ON FORM AND CLASSROOM PRACTICES
Example 2
Example 3
RESULTS
The results are reported in two parts. In each part, the teachers are listed in
order of teaching experience, beginning with the most experienced teacher.
First, the teachers' focus on form practices are described in terms of statistical
information relating to the various characteristics of focus on form episodes.
Second, the stated beliefs of each teacher, as re¯ected in the introspective
data, are reported qualitatively. Finally the extent to which the teachers'
beliefs matched their practices was assessed.
Steve 63 31 .49
Mark 58 26 .45
Rick 40 50 1.25
from teacher to teacher. Table 4 shows that Mark and Steve had similar ratios
of roughly 0.5 FFEs per minute, while Rick had a higher ratio of 1.25 FFEs per
minute.
Table 5 reveals the characteristics of focus on form that were present in each
class. Regarding type of FFE, we see that Rick had a signi®cantly higher
number of Reactive FFEs (82 per cent) than did the other two teachers.
Conversely, Steve participated in a signi®cantly higher number of student-
initiated FFEs (51.6 per cent) than did Mark and Rick. When examining the
complexity of the FFEs, the chi-square analysis indicated that Steve had a
signi®cantly higher frequency of complex FFEs (64.5 per cent) than the other
teachers. Although the complexity residuals for Mark do not reach the 2.0
threshold, they are very close (at 1.9), indicating a trend towards a
signi®cantly higher frequency of simple FFEs for Mark than for the other
teachers. For Response moves, the chi-square analysis indicated that Rick had
a signi®cantly higher frequency of Elicitation moves (26 per cent) than did the
other teachers. Again, Mark's residuals do not quite meet the 2.0 threshold,
but they come close (1.8), indicating a trend towards a signi®cantly higher
frequency of Provide moves for Mark than for the other teachers. With regard
to Linguistic Focus and Source, Table 5 indicates that there were no signi®cant
dierences among the three teachers.
Steve
Below are the main themes to emerge from the analysis of Steve's beliefs
about focus on form:
Table 5: FFE characteristics
Teacher Type Complexity Response Linguistic Focus*** Source
Reactive Student- Teacher- Simple Complex Provide** Elicitation Vocabulary Grammar Pronunciation Code Message
initiated initiated*
Recast Inform
Steve 13 16 2 11 20 9 20 2 13 10 6 25 6
(41.9%) (51.6%) (6.5%) (35.5%) (64.5%) (29%) (64.5%) (6.5%) (44.8%) (34.5%) (20.7%) (80.6%) (19.4%)
Residual ±3.8 3.8 ±2.4 2.4 1.6 ±1.6 .9 ±.6 ±.4 ±.4 .4
Mark 16 6 4 18 8 15 10 1 12 5 7 19
(61.5%) (23.1%) (15.4%) (69.2%) (30.8%) (57.7%) (38.5%) (3.8%) (50.0%) (20.8%) (29.2%) (73.1%) (26.9%)
Residual .2 ±.2 1.9 ±1.9 1.8 ±1.8 1.4 ±2.1 .8 ±1.6 1.6
Rick 41 6 3 28 22 29 8 13 13 24 10 45
(82%) (12%) (6%) (56.0%) (44.0%) (58%) (16%) (26.0%) (27.7%) (51.1%) (21.3%) (90.0%) (10.0%)
Residual 3.3 ±3.3 .5 ±.5 ±3.0 3.0 ±2.0 2.3 ±.4 1.8 ±1.8
* Teacher-initiated FFEs were not included in the chi-square analysis due to low cell counts.
** Recasts and Informs were combined as Provides for the chi-square analysis.
*** Spelling FFEs (n = 7) were excluded from this table.
HELEN BASTURKMEN, SHAWN LOEWEN, and ROD ELLIS 259
. Steve believed that the role of the teacher in communicative lessons should
be as a resource for students to consult if needed and not as a director. He
recalled one particularly successful lesson in which just a few students had
turned up to class. The lesson had been truly communicative, he felt,
because he had what he termed a `natural' conversation. He described this
as a discussion in which the students had led the discourse. It was
successful he felt because it was `directed by the students not by me'.
. Steve believed self-correction was the best form of error correctionÐ`self
correction is my favoured mode' (response to scenario 4). He felt his role
should be to indicate that an error had occurred, but not actually to repair
it. He reported the techniques he used, such as the `thumbs down' sign.
. Steve did not emphasize any particular aspects of language for focusing on;
however, he believed that special attention should be paid to language
forms that had been taught in previous lessons.
. Steve said that he was not keen on student-initiated language queries, but
felt he should answer such questions if asked.
. Steve had nothing to say about the need for focus on form to be
unobtrusive.
Steve was inconsistent about why to focus on form. He stressed several times
in the in-depth interview that communicative lessons should not be about
accuracy and that he was `very suspicious of error correction'. He talked of
wanting students to feel safe, expressed his view that learning takes time, and
pointed out that the research literature reports that correction of errors does
not lead to immediate acquisition. He stated a preference for refraining from
interfering during communicative activities unless there was a problem with
message comprehension. In his response to scenario 4, for example, he stated
that if diculties with language forms `were not impeding a ¯uent exchange,
I'd wait'. Yet Steve also stated his belief that errors related to target (that is
taught) structures should be corrected: `Target structures are fresh in the
students' minds and should be worked on while hot' (response to scenario 6).
Congruence between Steve's stated beliefs and practices was evident in that
the fact that, in accordance with his conviction that the teacher should serve
as a resource during a communicative task, in the Prisoner Task the students
were working on their own for approximately two-thirds of the time.
Otherwise there were a number of incongruences between stated beliefs
and practices. Table 5 shows that the proportion of student-initiated FFEs in
Steve's lesson was high (51.6 per cent), even though Steve's beliefs about this
issue were mixed. Also, in contrast to his stated beliefs, Steve rarely used
elicitation in response to FFEs (6.5 per cent) and mostly used informs (64.5
per cent). Finally, and in contrast to Steve's stated belief that he should refrain
from interfering in communicative activities unless there was a problem in
message comprehension, few FFEs in Steve's lesson arose from problems in
message comprehension (19.4 per cent).
260 INCIDENTAL FOCUS ON FORM AND CLASSROOM PRACTICES
Mark
The following are the main themes to emerge from the analysis of Marks'
beliefs about focus on form:
Rick
The main themes to emerge from the analysis of Rick's beliefs about focus on
form were as follows:
Rick expressed inconsistent beliefs about why to focus on form. He stated that
a legitimate time-out from communicative activities was when there was
incorrect use of a previously taught grammatical form. However, in relation to
one scenario he stated that the only legitimate time-outs from communicative
activities were when there was a problem with meaning. He stated that he
HELEN BASTURKMEN, SHAWN LOEWEN, and ROD ELLIS 263
would not interrupt the communicative activity, on the grounds that `the
success of interaction is more important to the students than accuracy'. Yet
elsewhere he talked of success in terms of students using the target structure
accurately and of the need to interrupt them if they do not do so. Rick also
said several times that he believed he should avoid stopping the ¯ow of a
communicative activity. Thus, there appeared to be a tension, or indeed
con¯ict, for Rick, between trying to get the students to use the targeted
language items correctly, and not stopping the ¯ow of a communicative
activity. Rick also had inconsistent beliefs about students raising questions
about language forms during communicative activities. At one point he stated
he thought he should try to `get around it. I'd prefer them to carry on talking'
(response to stimulated recall episode 2). However, elsewhere (response to
scenario 2), he stated he believed he should supply linguistic information
even if this was not essential for meaning, since `students look to the teacher
for guidance and recommendations of appropriate language'.
In two particular respects Rick's stated beliefs were congruent with his
practices. Table 5 shows that the preponderance of FFEs in Rick's lesson
concerned grammatical structures (51.1 per cent), a ®nding that was
congruent with his stated beliefs. Secondly he talked a good deal (if somewhat
inconsistently) about accuracy, which was in line with the high number of
reactive FFEs (82 per cent) in his lesson that con®rmed his preparedness to
address learner errors.
In other respects, however, there was incongruence. Table 5 shows that, in
contrast to his somewhat inconsistent beliefs regarding whether to address
code or communicative problems, the great majority of FFEs in Rick's lesson
(90 per cent) concerned code, and very few episodes arose as a result of
diculties in message comprehension. Also, although he stated that he should
attend to student-initiated problems, there were only a few student-initiated
episodes (12.8 per cent) in his lesson. Rick's stated belief about the purpose of
communicative activities (that is to provide opportunities for practising
grammatical structures) was also not clearly re¯ected in his focus on form
practices. There was no evidence of his using the Prisoner Task to focus on
pre-targeted structures; in fact he addressed a range of grammatical items (for
example, gender pronouns, conditionals, plurals, prepositions) in much the
same way as the other teachers. Further incongruences between stated beliefs
and practices were evident in how errors were corrected. Rick spoke strongly
and often of his belief that any focusing on form in a communicative activity
should be done unobtrusively, but, in spite of this, he manifested a fairly high
rate of complex episodes in his lesson. Table 5 shows that 44 per cent of the
episodes were complex (that is, longer and thus more obtrusive). Table 5 also
shows that, contrary to his stated belief, Rick used recasts (an indirect
correction technique) more often (58 per cent) than other error correction
types.
264 INCIDENTAL FOCUS ON FORM AND CLASSROOM PRACTICES
Summary
In order to highlight the main areas where the teachers' beliefs and practices
were congruent and incongruent, Table 6 summarizes their beliefs and
practices in relation to the main features of focus on form. There were three
aspects of focus on form (type, complexity, and response) where the teachers'
diered amongst themselves, and two aspects (linguistic focus and source)
where there were no dierences. There are some points at which the teachers'
practices clearly re¯ect their beliefs (for example their lack of enthusiasm for
student initiated FFEs) and several where they do not. A good example
concerns the source of FFEs. All three teachers focused on form predomi-
nately for its own sake (that is, where the motivation for the episode was the
`code', that is, source = code) whereas in fact all three stated that form should
be focused on only when there was a problem of comprehension (that is,
source = message). There are also several cases where one teacher's practice
diered from the other two, and this was re¯ected in this teacher's stated
beliefs (for example Mark used more recasts than either Rick or Steve and also
stated a preference for this type of response). But there were also cases where
a teacher's practice diered from the other two even though this did not
correspond to his stated belief (for example, Steve predominantly dealt with
student errors by providing linguistic information, whereas he stated that
student self-correction was preferable).
DISCUSSION
The ®rst research question concerned whether the teachers' practices
regarding focus on form diered. Overall, the results showed a number of
striking similarities in the three teachers' management of focus on form in the
Prisoner Task. However, there were also a number of dierences (see Loewen
2003). These dierences cannot be readily explained by contextual factors,
since, as stated earlier, the three teachers were teaching students of the same
pro®ciency level in the same institution and were using the same commun-
icative task. A more likely explanation for the dierences, then, is to be found
in the personal teaching styles of the three teachers and the belief systems that
underlie these styles, including their notions of what communicative
language teaching entailed. These dierences in practice are evident in the
summary provided in Table 6.
The second research question addressed the nature of the teachers' beliefs
about focus on form, and the extent to which their beliefs were internally
consistent. The teachers all expressed very de®nite beliefs about how to focus
on form. In a number of respects their beliefs were the same. For example, all
three teachers were unenthusiastic about students initiating FFEs, on the
grounds that this would interrupt the communicative ¯ow of the lesson. All
three teachers stated that it was important not to attend to form unless this
was needed to address a problem of understanding. Finally, they all expressed
HELEN BASTURKMEN, SHAWN LOEWEN, and ROD ELLIS 265
Table 6: cont.
FFE feature Teacher Observed Beliefs Comments
preferences
people's behaviour and are based primarily on their practical knowledge. The
two sets of beliefs may or may not be compatible and an individual may or
may not be aware of any incompatibility between the two. One teacher,
Steve, expressed his awareness of some incompatibility. In the stimulated
recall he said, `And it probably shows, I don't know if there is some
inconsistency here on my part or not when I'm looking at it [that is reading
the transcript]. Yeah I think that's probably what was going on.' How teachers
resolve such incongruences remains uncertain. Possibly, with experience,
they are able to reduce the mismatches between their espoused theories and
theories in use by proceduralizing their technical knowledge. Possibly, they
eradicate the mismatches by bringing their espoused theories more in line
with their practical theories, which they develop through experience of `cases'
of what works for them. However, this remains an issue in need of further
study.
The present research centred on a phenomenon (focus on form) that arose
incidentally as part of the process of accomplishing the task. The fact that we
found tenuous links between the teachers' stated beliefs and their practice of
focus on form might be because the links between stated beliefs and incidental
behaviours are weaker than those between stated beliefs and planned
behaviours (as evident in lesson plans and instructional materials). Pajares
(1992) has argued that inferences about teachers' beliefs should be based on
assessments of what teachers say, intend to do, and actually do. We would
agree but add that this may be particularly important for an understanding of
teachers' thinking in relation to on-line decision making in the classroom,
such as incidental focus on form. Enquiry into teachers' beliefs of unplanned
elements of teaching needs to be based on both stated beliefs and observed
behaviours.
Final version received January 2003
already. She was married, but her marriage broke up after she went to prison. She has
a mother who is very sick. Her father is dead. She has a large family and many nieces
and nephews whom she is very close to. They visit her regularly. She has continued to
study medicine in prison so she keeps up to date. She is very positive about the future
and thinks she should be allowed to practise medicine again once she comes out of
prison. She thinks she has paid for her mistake.
Episode 1
In this episode, a student makes a linguistic error and you respond to it.
T: have you ever betrayed China
S: no, never
T: never
S: yeah I'm a patriost
T: a patriot
HELEN BASTURKMEN, SHAWN LOEWEN, and ROD ELLIS 271
S: yeah,
S: not like this, like this
T: yes, a patriot
1. Can you tell me what this is about?
2. Is this how you prefer to deal with linguistic errors? If not ± why not?
NOTE
1 We are grateful to an anonymous reviewer for drawing our attention to this limitation.
REFERENCES
Argyris, C. and D. A. Schon. 1974. Theory in Ellis, R. 1997. SLA Research and Language
Practice, Increasing Professional Effectiveness. Teaching. Oxford: Oxford University Press.
San Fransisco: Jossey-Bass. Ellis, R. 2001. `Introduction: Investigating
Bailey, K. L. 1996. `The best laid plans: form-focused instruction' in R. Ellis (ed.):
Teachers' in-class decisions to depart from Form-Focused Instruction and Second Lan-
their lesson plans' in K. M. Bailey and guage Learning. Malden, MA: Blackwell,
D. Nunan (eds): Voices from the Language pp. 1±46.
Classroom: Qualitative Research in Second Lan- Ellis, R., S. Loewen, and H. Basturkmen.
guage Education. Cambridge: Cambridge Uni- 1999. Focussing on Form in the Classroom.
versity Press, pp. 15±40. (Institute of Language Teaching and Learn-
Borg, M. 2001. `Teachers' beliefs,' English ing, Occasional Paper, No. 13). Auckland,
Language Teaching Journal 55/2: 186±8. New Zealand: University of Auckland.
Borg, S. 1999. `Teachers' theories in grammar Ellis, R., H. Basturkmen, and S. Loewen.
teaching,' English Language Teaching Journal 2001a. `Learner uptake in communicative
53/3: 157±67. ESL lessons,' Language Learning 51: 281±318.
Borg, S. 2003. `Teacher cognition in language Ellis, R., H. Basturkmen, and S. Loewen.
teaching: A review of research on what 2001b. `Preemptive focus on form in the
teachers think, know, believe, and do,' ESL Classroom,' TESOL Quarterly 35: 407±32.
Language Teaching 36: 81±109. Ellis, R., H. Basturkmen, and S. Loewen.
Borko, H. and J. A. Niles. 1982. `Factors 2003. `Doing focus-on-form,' System 30:
contributing to teachers' decisions about 419±32.
grouping students for reading instruction,' Eraut, M. 1994. Developing Professional Know-
Journal of Reading Behavior 14: 127±40. ledge and Competence. London: Falmer.
Breen, M. P., B. Hird, M. Milton, R. Oliver, Fang, Z. 1996. `A review of research on
and A. Thwaite. 2001. `Making sense of teacher beliefs and practices,' Educational
language teaching: Teachers' principles and Research 38/1: 47±65.
classroom practices,' Applied Linguistics 22/4: Gass, S. and A. Mackey. 2000. Stimulated Recall
470±501. Methodology in Second Language Acquisition.
Burns, A. 1992. `Teacher beliefs and their Mahwah, NJ: Lawrence Erlbaum Associates.
influence on classroom practice,' Prospect 7/ Golombek, R. P. 1998. `A study of language
3: 56±66. teachers' personal practical knowledge,' Tea-
Cohen, L. and L. Manion. 1989. Research chers of English to Speakers of Other Languages
Methods in Education. London: Routledge. Quarterly 32/3: 447±64.
Doughty, C. and E. Varela. 1998. `Commun- Graden, E. C. 1996. `How language teachers'
icative focus on form' in C. Doughty and beliefs about reading instruction are
J. Williams (eds): Focus on Form in Classroom mediated by their beliefs about students,'
Second Language Acquisition. Cambridge: Foreign Language Annals 29/3: 387±95.
Cambridge University Press, pp. 114±38. Grotjahn, R. 1991. `The research programme
272 INCIDENTAL FOCUS ON FORM AND CLASSROOM PRACTICES