Introduction

The purpose of this paper is to consider methodological procedures for teaching tasks. These are of two basic kinds. Firstly, there are those procedures relating to how the tasks specified in a task-based syllabus can be converted into actual lessons. Secondly, there are procedures relating to how the teacher and learners are to participate in the lessons. This paper will address only the first of these.

The design of a task-based lesson involves consideration of the stages or components of a lesson that has a task as its principal component. Various designs have been proposed (e.g. Estaire and Zanon 1994; Lee 2000; Prabhu 1987; Skehan 1996; Willis 1996). However they all have in common three principal phases, which are shown in Figure 1. These phases reflect the chronology of a task-based lesson. Thus, the first phase is 'pre-task' and concerns the various activities that teachers and students can undertake before they start the task, such as whether students are given time to plan the performance of the task. The second phase, the 'during task' phase, centres around the task itself and affords various instructional options, including whether students are required to operate under time-pressure or not. The final phase is 'post-task' and involves procedures for following-up on the task performance. Only the 'during task' phase is obligatory in task-based teaching. Thus, minimally, a task-based lesson consists of the students just performing a task. Options selected from the 'pre-task' or 'post-task' phases are non-obligatory but, as we will see, can serve a crucial role in ensuring that the task performance is maximally effective for language development.

Figure 1: A framework for designing task-based lessons

Access to a clear framework for a task-based lesson is of obvious advantage to both teachers and learners. Richards (1996) shows how many experienced teachers adhere to a maxim of planning ('Plan your teaching and try to follow your plan') while Numrich (1996) reports on how novice teachers feel the 'need to be creative and varied in teaching'. A framework such as the one outlined in Figure 1 caters to both needs. It provides a clear structure for a lesson and it also allows for creativity and variety in the choice of options in each phase.

The pre-task phase

The purpose of the pre-task phase is to prepare students to perform the task in ways that will promote acquisition. Lee (2000) describes the importance of 'framing' the task to be performed and suggests that one way of doing this is to provide an advance organizer of what the students will be required to do and the nature of the outcome they will arrive at. Dornyei (2001) emphasizes the importance of presenting a task in a way that motivates learners. Like Lee, he sees value in explaining the purpose and utility of the task. This may be especially important for learners from traditional 'studial' classrooms; they may need to be convinced of the value of a more 'experiential' approach. Dornyei also suggests that task preparation should involve strategies for whetting students' appetites to perform the task (e.g. by asking them to guess what the task will involve) and for helping them to perform the task. Strategies in this latter category are discussed below.

Skehan (1996) refers to two broad alternatives available to

These alternatives can be tackled procedurally in one of four ways; (1) supporting learners in performing a task similar to the task they will perform in the during-task phase of the lesson, (2) asking students to observe a model of how to perform the task, (3) engaging learners in non-task activities designed to prepare them to perform the task or (4) strategic planning of the main task performance. We will consider each in some detail.

Performing a similar task

The use of a 'pre-task' was a key feature of the Communicational Teaching Project (Prabhu 1987). It was carried out as a whole-class activity with the teacher and involved the learners in completing a task of the same kind as and with similar content to the main task. Thus, it served as a preparation for performing the main task individually. For example, if the main task involved working out a class timetable from the timetables of individual teachers, then the pre-task would be the same but with different information in the teachers' timetables.

Prabhu explains that the pre-task was conducted through interaction of the question-and-answer type. The teacher was expected to lead the class step-by-step to the expected outcome, to break down a step into smaller steps if the learners encountered difficulty and to offer one of more parallels to a step in the reasoning process to ensure that mixed ability learners could understand what was required. The teacher was provided with a lesson plan that included (1) the pre-task and (2) a set of graded questions or instructions together with parallel questions to be used as needed. When implemented in the classroom, the plan results in a 'pedagogic dialogue'. Prabhu emphasises that the pre-task was not a 'demonstration' but 'a task in its own right'. It is clear from this account that the 'pre-task' serves as a mediational tool for the kind of 'instructional conversation' that sociocultural theorists advocate. The teacher, as an expert, uses the pre-task to scaffold learners' performance of the task with the expectancy that this 'other-regulation' facilitates the 'self-regulation' learners will need to perform the main task on their own.

Providing a model

An alternative is to ask the students to observe a model of how the task can be performed without requiring them to undertake a trial performance of the task (see Aston (1982) for an early example of such an approach). Minimally this involves presenting them with a text (oral or written) to demonstrate an 'ideal' performance of the task. Both Skehan (1996) and Willis (1996) suggest than simply 'observing' others perform a task can help reduce the cognitive load on the learner. However, the model can also be accompanied by activities designed to raise learners' consciousness about specific features of the task performance - for example, the strategies that can be employed to overcome communication problems, the conversational gambits for holding the floor during a discussion or the pragmalinguistic devices for performing key language functions. Such activities might require the learners to identify and analyze these features in the model texts. Alternatively, they might involve pre-training in the use of specific strategies. Nunan (1989) lists a number of learning strategies (e.g. 'Learning to live with uncertainty' and 'Learning to make intelligent guesses') that students can be taught to help them become 'adaptable, creative, inventive and above all independent' (p. 81) and thus more effective performers of a task. However, the effectiveness of such strategy training remains to be convincingly demonstrated.

Non-task preparation activities

There are a variety of non-task preparation activities that teachers can choose from. These can centre on reducing the cognitive or the linguistic demands placed on the learner. Activating learners' content schemata or providing them with background information serves as a means of defining the topic area of a task. Willis (1996) provides a list of activities for achieving this (e.g. brainstorming and mind-maps). When learners know what they are going to talk or write about they have more processing space available for formulating the language needed to express their ideas with the result that the quantity of the output will be enhanced and also fluency and complexity. Recommended activities for addressing the linguistic demands of a task often focus on vocabulary rather than grammar, perhaps because vocabulary is seen as more helpful for the successful performance of a task than grammar. Newton (2001) suggests three ways in which teachers can target unfamiliar vocabulary in the pre-task phase; predicting (i.e. asking learners to brainstorm a list of words related to the task title or topic), cooperative dictionary search (i.e. allocating different learners words to look up in their dictionary), and words and definitions (i.e. learners match a list of words to their definitions). Newton argues that such activities will 'prevent the struggle with new words overtaking other important goals such as fluency or content-learning' when learners perform the task. However, there is always the danger that pre-teaching vocabulary will result in learners treating the task as an opportunity to practise pre-selected words. In the case of task-supported teaching this can be seen as desirable but in the case of task-based teaching it can threaten the integrity of the task.

Strategic planning

Finally, learners can be given time to plan how they will perform the task. This involves 'strategic planning' and contrasts with the 'online planning' that can occur during the performance of the task. It can be distinguished from other pre-task options in that it does not involve students in a trial performance of the task or in observing a model. However, it may involve the provision of linguistic forms/strategies for performing the task but a distinction can still be drawn between the non-task preparation procedures described above and strategic planning, as the former occur without the students having access to the task they will be asked to perform while strategic planning involves the students considering the forms they will need to execute the task workplan they have been given.

There are a number of methodological options available to teachers who opt for strategic planning. The first concerns whether the students are simply given the task workplan and left to decide for themselves what to plan, which typically results in priority being given to content over form, or whether they are given guidance in what to plan. In the case of the latter option, the guidance may focus learners' attention on form or content or, as in Sangarun's (2001) study, form and content together. Skehan (1996) suggests that learners need to be made explicitly aware of where they are focussing their attention - whether on fluency, complexity or accuracy. These planning options are illustrated in Figure 2. Here the context is a task involving a balloon debate (i.e. deciding who should be ejected from a balloon to keep it afloat). The guidance can also be 'detailed' or 'undetailed' (Foster and Skehan 1996). The examples in Figure 2 are of the undetailed kind. Skehan (1998) gives an example of detailed planning for a personal task involving asking someone to go to your house to turn off the oven that you have left on. This involved instructions relating to planning content (e.g. 'think about what problems your listener could have and how you might help her') and language (e.g. 'think what grammar you need to do the task'). These options do not just provide for variety in planning activities; they also enable the teacher to channel the learners' attention onto different aspects of language use. For example, Foster and Skehan (1996) found that when students were given detailed guidance they tended to prioritise content with resulting gains in complexity when they performed the task.

Figure 2: Options for strategic planning (based on Foster and Skehan 1999).

Another option concerns the amount of time students are given to carry out the pre-task planning. Most of the research studies that have investigated this kind of planning have allocated between 1 and 10 minutes. An effect on fluency was evident with very short periods of planning in some studies but longer was needed for an effect on complexity (Skehan 1998 suggests 10 minutes is optimal). Finally, planning can be carried out individually, in groups, or with the teacher.

Summary and final comment

In these four ways, teachers can help to create conditions that will make tasks work for acquisition. As Skehan (1998) points out, they serve to introduce new language that the learners can use while performing the task, to mobilize existing linguistic resources, to ease processing load and to push learners to interpret tasks in more demanding ways. However, it is not yet possible to 'fine tune' learners' performance of a task through selecting specific pre-task options. At best, all that the research to date has demonstrated is the likely effects of some of the procedures referred to above. Important questions remain unanswered. For example, we do not know whether task preparation that involves an actual performance of the task is more or less effective than preparation that involves just observation. Nor is it clear to what extent linguistic priming subverts the 'naturalness' of a task resulting in teaching of the present-practice-produce (PPP) kind. Only in the case of strategic planning do we have some idea of how the different options affect task performance.

The during-task phase

The methodological options available to the teacher in the during-task phase are of two basic kinds. First, there are various options relating to how the task is to be undertaken that can be selected prior to the actual performance of the task and thus planned for by the teacher. These will be called 'task-performance options'. Second, there are a number of 'process options' that involve the teacher and students in on-line decision making about how to perform the task as it is being completed.

Task performance options

We will consider three task performance options that have figured in the research to date. The first of these options concerns whether to require the students to perform the task under time pressure. The teacher can elect to allow students to complete the task in their own time or can set a time limit. Lee (2000) strongly recommends that teachers set strict time limits. This option is important because it can influence the nature of the language that students' produce. Yuan and Ellis (2002) found that giving students unlimited time to perform a narrative task resulted in language that was both more complex and more accurate in comparison to a control group that was asked to perform the same task under time pressure. The students used the time at their disposal to monitor and reformulate their utterances. Interestingly, the opportunity to plan on-line produced a different effect from the opportunity to engage in strategic planning, which led to greater fluency and complexity of language. It seems, then, that if teachers want to emphasize accuracy in a task performance, they need to ensure that the students can complete the task in their own time. However, if they want to encourage fluency they need to set a time limit.

The second task performance option involves deciding whether to allow the students access to the input data while they perform a task. In some tasks access to the input data is built into the design of a task (e.g. in Spot the Difference, Describe and Draw, or many information gap tasks). However, in other tasks it is optional. For example, in a story retelling/recall task the students can be permitted to keep the pictures/ text or be asked to put them on one side as they narrate the story. This can influence the complexity of the task, as tasks that are supported by pictures and texts are easier than tasks that are not. Joe (1998) reports a study that compared learners' acquisition of a set of target words (which they did not know prior to performing the task) in a narrative recall task under two conditions - with and without access to the text. She found that the learners who could see the text used the target words more frequently, although the difference was evident only in verbatim use of the words not generated use (i.e. they did not use the target words in original sentences). Joe's study raises an important question. Does borrowing from the input data assist acquisition? The term 'borrowing' in this context comes from Prabhu (1987).

He defines it as 'taking over an available verbal formulation in order to express some self-initiated meaning content, instead of generating the formulation from one's own competence' (p. 60). Prabhu distinguishes borrowing from 'reproduction' where the decision to 'take over' a sample of a language is not made by the learner but by some external authority (i.e. the teacher of the text book). Borrowing is compatible with task-based teaching but reproduction is not. Prabhu sees definite value in borrowing for maintaining a task-based activity and also probable value in promoting acquisition. Certainly, from the perspective of sociocultural theory, where learning occurs through 'participation', borrowing can be seen as contributing directly to acquisition.

The third task performance option consists of introducing some surprise element into the task. Skehan and Foster(1997) illustrate this option. They asked students to complete a decision-making task that required them to decide what punishment should be given to four criminals who had committed different crimes. At the beginning of the task they were given information about each criminal and the crime he/she had committed. Half way through the task the students were given further information of a surprising nature about each criminal. For example, the initial information provided about one of the criminals was as follows:

However, this study failed to find that introducing such a surprise had any effect on the fluency, complexity or accuracy of the learners' language. This does not mean that this option is of no pedagogic value, as requiring learners to cope with a surprise serves as an obvious way of extending the time learners spend on a task and thus increases the amount of talk. It may also help to enhance students' intrinsic interest in a task.

Process options

Process options differ from task performance options in that they concern the way in which the discourse arising from the task is enacted rather than pedagogical decisions about the way the task is to be handled. Whereas performance options can be selected in advance of the actual performance of the task, process options must be taken in flight while the task is being performed.

The teacher's on-line decision about how to conduct the discourse of a task reflect his/her 'theory-in-use' (Schon 1983) and 'practical knowledge' (Eraut 1994). On the learners' part, they reflect the language learning beliefs (Horwitz 1987) they bring to the classroom and, more particularly, to a specific task. How teachers and learners conduct a task will be influenced, to a large extent, by their prior experiences of teaching and learning and their personal definitions of the particular teaching-learning situation. Thus, the options described below are primarily descriptive, reflecting an internal rather than external perspective (Ellis 1998) on the methodology of task-based teaching.

A common assumption of task-based teaching is that the texts, the discursive practices and the social practices of the classroom (Breen 1998) that are constructed by and through a task resemble those found in non-pedagogic discourse. To achieve this, however, is no mean feat, especially if the teacher is directly involved in the performance of the task. As Breen points out, the 'texts' of lessons (i.e. the actual language produced by the participants) are typically teacher-centred with learners 'not actually required to do much overt or explicit discursive work' (p. 123), while the 'discursive practices' (i.e. the means by which the text are produced) 'construct learners as primarily responsive and seemingly fairly passive participants in the discourse' (p. 124) and the 'social practices' (i.e. the organisational and institutional circumstances that shape the texts and discursive practices) are directed at the avoidance of 'social trouble'. Task-based teaching, however, seeks the converse - texts that are learner-centred, discursive practices that encourage the learner to actively engage in shaping and controlling the discourse, and social practices that are centred on allowing and resolving social trouble. This poses a problem, which teachers need to address.

Figure 3 contrasts two sets of classroom processes. The first set corresponds to the classroom behaviours that are typical of a traditional form-focussed pedagogy where language is treated as an object and the students are required to act as 'learners'. The second set reflects the behaviours that characterize a task-based pedagogy, where language is treated as a tool for communicating and the teacher and students function primarily as 'language users' (Ellis 2001). Thus, which set of behaviours arise is crucially dependent on the participants' orientation to the classroom and to their motives for performing an activity.

Figure 3: Stereotypical classroom processes in traditional form-focussed pedagogy and task- based pedagogy

Two questions arise. The first concerns what the participants in a task need to do to ensure that the interactions they engage in manifest the processes described in column B in Figure 3. Implicit in this question is an acknowledgement of the importance of these processes for task-based instruction. The second question, however, challenges this assumption by asking whether in fact these processes are criterial of task-based pedagogy and whether, minimally, they need to be complemented by processes from column A.

It has often been pointed out (see, for example, Nunan 1987) that the processes described in column B are a rarity even in classrooms where the teacher claims to be teaching communicatively. The main reason for this lies in the difficulty teachers and students have in achieving the required orientation. As Goffman (1981) has pointed out, classrooms are governed by an 'educational imperative' which dictates the kind of discourse that arises. It is for this reason that teachers and students find it difficult to consistently orient to language as a tool and to adopt the role of language users when they both know that the raison-d'etre for their being together is to teach and learn the language. In effect, task-based teaching calls for the classroom participants to forget where they are and why they are there and to act in the belief that they can learn the language indirectly through communicating in it rather than directly through studying it. This is asking a lot of them, especially if the social practices the participants bring to the classroom belong to a pedagogy of transmission rather than of interpretation (Barnes 1976). It is probably easier to achieve when students are interacting among themselves, without the teacher being present, as the greater symmetry of social roles this affords leads naturally to the kinds of risk-taking behaviour required of a task-based pedagogy (Pica 1987). This is one reason why pair and group work are seen as central to task-based teaching.

However, even when the participants in a task are oriented to treat language as a tool and to function as language users, the text of the task may disappoint, manifesting few of the characteristics facilitative of acquisition. Seedhouse (1999) has pointed out that the characteristics of task-based interaction do not always match those described in Figure 3. He illustrates how in some tasks the turn-taking system is conspicuously constrained, there is a tendency for students to rely on topic-comment constructions where verbal elements are omitted (a feature also noted in pidgins) and to produce highly indexicalised utterances. An even greater limitation in task-based interaction, according to Seedhouse, is the minimalization that characterizes some task-based interactions. This is illustrated in the extract below where the students were required to complete and label a geometric figure:

L1: What?
L2: Stop.
L3: Dot?
L4: Dot?
L5: Point?
L6: Dot?
LL: Point, point, yeh.
L1: Point?
L5: Small point.
L3: Dot
(From Lynch 1989, p. 124; cited in Seedhouse 1999).

Here all the utterances but one consist of a single word. Clearly, such interactions do not help the 'stretch' learners' interlanguages, one of the stated goals of task-based pedagogy (Nunan 1989). Seedhouse suggests that such limited interactions arise because 'learners appear to be so concentrated on completing the task that linguistic forms are treated as a vehicle of minor importance' (p. 154). In other words, the very nature of a task (i.e. the fact it is directed at accomplishing a specified outcome) may result in a restricted variety of communication.

Seedhouse overstates this limitation of tasks. First, it is possible to argue that the restricted nature of the talk shown in the extract above is well suited to the students' purpose. Second, the nature of the interaction depends crucially on the design characteristics of tasks and procedures for implementing them. Thus, richer varieties of communication characterized by more complex language use, are achievable if, for example, students are asked to perform open tasks with divergent goals and are given the opportunity to plan their performance before hand. Nevertheless, Seedhouse's critique needs to be addressed. Clearly, teachers need to monitor their students' performance of a task carefully, examining to what extent the processes described in Figure 3 arise and, crucially, whether the interactions manifest the minimalized and pidgin-like uses of language Seedhouse illustrates. The information obtained from such monitoring can be used to inform decisions about what tasks and procedures to use in subsequent tasks. In this way, teachers can build up a fund of experience of the task characteristics and methods of implementation that will ensure the kinds of interactions hypothesized to promote acquisition. Thus, the solution, to the problem Seedhouse identifies lies not in attempting to manipulate process options directly, which may well be impossible without imperilling the 'taskness' of the task, but through careful selection from the pre-task options and the performance options described above.

Where Seedhouse questions whether the kinds of behaviours shown in Figure 3 are achievable in task-based teaching, others have challenged whether they constitute appropriate goals for interaction in a classroom. Cullen (1998) has pointed out that the classroom context constitutes a communicative environment in its own right that is distinct from the communicative contexts of the world outside and on these grounds has challenged the basis for assessing the communicativeness of classroom discourse. In effect, then, Cullen disputes the assumption that underlies task-bask pedagogy - that classrooms need to replicate the kind of communicative behaviour found outside the classroom. He illustrates how 'what appears to be non-communicative teacher talk is not necessarily so in the classroom context' (p. 183) with an extract from an English lesson in Egypt. This interaction is teacher-led, is full of display questions, includes feedback that is form-focussed and contains a lot of echoing - all processes associated with a traditional form-focussed pedagogy. However, Cullen argues that in the context of the classroom, the interaction can be considered 'communicative' in that the entire sequence manifests a focus on message content, the teacher's questions are carefully structured, the feedback is clear and the use of echoing serves to ensure that the students' attention is not lost. He claims that the discourse is pedagogically effective because the teacher has successfully combined the role of 'instructor' and 'interlocutor'. Arguably, this is what a task-based pedagogy needs to strive for. How might it be achieved?

One way is by incorporating a focus on form into the performance of the task. Ellis, Basturkmen and Loewen (2001) report this can be achieved in either responding focus-on-form episodes, where one of the participants, usually the teacher, responds to a student utterance containing an error, or in initiating episodes, where either the teacher or a student elects to take time out from the exchange of message content to attend briefly to form, usually by means of a direct query about a specific form. Such attention to form differs from that arising in lessons of the traditional, focus-on-forms kind because, for, as Wilberg (1987) notes, 'the content is dictated by the student, the form only by the teacher' (p. 27). It also differs in another way. As Prabhu (1987) points out, correction during a task is 'incidental' rather than 'systematic' in nature. In incidental correction, only 'tokens' are addressed (i.e. there is no attempt to generalize the type of error), it is seen by the participants as 'a part of getting on with the activity in hand, not as a separate objective' (p. 63) and, crucially, it is transitory. Prabhu excludes preventive or pre-emptive attention to form but, as Ellis, Basturkmen and Loewen's study shows this too can be 'incidental'.

Teachers can employ both implicit and explicit techniques to achieve this focus on form. These techniques can be used when some kind of communication problem arises (as occurs in the negotiation of meaning) or they can be used when the teacher chooses to abandon his/her role as a language user momentarily in order to function as an instructor (i.e. to negotiate form rather than meaning). Teachers can play a very direct role by initiating this negotiation but they can also intervene to support a process that students have started for themselves, a technique that involves 'nudging' the learners towards a solution . Teachers can also allow or even encourage students to use the same techniques themselves - for example, by accepting and responding to students' queries about form.

Figure 4 describes some of the techniques that can be used by the task participants. Evidence from research (Ellis, Basturkmen and Loewen 2001) indicates that the use of these techniques, even when quite frequent, need not detract from the primary focus on message, which is the defining characteristic of a task. Thus, they serve as important process options for reconciling the roles of 'instructor/learner' on the one hand and 'interlocutor/language user' on the other. Furthermore, they potentially enhance the acquisitional value of a task by inducing noticing of linguistic forms that lie outside or at the edges of students' current interlanguages.

Figure 4: Implicit and explicit techniques for focussing on form during a task

To sum up, it is clear that process options cannot be prescribed. Nevertheless, it is possible to identify, in broad terms, the kinds of processes that the participants in a task performance need to strive for. These are:
1. Discourse that is essentially 'conversational' in nature (i.e. as described in column B of Figure 3). Such discourse can include 'instructional conversations'.
2. Discourse that encourages the explicit formulation of messages.
3. Opportunities for students to take linguistic risks.
4. Occasions where the task participants focus implicitly and/or explicitly on specific linguistic forms.
5. Shared goals for the task.
6. Effective scaffolding of the participants' efforts to communicate in the L2.

The post-task phase

The post-task phase affords a number of options. These have three major pedagogic goals; (1) to provide an opportunity for a repeat performance of the task, (2) to encourage reflection on how the task was performed, and (3) to encourage attention to form, in particular to those forms that proved problematic to the learners when they performed the task.

Repeat performance

Several studies (e.g. Bygate 1996 and 2001; Lynch and Maclean 2000) indicate that when learners repeat a task their production improves in a number of ways (e.g. complexity increases, propositions are expressed more clearly, and they become more fluent). A repeat performance can be carried out under the same conditions as the first performance (i.e. in small groups or individually) or the conditions can be changed. One interesting possibility examined by Skehan and Foster (1997) is that of requiring students to carry out the second performance publicly. As their study examined the 'threat' of such a requirement on learners' initial performance of the task, it technically constituted a during-task option. However, if students are not told to repeat the task publicly until after they have completed the first performance, it becomes a post-task option. There has been no research comparing the learner production that results from a second performance carried out under 'private' conditions, as in the initial performance, and publicly. Clearly, performing a task in front of the class increases the communicative stress (Candlin 1987) placed on the learner and thus can be predicted to lead to a reduction in fluency and complexity. However, it is not without value if students need experience in using English in front of an audience, as, for example, might be the case with foreign academics training to give oral presentations in the L2. Public performance is likely to encourage the use of a more formal style and thus may push learners to use the grammaticalised resources associated with this style.

Reflecting on the task

Willis (1996) recommends asking students to present a report on how they did the task and on what they decided or discovered. She considers this 'the natural conclusion of the task cycle' (p. 58). The teacher's role is to act as a chairperson and to encourage the students. The reports can be oral or written. Willis' examples make it clear that the reports should primarily focus on summarising the outcome of the task. However, it would also be possible to ask students to reflect on and evaluate their own performance of the task. For example, they could be invited to comment on which aspect of language use (fluency, complexity or accuracy) they gave primacy to and why, how they dealt with communication problems, both their own and others, and even what language they learned from the task (i.e. to report what Allwright (1984) has called 'uptake' [1]). Students could also be invited to consider how they might improve their performance of the task. Encouraging students to reflect on their performance in these ways may contribute to the development of the metacognitive strategies of planning, monitoring and evaluating, which are seen as important for language learning (O'Malley and Chamot 1990).

There is also a case for asking students to evaluate the task itself. Such information will help the teacher to decide whether to use similar tasks in the future or look for a different type. I have suggested that student-based evaluations of tasks can be carried out quickly and effectively using simple questionnaires (see Ellis 1997a for an example).

Focussing on forms

Once the task is completed, students can be invited to focus on forms, with no danger that in so doing they will subvert the 'taskness' of the task. It is for this reason that some methodologists recommend reserving attention to form to the post-task phase of the lesson. Willis (1996), for example, sees the primary goal of the 'task component' as that of developing fluency and promoting the use of communication strategies. The post-task stage is needed to counter the danger that students will develop fluency at the expense of accuracy. In part, this is met by asking students to report on their performance of the task, as discussed above, but it can also be achieved by a direct focus on forms. It should be noted, however, that this is the not the position taken in this paper. I have emphasised that a focus on form constitutes a valuable during-task option and that it is quite compatible with a primary focus on message content, which is the hallmark of a task. Furthermore, in some tasks (e.g. consciousness raising tasks) a linguistic feature is made the topic of the task. Attention to form, in one way or another, can occur in any (or indeed all) of the phases of a task-based lesson. In the pre-task and post-task phases the focus will be on forms while in the during-task phase it will be on form, to invoke Long's (1991) distinction .

Two obvious methodological questions arise regarding attention to form in the post-task phase. The first concerns which forms should be attended to. The answer is fairly obvious; teachers should select forms that the students used incorrectly while performing the task or 'useful' or 'natural' forms (Loshcky and Bley Vroman 1993) that they failed to use at all. In other words, teachers should seek to address errors or gaps in the students' L2 knowledge. Consideration also needs to be given to how many such forms a teacher should seek to address. Should the focus be placed on a single form that is treated intensively or a number of forms that are treated extensively? Both approaches are warranted and are reflected in the various options described below.

The second question concerns how the target forms should be dealt with. There is a whole range of options available to the teacher. It should be noted however that in many cases the effectiveness of these options has not been investigated.

1. Review of learner errors
While the students are performing a task in groups, teachers can move from group to group to listen in and note down some of the conspicuous errors the students make together with actual examples. In the post-task phase, the teacher can address these errors with the whole class. A sentence illustrating the error can be written on the board, students can be invited to correct it, the corrected version is written up, and a brief explanation provided. Lynch (2001) offers an interesting way of conducting a post-task analysis, which he calls 'proof-listening'. This involves three cycles based on repeated playing of a recording of the task. First, the students who did the task review and edit their own performance. Second, the recording is replayed and other students are invited to comment, correct or ask questions. Finally, the teacher comments on any points that have been missed.

2. Consciousness-raising tasks
CR-tasks constitute tasks in their own right and, therefore, can be used as the main task in a lesson. But they can also be used as follow-up tasks to direct students to attend explicitly to a specific form that they used incorrectly or failed to use at all in the main task. Willis and Willis (1996) and Ellis (1997b) offer descriptions of the various options that are available for the design and implementation of CR tasks. When used as follow-up tasks, CR tasks can profitably take their data from recordings of the students' performance of the task. For example, students might be presented with a number of their own utterances all illustrating the same error and asked to identify the error, correct the sentences and work out an explanation.

3. Production practice activities
An alternative or addition to CR tasks is to provide more traditional practice of selected forms. Traditional exercise types include repetition, substitution, gapped sentences, jumbled sentences, transformation drills, and dialogues. Willis (1996; pp. 110) offers a number of more novel ideas. The value of such production practice activities has been called into question (see, for example, VanPatten 1996) on the grounds that they have no direct effect on learners' interlanguage systems. However, they may help learners to automatize forms that they have begun to use on their own accord but have not yet gained full control over.

4. Noticing activities
A number of suggestions have been made for developing noticing activities as a follow-up to a task performance. Fotos (1994) used dictation exercises that had been enriched with the target structures that students had tackled initially in CR tasks to examine whether the subjects in her study subsequently attended to the structures. She found that they did so quite consistently. Lynch (2001) recommends getting students to make transcripts of an extract (90-120 seconds) from their task performance as a method for inducing noticing. After transcribing, they are required to make any editing changes they wish. The teacher then takes away the word-processed transcripts and reformulates them. The next day the students are asked to compare their own edited transcript with the teacher's reformulated version. In a study that investigated this procedure, Lynch found that students cooperated effectively in transcribing, made a number of changes (most of which resulted in accurate corrections of linguistic forms), and engaged in both self- and other-correction. Lynch also analysed the types of changes the students made, noting that the majority involved grammatical corrections, 'editing' slips (i.e. removal of redundancies, literal repetitions and dysfluencies) and 'reformulation' (i.e. changes directed at more precise expressions). Finally, Lynch comments that there was plenty left for the teacher to do after the students had made their changes.

Using the framework for designing a lesson

What constitutes the main activity of a lesson is largely a matter of perception and therefore, to some extent at least, arbitrary. For example, Prabhu (1987) talks of a 'pre-task' and a 'task'. The former is carried out between the teacher and the whole class. The latter is performed by the students working individually. But, such a sequence of activities could easily be described in terms of 'task' and 'post-task'. Indeed, Prabhu's 'pre-task' involves the type of activity that most task-based methodologists would consider to belong to the during-task phase of a lesson. Similarly, a sequence of activities consisting of 'task' and 'post-task' where the latter involves the kind of transcribing activity advocated by Lynch could also be described in terms of 'pre-task' and 'task', if the transcribing activity is viewed as the main activity.

However, this caveat does not detract from the usefulness of the design framework described above as a basis for planning task-based lessons. Teachers need to decide first on the basic format of the lesson. Minimally, it will consist of the during-task phase but it can also include either or both of a pre-task and post-task phase. Once the basic structure of the lesson has been decided, the specific option(s) to be included in each phase of the lesson can be considered. The description of the process options for implementing the during-task phase of the lesson also provides a guide for the navigation of the actual task and for the teacher's ongoing monitoring of the task performance.

Notes:
1. Allwright's (1984) use of 'uptake' differs from that of researchers who have investigated corrective sequences in classroom discourse. Allwright uses the term to refer to what learners are able to explicitly report having learned as a result of participating in a lesson.

References

Allwright, D. 1984. Why don't learners learn what teachers teach? The interaction hypothesis. In D. Singleton and D. Little (eds.). Language Learning in Formal and Informal Contexts (pp. 3-18). Dublin: IRAL.

Aston, G. 1982. Interact. Oxford: Modern English Publications.

Barnes, D. 1976. From Communication to Curriculum. Harmondsworth: Penguin.

Breen, M. 1998. Navigating the discourse: On what is learned in the language classroom. In W. Renandya and G. Jacobs (eds). Learners and Language Learning. Singapore: SEAMEO Regional Language Centre.

Bygate, M. 1996. Effects of task repetition: Appraising the development of second language learners. In J. Willis and D. Willis (eds). Challenge and Change in Language Teaching. Oxford: Heinemann.

Bygate, M. 2001. Effects of task repetition on the structure and control of oral language. In M. Bygate, P. Skehan and M. Swain (eds). (pp. 23-48).

Bygate, M., P. Skehan, and M. Swain. (eds). 2001. Researching Pedagogic Tasks, Second Language Learning, Teaching and Testing. Harlow: Longman.

Candlin, C. 1987. Towards task-based language learning. In C. Candlin and D. Murphy (eds.). Language Learning Tasks. Englewood Cliffs N.J.: Prentice Hall International.

Candlin, C. and D. Murphy. 1987. Language Learning Tasks. Englewood Cliffs N.J.: Prentice Hall International.

Cullen, R. 1998. Teacher-talk and the classroom context. ELT Journal 52: 179-187.

Dornyei, Z. 2001. Motivational Strategies in the Classroom. Cambridge: Cambridge University Press.

Ellis, R. 1997a. SLA Research and Language Teaching. Oxford: Oxford University Press.

Ellis, R. 1997b. Explicit knowledge and second language pedagogy. In L. Van Lier and D. Corson (eds.). Encyclopedia of Language and Education Vol 6: Knowledge about Language. Dordrecht: Kluwer Academic.

Ellis, R. 1998. Discourse control and the acquisition-rich classroom. In W. Renandya and G. Jacobs (eds). Learners and Language Learning (pp. 145-171). Singapore: SEAMEO.

Ellis, R. 2001. Focussing on form: Towards a research agenda. In W. Renandya and N. Sunga (eds). Language Curriculum and Instruction in Multicultural Societies (pp. 123-144). Singapore: SEAMEO Regional Language Centre.

Ellis, R., H. Basturkmen and S. Loewen. 2001. Learner uptake in communicative ESL lessons. Language Learning 51: 281-318.

Eraut, M. 1994. Developing Professional Knowledge and Competence. London: Falmer.

Estaire, S. and J. Zanon. 1994. Planning Classwork: A Task Based Approach. Oxford: Heinemann.

Foster, P. and P. Skehan. 1996. The influence of planning and task type on second language performance. Studies in Second Language Acquisition 18: 299-323.

Foster, P. and P. Skehan. 1999. The influence of planning and focus of planning on task-based performance. Language Teaching Research 3: 215-247.

Fotos, S. 1994. Integrating grammar instruction and communicative language use through grammar consciousness-raising tasks. TESOL Quarterly 28: 323-351.

Goffman, E. 1981. Forms of Talk. Oxford: Basil Blackwell.

Horwtitz, E. 1987. Surveying student beliefs about language learning. In Wenden, A. and Rubin, J. (eds). Learner Strategies in Language Learning. Englewood Cliffs, NJ: Prentice Hall.

Joe, A. 1998. What effects do text-based tasks promoting generation have on incidental vocabulary acquisition. Applied Linguistics 19:357-77.

Lee, J. 2000. Tasks and Communicating in Language Classrooms. Boston: McGraw-Hill.

Long, M. 1991. Focus on form: A design feature in language teaching methodology. In K. de Bot, R. Ginsberg, and C. Kramsch (eds.). Foreign Language Research in Cross-Cultural Perspective (pp. 39-52). Amsterdam: John Benjamin.

Loscky, L. and R. Bley-Vroman. 1993. Grammar and task-based methodology. In Crookes, G. and Gass, S. (eds.) Tasks in a Pedagogical Context: Integrating Theory and Practice. Clevedon: Multilingual Matters.

Lynch, T. 1989.Researching teachers: Behaviour and belief. In C. Brumfit and R. Micthell (eds). Research in the Language Classroom. London: Modern English Publications and the British Council.

Lynch, T. 2001. Seeing what they meant: Transcribing as a route to noticing. ELT Journal 55: 124-32.

Lynch, T. and J. Maclean. 2000. Exploring the benefits of task repetition and recyling for classroom language learning. Language Teaching Research 4: 221- 250

Lynch, T. and J. Maclean. 2001. Effects of immediate task repetition on learners' performance. In M. Bygate, P. Skehan and M. Swain (eds). (pp. 99-118).

Newton, J. 2001. Options for vocabulary learning through communication tasks. ELT Journal 55: 30-37.

Numrich, C. 1996. On becoming a language teacher; Insights from diary studies. TESOL Quarterly 30: 131-153.

Nunan, D. 1987. Communicative language teaching: Making it work. ELT Journal 41: 136-45.

Nunan, D. 1989. Designing Tasks for the Communicative Classroom. Cambridge:Cambridge University Press.

O'Malley, J. and A. Chamot. 1990. Learning Strategies in Second Language Acquisition. Cambridge: Cambridge University Press.

Pica, T. 1987. Second language acquisition, social interaction, and the classroom. Applied Linguistics

Prabhu, N.S. 1987. Second Language Pedagogy. Oxford: Oxford University Press.

Richards, J. 1996. Teachers' maxims in language teaching. TESOL Quarterly 30: 281-296.

Sangarun, J. 2001. The effects of pre-task planning on foreign language performance. Doctoral thesis, University of Toronto, Canada.

Schon, D. 1983. The Reflective Practitioner. New York: Basic Books.

Seedhouse, P. 1999. Task-based interaction. ELT Journal 53: 149-156. Oxford University Press.

Skehan, P. 1996a. A framework for the implementation of task-based instruction. Applied Linguistics 17: 38-62.

Skehan, P. 1998a. A Cognitive Approach to Language Learning. Oxford: Oxford University Press.

Skehan, P. and P. Foster. 1997. Task type and task processing conditions as influences on foreign language performance. Language Teaching Research 1: 185-211

Wilberg, T. 1987. One-to-One. Hove: Language Teaching Publications.

Willis, D. and J. Willis. 1996. Consciousness-raising activities in the language classroom. In J. Willis and D. Willis (ed). A Framework for Task-Based Learning. Harlow: Longman.

Willis, J. 1996. A Framework for Task-Based Learning. Harlow: Longman.

Willis, J. and D. Willis. (eds.). 1996. Challenge and Change in Language Teaching. Oxford: Heinemann.

Yuan, F. and Ellis, R. 2002. The effects of pre-task planning and on-line planning on fluency, complexity and accuracy in L2 oral production. Applied Linguistics