This paper takes as its point of departure the distinction between curriculum guidelines and syllabus accounts proposed in Candlin (1984a). The distinction drawn there between strategic and tactical planning provides a warrant for the introduction of tasks as the basis for classroom action. Such tasks, however, are not merely convenient units for the organisation of what have been called process or procedural syllabuses (Prabhu, 1983; Breen, 1984). They serve as compelling and appropriate means for realising certain characteristic principles of communicative language teaching and learning, as well as serving as a testing-ground for hypotheses in pragmatics and second language acquisition. Tasks must nonetheless be defined and their means of operationali sation explained. It will be necessary to offer ideas for their classification and their targeting. Above all, their centrality to the syllabus cannot be taken for granted without evaluating how they can be selected and sequenced in a principled fashion. What follows cannot be a definitive response to these issues. It will have been useful if it makes the point that task-based language learning is not only a means to enhancing classroom communication and acquisition but also the means to the development of classroom syllabuses. The issues raised are themselves questions which task-based language learning is especially equipped to address.

In Candlin (1984a) I argue for the planning of language learning and teaching at two levels, neither of which is at the level of syllabus, as traditionally defined, and where plan has two distinct interpretations. At the level of curriculum guidelines (see Figure 1) we would find statements about learning in general and learning of particular subject matter, indications of learning purpose and experience, targets and modes of evaluation, role relationships of teachers and learners, as well as banks of learning items and scripts with accounts of procedures for drawing on these, exploring and expanding their boundaries. In this process, carried on in the classroom, syllabuses emerge as joint plans of the teachers and learners, recording the what, the how and the why. This is the level of the tactical exploitation of strategic principle. Such planning addresses matters of what is to be done, what questions suggest themselves, what processes are most conducive to exploring the given problem, what additional information is needed, which particular actions are worthwhile. This level of everyday decision-making, co-operatively carried out, leads to three kinds of syllabus, not one. A syllabus of learning (and language-learning especially), a syllabus of content and a syllabus of actions, of what was explored and how that was accomplished. What we have, then, as Figure 1 indicates, is a dialectic process between the level of guidelines and the level of syllabuses, by means of which the accounts of classroom work can effect curriculum change. These ethnographic and ethnomethodological accounts not only

evaluate classroom action, they also provide the data for evaluating the viability of the curriculum guidelines themselves. They are formative at both levels. We may expect from such accounts information on agreed learning goals, the content chosen for work and the manner of such work, what information and what resources were needed and drawn upon, how the work was evaluated and what modes of documentation selected. In sum, an account of what went on, subject only to the constraints of Stenhouse's comment that 'there is no telling it as it is' (Stenhouse, 1975). It is at this tactical level of syllabus planning that the concept of task becomes significant, suggesting a means whereby the different syllabuses can be accounted for, as it were retrospectively. Curriculum and syllabus plans not only differ in their level of specificity and authorship, they differ in their directionality and their accountability.

Each of the types of syllabus referred to above poses its own demands on the design and operation of language learning tasks. Exploring these demands allows us also to reiterate some of the essential principles of a communicative curriculum in language teaching, which have, of course, been extensively discussed elsewhere (cf. Breen and Candlin, 1980; Breen, 1983; Johnson and Porter, 1983, among others).

It is only where we attempt accounts of learners and their learning in the language classroom that the easy assumption that 'learners differ' begins to impose considerable demands on task design. Learners after all differ in a wide variety of ways, among which we could list: their intelligence and language learning ability; their interests, needs, wants and lacks; their strategies in communicating and interpreting what others say and write; their styles and modes of learning and the rates at which they learn; their desire to participate in the management of their own learning; their need for monitoring and supervision; in their socio-cultural backgrounds and educational histories; in their attitudes to the language in question and their motivation for learning it. To this differentiation should be added our practical experience as teachers (and learners) of those conditions which enhance rather than inhibit learning, each of which will also impose its own conditions on task design. For example, if we believe that learners learn most when they are aware of what it is that they have to achieve and how their efforts are to be judged, and have been party to the specification of both. If we believe that we cannot expect each learner to learn in the same way the same things at the same rate, or that learner preference will influence both choice of learning task, its procedures and its evaluation criteria.

Characterising the development of communicative competence as the 'content' of a language teaching curriculum now requires us (following the arguments set out in Widdowson, 1983; Swain and Canale, 1980; Canale, 1983; Candlin, 1985, among others) to subsume under the term both knowledge and procedures, rather than knowledge alone, and to reassess what is meant by knowledge. As we shall see, the definition we come to will have its own effect on task design and operation. In brief, our communicative competence is part pragmatic and part cognitive. It includes our ability to create meanings by exploring Halliday's 'meaning potential' in doing which we make use of our capacity to negotiate, interpret and express value in what is an educational and creative process. It views what is learned both as organised structures of knowledge and those procedures which are needed if we are to adapt that knowledge in the solution of new meaning-making problems, not hitherto confronted. Such knowledge, as Widdowson (1983) indicates, is best seen less as lists of items than as a set of structures which include forms but also encompass textual organisations and social routines, themselves deeply connected to our systems of value and belief and properties of particular cultural and social positions. Procedures are what we set to work on knowledge structures, testing their boundaries in the process of communication, making use of our inferencing and problem-sensing capacity and drawing on our strategies for problemsolving. Such interaction between knowledge and procedure is itself a task, one which, like learning, is sensitive to situation and inherently variable and differentiated.

Taken together, the two sets of demands above suggest an optimal environment for communicative language learning. A classroom where learners are encouraged to respond actively, to explore and take part in purposeful communication with their fellow learners. As Breen (1983, 1985a) indicates, such learning is a highly sociallsed activity where learners are engaged in a negotiative process, with themselves in terms of what they already know, with others in terms of sharing and refining knowledge and with the curriculum content in terms of what has to be learned. In such a classroom, one clear condition for action is preeminent. The need to offer learners a variety of alternative and differentiated options in the choice of classroom activity, subject-matter and modes of working. It is, after all, important to be consistent: if we are maintaining that most effective language learning arises when learners are engaged in collective exploration and experiment, making use of their own personal contributions, then we cannot in our organisation of those classroom activities deny them the variety that such contributions entail. However, the demands of the learning and the content syllabuses impose a more specific constraint on classroom action. What limitations do the resources of classroom management impose on possible permutations of variable data, different demands on information and varied activity? How can tasks be designed which can allow the maximum opportunity for learner and teacher choice? Will such tasks provide that accountability which institutions require and which motivates curriculum change?

From what has been said so far, a number of characteristics of communicative language teaching have emerged which impose conditions on task design. It may be useful, at this stage, to offer a summary:

From the conditions above, and the criteria illustrated, we might offer the following as a working definition of language- learning task:

No doubt there will be reader reaction and variable emphasis given to any such definition; this is as it should be, since different classrooms and different syllabuses will highlight their own features, both in general and at different points within their practice. Indeed, a useful in-service activity is to explore just such variable evaluation, both of the definition above and the conditions on task-based language learning listed earlier. Whatever the emphasis, however, and whatever the orientation of the tasks, it would seem that specification of the following key features will be required in each case:

By this I mean whatever data are presented or selected by the learners and teacher for work, in whatever medium/media and associated with whatever personal experiences of the participants. What resources are needed for accomplishment of the task?

By this I mean the specification of roles of participants in relation to the accomplishment of the task and their roles in respect of their relationships with each other. What do participants have to do and who assigns such duties? How are the participants to co-operate? What is to be their relative distance and relative powerfulness? Who acts and who monitors, who instructs, who guides?

By this I mean the classroom arrangements specified in the task. Does it imply individual, pair or group work? What combinations or sequences of these? Is choice among work-settings left to the participants or stipulated? What links are there between the overall classroom setting and out-of-class activity?

By this I mean the procedures to be followed in the understanding, execution and accomplishment of the task. How are these set down? Is choice among them left open or fixed in advance? How is the work to be shared? What behaviours of various participants are expected, tolerated? Who does what with what with whom and how?

Reflections in such a report of the design features for language -learning tasks are both clear and corroborative. Similar support can be derived from the actual comments of pupils themselves (cf. Dam, 1983, 1984), and from the list provided by Raths (1971), cited in Stenhouse (1975), of worthwhile activities. These latter are not specifically related to any subject-matter, and it would be educative for readers to discuss their language-learning relevance, in the light of what has been offered above:

Three comments are appropriate to our discussion: firstly, that many of Raths' criteria are mirrored in those of language teachers' own task experience, and that of their learners (at least in one case); secondly, that for all their non-subject-specificity, Raths' criteria are in several cases directly relevant to language learning (note, for example, the link to the well-known OISE study of The Good Language Learner (Naiman et al., 1978) and to Rubin's (1981) paper drawing language teaching consequences from this study); thirdly, how these criteria offer the basis for a research programme into the communicative (pragmatic and cognitive) behaviour of language learners and its associated acquisition.

Although Raths, for example, offers a means whereby we might both design tasks and discriminate among them, the criteria (and those we have suggested earlier) do not offer anything other than implicit suggestions that tasks might be catalogued under several distinct types. Such a typology is, however, necessary for any curriculum with differentiated goals. It will be important in language teaching as in other subjects where there can be a variable focus of the participants on different aspects of the curriculum. For example, we may wish to focus on the process of problem-solving, or on learning-training, or on information distribution, or on experimentation, or on discoursal outcomes, among many other possible contenders. In a classroom resourced by tasks, we shall need some way in which we can make informed and appropriate selections among the banks of tasks available. If one likens the classroom to the integrated language education network suggested in Leech and Candlin (1986), a bank of data (texts) would be augmented by a set of informational resources, to aid access to such texts, and drawn upon where relevant through the agency of a bank of process tasks, making use of a master programme to effect the necessary interlinking at learner demand.

Here tasks focus on ideas and information corning from individual learners in the group themselves. These ideas are pooled without any process of selection or synthesis.


Here the emphasis is on the importance of the outcome: either a consensus decision of the group or a single acceptable answer.


Here different information is given to different learners. The solution to the problem relies on the piecing together of different information from all the learners in the group or the class.


Here each learner is individually responsible for discovering information on a problem or a topic and reporting back to the group with his or her findings. The final product from the activity compiles different results and information from different learners.

Type3. Focus on Research and Experimentation (cf.Greve,1981)

(a) Tasks aimed at raising initial questions and problems.---What?

(b) Tasks aimed at formulating hypotheses.---Why?

(c) Tasks to decide on methods and procedures.---How?

(d) Tasks to decide on a precise workplan.---When?

(e) Tasks aimed at evaluating results.---How well?

(f) Tasks aimed at presenting results.---Why? What? How? For whom? Where?

Type4. Focus on Learner Strategy (cf. Bruner, 1974)

(a) Tasks focused on recognising and perceiving problems.---Attending
(b) Tasks focused on selecting, classifying, patterning.---Making Sense
(c) Tasks focused on inferring, deducing, judging.---Going beyond the Given
(d) Tasks focused on application of findings to new data.---Transferring and Generalising

As I point out above, other types can readily be imagined. For example, the typology of communication strategies available in Faerch and Kasper (1983) offers a ready basis for communication-oriented task design, while the proposals for consciousness- raising of grammatical processes given in Sharwood-Smith (1981) and Rutherford (forthcoming) indicate how tasks can focus on particular aspects of the language system.

7. Tasks and Educational Goals

Whatever tasks are devised, and whatever typologies established, selection from such a resource will need some higher order rationale as well as the tactical demands of immediate classroom need. In part this rationale has been provided by the criteria for task design outlined above, especially through the illustration from Raths (1971). Given the frequent failure, however, of language teaching curricula to emphasise educational goals at all in their pursuit of cost-effective training, it may just be worth reiterating what such a set of educational goals might be. After all, targets for language learning are all too frequently set up externally to learners with little reference to the value of such targets in the general educational development of the learner. Because we are concerned with language learning, it is very easy to forget that we should be equally if not more concerned with the developing personalities of our learners. How can these educational goals be defined and made the focus of language learning tasks? We might begin by identifying the following.

Awareness

The opportunity offered by tasks for learners to become more aware of their own personalities and social roles, and of those of their fellow learners, as they take part in communication - Ways in which tasks can make us focus on how language is used to reflect and reinforce our value and belief systems, critically exploring how language can both act to unmask as well as to obfuscate this process. An awareness, also, of how we and others go about communicating and learning.

Responsibility

If we assert (as we have repeatedly done here) the need for alternative action and differentiated goals, we cannot ignore the problems posed by the need to prioritise and select. Asserting the need for interdependence among learners and teachers implies an ability of all to exercise responsible choice of activities to undertake, problems and support they might need. How can tasks be designed to develop this ability? It would be ironic if a task-based syllabus merely made learners expert at following pre-set paths and did not promote their own capacities to draw their own maps.

Tolerance

What connections do we forge between the social and cultural worlds of learners outside classrooms and the world within? The viability and amelioration of both such worlds depend upon mutual acceptance and tolerance of their members, and overcoming the barriers raised by ideology and prejudice. The challenge here for tasks is to take a critical stance, both in terms of content choice and process preference. It ought not to be a choice solely governed by the often asocial considerations of what is appropriate for laboratory experiment, if to do so conceals the regular aysmmetries and ]inequalities of non-native speaker/native speaker interaction in the real world outside. This would be as true for process as it is for content.

Self-realisation

It would be difficult to find any set of curriculum guidelines which did not have as a priority the self- realisation and self-fulfilment of the learner. Nonetheless, we can imagine tasks which do not honour those priorities: where learners are not encouraged to make judgements and exercise choice, where they are not encouraged to profit from error or partially completed actions, where they find too little opportunity of impressing their own stamp on the tasks that are presented to them. This is particularly important in tasks concerned with the development of communicative competence, where the realisation by the learner of what can be achieved with language and the flexibility of language in the expression of attitudes and ideas is at the heart of the curriculum.

Self-confidence

Any set of task-based materials runs the risk of demoralising as well as enhancing the self- confidence of learners, in that it is impossible for task designers to gauge accurately in advance the thresholds of competence of different learners. What one can do, as I have suggested, is to offer learners alternative routes, the choice of which by learners is already the occasion for negotiation among learners and teachers. Once embarked upon a particular selection of related and dependent tasks, opportunities arise for extending and testing learners' self-confidence as they evaluate their own performance. The goal for tasks here is to assist learners in becoming aware of their growing capacity to communicate with others, at the same time as developing a self-evaluation of their own state of learning. It is only when such self-evaluation can be achieved in an atmosphere of confidence and lack of threat that learning can be effectively promoted. To that end tasks must present personal challenges to learners while at the same time offering them support through available resources and variably controlled and monitored stages.

8. Sequencing Tasks

To speak of syllabuses drawing on curriculum guidelines and critiquing them, as we have done, implies at least that some planning must take place at the level of classroom tactics, however short-range that must necessarily be. We must address, therefore, the issue of how tasks can be appropriately sequenced. If tasks were merely vehicles for the presentation of packaged ,units' of syntax, however attractively designed, it would be possible to argue strongly against any predetermined sequencing on the grounds of the research evidence from second language acquisition. For even though (for some albeit restricted set of items) there is evidence for a natural order, there is no evidence to confirm that learners are necessarily at the same point in such an order at any teaching moment. The consequent variability would preclude any such simple solution. Similarly, if tasks were to be designed on the basis of componential analysis of some target communicative behaviour (as for example in much ESP curriculum design), there is nothing which might lead us to insist for all learners that the order of such an assembly of components should be consistent. If, on the other hand, we were to abandon all pretence at sequencing in advance, and rely on feedback, this calls for great sophistication in self-evaluation by learners. Moreover, the guidance that would be needed from any teacher in such a dynamic world would itself have to rest on certain principles if it were not to be quite arbitrary and capricious.

What kinds of principled guidance might be forthcoming? In what follows, I offer a variety. No doubt, once more, these are interdependent and not exclusive. They will be more or less amenable to clear exemplification. They ought to promote both discussion and experiment, desirably within task- based language learning.

Cognitive Load

It will be difficult easily to distinguish what is cognitively difficult for learners to accomplish from what is communicatively difficult. Nonetheless, we may be able to design tasks in which there is a gradual increase in cognitive complexity without dramatically raising the communicative load. For example, tasks which require learners to follow a clear chronological sequence, referring to individual actions of individual characters, will clearly be cognitively less demanding than a task in which there is no such clear development and where the picture is complicated by multiple actions and multiple actors.

Communicative Stress

In a recent handbook for teachers of spoken English as a foreign language, Brown and Yule (1983) offer a description of the most complex situation that a non-native speaker of a language might be expected to face: where he or she has to converse on a topic known more to his/her interlocutor than him/herself, where there is more than one such interlocutor, where the interlocutors are more communicatively competent, where the interlocutors know more about the subject- matter, where the communicative task does not follow a clear structured organisation but focuses on reasons for actions rather than the actions themselves. Armed with such a worst scenario, we might, with Brown and Yule (op. cit.) construct a sequence of gradually more complex communicative tasks, upping the communicative ante, as it were, task by task.

Particularity and Generalisability

We might want to assert that tasks which follow some generalised pattern, say of assembling components, or some ritualised interpretive schema, would be easier to manage than those where the order of assembly or the norms of interpretation are unclear and to be negotiated. For example, asking someone original questions about hypothesised future events is likely to pose a more difficult task than the recounting of familiar experience. Such ease or difficulty could be readily transposed to reading or writing tasks, and would offer parallel possibilities for sequencing. We might indeed want to develop in learners a capacity for handling such routinised tasks before facing them with original departures fro in those goals and procedures at which they had become adept.

Code Complexity and Interpretive Density

There is no absolute reason whereby a complexifying of the code inevitably involves a corresponding increase in the interpretive density of the text, but we might want to construct tasks, particularly those focused on reading or listening, where sequencing was possible, playing off lexico- syntactic complexity with the 'depth' of the interpretive questions posed within the task. For example, we might want, with textually elaborate texts to ask more straightforward questions, while with textually simple texts asking questions which required the interpretive and explanatory analysis of the learner. Such tasks would clearly offer maximum scope of differentiated questioning of the same text for different levels and styles of learner.

Content Continuity

I have referred above to the suggestion of Long (1983f) that tasks be constructed on the basis of socio-linguistic descriptions of the target communicative encounter. Such a procedure has become commonplace in work-related tasks for ESP. Although such a task design procedure has clear apparent advantages in terms of target authenticity, being drawn as it is from 'real' life, there is a frequent practical problem within the classroom where professional participants reject such simulated encounters on the grounds of their perceived non-verisimilitude (Candlin, Bruton and Leather, 1976). This social-psychological alienation is matched often by the realisation that such representations act as models which preclude much active engagement and learner involvement (Breen, 1985). Nonetheless, it would seem for some learners, especially those engaged in short-term training programmes, that task design might fruitfully begin from such content tasks, allowing, however, for the variation in mode of work, perhaps, which could counter Breen's criticism.

Process Continuity

Within the Challenges teaching materials (Candlin and Edelhoff, 1982) the means are offered for learners to create their own continuity and sequencing among the available tasks. Learners are encouraged collectively to sequence their own tasks by examining what needs to be known and experienced before a particular, for some, more complex task can be achieved. All the time, in practice, learners receive feedback on the success of the process of learning: from him- or herself ('Can I now do this new task?') and from colleagues, and from the teacher. Learners come to know that particular tasks need adequate preparation, in terms of language forms, functions and discoursal strategies, ways of thinking and organising ideas, and knowledge of content. Learners can be enabled, through such flexibility, to orgamse their own learning orders and to construct their own continuity.

9. Evaluating Tasks

There has been some criticism of the original notion (Breen and Candlin, 1980) that a communicative curriculum realised in classroom tasks ought to be self-evaluating. The assumption has been that evaluation was too precious a process to be left, like war, to the teachers and the learners. There was a need (Rea, 1985) for external agency. I reject this necessity as some given good. Indeed, a careful reading of the original proposals for a communicative curriculum, together with the suggestions here, would make it plain that such external evaluation, while no doubt valuable as corroborative information, was strictly unnecessary where the syllabus action of the classroom had the provision of a formative evaluation as a main objective. Nonetheless, in order to provide this formative evaluation, illuminative of the process of learning, tasks will need to be ratable according to stipulated criteria. What might such criteria be?

Problematicity

To what extent does a given task reveal variation in learners' abilities and knowledge? How diagnostic is it and how explanatory? Does the task provide not only monitoring (cf. p. 11 above) but also feedback in a form which reveals possible courses of future action for particular learners, or for the class? To what extent does the task offer potential for appropriate future action?

Implementability

What are the demands of the task in terms of resources required? What complexity of classroom organisation and teacher-learner management is presupposed? What logistic demands does the task impose? Can we evaluate the implementability of tasks at two levels: a level of the task content and a level of the task action? We can envisage tasks where the variable amendment to the process of work within the task has an effect on the nature and extent of the content processed. In other words, tasks may be evaluated according to their adaptability and conformability to learner contribution.

Combinability

Tasks will differentiate themselves according to their combinability with other tasks. Some tasks will be truly process tasks, neutral as to content, able to be adaptably introduced into the syllabus at a wide variety of points, on demand by learners. Such free-floating tasks lay at the heart of the suggestions for materials design in a communicative curriculum made some time ago now by Breen, Candlin and Waters (1979), and labelled there, process tasks. Others will be more bound to some sequence with other tasks, acting as enabling devices to the accomplishment of more complex goals. Over time, of course, and benefitting from the retrospective accounts advocated at the outset of this paper, some tasks will come to occupy particular places in the sequence of learning for particular learners, perhaps associated by effective practice with particular learning cruces, or special content difficulty. Combinability is not, then, some given: it emerges through the guaranteed formative evaluation of the syllabus account, as a means of evaluating tasks.

10. Postscript

I am conscious that the foregoing has been programmatic and issueraising, rather than conclusive and encompassing. There are many questions left unaddressed. I have not broached the problematic of the roleassignment of participants within particular tasks, nor have I adequately explored the contribution tasks can make to the definition of the subjectmatter as well as to the syllabuses and the curriculum. In the teaching of communication, after all, conventions are being negotiated in the process of learning, and tasks have there a codifying role. What I hope I have done, here, however, is to offer, incidentally perhaps, the following rewards from task-based language learning:

*a restructuring of the relationship between curriculum and syllabuses;

*a re-evaluation of the relationship between data, resources and processes in task content and task action;

*a re-estimation of the relationship between goal-based and normreferenced evaluation;

*a review of the role relationship of 'teacher' and 'learner' in task co-participation;

*a re-emphasis of the potential of the classroom as a place for experimentation into language and language learning;

*a renewing of a critical perspective on the understanding of second language acquisition and the place of language learning in the social identity of the learner.

---

Topic 10 of 12

GO BACK TO: Topic 10: Selecting, Grading and Sequencing Tasks

---

[Home] [Topics Index] [WebBoard] Copyright by University of Illinois at Urbana-Champaign, 1998