Henry G. Widdowson The pedagogic relevance of language awareness Abstract. Awareness (of language as grammatical code) was central to language learning in grarnmar-translation methodology, but was played down in both structural and communicative language teaching. A communicative approach implied however a need to become aware of aspects of language in contextual use, though it was not always clear what the facts were of which one needed to be aware. Now that aspects of language use and of usage have been described, it is valuable that language teachers be aware of these developments. However, whether and in what sense such awareness is of pedagogical relevance cannot be determined by linguistic discovery procedures, or by investigating native-speaker norms. Classroom learning has its own agenda, and we need above all to be aware of this. Language teaching, like any other human activity, needs the stimulation of novelty to keep it going. Where there's life there's change, andin a sense it does not much matter whether the change is really innovative or not. lt is the appearance that counts. The currently fashionable notion of language awareness is a case in point. Like all fashions, it serves the important purpose of providing people with a sense of renewal and progress, something different they can identify with. But it would be as well for us to go beyond appearances and try to establish just what this notion is, where it comes from, and how far it is pedagogically valid. Fashions come and go. There is much talk in the language teaching profession about bandwagons and swings of the pendulum as if these were the result of forces beyond the teachers's control. But it is possible to take a more positive attitude and subject these influences to critical examination so as to establish the grounds of their validity and the extent of their relevance. Fashions are inevitable, but this does not mean that they have tobe conformed to without thought. There may, of course, be good reasons to go along with the fashion. But at least we should ask ourselves what these reasons are. Perhaps the first point to be made about language awareness is that, in one sense at least, it is not a new idea at all. lt has been the traditional staple of language teaching for generations. Thus, the explicit teaching of grammar is designed to make students aware of the formal properties of the language being learned, and when associated with translation, it makes them aware of how these properties differ across languages. The pedagogic assumption here is that such conscious awareness leads to linguistic knowledge which can be acted upon subsequently as actual behaviour as and when required. Essentially the idea is that you invest in an awareness of how meanings are semantically encoded, and you get retums in the form of an ability to realise these meanings pragmatically in communication. This FLuL 26 (1997) 34 Henry G. Widdowson idea has, of course, since fallen out of fashion, and the pedagogic approach based on it superseded. But we need to be clear, I think, that it was superseded in two very different ways. The structuralist approach, itself a descendent of the Direct Method, denied the validity of awareness as such. You did not leam language by leaming about it but by engaging with it directly. So classroom activities were designed to deflect conscious attention away from the formal features. These were to be made habitual by practice. Thinking about what you were doing would, it was supposed, inhibit this habitual intemalisation. Declarative awareness gave way to procedural ability. This approach was itself subsequently called into question in the early seventies with the ascendency of communicative language teaching. But we should note that what was called into question was not the primacy given to behaviour. There was no restoration of the belief in conscious awareness as the investment from which practical ability would be realised. Communicative language teaching, at least as practised under the influence of the Council of Europe and its Threshold Level specifications, still favoured the development of instinctive ability by the direct experience of doing things with the language. The difference was in what kinds of thing learners were required to do. Instead of manipulating forms, they manipulated functions, but the manipulation was still a matter of developing procedural ability. Declarative analysis did not come into it. Learners were still required to leam by direct exposure and experience without the intervention of conscious intellectual awareness. In such a scheme of things, there was, naturally enough, little place for grammar. This was not only because grammar represented the formal properties of language, whereas the communicative focus was on its functions, but also because grammar was an explicit and declarative· description which engaged intellectual awareness, and this was not congruent with a view of learning which, like the structuralist approach which preceded, was bent on encouraging intuitive ability by behavioural means. Even if there had been explicit descriptions of functions available for reference, their use in dass to develop conscious awareness would not have been consistent with the basic pedagogic principle of intuitive 'natural' learning by experience which was predominant in the communicative approach. In an important sense, this approach was as fundamentally behavourist as the structuralist approach_ it rep-Iaced. In neither case was language awareness, either of forms or of functions, part of the agenda. When and where then does language awareness make its re-appearance? lt does so, I think, as certain implications of the communicative position in language pedagogy begin to become apparent. And they do so in respect to two kinds of pedagogic consideration. The first concems the objectives of leaming as specified by a syllabus, and the second the process of learning as activated by methodology. With regard to communicative objectives, these were held to be accounted for in the specifications of the notional (or notional/ functional) syllabus. As Wilkins put it in his highly influential book: FLuL 26 (1997) The pedagogic relevance of language awareness "The advantage of the national syllabus is that it takes the communicative facts of language into account [....] it will produce a communicative competence" (Wilkins 1976: 19). 35 But what are the communicative facts of language? And what is this communicative competence that can seemingly be produced by the straightforward expedient of redefining language content in notional/ functional terms? As described in Hymes (1972), routinely invoked as providing the theoretical framework for communicative language teaching (see, for example, Brumfit & Johnson 1979), the concept of communicative competence was rather more complex than teachers at the time were led to believe. Hymes proposed that to be communicatively competent in a language you had to be capable of making four kinds of judgement about a particular instance of use, namely: 1. Whether and to what degree it was possible. 2. Whether and to what degree it was feasible. 3. Whether and to what degree it was appropriate. 4. Whether and to what degree it was actually done ( performed, attested). Let us call these communicative conditions. The possibility condition has to do with conformity with the linguistic code, however that is established. Judgement on possibility is based on notions of correctness, and especially on grammaticality. lt was this condition, clearly, that was paid particular attention to in structuralist language teaching. But not, we should note, at the expense of the second condition. Feasibility has to do not with linguistic correctness, but with cognitive intelligibility, the extent to which an instance of language, no matter how well formed, can be readily processed. lt has to do with how far the encoded meaning is accessible. And, of course, in the structuralist syllabus, possible items (in this Hymes' sense) were selected and ordered in ways to make them easy to process. In these respects, one can argue that the structuralist approach did take aspects of communicative competence into account. lt did not, however, do so well on the other conditions; they were subordinated to the first two. Thus the pedagogic purpose was seen not as a matter of teaching language appropriate to 'real-life' contexts, but of making language fit classroom contexts specially designed to be appropriate for teaching the structural pattems of the possible. And attestedness in the sense of naturally occurring language usage was hardly considered at all. Leamers manipulated elements of the language code in combinations which were meant to facilitate learning, and there was no requirement that these should correspond with those that were actually produced by native speakers. With communicative language teaching it was these two conditions of attestedness and appropriateness that now came into focus. Council of Europe Threshold Level inventories, and the syllabuses drawn from them, were based on the idea that what leamers ought to be taught was the language as actually attested in real life, rather than the arid abstractions of its generalised linguistic properties. But what was the language of real life? The Threshold Level inventories were based on a needs analysis which identified those transac- FLuL 26 (] 997) 36 Henry G. Widdowson tional and interactional uses of language which seemed to occur most naturally and commonly in contexts of native speaker communication. But this identification, and the specification drawn from it, was somewhat intuitive, and not informed in detail by the facts of actual usage. These facts have since become available on a large scale through the computer analysis of actually occurring language. Such analysis continually takes intuition by surprise, and reveals that a good deal of what is presented in language textbooks does not correspond with the 'real' language of native speakers. As John Sinclair has put it: "We are teaching English in ignorance of a vast amount of basic fact. This is not our fault, but it should not inhibit the absorption of new material" (Sinclair 1985: 252). Sinclair is talking about English in particular, but it is clear that his remarks would apply to the teaching of any foreign language. Now, obviously, if you are claiming to be teaching the ability to communicate you need to be aware of the pattems of language which people actually produce when they communicate, rather than some invented version of them. With the findings of corpus analysis at our disposal, we are now in a position to specify the objectives of learning based on an awareness of actual language usage. But an awareness of language usage is only part of the story: it only accounts for Hymes' fourth condition. A corpus analysis will reveal pattems of texts, attested features of the language that people have actually produced. But it will not of itself reveal what makes this language appropriate as use. lt will show what language is used but not how it is used in the conduct of social life, for the expression of identity and the management of human relations. lt cannot show what people mean by the language they use in particular contexts. And these contexts are of course culturally embedded. Communication as the contextually appropriate use of language is inextricably bound up with culture, and learning how to use another language must to some degree be a matter of qualifying for membership of a different community. To learn a language is clearly not just to pick up the most common pattems of usage but to realise these as appropriate use, to know what communal and cultural significance they carry as social action. Language awareness in respect to appropriateness, therefore, involves cultural factors of a complex and subtle kind. If you are learning somebody's language, you are inevitably being initiated into some kind of social membership of somebody else's community; its attitudes, values, modes of thinking. You ought to know what you are letting yourself in for, what the socio-cultural concomitants of learning a language are. When the communicative approach came on the scene in the early seventies, teachers did not know what they were letting themselves in for. At that time, there was little in the way of well-grounded awareness as to what the communicative facts of particular languages actually were. Certainly, even abrief look at Hymes' original formulation of the concept of communicative competence suggests that it FLuL 26 (1997) The pedagogic relevance of language awareness 37 is unlikely that it can be 'produced' by simply restating syllabus units in terms of intuitively conceived notions and functions rather than grammatical structures. What the communicative approach did, in the shape of the notional/ functional syllabus, was to give priority to the third of Hymes' conditions and make the first two subordinate. The difficulty was that this was done without an informed awareness of the complex socio-cultural factors involved in establishing appropriateness in actual contextual use. The extensive work on discourse analysis and pragmatics which might inform such awareness was yet to be done. As to the attestedness condition, this (as Sinclair indicates) could not be met in any informed way either, since at that time the data on usage was not available. In short, although the programme proposed at that time was the teaching of communicative competence, there was little in the way of description of appropriate or attested language that could be drawn on to sustain such a programme. To quote Sinclair again, teachers were "teaching English in ignorance of a vast amount of basic fact", and this refers as much to the appropriate use of language as a socio-cultural process as to the attested usage of language as its product. To say this is not to apportion blame. lt was not, as Sinclair says, the teachers' fault, for they had only their intuitions to go by. These days, however, the case is different, and there is no excuse for remaining in ignorance of the nature of communication. There is an abundance of literature available on the matter which teachers need to be aware of if they are to claim authority in their profession. But it is not just a question of being well-informed about pattems of usage as revealed by corpus analysis by reference to, say, Sinclair (1991), and the socio~cultural conditions of appropriate use as revealed by discourse analysis and pragmatics, by reference to, say, Cook (1989), McCarthy & Carter (1994), Yule (1996). lt is also to be aware of the possible ideological implications of furthering communicative objectives. For if communication is so intricately bound up with the socio-cultural values of particular communities, then the acquisition of competence in it must mean the acceptance, to some degree at least, of the membership conditions of such communities. Language leaming then becomes a process of acculturation and questions naturally arise about the desirability of this, of whether, and to what extent, conforming to the cultural norms of another community involves identifying with them. With regard to some languages, most notably English, questions also arise as to whether its promotion as a universal means of communication does not have the hegemonic and neo-colonial effect of maintaining the power and prestige of its speakers (see Phillipson 1992, Pennycook 1994). So long as teachers are concemed only with possible and feasible linguistic units abstracted from any considerations of contextual appropriateness, these questions about acculturation are not evident. Indeed we can think of these units as malleable linguistic forms which can be impressed with whatever socio-cultural values seem subsequently to be suitable. But if you are committed to teaching communication, then there are implications about acculturation and ideology which it would be as well to be aware of. FLuL 26 (1997) 38 Henry G. Widdowson What has happened over recent years is that the scope of linguistics has been extended to take into account patterns of actually occurring usage and socio-cultural conditions of use which were previously excluded from consideration. The assumption has been that this redefinition of language study as a discipline must have direct relevance to the definition of language teaching as a subject. If linguists tell us that this is what language is really like, then this, surely, is what teachers should teach. But things are not, I think, quite so simple. Disciplines and subjects are different things. And this too is something that language teachers should be aware of. A subject is a pedagogic construct, and it needs to take into account not only the eventual objectives of learning but the interim processes whereby these objectives are to be achieved. One can accept that teachers of a foreign language need to have an awareness of the language they are teaching as it is used in normal social circumstances by its users. One can accept too that the formulation of objectives will itself be informed by this awareness. So it is that it seems reasonable to say that if you are teaching French or German to foreigners you should define your pedagogic objectives by reference to the way these languages are used by native French and German speakers. But this is only part of the story. The language subject, whether it be French, German, English or whatever, needs also to be defined in terms of the process of achieving these objectives. lt is clearly not enough to specify a desirable destination. You have to design a feasible route, and this implies taking the starting point into account as well. This means that we need to consider these languages not only for what they are to their users, but also for what they are to the learners. And crucially what they are for learners is that they are foreign. What is real for the users is not real for them. They cannot appreciate the socio-cultural authenticity of actually occurring language quite simply because they are not, by definition, familiar with the socio-cultural conditions which would enable them to authenticate it. The task for language pedagogy, as for the pedagogy of any other subject, is to select and organise what is to be learned so that learners will engage with it, make it real for themselves, make it their own, and in so doing gradually approach the objectives set for them by a process of approximate authentication. The language activities appropriate to that end w~ll in many respects not correspond at all with native speaker norms. lt might be objected that as a result the learners will be encouraged to indulge in all kinds of curious non-conformist language use which will settle into patterns of unacceptable behaviour. The same argument has been used, of course, against the tolerance of learner error in respect to linguistic forms: if errors are not immediately corrected, it was supposed, then they will take permanent root in the learners' minds. Against this, it is now widely accepted that errors are evidence of the learner engaging with the language, conceptually appropriating it indeed, and so a necessary feature of the learning process. If this is true of language forms, there seems no reason why it should not be equally true of the communicative functions that they realise. In both cases they are interim stages in the process of development and serve as catalysts for further learning. What FLuL 26 (1997) The pedagogic relevance of language awareness 39 leamers do with the language in the classroom does not have to correspond with native speaker norms to be effective for leaming. lt does not have to replicate authentic use to activate the authentication process. On the contrary, in fact: attempts to replicate authentic use can actually undermine the appropriate conditions for leaming. And it is not difficult to see why. If leamers are presented with language, or required to enact language, which they cannot engage with, they are likely to be frustrated and alienated. One of the problems with structuralist language teaching, at least as often practised, was that leamers were required to manipulate forms of the language without there being any communicative (or other) purpose which would give any point to the exercise. Leamers were required to participate in classroom activities with which they could not engage. The foreignness of the language to be leamed was therefore not only that it was a different language from their own, in that it represented an alternative encoding of communicative resources, but that it appeared to be a quite different phenomenon in that it consisted of linguistic forms whose communicative potential remained unrealised. They were even denied the natural recourse to translation which would have helped to establish the communicative commonality of the two languages. But the same kind of alienation is likely to occur if the leamers are confronted with language which they cannot engage with because its communicative character is dependent on contextual conditions they cannot meet. All this is not to question the importance of teachers being aware of how the particular language they are teaching is actually put to use by its users, but only to point out that they need also to be aware of the need to establish what pedagogic relevance this rnight have. There has been a tendency in the past to suppose that it is self-evidently the case that features of usage and use which are revealed by linguistic description must necessarily constitute the content to be taught in language classes. But descriptive validity does not directly transfer to pedagogic value. With respect to language usage, for example, Sinclair makes the following point: "[...] it should never ever be necessary for students to 'unlearn' anything they have been taught. They cannot be taught everything at once, and' because our knowledge of the textual detail of language has been so vague, they have been taught half-truths, generalities which apply only in some circumstances. However, we now have the information on which accurate selection can be made. Students may have to unlearn some of their own projections and hypotheses, but that is a different matter from unlearning what has been authoritatively put forward as an accurate observation about the language" (Sinclair 1991: 500-- 501). The difficulty with this position is that the very process of leaming involves the recurrent reformulation of knowledge through successive stages of unleaming, and it is this that is recognised in the positive view of leamer error that I referred to earlier. Of course the teacher will not want to deliberately mislead students into thinking that what they produce in the classroom is as authentic as the 'real' language of native speakers, to be replicated in contexts of use. But there is no reason why they should be rnisled since they are schooled in instructional convention in FLuL 26 (1997) 40 Henry G. Widdowson their very role as leamers, and become themselves aware that activities in classrooms are designed to facilitate leaming and do not directly replicate conditions in the world outside. They know from their experience of instruction in other subjects that teaching progresses through half-truths and partial generalities which are gradually reformulated and revised. This is the necessary consequence of the fact that you cannot leam everything at once, whatever it is, so that subsequent revelations lead to a re-alignment and correction of what has been previously leamed. As I suggested earlier, this is what defining a subject means. Awareness of this is a crucial element in the role of the leamer, and language pedagogy (or pedadogy of any other subject) cannot get off the ground at all unless students co-operate in recognising that what they do has this interim instructional character, that they are expected to leam from the language they are presented with, and not to accumulate it as tokens of actual usage. To say this is not to deny that objectives need to be specified in reference to 'accurate observations about language'; but objectives are achieved as the consequence of the recurrent reformulations and revisions of the leaming process. Similar points can be made about language use and the socio-cultural conditions of appropriateness. lt has been argued that language leamers need to be initiated into the kinds of pragmatic knowledge that language users bring to bear in the realisation of communication in context. This is what Gabriele Kasper has to say on the matter: "The learner's task is not very different from that of the pragmaticist: she has to discover the contextual (situational) and co-textual (linguistic) constraints governing SA (Speech Act) selection and modes of realization in the target language and culture. In Hymes' terms, she has to discover what is possible, feasible, appropriate and done in carrying out SAs in L2" (Kasper 1989: 42). But one could also argue that the leamer's role is not like that of the pragmaticist in that her task is not to analyse language use but to engage with it. Quite apart from the fact that a conscious enquiry into complex contextual constraints would be beyond her, and would itself indeed be impossible and unfeasible, it could be instructionally inappropriate anyway, and unlikely to be done. If the discovery of such constraints is to be a condition on language leaming, it is a condition that few if any classrooms could ever effectively meet. As I have pointed out elsewhere (Widdowson 1993) the discovery of these constraints on speech act realisation by pragmaticists themselves has been limited, and when we consider the range of possible contexts of use and the complexity of the interplay of socio-cultural factors this is not surprising. Furthermore, it is not only a matter of knowing about them, but of knowing how far they can be effectively acted upon. And this, of course, brings up the central pedagogic question that concerns us: namely how is awareness of language an enabling condition for leaming and use; how far can students access it and act upon it in the process of learning and in achieving the objectives of that process. FLuL 26 ( 1997) The pedagogic relevance of language awareness 41 This question is commonly raised in respect to the forms of the linguistic code, but it applies equally to communicative functions in context. As I pointed out earlier, structuralist language teaching developed language awareness by directing explicit attention to the formal properties of the language and their semantic potential. Tue objection raised against this was that students thereby acquired knowledge they could not act upon. The acquisition of knowledge about contextual constraints on communicative functions is obviously open to the same charge. If we question the assumption in the case of forms, we should also do so in the case of functions. And at least with forms, one might argue, we had a source of information in grammars and dictionaries that we could draw upon, whereas with functions we have very little by way of description to go on. We now know more about usage, but this has to do with the way linguistic forms manifest themselves as actually occurring linguistic behaviour, and this, as John Sinclair and other corpus linguists have pointed out, rnight lead us to a revision of our grammars and dictionaries as sources of information about these forms. In respect to use, however, the complex and subtle ways in which these forms realise pragmatic meaning in socio-cultural contexts, we have very sparse evidence indeed. What makes matters worse is that knowledge about the socio-cultural contexts of language use is often represented as a crucial condition on leaming. If you do not know about them, you do not know the language. But since such contexts are so complex, and information about them so hard to come by, what happens is that the required knowledge is reduced to a compendium of cultural facts. So it is, for example, that the Longman Dictionary of English Language and Culture (with its ambiguous title) claims "to get to the very heart of the English language and into the head of the native speaker" (thereby confounding English with the English) and implies that if students want to achieve fluency in the language they have only to enquire within. This thereby equates contextual conditions with cultural facts and awareness with ability. Elsewhere, it is true, there is a recognition that things are not so simple. In particular, the work on what has been called critical language awareness (eg in Fairclough 1992) seeks to make explicit how social meaning is subtly implicated in language. Tue assumption here is that the leaming of a foreign language is bound to involve the ingestion of the socio-cultural values of its users, and to make leamers prey to the hegemonic exercise of power through the insinuation of ideological influence. lt is not, then, a matter of the leamer discovering what the contextually appropriate conditions of use are so they can conform to them, but also so that they can question them, and indeed resist too ready a compliance with them. In this view, communicative competence necessarily implies some degree of communal commitment, and this, it is argued, is something that both teachers and students should be aware of, and wary of: leaming a language is not so innocent an activity as it might appear. The work of this persuasion on critical language awareness raises a number of interesting issues, but again the question arises as to its relevance to second lan- FLuL 26 (] 997) 42 Henry G. Widdowson guage pedagogy. One might accept the general argument that students should subject their own communicative experience to critical appraisal so as to be made aware of the covert socio-cultural influences that inform it. But this appraisal works on what is familar, and calls for a degree of detachment, therefore, such that in a sense critical language awareness involves defamiliarizing linguistic experience and making it foreign. For second language kamers, however, the language is not familiar to start with: on the contrary, as I suggested earlier, its most salient feature is its inherent foreignness. So the task of the foreign language student is quite the opposite: it is to get engaged with the language so as to make it familiar. Clearly, to the extent that students have not acquired an experience of it, it is hard to see how they can subject it to any critical appraisal. The obvious danger is that they may be schooled in critical response before they themselves are capable of making any such response on their own account. This pedagogic phenomenon is not unknown, of course, and has in the past been particularly evident in the teaching of literature, which in many ways the techniques of critical language awareness quite closely resemble. Again, all this is not at all to deny that there is a good deal in language that we need to be aware of, nor that work over recent years in corpus descriptions, pragmatics, and critical analysis have yielded fascinating facts and findings about language in use that we need to make educational provision for. These could well, it seems to me, provide the disciplinary source of a subject language, designed to develop awareness of the nature of language in general as exemplified by the usage and use of different languages, including, obviously enough, those that are immediate to the students own experience. Such a subject would, I believe, bring enormous educational benefits. But it is not the subject that foreign language teachers are concemed with. They are concemed not with language but particular languages (French, German, English, Russian or whatever) and furthermore particular foreign languages. lt may be that an awareness of the different aspects of language that I have touched on here might be tumed to pedagogic advantage in developing the required ability in the particular language being learned. But this cannot just be taken on trust. We need always to enquire about the relevance of such awareness in principle and its applicability in practice to the objectives and processes of particular courses of language instruction. And this is also a matter of awareness: not of language in general, but of how it can best be defined as a teaching and learning activity, how, in short, it constitutes a pedagogic subject, and how, as a subject, it might be defined in terms of what is possible, feasible, appropriate, and actually done in different teaching/ leaming contexts. FLuL 26 (1997) The pedagogic relevance of language awareness 43 References ALATIS, James E. (ed.) (1991): Linguistics and Language Pedagogy: the State ofthe Art. Georgetown: University Press. ALATIS, James E. (ed.) (1993): Language, Communication and Social Meaning. Georgetown: University Press. BRUMFIT, Christopher J. / JOHNSON, Keith (eds.) (1979): The Communicative Approach to Language Teaching. Oxford: University Press. COOK, Guy (1989): "Discourse" [title in Language Teaching: a Scheme for Teacher Education]. Oxford: University Press. FAIRCL0UGH, Norman (ed.) (1992): Critical Language Awareness. Longman. GASS, Susan C. [et al.] (eds.) (1989): Variation in Second Language Acquisition. Clevedon: Multilingual Matters. HYMES, Dell (1972): "On communicative competence". In: PRIDE/ H0LMES 1972, 269-293. KASPER, Gabriele (1989): "Variation in interlanguage speech act realization". In: GASS [et al.] 1989, 37-58. McCARTHY, Michael J. / CARTER, Ronald A. (1994): Language as Discourse: Perspectives for Language Teaching. London: Longman. PENNYCO0K, Alastair (1994): The Cultural Politics of English as an International Language. London: Longman. PHILLIPS0N, Robert (1992): Linguistic lmperialism. Oxford: University Press. PRIDE, John/ HOLMES, Janet (eds.) (1989): Sociolinguistics. Penguin Books. QUIRK, Randolph/ WIDD0WSON, Henry G. (eds.) (1985): English in the World. Cambridge: University Press. SINCLAIR, John M. (1985): "Selected Issues": In: Quirk/ Widdowson 1985, 248-254. SINCLAIR, John M. (1991): "Shared Knowledge". In: Alatis 1991, 489-500. WIDD0WSON, Henry G. (1993): "Communication, community and the problem of appropriate use". In: ALATIS 1993, 1-12. WILKINS, David (1976): National Syllabuses. Oxford: University Press. YULE, George (1996): "Pragmatics" [title in Oxford Introductions to Language Study]. Oxford: University Press. FLuL 26 (I 997)