Geoff Thompson
ELT Journal Volume 50/1 January 1996, pp. 9 - 15
Although
communicative language teaching is accepted by many applied linguists
and teachers as the most effective approach among those in general use,
there are still a number of misconceptions about what it involves. This
article sets out four of the main misconceptions, discusses why they
have arisen, and why they can be so described. In doing this, the
article attempts to define some important characteristics of
communicative language teaching as it is practised at present.
Whatever the
situation may be as regards actual teaching practices, communicative
language teaching (CLT) is well established as the dominant theoretical
model in ELT. There have been recurrent attempts to take stock of CLT
and to identify its characteristic features (e.g. Richards and Rodgers
1986) and in areas such as teacher training the principles of CLT are
largely treated as clearly understood and accepted (see, for example,
Harmer 19911).
Despite this apparent unanimity, many teachers remain
somewhat confused about what exactly CLT is. At the more abstract end,
there is general agreement that CLT involves an emphasis on
communicating by means of the foreign language (the way in which this
idea is expressed tends, as here, to be so vague as to make it difficult
to disagree with): at the practical classroom end, CLT is strongly
associated with a number of particular activity types, such as
problem-solving and pair work. But in the middle ground, the area where
theory meets practice, things become less certain, For example, what
exactly does CLT set out to teach? Is there such a thing as a
communicative language syllabus? If so, what does it consist of? Is it
simply a notional-functional syllabus under a new name? Or does CLT only
exist as a methodological approach, a way of helping learners to
practise the skills needed to put their knowledge of the foreign
language into use?
In working with colleagues around the world, with
teachers and trainees on MA and initial TEFL courses, and with modern
languages teachers in the UK, I am constantly struck by the very
disparate perceptions they have of CLT. There are, I think, a number of
reasons for the confusion, not least the fact that CLT has developed
extremely rapidly over the past fifteen or so years and has now moved a
considerable distance from its original practices (though without
substantially changing its original principles).
I believe that an ‘orthodox’ and practical form of
CLT has emerged, not only in the writings of applied linguists such as
Littlewood (1992) and McDonough and Shaw (1993) but, perhaps more
importantly, in mainstream language textbooks, such as the Headway
series and the New Cambridge English Course, which represent good
contemporary practice. However, certain misconceptions about CLT
continue to survive, making it difficult for many teachers to see
clearly what is happening and to identify the useful innovations that
CLT has brought. A surprisingly large number of teachers that I have
spoken to criticize or reject CLT for what seem to me to be the wrong
reasons. In this article, I would like to set out the four
misconceptions that I most frequently hear voiced, discuss why they have
arisen, and explain why I think that they are misconceptions.
Misconception 1: CLT means not teaching grammar
This is the most persistent - and most damaging -
misconception. It must be admitted, however, that there are good
reasons for its existence. There have been a number of applied linguists
who have argued strongly and in theoretically persuasive terms that
explicit grammar teaching should be avoided. One line of argument is
that grammar teaching is impossible because the knowledge that a speaker
needs in order to use a language is simply too complex (Prabhu 1987).
Another is that grammar teaching is unnecessary because that knowledge
is of a kind which cannot be passed on in the form of statable rules,
but can only be acquired unconsciously through exposure to the language
(Krashen 1988).
For most teachers, the effects of these ideas have
been felt through their practical application in language textbooks and
syllabuses. In the early days of CLT, pioneering textbooks such as
Functions of English included no explicit teaching of grammar (although
Functions of English was
aimed at students who had typically already been
through a more conventional grammar-based course). Syllabuses were
developed (and are still in force in many places) which expressed the
teaching aims purely or predominantly in terms of what the learners
would learn to do (‘make a telephone call to book a hotel room’; ‘scan a
written text to extract specific information’), and which ignored or
minimized the underlying knowledge of the language that they would need
to actually perform those tasks.
However, the exclusion of explicit attention to
grammar was never a necessary part of CLT. It is certainly
understandable that there was a reaction against the heavy emphasis on
structure at the expense of natural communication. It is worth looking
back with hindsight at textbooks such as New Concept English, in its day
- the late 1960s and the 1970s - enormously and deservedly popular, to
see how narrow and constraining the approach was in many ways. But there
have always been theorists and teachers pointing out that grammar is
necessary for communication to take place efficiently, even though their
voices may for a time have been drowned out in the noise of learners
busily practising in pairs. This is such self-evident common sense that,
from the vantage point of the present, it seems odd that it should ever
have been seriously questioned.
Learning grammar through CLT: the retrospective approach
Of course, the question of how learners are to
learn the necessary grammar remains. Although, in the consensus view of
CLT that I have mentioned, it is now fully accepted that an appropriate
amount of class time should be devoted to grammar, this has not meant a
simple return to a traditional treatment of grammar rules. The view that
grammar is too complex to be taught in that over-simplifying way has
had an influence, and the focus has now moved away from the teacher
covering grammar to the learners discovering grammar.
Wherever possible, learners are first exposed to new
language in a comprehensible context, so that they are able to
understand its function and meaning. Only then is their attention turned
to examining the grammatical forms that have been used to convey that
meaning. The discussion of grammar is explicit, but it is the learners
who are doing most of the discussing, working out - with guidance from
the teacher – as much of their new knowledge of the language as can
easily and usefully be expressed. Behind this strategy lies the
recognition that the learners may well have ‘understood’ more about the
language than they - or the teacher - can put into words. If the new
language were introduced in the form of an apparently all-embracing (but
actually pitifully incomplete) rule from the teacher, this would convey
the unspoken message that the learners had nothing further to
understand about the language point and simply needed to practise it.
If, on the other hand, talking about grammar is postponed until the
learners themselves can contribute by bringing to light what they
already in some sense ‘know’, the unspoken message is that the process
of acquiring the new knowledge is one which takes place inside them and
over which they have some control. Indeed, with the recent emphasis on
training learners to learn efficiently, this message is likely to be
explicitly discussed.
This ‘retrospective’ approach to grammar is a natural
development from the original CLT emphasis on viewing language as a
system for communication; it also takes into account the fact that
learning is likely to be more efficient if the learners have an
opportunity to talk about what they are learning. Ellis (1992) argues
that while looking explicitly at grammar may not lead immediately to
learning, it will facilitate learning at a later stage when the learner
is ready (in some way that is not yet understood) to internalize the new
information about the language. The retrospective approach also has the
advantage that, if the lesson is conducted in English, it encourages
the learners to communicate fairly naturally about a subject that is
important to what they are doing: the language itself.
Misconception 2: CLT means teaching only speaking
Again, there are reasons why this misconception
is fairly widespread. CLT was influenced. as earlier approaches had
been, by the general movement in linguistics towards giving primacy to
the spoken language. In addition, a focus on encouraging learners to
communicate leads naturally towards thinking about what they will need
to communicate about, and why; this is part of the wider
tendency in CLT to look beyond the classroom. For many learners, the
main uses that they are likely to make of the language are oral: getting
around in the foreign country if they visit it, talking to visitors
from that country, etc. Even if they are unlikely in reality to use the
language outside the classroom, learners are often willing to suspend
their disbelief and act as if they might need the language for personal
contacts.2 Therefore, the emphasis is likely to be on speaking and listening skills.
A further reason
for this misconception is that CLT stresses the need for the learners to
have sufficient practice, of an appropriate kind. This is often
translated, especially by teacher trainers, into the principle that TTT
(teacher talking time) is to be reduced, and STT (student talking time)
is to be maximized - chiefly by putting students into pairs and telling
them to talk to their partners. At the same time, while the slogan ‘TTT
bad, STT good’ almost certainly represents a useful (though perhaps
rarely attained) goal for most teachers, it is also important to
recognize that communication does not only take place through speech,
and that it is not only the speaker (or writer) who is communicating.
Communication through language happens in both the written and spoken
medium, and involves at least two people. Learners reading a text
silently to themselves are taking part in communication (assuming that
the text has something of relevance to them) just as much as if they
were talking to their partner. No doubt this seems too obvious to be
worth saying; and yet I have heard the complaint that CLT ignores
written language surprisingly often, from experienced teachers as well
as trainees. Learners are probably likely to talk more in a successful
CLT class than in classes using ‘traditional’ approaches; but a glance
at recent mainstream textbooks will
immediately show that they are also likely to be reading and writing a
more varied range of texts than those in more traditional classes. CLT
involves encouraging learners to take part in - and reflect on -
communication in as many different contexts as possible (and as many as
necessary, not only for their future language-using needs, but also for
their present language-learning needs). Perhaps, rather than student
talking time, we should be thinking about the broader concept of student
communicating time (or even just student time, to include necessary
periods of silent reflection undistracted by talk from teacher or
partner).
Misconception 3: CLT means pair work, which means role play
The misconception here is not so much in the
emphasis on pair work itself as in the narrowness of the second
assumption concerning the ways in which it is used. Role play can
certainly be a useful technique - though personally my heart sinks a
little when I see yet another instruction along the lines: ‘One of you
is the shopkeeper/hotel manager/doctor’s receptionist; the other is the
customer/guest/patient. Act out the conversation’. However, pair work
(and group work) are far more flexible and useful techniques than that
suggests.
One
of the constant themes of CLT is that learners need to be given some
degree of control over their learning (since language is a system of
choices, the learners must be given the opportunity to learn how to make
choices). Looking back, again with hindsight, at popular textbooks of
even the fairly recent past, such as
Kernel Lessons Intermediate from the 1970s it is immediately noticeable
that the content of what is said by the learners is controlled at every
point by the book: make a question using these prompts; answer these
questions about the text; read this dialogue, and so on. Even when pair
work is used, the learners never choose what to say, they simply work
out how to say what they are told to say.3 The use of pair
work is a physical signal of some degree of control and choice passing
to the learners; but that needs to be complemented by real choice -
which role play, particularly at simpler levels, may not encourage as
much as other uses of pair work. It is helpful to start from considering
how learners working together can actually help each other. They can
provide each other with a relatively safe opportunity to try out ideas
before launching them in public: this may well lead to more developed
ideas, and therefore greater confidence and more effective
communication. They can also provide knowledge and skills which may
complement those of their partners: this can lead to greater success in
undertaking tasks.
Alternative uses of pair work
Instead of just seeing pair work as a useful
follow-up, a way of getting everyone practising at the same time after a
new language point has been introduced, we can see it as a potential
preliminary stage to any contribution from the learners. They can work
together to do a grammatical exercise, solve a problem, analyse the new
language structures in a text, prepare a questionnaire for other members
of the class, or agree on the opinion they want to present to the
class. Once pair work is seen as a preparation as well as (or more than)
an end-point, the range of possibilities increases dramatically. It is
less a question of: ‘When in my lesson do I get to the freer practice
stage so that I can fit in a role play in pairs?’ and more a question of
‘Is there any reason why I can’t use pair work as part of whatever I’m
planning to do now?’ (Of course, one reason for not using it may be
simply variety - even the best techniques can be overused.)
Misconception 4: CLT means expecting too much from the teacher
It is perhaps cheating to label this a misconception,
since there is a great deal of truth in the argument - voiced most
persuasively by Medgyes (1986) - that CLT places greater demands on the
teacher than certain other widely-used approaches. Lessons tend to be
less predictable; teachers have to be ready to listen to what learners
say and not just how they say it, and to interact with them in as
‘natural’ a way as possible; they have to use a wider range of
management skills than in the traditional teacher-dominated classroom.
In addition, non-native speakers of English probably need a higher level
of language proficiency - or rather, a different balance of proficiency
skills - to be able to communicate with ease, and to cope with
discussing a broader range of facts about language use than they are
accustomed to. Perhaps most importantly, teachers may have to bring to
light deeply-buried preconceptions about language teaching (mostly based
on their own language learning experiences at school and university),
and to compare them openly with alternative possibilities that may be
less familiar but perhaps make better pedagogic sense.
In some ways, there is no answer to these points. It
is certainly difficult, for example, to ignore the charge that CLT is an
approach developed by and for native speaker teachers. Nevertheless,
the label of misconception is perhaps valid for two reasons. Firstly,
the points are presented as defects of CLT, as reasons for rejecting it,
but they can equally well be presented as reasons for embracing it.
Teachers have the opportunity to re-evaluate their beliefs and
practices; they have an incentive to develop their skills; they are
encouraged to enjoy themselves in their work, to avoid dull repetition
of the same predictable set of materials, activities, and answers year
in, year out. This view may appear unduly optimistic to some, but there
seems no reason to assume that the majority of teachers do not welcome
such opportunities - if they are recognized as such. Secondly, the
extent of the demands can easily be exaggerated - indeed, this
misconception may sometimes be fostered by teachers who may have other
reasons for not wishing to change their current practices. Even Medgyes
(1986), in order to make his point more forcefully, ends up by
describing as the CLT norm an unrealistically superhuman teacher that
few CLT teachers would recognize. It can, admittedly, be difficult to
use a communicative approach if you are obliged to use resolutely
uncommunicative materials; but that is increasingly not the case. Many
textbooks now provide very practical, straightforward CLT guidelines and
activities which place few demands on the teacher beyond a willingness
to try them out with enough conviction. The majority of non-native
teachers of English that I have worked with have a high enough level of
proficiency to cope fairly easily with the required shift towards more
fluent and less pre-planned use of the language. And it seems very odd
for language teachers to argue that listening responsively to what other
people say is not part of their job - perhaps teachers who do argue
that should be thinking of going into politics instead?
Conclusion
Given the fairly dramatic change in attitudes not
only to language but also to learners and teachers that came with the
development of CLT, it is not surprising that it has taken some time to
work out the implications for all aspects of the teaching/learning
process. It is, however, worrying that many people’s perceptions of CLT
seem to have got stuck at its early stage of questioning and
experimentation (admittedly sometimes over-enthusiastic), before some of
the key issues were fully resolved. CLT is by no means the final answer
- no doubt the next ‘revolution’ in language teaching is already under
way somewhere. But whatever innovations emerge, they will do so against
the background of the changes brought about by CLT, and will need to
accommodate or explicitly reject those changes. Certain of them are too
important to lose: the concern with the world beyond the classroom, the
concern with the learner as an individual, the view of language as
structured to carry out the functions we want it to perform. In order to
ensure that these changes are not pushed aside in future developments,
it seems essential to attempt to clear away misconceptions that might
otherwise be used to damn them and CLT as a whole.
Notes
1
Interestingly, Harmer rejects the term ‘communicative’ for the approach
outlined in his book. He prefers to call it a ‘balanced activities
approach’, because of the inclusion of controlled, non-communicative
activities as an integral part of learning. However. since the approach
takes communicative activities as the point towards which the other
activities are designed to lead, there seems no reason not to accept
Littlewood’s (1992) term ‘pre-communicative’ for the controlled
activities, and to keep ‘communicative approach’ as the general term.
2 An alternative approach to setting up goals for
language learning is to hold out as the final destination some kind of
abstract mastery of the language (perhaps with a structure-oriented
examination as the final validation). This runs counter to the basic
principles of CLT because it treats the language merely as a
classroom-bound object of study, a pedagogic dead-end. Another
alternative, which does provide an outside, authentically communicative
goal, is to teach the language as a means of preparing to read
literature. This is still accepted as the main aim in many university
courses in particular. However, it represents a demoralisingly difficult
and remote goal for a great many learners. Conversation has the
advantage that it is possible to take part in it reasonably successfully
at many levels, including elementary.
3 This is essentially no different from the way in
which translation is used in the grammar-translation method: the
sentence or text to be translated provides the content, and learner and
teacher only have to worry about how to express that content. This
control of content simplifies the teacher’s task, of course, in that he
or she does not need to judge - or respond to - the appropriacy,
interest, relevance, etc., of what the learners say, but only whether or
not the responses are grammatically correct (see Misconception 4).
