OXFORD

APPLIED
LINGUISTICS

Aspects
of Language
Teaching
H. G. Widdowson

Oxford University Press

J

i
Aspects of Language Teaching

H. G. Widdowson

ESOLCENTRi
WIGAN & LEIGH
COLLEGE

- 4 MAR 200^

Oxford University Press

Oxford University Press
Great Clarendon Street, Oxford 0X2 6DP

Oxford New York

Auckland Bangkok Buenos Aires Cape Town Chennai
Dar es Salaam Delhi Hong Kong Istanbul Karachi Kolkata
Kuala Lumpur Madrid Melbourne Mexico City Mumbai Nairobi
Sao Paulo Shanghai Singapore Taipei Tokyo Toronto

and an associated company in Berlin

Oxford and Oxford English are trade marks of

Oxford University Press

ISBN 0 19 437128 X

© H. G. Widdowson 1990

First published 1990

Fifth impression 2002

All rights reserved. No part of this publication may be reproduced,

stored in a retrieval system, or transmitted, in any form or by any
means, electronic, mechanical, photocopying, recording, or
otherwise, without the prior permission of Oxford University Press.

This book is sold subject to the condition that it shall not, by way
of trade or otherwise, be lent, resold, hired out, or otherwise
circulated without the publisher’s prior consent in any form of
binding or cover other than that in which it is published and
without a similar condition including this condition being
imposed on the subsequent purchaser.

Set in 10/12 pt Sabon.

Typeset by Pentacor Ltd, High Wycombe, Bucks.
Printed in Hong Kong
In memory of Peter Strevens
 i

>•

t tie,.**' *

o
Contents

Acknowledgements vii

Preface ix

PART ONE The theory and practice of language teaching

1 Introduction 1
2 Problems with solutions 7
3 The pragmatics of language teaching 29
4 Pedagogic research and teacher education 55

PART TWO Aspects of language

5 Preliminaries: approaches to description 73
6 Grammar, and nonsense, and learning 79
7 The negotiation of meaning 99

PART THREE Aspects of teaching

8 General perspectives on pedagogy 117
9 TTie problems and principles of syllabus design 127
10 Methodology for the teaching of meaning 157
11 The roles of teacher and learner 181

Conclusion 193

Bibliography 197

Index 209
 n - ■ 'I «■. ;; i.'

xTvsfU .' J TJIAH

, ' ' twc ..t'bi>i*n’ f
r*ifw tny * “ o'!' ( ^

' m. '

” ■ 1 ■' if^imm'*>•.■■: r. i-.P. \,yf/- }i/]

S*#l - '! ^
^ '•. '.' . tq :ji^r ^
., 'N'Lni^T'sM 01
' J' 1' OiOl i/fT fl

f>qUMait>00 >

’‘■’fc,

O'
Acknowledgements

A number of the chapters of this book have developed from

presentations at conferences and published papers. Chapter 2
started life as a plenary address at the 11th Annual Ontario
TESL Conference in Toronto in November 1983. Chapters 3
and 4 grew out of a paper originally commissioned by the
Council of Europe and submitted in September 1986. An earlier
version of Chapter 6 was read at the Fifth National LEND
conference in Rimini in November 1985, and appeared in print
in a collection of papers edited by William Rutherford and
Michael Sharwood-Smith entitled Grammar and Second Lan¬
guage Teaching published by Newbury House in 1988. Chapter
9 is an elaboration of a paper written for a book called Language
Syllabuses: The State of the Art edited by M L Tickoo and
published in 1987 by the Regional English Language Centre in
Singapore. Parts of Chapter 10 began as a presentation entitled
Design Principles for a Communicative Grammar, given at a
TESOL symposium and subsequently published in an ELT
Document (124) edited by Christopher Brumfit. The original for
Chapter 11 was a plenary address at the lATEFL Annual
Conference in Brighton, April 1986. It was published in English
Language Teaching Journal Volume 41, Number 2 just one year
later.
I am grateful to all those people who have given me the
opportunity to make my thoughts public in talk and print.
I would like to express my appreciation, too, to all those
colleagues and students who have stimulated and guided my
thinking over the years by pertinent observation and critical
comment and who have supported me, more than they know, by
their approval and the sense of community that they provide. I
should like to make particular mention of Simon Murison-
Bowie, not only because of the valuable comments he made on
an earlier draft of this book, but also for twenty years of
companionship in the profession.
viii Acknowledgements

The content of this book, then, owes a great deal to the ideas
of many other people. Its compilation owes a great deal to one
person, Sybil Spence. For her unfailing patience, dedication and
care in preparing copy for publication I am extremely grateful.
Acknowledgements are made to the following publishers from
whose texts the extracts and diagrams below have been taken:

Ann Brumfit and Scott Windeatt, Communicative Grammar

2 and 3 (ELTA/OUP 1984)

Emily Dickinson, Selected Poems of Emily Dickinson, ed.

Conrad Aiken (Random House, The Modern Library: New
York 1924)

W. Somerset Maugham, ‘The Unconquered’, Complete Short

Stories (Heinemann 1951)
Preface

This book has been made out of a number of papers written over
the past five years or so. Most of them were originally prepared
as conference presentations, some appeared subsequently in
print. My first intention was simply to put them together as a
compilation with a minimum of additional comment. Then I saw
that there were matters which called for further expounding and
for explicit cross-reference. So I went to work elaborating and
linking. The papers grew into chapters. And as chapters, they
carry the implicit claim that they constitute parts of a reasonably
coherent and uniform whole.
But this manner of composition leaves its traces. The original
papers were designed for particular occasions and purposes as
self-contained statements. Certain ideas and arguments naturally
appear therefore in several places, variously emphasized and
formulated as appropriate to the perspectives of different
discussions. When the papers are brought together and fash¬
ioned as chapters they cease to be separate episodes and become
elements in a sequence. While recurrence is necessary for the
independence of each paper as such, when it appears in chapters,
which are linked in mterdependent continuity, it can read like
needless repetition. What I have done is to pare away some of
the repetition while retaining recurrence where I feel it is
necessary to restate ideas in demonstration of their more specific
relevance to the chapters concerned.
The chapters are assembled into three parts, each of which
deals with a central theme. The first of these perhaps calls for
particular comment because it sets the key, so to speak, for the
discussion in the rest of the book. In it I enquire into the nature
of language teaching as a professional enterprise. What does the
process of teaching actually involve? What do teachers need to
know in theory and what procedures do they need to employ in
practice to actualize that knowledge as effective action? How
can they learn from experience? These questions have to do with
X Preface

the education of teachers and their status as professional people.

They are questions which I have raised on occasions elsewhere,
but I make no apology for giving them prominence here. They
take precedence over all others within the scope of this book.
Unless they are seriously considered, proposals for general
curriculum change or particular classroom techniques are
effectively meaningless, for such proposals have to be mediated
by teachers and such mediation presupposes a degree of
awareness and expertise which only a thoroughgoing and
continuing teacher education can provide.
This might seem an obvious point, but it is one which those
responsible for determining current educational policy in Britain,
for example, seem to have some difficulty in grasping. They are
busy planning a National Curriculum which necessarily requires
a degree of professionalism for which little provision is being
made. Teachers tend to be referred to as if they were factory
workers to be provided with minimal practical skills and
required to pick up on the job whatever extra expertise is
necessary to keep the pedagogic production line going. The
result is that teachers’ morale declines with their status. Many
leave a profession officially treated with such disdain. Further¬
more, expedient stop-gap attempts to provide for the lack of
qualified teachers only make matters worse. I read in my
newspaper today (6 July 1989), for example, of a teacher with
only rudimentary French and no German being engaged to teach
both languages in a secondary school. I^read that it has been
calculated that 25 per cent of teaching in British secondary schools
is being carried out by teachers with an inadequate knowledge of
the subject. Those in authority seem not to be particularly troubled
by this situation. So long as there are people available to stand in
front of classes it does not' seem to matter very much about their
competence as teachers. If it is expedient and cheap to employ
licensed teachers rather, than qualified ones then this solution will
be preferred whatever the consequences for education.
In the light of such attitudes it is particularly important to
assert the professional status of teaching. But the assertion of
status has to be supported by a corresponding commitment on
the part of teachers to standards of professionalism. These
standards, as I argue in Part 1, depend on a continual process of
self-education through an evaluation of practice in reference to
theory. Unless teaching is informed by principled pragmatism in
this way, it can make no claim to be a serious professional
 Preface xi

activity. It becomes hack work. I would argue that teachers who

reject theory as being irrelevant to practice not only misunder¬
stand the nature of their work, but at the same time undermine
the profession. Furthermore they lend support in this way to the
enemies of education and so ultimately act against their own
interests.
Part 1, then, investigates the nature of language teaching in
general as a principled professional activity. It deals with the
interdependence of theory and practice and the appropriate
exploitation of ideas. Part 2 then looks at one theoretical area
which it seems reasonable to suppose is relevant to language
teaching, namely enquiries into the nature of language. Here I
look at the relationship between words, grammar, and context,
at how meaning is encoded semantically within the linguistic
system on the one hand and achieved pragmatically by con¬
textual negotiation on the other. I then seek to show in Part 3
how these perspectives on language description lead to different
approaches to the teaching of language. Thus Parts 2 and 3 are
intended to exemplify the kind of critical enquiry which, as I
argue in Part 1, should direct the work of language teachers. I
say the kind of critical enquiry. It would run counter to my own
position to suggest that the teachers’ professional salvation
depends on their accepting the ideas and arguments proposed in
these pages.
For the validity of ideas and arguments is always relative. To
begin with, whether they are acceptable or not will depend on
the extent to which one accepts beliefs of a general ideological
kind on which they are based. I have already expressed one
belief of this sort, namely that to deny the relevance of theory is
unprofessional practice. If readers do not accept this belief there
is no point in their reading this book—or any other book on the
principles of language teaching. There are two other and related
beliefs which direct the thinking of these chapters and, by the
same token, limit their validity for other minds. It would be as
well for me to make them explicit too.
First there is the belief in the importance of analytic thought. It
seems to me that it is the purpose of education (teacher
education included) to develop ways of applying the intellect to
experience, of pushing rational enquiry as far as it will go. This is
not to deny the value of experience but only to say that this value
is not intrinsic to experience itself but is derived from it by
reflection. Experience, one might say, is the sensation which the
xii Preface

mind must make sense of. It provides the data for analysis. This
should not be taken to mean that all experience is explicable by
reference to reason. Heaven forbid that this should be so. But we
can only really tell what is genuinely inexplicable when
explanation fails. To accept that something is mysterious and
beyond the reach of reason before trying to subject it to rational
analysis is simply to diminish the mystery and make it
commonplace. So, in reference to this book, one can acknow¬
ledge that there are aspects of language teaching which will
remain mysterious, that, irfthe last analysis, teaching is an art
depending on the intuitive flair of individual personalities. In the
last analysis, agreed: for it is only as a consequence of analysis
that one can arrive at such a conclusion. Of course, there
will be differences of opinion about which aspects of teaching
should be analysed and about the validity of the analysis. All
enquiry, as I have said, is limited by preconceived assumptions,
and readers are likely to have some of their own which are at
variance with mine. That is all to the good. Readers may then be
induced to review their own position by the kind of critical
thinking it is the purpose of this book to provoke.
Finally, much of the argument in this book is based on a belief
in the need for preconceived ideas as a condition for effective
language teaching and learning. Teaching and learning: the
order is significant. For the preconceived ideas are used by the
teacher to control the learning process. They define pedagogic
principles. This view is not, I know, a popular one. The notion of
teacher control is anathema in many quarters. It sounds illiberal.
It smacks of prescription and even perhaps suggests the
suppression of human rights. The view which prevails in many
places is one which holds that the description of language use
and the promotion of language learning should proceed without
preconceived ideas, because otherwise the language behaviour of
real people, users and learners, is cramped into conformity and
so misrepresented on the one hand, inhibited on the other.
Instead, it is argued, we should let the people speak, as it were,
for themselves. If they are learners we should let them find their
own natural way as they go, instead of confining them to an
itinerary fixed in advance. This is a seductive doctrine and one of
which it would be wise to be wary. As I have already suggested,
all enquiry presupposes a purpose of one sort or another and so
is primed by a set of ideas. Otherwise, there is no way in which
one can make sense of experience. Language learning is no
 Preface xiii

different. There must always be some points of reference to give

direction to the process and it is the teachers’ task to provide
them. The idea that learners will learn efficiently for themselves
if they are left alone is, I believe, misconceived. If natural learning
was so effective there would be no need for education at all.
Classrooms exist to provide opportunities which would other¬
wise be denied by controlling conditions for learning which
would not otherwise take place. Pedagogy presupposes control
and control presupposes preconceived ideas. The central question
is how this control is to be exercised tactically, tightened, or
relaxed so as to facilitate the learning process: how preconceived
ideas are to be evaluated and modified to accommodate
unpredictable developments in the classroom. It is when control
is inflexible and ideas fixed that they become constraints on
learning. Otherwise they are necessary conditions. It is in the
setting up of such conditions that teachers apply their special
knowledge and expertise and discharge their professional
responsibility.
It is because learners do not learn effectively without the
intervention of properly educated teachers that we need to insist
on the proper professional standards and status of teaching. It is
to those who take the profession of language teaching seriously
in this way that this book is addressed and dedicated.

HGW
London, July 1989

Note
On the matter of controversial pronominals, I have either sought
to avoid them by plurality, or to imply dual or neutral reference
by the random distribution of he, she, and so on. If I have lapsed
I can only hope that readers will neutrally piece out my
imperfections with their thoughts.
 T V. *;

.11. t(I >**t'/'•' '■(iti

r .-t ‘ rj/t-t •*

'•■ I iVl '■

' i
i ■' '■/

’ o'*

. ■' ‘ ' .’•.■"

’ I , ^'' - V '.I-

._ ^ li."

Un'
Iff’ I ‘il I jl

/. ..l.irrjj r' • ••
•Kr^ • ■■r'i
'< <•] ‘ o

il: V***.

lECW
• fl furti
A I'i * . ^ If
■•I'tiS* ■ •>

Vi1j» ■ t ' . • '. pr ■' • I '

. •/• .
'i, .TtJt ■|r'' C. ■}^* ^ I ML. ij]
' ' (•■I 111 ' ••••■. /fi
' ''
il»» m ! w M. •" ;•; •if. '1 o rti^ ' I
.'v *1^1.^. ■ • I “•'t)'',
i: ; ^ \ '
■;Iy» I l. M
■ ‘ IV -.iw) *_« ^
PART ONE

The theory and practice of

language teaching
 MOTJIA^

: %
r
”vbr^)i
1 Introduction

General matters of principle

The purpose of this first part of the book is to establish a
perspective on language teaching. It provides a conceptual
setting for what follows in the succeeding chapters.
This perspective aims to present teaching as a self-conscious
enquiring enterprise whereby classroom activities are referred to
theoretical principles of one sort or another. These principles
essentially define the subject: they are the bearings that teachers
need to take in order to plot their course. The theory which
provides such bearings may come from a variety of sources: from
experience or experiment, from sudden inspirational insight, from
the archives of conventional wisdom. But wherever it comes from,
the theory needs to be made explicit and public if its relevance to
pedagogy is to be effectively assessed.
This is not to deny that individual teachers may be highly
effective in making their own way by an intuitive sense of
direction. The effectiveness of teaching cannot be equated with
its rational accountability. In any classroom there will always be
aspects of the classroom encounter, the play of personality, the
tactics of expedient interaction, which will defy the reduction to
generality.
But if we are to talk about pedagogy, individual effort must be
referred to more general ideas, otherwise there is no way for
experience to be communicated, no way in which others can
derive benefit from the particular successes of the individual.
What principles do is to make private experience publicly
accessible, open to discussion and capable of wider relevance. In
our case, they enable us to claim that there is indeed a pedagogy
of language teaching and a profession which practises it.
So an insistence on the importance of principle in no way
denies the value of experience or customary practices but simply
requires that they are subjected to evaluation, and not just taken
2 The theory and practice of language teaching

on trust. Experience of itself has no significance but can only

have significance attributed to it. Custom of itself is no surety of
effective action. It may of course turn out that there are after all
good independent reasons for respecting the intuitive judge¬
ments which come from long experience. But it does not seem
sensible to accept them unless, they can be given rational
sanction. The contexts of language teaching, like the more
general social contexts within which they are located, are
continually changing, continually challenging habitual ways of
thinking and the patterns-of past certainty. Unless there is a
corresponding process of critical appraisal, there can be no
adaptation, no adjustment to change.

Principle and technique

Principles are abstractions. They have to be actualized as
techniques in the particular circumstances of different class¬
rooms. The teaching task is to see that the techniques that are
used are effective in promoting learning objectives, so they have
to be designed to account for specific contexts of instruction. A
technique may be consistent with a principle but ineffective for a
particular group of learners. This may be a case of inadequate
actualization, and this would call for a change of technique. On
the other hand, of course, it may be that the principle itself needs
to be questioned. How can the teachers tell?
The teachers’ dilemma is the same as that of the researcher
outside the classroom. The researcher has a hypothesis and sets
up experiments to test it. If these do not yield results which
support the hypothesis, it may either be because the experiments
were ineffectively designed, in which case the researcher will
design some more, or because the hypothesis was invalid as
formulated in the first place, in which case the researcher will
reformulate it, and then work out what new experiments are
needed to test it. Language teachers can be seen as involved in
very much the same sort of process—their principles correspond
to hypotheses, their techniques to experiments. Teachers too are
faced with methodological decisions as to where adjustments are
to be made in the matching up of abstraction with actuality.

Teaching as a research activity

Teaching, then, can be conceived of as a research activity
 Introduction 3

whereby experimental techniques of instruction are designed to

correspond with hypothetical principles of pedagogy, with
provision made for mutual adjustment so as to bring validity of
principle into as close an alignment as possible with the utility of
technique.
But of course teachers have extra commitments. They cannot
just assume the researcher role and use students as experimental
subjects, observing how they learn under varying conditions
with detached interest to satisfy an intellectual curiosity. The
teacher’s business is to induce learning and the techniques that
are used have to work to that end. In effect, teachers become
intervening variables in their own experiments. Their research
has to be applied in the very process of enquiry: it has to be
directly accountable in terms of practical pay-off.
This being so, we can regard the classroom as the context for
two related kinds of activity. In one, techniques are devised with
regard to their practical effectiveness in the promotion of
learning. They are directed at the benefit of learners. This we
might call the instructional activity, with the teacher engaged as
participant mediating the techniques concerned. In the other
activity, techniques are related to principles with a view to
enquiring into the relationship between the two. Here they are
directed at the benefit of the teachers’ own understanding of
their craft. This we might call the experimental activity, with the
teacher acting as observer manipulating the techniques con¬
cerned.
Thus the experimental activity and the instructional activity
are reciprocally enhanced, and the most effective pedagogy is
one in which the two act together, each informing and
reinforcing the other. In this way, teaching which provides for
learner development serves the cause of teacher development at
the same time. It fulfils a dual educational purpose.
The view of pedagogy proposed here, then, makes teachers
responsible for defining their own problems and providing their
own solutions. Research from outside, whether descriptive,
experimental or speculative, cannot therefore be directly trans¬
posed to the classroom context. It does, however, have a crucial
role to play in two respects, theoretically and methodologically.
Theoretically, it can serve as a source of ideas and insights which
are of potential relevance for the formulation of principles: ideas
emerging from disciplines devoted to the study of language and
learning which might bear upon the definition of language as
4 The theory and practice of language teaching

subject. Methodologically, it can provide precept and example

of what is involved in critical enquiry, of how intuition can be
subjected to conceptual and empirical evaluation. It can raise
consciousness of the whole process of continuing self-appraisal.
Some people will perhaps feel uneasy about the definition of
pedagogy as operational research' in which experience is pressed
into partnership with principled enquiry. It may seem too much
like confinement, a denial of individual enterprise and the
constraining of intuition into patterns of conformity. It looks as
if it might cramp the teacher’s style. This raises an issue which is
of general relevance to the matters discussed in this book. It has
to do with the relationship between individual initiative and
conventional constraint, with the limits social conditions put on
the freedom of thought and action. This is, of course, an issue of
much wider social and political significance which it is not part
of my brief to explore. But it finds expression in current ideas
and attitudes within the theoretical and practical domains of
language study, language teaching and language teacher educa¬
tion. Since these are within my brief in this book, the issue
warrants consideration here.

The limits of initiative

In very general terms, what has characterized recent tendencies
in the theoretical study and the practical teaching of language
has been a distrust of authority and of the rules and conventions
through which it is exercised. Thus in the theoretical study of
language the deference previously accorded to analysis and
explanation by the informed observer has been questioned in
favour of an uncommitted approach to enquiry, without
preconception and without privilege, into the ways in which
participants negotiate their own conditions for achieving their
purposes. And in language teaching, the idea that the teacher
should direct the progress of learners has been questioned on the
grounds that such direction impedes the natural process of
learning. In both theoretical and practical domains, therefore,
the exercise of authority is seen to result in the artificial
manipulation of the actuality of experience. The description of
language use is thereby distorted. The development of language
learning is thereby disrupted.
While acknowledging the danger of ideas casting reality in
their own image and serving the cause of suppression, one needs
 Introduction 5

to be wary, it seems to me, of being too readily persuaded by the

heady vision of individual freedom which these tendencies
invoke. To begin with, one can discern in this thinking some
confusion between what is authoritative and what is authorit¬
arian, and to suppose that because some preconceptions are used
in the exercise of power, all preconceptions, no matter how
conceived, and no matter how prudently used, are to be rejected
out of hand.
There is here, I believe, a fundamental misunderstanding
about the nature of knowledge and learning. For these are, of
their very nature, dependent on the acknowledgement of existing
conceptual frameworks. Their definition presupposes delimita¬
tion. To know is to have formulated experience in reference to
given categories; to learn is to engage in the process of such
formulation. To be sure, these categories are not inscribed
immutably in the mind: they can be altered to accommodate new
experience. But they cannot be altered unless they are first
apprehended. What is new is only recognizable in relation to
what is given, individual inspiration only recognizable in
relation to accepted social convention. The freedom of inde¬
pendent initiative only exists as a meaningful notion if there are
limits to define it. So the rules and conventions which set these
limits are effectively, if paradoxically, the surest guarantee we
have of freedom of thought and action. They operate negatively
as constraints, but positively as enabling conditions.
The central question, then, is not whether or not we should
tolerate the rules and conventions, the systems of thought, the
preconceptions that regulate enquiry and instruction for if our
enterprise is to have any significance at all we have to—but
which rules, conventions, and preconceptions are likely to offer
us the most relevant and reliable set of bearings for our work,
and how we are to use them so that we can allow for their
modification, or even their complete replacement, when new
insights and experiences need to be accommodated. The
identification of appropriate preconceptions is precisely the
recognition of hypothetical principles I referred to earlier, and
their effective actualization in practice is a matter of experi¬
mental technique.

Applied linguistics
The mediation between theory and practice that I have been
 6 The theory and practice of language teaching

discussing here defines the domain of applied linguistics as this

relates to language teaching.
Its scope delimited in this way, applied linguistics is in my
view an activity which seeks to identify, within the disciplines
concerned with language and learning, those insights and
procedures of enquiry which are Relevant for the formulation of
pedagogic principles and their effective actualization in practice
(Widdowson 1979, 1984a). In effect what applied linguistics
does is to enquire into cross-cultural accommodation: it trans¬
fers ideas and methods from different disciplinary cultures and
seeks to demonstrate how they can be made coherent and
effective in the different conditions of pedagogic practice. Some
people involved in this enterprise concentrate on making ideas
coherent; others focus attention on their relationship with data;
some speculate, and some experiment, and some do both. But
these efforts are of no eventual pedagogic value unless they can
be carried through into the classroom context. Applied lin¬
guistics in this sense must be practised by teachers too if it is to
have any effective operational relevance at all. If applied
linguistics is left exclusively to an elite band of researchers, then
the whole object of the exercise disappears.
The perspective on pedagogy that informs this book, then,
makes considerable demands on the practising teacher. But these
are no greater than are required to substantiate a claim to
professional status. And while they charge the teacher with more
responsibility than has sometimes been customary in the past,
they also bring corresponding rewards in the way of job
satisfaction and self-esteem.
2 Problems with solutions

From the perspective outlined in the preceding chapter,

language teaching can be seen as a principled problem-solving
activity: a kind of operational research 'which works out
solutions to its own local problems. But it is not always seen in
this way. There is a very pervasive belief that it is research in
theoretical and applied linguistics which provides the solutions.
My view would be that it does nothing of the sort and that if we
think it does we delude ourselves. Disciplinary research has a
role, but it is not as a purveyor of solutions. Much of it is
engrossed with problems internally generated within the re¬
search itself, sometimes because of the need for experimental
precision, sometimes because of conceptual ineptitude. But I am
running on too fast. Let me approach the theme of this chapter
more circumspectly by way of literary reference.
Samuel Johnson, a notable figure in the field of language
study, took upon himself the task of compiling a dictionary of
the English language. One of the problems of lexicography, as he
recognized, is how to provide explanations of words by the use
of other words whose meanings are self-evident. As Johnson
himself put it in his Preface:
To explain, requires the use of terms less abstruse than that
which is to be explained, and such terms cannot always be
found; for as nothing can be proved but by supposing
something intuitively known, and evident without proof, so
nothing can be defined but by the use of words too plain to
admit a definition.
The use of simple terms to meet the purpose of a dictionary as
a work of reference for the general public can, however, conflict
with semantic precision. Not all of the entries in Johnson’s
dictionary conform to this principle of plainness. Here, for
example, is the definition of the word network-.
8 The theory and practice of language teaching

NETWORK: Any thing reticulated or decussated, at equal

distances, with interstices between the intersections.
How do we account for this obfuscation? It seems clear that
what has happened here is that Johnson has forgotten about the
problem of finding ways of making meaning plain and has
shifted to a different lexicographical problem, namely how to
provide an exhaustive definition of words so that their entire
meaning is made explicit. In seeking to solve this second problem
of semantic exactitude, he not only fails to solve the problem of
making meanings plain to the public but actually makes that
problem a good deal more complex.

The matching of solution to problem

What I want to point out by this example is that the nature of a
solution is determined by the prior definition of the problem that
goes with it. Problem and solution are a kind of conceptual
adjacency pair, comparable with question and answer in
conversation. It makes no more sense to talk about solutions
without being absolutely clear about what problems they solve,
than it does to talk about excellent answers to questions that
have never been posed. Solutions presuppose problems as
answers presuppose questions, as the sentence (in the legal sense)
presupposes a verdict. ‘Sentence first, verdict afterwards,’ cries
the Red Queen in Alice in Wonderland. However, this is not the
acceptable order in our ordinary world.
But surely all this, you may say, is self-evident. That may be
so, but it does not prevent people in our field from continually
falling into the error of supposing that a solution designed to
match one problem must be applicable to quite a different
problem as well. Let us consider a few cases.

Case 1: Linguistics

A linguist devises a model of grammar in the process of

developing a theory of human cognition. Among those con¬
cerned with language teaching some assume that such a model
must also be directly relevant to problems of pedagogy, that in
general the principles and objectives which define the discipline
of linguistics must necessarily also apply to the subject of
language teaching in the school curriculum. Others assume that
such theoretital enquiry can have no relevance whatever.
Neither position is tenable.
 Problems with solutions 9

We cannot assume that what the linguist identifies as

significant should correspond with aspects of language to be
focused on in the teaching and learning of a language as a school
subject. But this is not of course to deny that relevance or
significance can be inferred, that insights can be drawn from the
discipline and their implications for the subject explored. Indeed
as I have suggested in Chapter 1, it can be seen as the business of
applied linguistics to do just this. But the linguist cannot
determine relevance.
Chomsky, the linguist referred to here, recognizes this. He has
expressed scepticism about the significance of linguistics for
language teaching. But since he acknowledges that he has no
expertise in the field of language teaching, there is no reason to
take his remarks as authoritative. They do, however, raise issues
that have a direct bearing on the theme of this part of the book
and are worth quoting at some length:

I should like to make it clear from the outset that I am

participating in this conference not as an expert on any aspect
of the teaching of languages, but rather as someone whose
primary concern is with the structure of language and, more
generally, the nature of cognitive processes.
Furthermore I am, frankly, rather sceptical about the
significance, for the teaching of languages, of such insights and
understanding ^s have been attained by linguistics and
psychology. Certainly the teacher of language would do well
to keep informed of progress and discussions in these fields,
and the efforts of linguists and psychologists to approach the
problems of language teaching from a principled point of view
are extremely worthwhile, from an intellectual as well as a
social point of view. Still, it is difficult to believe that either
linguistics or psychology has achieved a level of theoretical
understanding that might enable it to support a ‘technology’
of language teaching. Both fields have made significant
progress in recent decades, and, furthermore, both draw on
centuries of careful thought and study. These disciplines are,
at present, in a state of flux and agitation. What seemed to be
well-established doctrine a few years ago may now be the
subject of extensive debate. Although it would be difficult to
document the generalisation, it seems to me that there has
been a significant decline, over the past ten or fifteen years, in
the degree of confidence in the scope and security of
foundations in both psychology and linguistics. I personally
10 The theory and practice of language teaching

feel that this decline in confidence is both healthy and realistic.

But it should serve as a warning to teachers that suggestions
from the ‘fundamental disciplines’ must be viewed with
caution and scepticism. (Chomsky 1965, quoted in Allen
and van Buren 1971 [my italicsf

These remarks of Chomsky have often been quoted, and

adduced more often than not, almost gleefully at times, as
evidence in support of the view that linguistics has no relevance
to language teaching and that therefore applied linguistics, as it
relates to pedagogy at least, is vacuous. But if one troubles to
read what Chomsky actually says here, it is apparent that he
recognizes that linguistics and psychology are associated with
ways of approaching ‘the problems of language teaching from a
principled point of view’. What he questions is whether these
disciplines can ‘support a “technology” of language teaching’,
that is to say, presumably, whether they can directly inform
pedagogic technique.
I see Chomsky’s position as consistent with the view I
expressed earlier, namely that the theoretical disciplines provide
a reference for establishing principles of approach, but they
cannot determine techniques. So ‘suggestions from the “funda¬
mental disciplines” must be viewed with caution and scepticism’.
This is very different from disregarding them altogether. What
Chomsky is indicating, it seems to me, is that the significance of
these disciplines is debatable: their relevance has to be demon¬
strated.
One might (in passing) compare Chomsky’s comments, which
raise issues it is ‘healthy and realistic’ to debate, with the
following declaration of prejudice which simply pronounces the
dogma that there are no issues worth debating:

I do not believe that linguistics has any contribution to make

to the teaching of English or the standard European lan¬
guages. The many people who claim that it has seem to me to
deceive themselves and others. (Sampson 1980:10)

Sampson, of course, has no more claim to authority on

pedagogic matters than Chomsky does. The difference is that
Chomsky acknowledges this fact and accordingly expresses his
views with appropriate caution. Sampson, on the other hand,
does not seem to feel in the least inhibited by his own lack of
competence to pronounce on the matter.
 Problems with solutions 11

Case 2: Behaviourist psychology

A psychologist develops an approach to the shaping of animal
behaviour and demonstrates its effectiveness on rats and pigeons
and then takes the reckless step of assuming that similar
solutions apply to human behaviour. Language teachers, im¬
pressed by the apparent authority of science, follow suit and try
to teach by means of operant conditioning. The solution shifts
from pigeons to people in general to pupils in particular. I refer
here, of course, to B. F. Skinner, whose views on language
learning as essentially a matter of behaviour being shaped by
stimulus control provided theoretical warrant for an approach
to language teaching which focused on habit formation. The
extreme positivist version of behaviourism which Skinner
propounded was effectively demolished by Chomsky in his
review of Verbal Behaviour (Chomsky 1959 in Allen & van
Buren 1971: 136-9, 147-8). In current pedagogic fashion,
behaviourist practices have been largely superseded by cognitive
and communicative perspectives on learning. As a result, there is
now a widespread supposition that there is nothing whatever of
any theoretical validity or pedagogic value in behaviourist
thinking. Its day is done: another day has dawned. Again we see
how much easier it is to be absolute in allegiance to a doctrine
rather than to enquire into the beliefs upon which it is founded.
(For the kind of reasoned enquiry needed into the relevance of
behaviourist ideas to language teaching see Rivers 1964.)
Total rejection of behaviourist theory is no more reasonable
than total acceptance. For when one considers the matter, it is
clear that there must be some aspects of language learning which
have to do with habit formation. Effective communication
depends on the immediate and automatic access to linguistic
forms so that the mind can consciously engage in the more
creative business of negotiating meaning. If these forms were not
internalized as habitual mental patterns independent of thought,
they could not be readily accessed and language could not
function effectively as a means of thinking and communicating.
It is just this point that is made by Lado, whose approach to
language teaching is so often represented as directly opposed to
the development of communicative abilities:

Nothing could be more enslaving and therefore less worthy of

the human mind than to have it chained to the mechanics of
12 The theory and practice of language teaching

the patterns of the language rather than free to dwell on the

message conveyed through the language. It is precisely because
of this view that we discover the highest purpose of PAT¬
TERN PRACTICE: TO REDUCE TO HABIT WHAT
RIGHTEULLY BELONGS TO HABIT IN THE NEW LAN¬
GUAGE, so that the mind and personality may be freed to
dwell in their proper realm, that is on the meaning of the
communication rather than the mechanics of grammar.
{Ladol957)

Just how these habitual patterns are to be internalized for easy

access is another matter: direct inducement by repetitive drill
may well not be the best way, in which case the thing to do is to
think of other techniques. But this does not entail an absolute
denial of the principle. However, the tendency has been to
assume that if behaviourist notions cannot account for all
aspects of language learning it cannot account for any. Both the
uncritical acceptance of Skinner’s concept of the problem and
solution of learning and its equally uncritical rejection in favour
of another serve only to bring pedagogy into disrepute.

Case 3: The Threshold Level

A group of experts in Europe set out to solve problems to do

with language learning in continuing education for adults
beyond formal schooling. Their proposed solutions are trans¬
ferred to schools in Europe and subsequently to schools
everywhere else, and notional/functional syllabuses are peddled
as nostrums all over the place as suitable for every situation.
I am referring to the work of the Council of Europe (see Trim
et al. 1980). This has been an important influence in promoting a
communicative approach to the teaching of languages. It should
be noted, however, that the rationale for such an approach
relates primarily to the ends and not the means of learning, and
is directed at furthering the cause of cultural co-operation in
Europe. The original concern for out-of-school education is
evident in the way objectives are defined in terms of utility:
threshold level specifications are drawn up with an eye to
meeting the needs of learners as eventual participants in contexts
of communicative interaction, rather than with a concern to
activate the actual learning process itself. Their relevance outside
the situation ^or which they were designed, where eventual aims
 Problems with solutions 13

cannot be so readily related to learning objectives, should not

therefore be taken on trust (see Widdowson 1983). But it has
been commonly supposed that specifications along threshold
level lines must be universally relevant, without reference to the
definition of objectives for learning on which the original
threshold level specifications are based. Thus conclusions are
adopted without a consideration of the local validity of the
rationale from which they derive.

Case 4: Humanistic learning

If people learn by caring and sharing and linking hands in
Southern California, it does not follow they will learn by similar
therapeutic techniques in Thailand and Tanzania. I refer here to
so-called humanistic approaches to language teaching, and make
specific allusion to Gertrude Moskowitz’s book Caring and
Sharing in the Foreign Language Classroom (1978). In contrast
with the Council of Europe proposals, which seek to specify
what learners are meant to achieve and so bear mainly upon
questions relating to syllabus design, humanistic approaches are
centrally concerned with the actual process of learning and
therefore have a bearing primarily on methodology. (For a
perceptive review, see Brumfit 1985: 79—95.)
The main aim of humanistic approaches is to draw the learner
into an affective engagement with the learning process, to make
classroom activities meaningful as experiences which involve the
individual as a whole person. They serve as a valuable corrective
to approaches of the kind that a behaviourist view might
encourage, approaches which impose conformity on learners,
reduce the scope of their participation as persons, and deny them
the exercise of individual initiative in the learning process.
However, it is not always recognized that individuality is itself a
cultural concept: there can be no private independent real person
dissociated from the cultural values which define the society in
which the individual lives.
Furthermore, the individual may not want to reveal his private
life in a public role. Thus, encouraging learners to explore and
share their own personality can actually be seen as an un¬
warranted intrusion on privacy, and as the imposition of alien
attitudes, in some cultures and for some individuals. In which
case, it may lead to a disengagement from learning. Activities
which may resolve problems in one cultural (and therefore
14 The theory and practice of language teaching

educational) setting may create problems in another. Pedagogy

in this case actually makes the learning task more difficult.
It is also worth noting, I think, that although ‘whole-person’,
‘humanistic’ approaches seem superficially to be opposed to
behaviourism in their recognition of the individual’s personality,
on closer inspection the two views on learning seem in one
fundamental respect to be essentially alike. Both focus on
affective rather than cognitive factors as central to the process.
One of the main tenets of behaviourism is that behaviour can be
shaped through reinforcement by reward. Thus pigeons learn
patterns of pecking behaviour hy being given grain; students
learn patterns of language behaviour by being given approval. In
both cases acceptable responses are reinforced and learning is
controlled by affective regulation. In the ‘humanistic’ view,
learning is also brought about by the promotion of experiential
well-being, but in this case by affective se//-regulation.
What is underrated in both views is the cognitive dimension.
Regulation can after all be a matter of acting on positive or
negative evidence, the adjustment of behaviour by reference to
feedback information. This evidence and ^adjustment can be
quite free of affective associations. What we have here is
cognitive regulation. Both behaviourism and ‘humanistic’ views
would have us believe that to be affective is to be effective in
setting up conditions for learning. But such a simple formula
really will not do.

>

Case 5: Immersion

If children become bilingual by being immersed in French in

Canada, then perhaps all we need to do is to immerse Chinese¬
speaking children in English in Hong Kong to achieve the same
effect. Immersion in this view would seem to be a kind of
baptism which mysteriously induces the gift of tongues.
The immersion programme in Canada (described in Stern
1978, 1983; Swain 1978, 1982; Swain and Lapkin 1981)
involves the teaching of French contingently by using it as a
medium of instruction for other subjects on the curriculum. It
can he seen as applying principles of ESP to the general
curriculum of school education, presenting language as a service
for the achievement of other than language objectives. It puts
into operation on a more comprehensive scale a proposal for
English teaching pedagogy that I myself put forward, specuht-
 Problems with solutions 15

ively, as appropriate for the Indian context (Widdowson 1968).

How far it would be appropriate to the Hong Kong context I do
not know. The point is that it cannot be presumed to be, without
careful enquiry into the whole range of factors that bear upon
educational decisions. Stern, one of those most closely concerned
with the Canadian immersion programme, sounds a suitably
cautionary note in respect to some of these factors:

Efforts to create bilingualism by means of bilingual schooling

—as, for example, in the immersion programme in Canada—
are likely to be more successful than conventional language
teaching as a subject because the language is treated in school
as a medium rather than as a subject. But even in these cases
the success is likely to be shortlived if it is not backed by
bilingual contacts and exchanges in the community at large.
Thus, the success of language teaching is dependent upon
major forces in society, such as the role, or perception of,
language in that society. {Stern 1983:426)

It should also be noted that the success of such immersion by

medium teaching is not complete. Students appear to acquire
more in the way of fluency than accuracy:

There is evidence to show that after six to seven years of an

immersion program, productive use of the second language
still differs considerably in grammatical and lexical ways from
that of native speakers. (Harley and Swain 1984:291)

It would seem that students need something in the way of

formal instruction as well as acquisition by natural exposure and
engagement. It is not just that one supplements the other:
effective learning would appear to be a function of the
relationship between formal instructional and natural use.

Learners require opportunities for both form-focused and

function-focused practice in the development of particular
skill areas, and if one or the other is lacking they do not
appear to benefit as much . . . Learners who live in what
Krashen has referred to as ‘acquisition-rich’ environments and
take advantage of such settings to use their communicative
skills in the L2, also need opportunities to focus on the
functional properties of the language and attend to form.
[Spada 1986:181-99. See also Spada 1987)

All this would seem to cast some doubt on the idea that in
16 The theory and practice of language teaching

using the language for communicative purposes, the learner

automatically and without other instruction internalizes the
detailed knowledge of language as a generative system available
for general use.

The problem of empirical evidence: an example

Surrounded as we are by solutions of all kinds, each one
supported by persuasive evidence of attested success, we cannot
but be tempted into the belief that somewhere among them there
will be one which matches our particular teaching problem, and
which can therefore be slotted into our situation like a cassette
or a computer programme. But this is a temptation which should
be strenuously resisted. One must beware especially of solutions
that are presented as if they were based on the results of rigorous
experimental tests and bear the stamp of proof.
Consider how easy it is to be misled by the persuasive power
of apparent proof. A group of researchers wish to find out
whether there is any substance in the idea that second language
learners follow a natural predetermined path in their internal¬
ization of the language system they are learning. A test is
devised, known as the Bilingual Syntax Measure (see Burt,
Dulay, and Hernandez 1973), which will require an appropriate
number of subjects to provide certain linguistic responses by
reference to pictures. These responses are then examined to see
how the subjects perform on the production of certain given
morphemes (the progressive morpheme -ing, the plural morph¬
eme, the past tense morpheme, etc.) in linguistic contexts which
would, in native speaker speech, require their obligatory
occurrence.
Now by looking at how the subjects perform on these tests,
our researcher will be able to determine different degrees of
accuracy for each of the morphemes focused upon in this way.
Thus it might be established that the progressive -ing has a high
accuracy rating and the past tense morpheme -ed a low one, that
the plural morpheme -s rates high on this accuracy measure as
compared with the singular third person morpheme -s, and so
on. An interesting finding. But we will suppose that our
researcher is cautious in making claims, and so might wish to
extend the enquiry and see whether the same results can be
replicated with different subjects doing different tasks. After all,
this accuracy order might be a function of this particular test
 Problems with solutions 17

design, or of the particular groups of subjects. More experiments

follow, all, we will suppose, impeccably designed, and the results
seem to confirm the persistence of this accuracy order across a
wide range of experimental conditions. So far, so good.

Accuracy and acquisition

But how do these results relate to the problem we began with?
Do they show that learners follow a natural order in the
internalization of the language system? The answer is: no, they
do not. They provide evidence for an accuracy order in
performance. Now we may wish to assume that an accuracy
order is the same as an acquisitional order, but such an
assumption is not warranted by the evidence, it is based on the
speculation that language learners will reveal what they know,
that their performance will be a reflection of their competence.
But why should we suppose this to be the case? It may be that
learners have internalized aspects of the system which for one
reason or another they cannot access on particular occasions,
that circumstances of different kinds prevent them from acting
on this knowledge. This may seem a somewhat fanciful
proposition, but no less a person than Chomsky has suggested
such a possibility. He expresses it in the following way:
Consider ... a child learning English. Suppose that the child is
at the stage at which he produces so-called ‘telegraphic
speech’, that is, a series of content words without grammatical
elements. Imagine the following (not entirely hypothetical)
sequence of events. At one point the child produces only
telegraphic speech. Shortly after, he makes correct use of
grammatical particles such as do and, let us say, the whole
auxiliary system of English, and does so across the board, that
is, in questions, assertions, negations, etc. At the earlier stage,
the child did not have the capacity to use such items, so his
behaviour indicates. Did he have the knowledge of the
appropriate rules and forms? In the framework I am suggest¬
ing, the answer might be that he did. That is, it might be that
he had fully internalized the requisite mental structure, but for
some reason lacked the capacity to use it. (Chomsky 1980:
53)
He goes on to add:
I see no reason to deny . . . that behaviour is only one kind of
18 The theory and practice of language teaching

evidence, sometimes not the best, and surely no criterion for

knowledge. (Ibid: S4)

Accuracy has to do with behaviour, acquisition has to do with

knowledge. One cannot directly infer one from the other. The
relationship between them is problematic. This point is elaborated
in a recent paper (1985) by Bialystok and Sharwood-Smith (a
development from Bialystok 1982), Here it is pointed out that
there is a difference between knowledge of language and the ability
to access that knowledge effectively in different contexts of use.
Thus a learner may achieve a high accuracy profile in one context,
which might suggest acquisition of certain forms, but perform
badly on the same forms in a different context. This variation
might either be because these forms are tied in some way to a
particular kind of context and so are not freely transferable, or
because the second context imposes inhibiting conditions which
prevent learners from accessing and applying what they know.
Either way, accuracy and acquisition do not match.
There is empirical evidence, too, against inferring acquired
competence from relative accuracy of performance. Examples
have been widely attested of learners who exhibit correct
performance on certain forms, and then lapse into deviance later
on. This may either be because of ‘backsliding’—a reversion to a
previous state of interlanguage—or it may be that the increased
transferability of knowledge, which must to some degree involve
analysis, calls for the recurrent dismantling and reassembling of
linguistic forms which the learner may have internalized as
complete formulaic units. Hence we might find instances of
accuracy in the production of such formulas as fixed phrases but
it does not follow that the learner can transfer the accurate
production of forms to a wider range of constructions and
contexts. Again, particular accuracy is not evidence for general
acquisition (see Ellis 1985; Tarone 1983,1988).
Now one would not wish to deny the value of extrapolation,
of drawing general conclusions from the evidence of particular
instances. But clearly we have to be careful to ensure that
evidence for one phenomenon is not, either by accident or
design, adduced in respect of another. The shift from evidence of
accuracy in performance under certain conditions to general
conclusions about internalization of competence is in principle
no more justified than is the extrapolation from pigeon pecking
to human verbal behaviour.
 Problems with solutions 19

The Fundamental Pedagogical Principle

My reason for dwelling at some length on this matter of
accuracy and acquisition is that the assumption that they are
identical is a corner-stone to the whole edifice of a theory
recently propounded which claims to provide a solution to the
problems of language teaching which have plagued us for so
long. I refer of course to the Monitor theory and its supposed
corollaries. In a recent paper, Krashen, the sole begetter and
main publicist of the Monitor theory, seeks to show how his
theory leads to the postulation of ‘The “Fundamental Pedago¬
gical Principle” in second language teaching’ (Krashen 1982,
Krashen and Terrell 1983). The Fundamental Principle, pro¬
claimed in capital letters with definite reference implying
definitive solution. Is this what we have been waiting for all these
years, the answer that has so far eluded us? Some teachers will
believe so, particularly since it seems to be well founded on
experimental findings. We should take a close look at what is
being offered here before accepting it too readily.
Let us chart the way by which we arrive at the Fundamental
Principle. Experimental studies have shown widespread consist¬
ency in accuracy orders by subjects of different kinds performing
tasks requiring the production of certain linguistic forms. By a
kind of rational conjuring, this is taken to mean that there is a
natural order of internalizution for these forms, a function in all
probability of a universal Chomskyan Language Acquisition
Device. Let us call this acquisition. All the empirical evidence in
favour of accuracy order can now be shifted by sleight of hand in
support of the notion of natural acquisition. That, briefly, is how
the trick is done.

Acquisition and learning

But sometimes this natural acquisition order fails to make an
appearance. On some tasks, subjects will perform with a degree
of accuracy beyond what their state of acquisition should allow.
Flow do we account for this? One might speculate, as I did
myself several years ago (N)(^iddowson 1979: Paper 15), that it is
because the learner draws variably on his interim competence
according to the situational demands made upon it; that his
mterlanguage, m other words, contains variable rules as well as
invariant categorial ones and in this respect resembles fully
20 The theory and practice of language teaching

fledged languages (see Ellis 1985, Chapter 4). Such a suggestion

has at least the face validity of being consistent with the research
done by Labov and others on language variation (see Hudson
1980: Chapter 5 for a review). But this is not the Krashen
explanation for variable performance. Instead we have a
different speculation: since learhers do not always conform to
the natural order (the argument goes), there must be an
Mwnatural order disrupting it. There must therefore be another
process at work affecting performance, a process which is
engaged for certain tasks but not for others. Let us call this
learning, as distinct from acquisition. Learning is then a process
of conscious intervention whereby performance initiated by the
natural and unconscious process of acquisition is monitored, so
that elements which have been learned as formal rules are
grafted on to elements which emerge spontaneously from the
domain of the unconscious. Two processes, then: acquisition,
which is natural, unconscious, primary, and causative; and
learning, which is unnatural, conscious, auxiliary, and corrective.
Now it follows by definition that learning can only be brought
into operation when the occasion allows leisure for conscious
thought about the language being used and its conformity to
rule, when there is time to ‘focus on form’. When learners are
caught up in communication, concerned with making meaning,
they have neither the time nor indeed the inclination to monitor
their performance, which in consequence reveals what they have
acquired without, as it were, the artificial additives of learning.
What empirical evidence is there that might persuade us to
give credence to this sharp and absolute distinction? There are
plenty of references to various studies which are pressed into
support of this theory, giving it the appearance of authority. But
on closer inspection it becomes clear that their validity as
evidence depends crucially on how the key concepts in the theory
are to be defined, and on what empirical conditions they have to
meet to be sustained.
Consider, for example, two of the conditions which are said to
be necessary for Monitor use: time and focus on form. We are
told that ‘in normal conversation, both in speaking and in
listening, performers do not generally have time to think about
and apply conscious grammatical rules’ (Krashen 1981: 3). How
do we know? Such an assertion presupposes a well-founded
theory of performance, one which was able to assign periods of
time to mental processes. The mind would appear to be fairly
 Problems with solutions 21

deft and rapid in its on-line operations: is it really incapable of

referring to rules even when under conversational pressure?
How long does it take to refer to a grammatical rule? Does it
depend on the rule? Does it depend on who is doing the
referring? Does it depend on what kind of conversation is in
progress? I do not know the answers to any of these questions.
But neither, I submit, does Krashen. And then, what does ‘focus
on form’ mean? Is it not possible to focus on form because you
want to make your meaning clear, because you want to be more
communicatively effective? Why should focusing on form not be
consistent with focusing on the message or on content, which is
supposed to preclude the use of the monitor? The sharp dualism
that is proposed whereby acquisition and learning are two quite
distinct processes would seem to force the conclusion that if you
think carefully, choose your words, take your time before
making your conversational contribution you cannot com¬
municate, or at any rate not very effectively, because you are
interfering with the natural function of the acquired system. And
since acquisition depends on communication, your deliberate
delivery will impede your progress in learning the language as
well.
In Krashen’s theory, acquisition is the grand initiator of
messages and the prime mover in communication. Communica¬
tion, reflexively, creates conditions for the process of acquisition
to take place. Learning plays a minor role in communicating and
apparently is not itself affected by it. You cannot communicate
with what you have learnt, nor it seems can you learn by
communicating. But what is meant by communication? This,
clearly, is another key concept in the theory, so how is it defined?
The short answer is: it is not defined at all. But what indications
we do have as to what is meant by communication are not very
reassuring. We are told that the SLOPE test, for example, which
yields results favourable to the natural order hypothesis, is
communicative in character (Krashen 1981). It requires learners
to provide a word to fill a blank in reference to a picture. For
example:

Here is a ball.
Here are two . . . (Picture)
What is she doing here? (Picture)
She is . . . (running, swimming, sitting down)

To suggest that such a blank-filling exercise constitutes com-:

22 The theory and practice of language teaching

municative behaviour is to generalize the concept so much as to

make it almost meaningless. Does the whole conceptual per¬
spective of the Monitor theory narrow down to a focus on filling
in the blanks?

Communication and comprehensible input

But we have not yet arrived at the ‘Fundamental Pedagogical
Principle’. Acquisition is triggered by communicative activity.
But this does not mean, apparently, the active negotiation of
meaning, the realization of speech acts, discourse enactment,
social interaction, or indeed any aspect of language use which
those who have been concerned with communicative approaches
to language teaching have found it necessary to invoke. There is
no mention in the copious references attached to this theory of
Austin or Searle, or Gumperz, Hymes or Labov, or of prag¬
matics, discourse analysis, ethnography or ethnomethodology.
What communication reduces to is comprehensible input.
Exactly what is left out of account in this highly impover¬
ished concept of communication is indicated by Richards and
Schmidt in Language and Communication (1983). This is a
collection of papers which makes extensive reference to the work
I mention here. The beginning of the editors’ introduction runs
as follows;

This book is intended for language teachers, teachers in

training, and students of applied linguistics. The purpose of
this book is to present a coherent survey of major issues in the
study of language and communication, and to show how these
are related to questions of practical concern in the learning
and teaching of second and foreign languages. (1983: xi)

None of these major issues figures in Krashen’s scheme of

things, so he presumably does not regard them as having any
relation at all to the learning and teaching of languages.
Morrison and Low, in one of the papers in Language and
Communication, refer directly to the distortions brought about
by the narrowness of Krashen’s perspective. In particular, they
point out that the phenomenon of monitoring cannot be
confined within the restricted role it plays in Monitor theory:

While the monitoring function is clearly an interesting and

important' aspect of language learning and language use, one
 Problems with solutions 23

feels it deserves to be represented as a complex and subtle

activity, responsive to a variety of social and psychological
pressures. (1983:231)

The authors argue that monitoring, properly understood, is a

crucial feature of communicative interaction. Monitoring, as
understood by Krashen, would appear to play no positive part in
communication whatever.
One sees once more how a failure in the proper conceptual¬
ization of the issues to be investigated can compromise an
enquiry, no matter how much empirical evidence is on display.
The narrowness of vision that Morrison and Low pinpoint is
actually not unlike that of Skinner (referred to earlier in this
chapter), which Chomsky attacked to such deadly effect.
Monitor theory and reinforcement theory would seem to have a
good deal in common.
So it is the provision of comprehensible input which is the
‘Fundamental Pedagogical Principle’. Learners acquire a particu¬
lar language by receiving comprehensible input, that is to say by
being exposed to messages expressed in language which is within
the current acquired competence of the learners, together with
language which is due to appear in the next stage of acquisition,
and which can be eased into the mind by the help of context,
knowledge of the world and so on.
The solution to our problems as teachers, then, is to provide
comprehensible input. It does not appear to matter whether the
learners are particularly interested in this input, or what they
actually do with it, or whether they are actively engaged in
achieving purposeful outcomes. They do, it is true, have to lower
their affective filter to allow the input to flow in, but this might
be achieved in all manner of ways. The theory only requires that
the learner should be a kind of humanoid receptacle in a
maximal state of receptivity so that the input can enter to work
its mysterious way.
In recent Second Language Acquisition (SLA) research (for
example, Long 1983a) there has been a recognition that the
learner does not naturally assume the role of passive recipient
but is actively involved in the process of manipulating the input
so that it is optimally comprehensible. Long points out that in
native speaker/non-native speaker (NS/NNS) interactions, both
participants collaborate in recipient design by employing strat¬
egies for negotiating meaning. In this respect, the learner
24 The theory and practice of language teaching

exercises his own initiative to ensure the proper supply of

comprehensible input.
But the question then arises as to whether all this interactive
effort serves only to facilitate the internalization of linguistic
knowledge, as Long appears to imply, or whether it does not
also develop the executive ability referred to earlier, whereby the
learner can access that knowledge in a range of communicative
contexts. The emergence of this question (which I shall return to
in subsequent chapters) shows how an enquiry which seeks to
provide a solution to one problem generates issues which, if
taken up, can lead to a reconsideration of the problem. In this
case, instead of accepting that comprehensible input is indeed
validly conceived as ‘The “Fundamental Pedagogical Principle’”
and theti looking at how it might be achieved by interaction, one
might more profitably ask whether these strategies of interaction
on the part of the learner do not suggest that the notion of
comprehensible input is inadequate for explaining how language
is learned, and therefore is not so very fundamental after all.
The reduction of the concept of communication to compre¬
hensible input, in association with the absence of any clear
definition of what is meant by comprehensible or any criteria for
knowing whether language has been comprehended or not,
means that almost any approach to teaching can claim to have
the blessing of the theory. Direct Method, for example, provides
an abundance of comprehensible input: Hand. This is my hand.
Ear. This is my ear. This is my ear. Two ears. I have two ears. So
does direct translation. After all, the presence of the first language
is likely to allay anxiety, increase the sense of security and
generally dispose learners to lower their affective filter. At the
same time, of course, it assures them of comprehensible input.
These methods are, in the terms of the theory, communicative
approaches to language teaching. All the problems that we have
been wrestling with over recent years in attempting to formulate
a set of principles for communicative language teaching sud¬
denly vanish as if at the waving of a wand. All our efforts are
rendered irrelevant.
At this point I want to make it quite clear that I have discussed
Monitor theory in some detail not to criticize it as such (cf.
McLaughlin 1978, 1987; Gregg 1984; White 1987), but what it
represents. In fact, I find the ideas of considerable interest, being
myself naturally disposed to speculation. But my purpose in
dwelling on this theory is to demonstrate how ideas are sprea,d
 Problems with solutions 25

by the action of persuasion on uncritical acquiescence and get

converted into solutions, which are assumed to be valid
everywhere, like American Express traveller’s cheques. Monitor
theory is not, of course, the only theory which has laid claim to
such universal solving powers. Behaviourism, with its reduction
of language learning to habit formation, is another example.
There are indeed, as I have already suggested, certain similarities
between the two theories, and they have similar powers of
deception and similar capacities for making mischief.

The relativity and relevance of research

The essential point is that there are no universal solutions. We
should not expect that research will come up with recipes and
remedies which will work whatever the circumstances. We
should recognize that the validity of research findings is always
relative, and relative in two quite different ways. First, it is
relative to the conditions which are imposed on a particular
empirical enquiry. It does not follow that if subjects are induced
to behave in a certain way within the idealized limits of
experimental control, they will behave in the same way when
these limits are relaxed. In a sense, experimental subjects are
only partially real people. ^JC^e come back here, of course, to the
question of extrapolation which I mentioned earlier.
The first kind of relativity, then, has to do with the conditions
that define the methodology of enquiry. The second kind has to
do with the conceptual coherence of the theory that the
empirical investigation is designed to support. If, for example,
we have no clear definition of what exactly is meant by
communication or comprehensible input, how can we recognize
what is evidence and what is not? If we are not provided with a
reasoned argument in its support, why should we accept, as if it
were self-evident, that accuracy and acquisition orders are the
same? No matter how rigorously designed empirical studies may
be, their value is always relative to the explicit formulation of the
theory they are designed to serve. Excellent technology will not
compensate for poor science. The value of empirical research
ultimately depends on the quality of conceptual analysis that
defines the objects of enquiry.
But if research is so unreliable as a source of solutions, what
good is it? In my view, the value of research is that it can help
teachers to define more clearly the problems that they themselves
26 The theory and practice of language teaching

must solve. What it can do is to stimulate interest and encourage

teachers to think about the implications of their practices. It can
also provide them with a conceptual context within which to
work, in the form of hypotheses to test out in the conditions of
their particular classrooms. In short, theory can help practi¬
tioners to adopt a theoretical orientation to their task, whereby
they seek to refer the particular techniques they use to more
general principles, and, reciprocally, test out the validity of such
principles against the observed actualities of classroom practice.
Some years ago, Evelyn Hatch gave characteristically wise
counsel to teachers about applying research findings to class¬
room teaching: apply with caution (Hatch 1979). I would go
further and say: do not apply at all. We cannot exercise caution
unless we know how to interpret such findings in terms of actual
conditions, other than those which define the relative validity of
these findings. This suggests that it is not the findings, the
products of enquiry, that we should apply but the process of
enquiry, the conceptual analysis, observation and experimenta¬
tion which research exemplifies. Simply to apply findings,
without regard to their particular conditions of validity, is to
impose prescribed patterns of behaviour on learners as if they
were subjects rather than people and to make them submit to
solutions which correspond to problems other than their own.
To adopt such a pedagogy is to follow the example of Procrustes
who, it will be recalled, would either stretch his guests on a rack
or cut pieces off their lower limbs so that they would fit his bed.
The pedagogic relevance of research outside the classroom can
only be realized by research inside the classroom. This is not to
diminish the importance of controlled empirical study and the
value of its findings, but to suggest that the extent of its
usefulness can only be established by continuing enquiry and
experimentation in the classroom. This too is research, even
though we may not honour it with the name. I think it is
convenient, and not too misleading, to conceive of research
outside the classroom, both of the soft speculative and the hard
empirical varieties, as resulting in kinds of stereotypes. As with
the stereotypes we refer to in the business of everyday life, we
know they are not, and cannot be, comprehensively true or
correct, but they provide us with an indispensable framework
within which we can interpret particular instances. They are
frameworks for assessment, not formulas to be rigidly applied.
And if people and pupils do not fit the stereotypic formula, we
 Problems with solutions 27

will, I hope, be disinclined to maim them into conformity, but to

adjust the dimensions of the bed to accommodate them.
Over the past ten years or so there has been a great deal of
innovative thinking and impressive empirical investigation in
our field and the theory I have been reviewing has played its part.
All this enquiry has led not to the definitive ‘Fundamental
Pedagogical Principle’, but to a quite fundamental reappraisal of
pedagogic principles in general. That is to say, it has indicated
ways of redefining our problems. But it is idle, and indeed
pernicious, to suppose that it has provided solutions of any
worthwhile generality. We cannot expect that the experience
and experiments of other people in other places occupied with
other problems will produce answers off the peg which will fit
our particular requirements.
This is just as well, otherwise the practice of language teaching
would have no claim to any professional status, for such a claim
is not consistent with a reliance on ready-made solutions. It rests
on a readiness on the part of practitioners to pursue their own
enquiries in the very process of classroom teaching. How such
enquiries might be conducted is a question which is considered
in the next chapter.
 .<• " i: ■ I »■' r < ’f -,n>
/I ' j, . .1 / j (

, • ' ' fi ' ' Hi?-'

■ . ■ I j W > "<^45
f,

. A 4 ■*' •.,
/ ■ •
.> A i'l , .,1 ^ !/• '' .1 r

j # •>71

l*ii A ' . <■

*
mBbI <rr.^. ..
3 The pragmatics of language
teaching

The question of relevance

How should language teachers react to the besetting influence
of ideas about language and learning which emerge from the
domains of disciplinary enquiry, from research in linguistics,
psychology, sociology, philosophy, and education? They can, of
course, declare their independence, refuse, so to speak, to be
beset, and carry on as if these ideas were not there. This does not
seem to me a very sensible strategy, because on the face of it
these ideas would appear to be of potential relevance to
pedagogic practice. The question is, how is this potential to be
realized? How is relevance to be determined? Different domains
of enquiry and action work to different criteria of significance.
There is no reason to suppose that what goes on in one domain is
necessarily relevant to what goes on in another. Relevance is a
matter of significance to one’s own concerns. But the concerns of
pedagogy are the business of teachers. It is surely they who have
to determine relevance in this case, they who have to be
convinced that what research has to say has a bearing on what
they do. In other words, it is they who have to act as mediators
between theory and practice, between the domains of disciplin¬
ary research and pedagogy.
How is the mediating role of teachers to be defined and how
can they be guided to adopt it? Equally, how is language
pedagogy to be defined in relation to other areas of enquiry that
impinge upon it? Unless these issues are satisfactorily resolved,
ideas emerging from disciplinary study cannot be effectively
assessed or acted upon in the pedagogic domain.
This chapter seeks to set up a scheme for language teacher
education, a pragmatics of pedagogy, which incorporates these
conditions for establishing relevance and provides for the
furtherance of proper and profitable relations between theory
and practice.
30 The theory and practice of language teaching

The pragmatics of language pedagogy

By the pragmatics of pedagogy, I mean the working out of a
reflexive, interdependent relationship between theory and prac¬
tice, between abstract ideas deriving from various areas of
enquiry and their actualization in the achievement of practical
outcomes. The realization of this relationship in the act of
teaching (which I maintain is the only way it can be effectively
realized) has reciprocal effects. On the one hand it provides for
the possibility of improved techniques for bringing about
learning; on the other it provides a rationale whereby such
techniques can be explicitly identified as exemplars of more
general principles of teaching. Seen in this way, the reflexive
nature of pragmatism, with theory realized in practice, practice
informed by theory, brings mutual benefits in that it serves the
cause both of effective learning and, as a corollary, of the
professional development of the teacher.
As I indicated in the preceding chapter, innovative approaches
to language teaching that have been recommended in the past
have not, generally speaking, been subjected to this kind of
pragmatic treatment. Such recommendations have a non¬
reflexive and unilateral character: they derive from theory and
determine practice. Typically, the teacher is called upon to
implement the second order realization of new ideas in the form
of teaching materials rather than to consider the ideas them¬
selves and how they key in with his own experience of teaching.
Teaching materials usually provide very little in the way of
explicit rationale which would enable teachers to modify them in
a principled way with reference to the ideas which inform them.
They are designed not for experimentation but for implementa¬
tion. The teacher acts as medium.
But pragmatism, as I am using the term here, is a function of
pedagogic mediation whereby the relationship between theory
and practice, ideas and their actualization, can only be realized
within the domain of application, that is, through the immediate
activity of teaching. In this view, teaching materials are to be
seen as hypothetical constructs, models or exemplars of abstract
principles from which actual instances of pedagogic activities
might be developed in the light of particular classroom con¬
ditions. They have the same ideal character as does any abstract
model of reality and its canonical exemplification, and their
relationship with actual states of affairs is a matter of continual
interpretation and reappraisal.
 The pragmatics of language teaching 31

The dangers of disregarding, by ignorance or design, the

essentially conditional nature of abstract models and of making
data fit into preconceived categories are well attested in the
theoretical domain. The dangers are no less apparent in the
practical domain: they are attested in countless classrooms
where teaching is confined to the transmission of textbook
material rather than its exploitation for the negotiation of
appropriate activities for learning. The relationship between
linguistic theory, the description of a particular language based
upon it, and the way that language is actualized as behaviour in
contexts of use is analogous to the relationship between a
pedagogic theory of language learning, the devising of teaching
materials based upon it, and the way that language is most
effectively actualized for learning in the contexts of particular
classrooms. The relationship is a pragmatic one in both cases:
the connection between the ideal and the real needs to be
established by mediation.

A model of mediation
With regard to pedagogic mediation (teaching as a pragmatic
activity) two interdependent processes seem to be involved. The
first of these we might refer to as appraisal. This focuses on
theory and consists of the interpretation of ideas within their
own terms of reference, within the context of their own
theoretical provenance, and the evaluation of their relevance or
validity in principle with reference to the domain of enquiry
which constitutes the context of application. Evaluation, then, is
the process of specifying what might be called the transfer value
of ideas.
The second pragmatic process might be called application.
This connects with practice and also can be conceived of as a
two-stage operation. First, ideas are actually put into operation
in the practical domain arid then the consequences are mon¬
itored in a second process of evaluation, this time directed at
establishing the practical effect of the ideas as operationally
realized.
Appraisal, then, is a conceptual evaluation based on a proper
understanding of the ideas proposed, and it is directed at
establishing a set of valid principles of general relevance.
Application is an empirical evaluation based on teaching
experience and has to do with the devising of effective
techniques specific to particular circumstances. Application can,
32 The theory and practice of language teaching

of course, lead to reappraisal, just as empirical evidence in other

areas of research can lead to a reformulation or abandonment of
initial hypotheses. There is, however, always the problem of
knowing whether the fault in the mediation process lies in the
validity of the principle or in the particular techniques that have
been used to operationalize it. ^This problem is very much in
evidence in language teaching too. It is a problem that can never
be finally resolved. And it is the impossibility of ultimate
resolution that makes pedagogic mediation a continuous process
of enquiry and experiment. If this process is arrested and
teaching reduced to the manipulation of a set of techniques, or
conformity to a fixed method, then pedagogy as such ceases to
exist.
We might make a diagram of the mediation process in the
following way:

Theory
APPRAISAL
Interpretation Conceptua evaluation
(in principle)

Practice
APPLICATION
Oper ation — Empirical evaluation
(of technique)

Figure 1

Applied linguists and language teachers

It is, I think, commonly supposed that there is a division of
responsibility in respect to these two main aspects of mediation.
Appraisal is seen to be principally the activity of applied
linguistics (or that branch of it that concerns itself with language
teaching) and application (as defined here) the principal activity
of language teaching. In some respects, this is a reasonable
supposition to make. The roles of applied linguist and language
teacher are different; they work to different professional briefs.
The former does not have a full-time commitment to the
classroom and so has no occasion to engage in operational
activities and empirical evaluation as intrinsic to the pedagogic
 The pragmatics of language teaching 33

process. The latter, fully engaged in this process, has little

opportunity to go in pursuit of relevant insights from the
informing disciplines. If one is to recognize that there are applied
linguists and language teachers in the world identifiable by the
difference of role they play (even though these roles may exist in
different degrees of convergence) then the division of respon¬
sibility which corresponds with appraisal and application seems
to be just about the only one to make. But it is the relationship
between these roles, as between the two aspects of mediation,
which is crucial.
I have argued that it is the teachers’ task to mediate through
everyday pedagogic activity: it is their exercise of pragmatism
which should achieve the double objective of learning outcome
and professional development. It follows that it is the applied
linguists’ task to facilitate mediation since their role precludes
any direct engagement with this process on their own account.
Their contribution is to identify ideas of likely relevance and to
present argument and evidence for validity in an accessible way.
Their business is to propound ideas in such a way that their
claimed transfer value is made explicit for the consideration and
possible operationalization by the teacher. Of course, the
propounding may be very persuasive and the teacher needs to be
wary. But the teachers are in charge in the scheme of things
proposed here. The applied linguists have the subordinate and
supporting role. What they have to say by way of appraisal has
no effective force unless it is incorporated into the mediation
process enacted by teachers and under their control.
The relationship between applied linguistics and language
pedagogy is explored in a collection of papers presented at the
Georgetown Round Table on Languages and Linguistics in 1983
(Alatis, Stern, and Strevens 1983). It is of interest to compare the
positions of Brumfit and Krashen in this collection, positions
which might be said to typify to some degree the attitudes to
applied linguistics and language teaching on each side of the
Atlantic.
For Brumfit, applied linguistics is seen as accountable to
pedagogy in the sense that it is the teacher who is the ultimate
arbiter of relevance. The starting point of theory is in the
problems of classroom practice: the researcher’s domain is that
of the teacher.

They [i.e. researchers] must be concerned with providing a

34 The theory and practice of language teaching

model of interpretation of evidence from theory and experi¬

ment, related to the solution of the practical problems which
immediately present themselves to classroom teachers. They
will be demonstrating, in other words, the continuing process
of adaptation and adjustment without which, as we well know
from observation, teaching can easily stagnate into a set of
half-understood routines, performed irrespective of the con¬
ditions of the class or the needs of the learners. (Brumfit in
Alatis, Stern, and Strevens 1983: 62)

Krashen too sees the value of theory as providing for

adaptation:

If we provide teachers with only one method, we are doing

them a disservice. The teacher will be unprepared for change
and will not have the flexibility needed to adapt to new
situations . . . When we provide theory, we provide them with
the underlying rationale for methodology in general. This
permits adaptation for different situations, evaluations of new
techniques and evaluation of materials. (Krashen in Alatis,
Stern, and Strevens 1983:261)

Thus far, Krashen’s position would seem to correspond

closely with that of Brumfit, and indeed to the pragmatic
rationale which I am proposing here. But it becomes clear that
Krashen is not thinking of theory in general, that is to say of a
theoretical perspective on pedagogy, but of a theory in particular
which can be applied directly; not, therefore, of the process of
referring actual problems to abstract ideas but the process of
making practice conform to a preconceived conceptual pattern.
He is seeking to impose a method. There is no pragmatic
reciprocity here. He goes on:

The theory must be a theory of second language acquisition,

not a theory of grammatical structure, and it must be
consistent with all known research, not merely armchair
speculation. (Ibid: 1983:261 [my italics])

The theory which Krashen has in mind is of course his own.

As I have indicated in the preceding chapter, this theory is not in
fact consistent with all known research (it is doubtful if any
theory could be that) nor is it free of speculation.
But the important point to note is that the adherence to a
theory which Krashen proposes actually precludes the exercise
 The pragmatics of language teaching 35

of pragmatic options and the quest for relevance which

characterize a theoretical orientation to pedagogy of the kind
outlined in this chapter. In other words it does not use theory to
activate the crucial process of mediation.

Mediation failure
Mediation can of course fail (and has failed in the past) at
different phases in the process. There can be evaluation of
validity, positive or negative, without the proper interpretation
of ideas. This can lead to fervent acceptance or rejection
uninformed by understanding, the imposition of one mode of
thinking on another. All this raises the whole issue of the
relationship between interpretation and conceptual evaluation
which I touched on in Chapter 2. A failure to recognize that
interpretation and conceptual evaluation are different may lead
to the assumption that ideas from one area of enquiry are self-
evidently relevant to another. It has been assumed, for example,
that the units of language analysis that figure in models of
linguistic description are necessarily valid as units of language
for pedagogic purposes. The central issue here is the relationship
between the description of language and the prescription of
language for pedagogic purposes. The shift over recent years to
the data-based study of actually occurring language, as distinct
from a reliance on the linguist’s intuition as a representative
user, has yielded new information about English—for example,
new facts about frequency of usage and about the use of
language in the achievement of meaning in context. A number of
scholars have shown that these facts are left unaccounted for in
standard descriptions of the language and in the prescriptions
which are provided in syllabuses (Coates 1982; Labov 1984;
Sinclair 1985; Stubbs 1986a). This is what Sinclair has to say:

I think that we are on the verge of a major reorientation in

language description—one that will create problems for
anyone who thinks that the facts are known. I am compelled
to take this view by the early results of computer processing of
language text. The picture is quite disturbing. On the one
hand, there is now ample evidence of the existence of
significant language patterns which have gone largely un¬
recorded m centuries of study; on the other hand there is a
dearth of support for some phenomena which are regularly
36 The theory and practice of language teaching

put forward as normal patterns of English. (Sinclair 198S:

251)

It should be noted, however, that the relationship between

descriptive fact and pedagogic prescription cannot be one of
determinacy. Before the advent of the computer, H. E. Palmer,
Michael West and others carried'out word frequency studies on
a corpus of language data with a view to determining the content
of language courses. They discovered that compromises had to
be made between the descriptive criteria of frequency and range
of language items and pedagogic criteria which were adduced
from the assumed purpose or process of learning (see Mackey
1965: Chapter 6; Widdowson 1968: Chapter 1). One such
pedagogic criterion is coverage:

The coverage or covering capacity of an item is the number of

things one can say with it. It can be measured by the number
of other items which it can displace. (Mackey 1965:184)

The criterion of coverage is likely to prevail over frequency in

cases where the purpose of learning is to acquire a minimal
productive competence across a limited range of predictable
situations.
There is a process version of this criterion which we might call
valency. By this I mean the potential of an item to generate
further learning. Thus we might wish to teach a particular
structure or word meaning not on the grounds of frequency of
occurrence but because its acquisition provides a basis for the
learner to understand and learn other structures or meanings by
extension. The item which is most used is not necessarily the
item which is most useful for learning.
Consider an example from the new Collins Cobuild dictionary
(1987), of which Sinclair is editor-in-chief and which is based on
just the kind of computer analysis of text that he refers to. With
respect to the lexical item bet, analysis reveals the descriptive
fact that what we might call the canonical meaning of the word,
‘to lay a wager’, is relatively rarely attested as compared with its
very frequent informal occurrence as a modal marker indicating
conviction as in expressions like ‘I bet he’ll turn up tomorrow’,
‘There’s no milk in the fridge, I bet’. As Sinclair points out in his
introduction (p. xix), this fact about frequency does not
determine the order of appearance of dictionary entries: there is
the factor of the native speakers’ sense of what is the core
 The pragmatics of language teaching 37

meaning, for example. Nor does it follow that this second

meaning should be given pedagogic preference over the first.
Here too there are reasons for reversing the priority.
In the first place, the informal modal meaning tends to be
restricted to the grammatical environment of first person
singular and present tense and in this respect to have limited
productive generality. It takes on something of the character of
an idiom. With other persons and tenses, the word assumes its
other more fully lexical meaning as in;

I bet her he would turn up yesterday.

He’ll bet there’s no milk in the fridge.

Or the result is an ungrammatical expression, as in:

^There’s no milk in the fridge, he bets.

There are good pedagogic reasons for avoiding idiomatic

idiosyncrasy of this kind.
In the second place, the modal meaning is derivable from the
canonical lexical meaning but not the other way round. Thus if
the learner knows that bet means ‘to lay a wager’, it is possible to
infer the modal sense by extension when it is encountered in use.
It is in this sense that the former has a greater valency and so
constitutes a better learning investment. We might indeed
propose a general principle: high valency items are to be taught
so that high frequency items can be more effectively learned.
Techniques for the analysis of textual data have become very
sophisticated and these lead to descriptive refinements beyond
the scope of Palmer and West. But the relationship between
description and prescription has still to be worked out; the
principle of pedagogic accountability remains in force. This
matter will come up again in later parts of the book. At present
what we need to note is the recognition of the validity of a
description in its own terms does not commit us to acknowledge
its pedagogic relevance in principle. It still needs to be evaluated
in other terms to be effectively mediated into practice.
Again, mediation failure may occur when ideas are put into
operation in the application phase without being subjected to
adequate appraisal. This is what happens in the cases of
textbook transmission that I referred to earlier; the teacher is
required only to put into operation ideas which have already
been realized as materials and is given no guidance in the
evaluation of the validity of the principles on which the materials
38 The theory and practice of language teaching

have been designed even when these principles are clear to the
textbook writers themselves.
A number of implications arise from the model of pragmatic
mediation that I have sketched out here. First, the process, as
outlined in the diagram, is a scheme for research as well as for
teaching conventionally considered: each is seen as a concomit¬
ant of the other, and it is this which provides for the professional
development of the teacher. The question then arises as to how
this pragmatic enterprise differs, if it differs at all, from the kinds
of activity which are customarily carried out under the name of
research. Related to this question of pedagogic/pragmatic
research as an integrated element of classroom practice is the
matter of teacher formation. What kind of preparation or
priming do teachers need in order to exploit their classroom
experience in the manner I have suggested? The concepts of
education and training in the professional development of
teachers are relevant here. These implications are explored in
some detail in the next chapter, but in this one I want to
elaborate a little on the model itself by considering more closely
how it bears upon certain current issues in the pedagogy of
language teaching.

Issues in appraisal and application

1 Interpretation

Interpretation must always be a matter of matching up what is

new to what is familiar: ideas can only be understood in
reference to established categories of thought. There is always
the tendency, therefore, to adjust ideas so that they conform to
what is conventional and customary. Yet conventional cat¬
egories can be modified also to allow for change in concepts and
attitudes. We have, then, two opposing processes at work whose
tensions create the dialectical conditions for gradual change.
They can be seen as large-scale social variants of Piaget’s
psychological processes of assimilation and accommodation
(see, for example, Piaget 1955). New ideas may be assimilated
into given modes of thinking or given modes of thinking may be
altered to accommodate new ideas.
Generally speaking, it has been assimilation rather than
accommodation which has characterized interpretation in the
domain of language pedagogy over recent years. This has had
 The pragmatics of language teaching 39

two effects. On the one hand, ideas have been simplified into
reduced versions which often bear little resemblance to their
originals in the disciplinary contexts from which they have been
taken. On the other hand, the potential of these ideas for the
development of new lines of thought in pedagogy has been
under-exploited.
Take, for example, the idea of ‘communicative competence’.
This is a complex and still unstable concept whose understanding
involves a consideration of a range of issues within discourse
analysis, pragmatics, and the theory of grammar (see, for
example, Canale and Swain 1980; Richards and Schmidt 1983).
Yet the idea is frequently interpreted to mean simply the ability
to produce spoken utterances which are marked for illocution¬
ary function: promising, warning, recommending, agreeing,
predicting, and so on. And a syllabus which defines its content in
functional terms is supposed to account for communicative
competence in a way which syllabuses designed on other
principles cannot.
Consider the following remarks:

The advantage of the notional syllabus is that it takes the

communicative facts of language into account from the
beginning without losing sight of grammatical and situational
factors. It is potentially superior to the grammatical syllabus
because it will produce a communicative competence.
(Wilkins 1976:19)

I would argue that no syllabus, however conceived and

designed, can produce a communicative competence. A syllabus
is simply an inert specification. Only when it is actualized
through classroom activity can it have an effect on learning. The
syllabus is a scheme for teachers and its influence on learners is
only indirect, mediated by methodology.
Whether or not a notional syllabus will help to promote a
communicative competence will depend on just how it is used,
how grammatical and situational factors are taken into account
in the manner of its implementation. Wilkins’ proposals, for
example, have been widely taken to sanction the teaching of
more or less fixed phrases as standard expressions of function
(requesting, inviting, describing, agreeing, disagreeing, etc.). But
the ability to produce such phrases does not constitute com¬
municative competence in so far as I understand the concept
40 The theory and practice of language teaching

(Widdowson 1978, 1979, 1983, 1984a, 1989), or as discussed in,

for example, Canale and Swain 1980, and Bachman 1990.
Proposals for syllabuses based on this reduced notion of
communication also illustrate how the potential for change
inherent in new ideas is stifled by assimilation into established
patterns of thinking. The syllabps is still conceived of as a
collection of atomistic linguistic elements even though they are
now functionally rather than formally defined. It is still regarded
as a preconceived construct which incorporates goals for
learning, and its relationship with a methodology which activates
the immediate process of learning remains as undefined as
before.
But the idea of communicative competence arises from a
dissatisfaction with the Chomskyan distinction between com¬
petence and performance and essentially seeks to establish
competence status for aspects of language behaviour which were
indiscriminately collected into the performance category. Now it
seems reasonable to suggest that the competence/performance
distinction might also mark the division of responsibility
between syllabus and methodology, with the former specifying
the knowledge to be acquired and the latter providing conditions
for its behavioural realization. One might have expected,
therefore, that this issue about the nature and scope of linguistic
description would have been interpreted as a derived issue, so to
speak, in respect of syllabus design and methodology. There is
little sign that this has happened. Over recent years there has
been a growing realization that the development of communicat¬
ive competence is crucially a methodological matter, but we also
need a rationale in support of such a shift of focus. Similarly
there is a recognition that grammar has an importance in
language learning which a. too enthusiastic pursuit of functions
has tended to ignore, and grammar reference books and
exercises of all kinds (and all qualities) are now springing out of
publishing houses. But a proper understanding of the concept of
communicative competence would have revealed that it gives no
endorsement for the neglect of grammar. And now that
grammar is coming back into favour, there does not seem to be
much in the way of reasoned argument for its return. It was left
out before, it is put back now. Why? Many seem to be content to
say that fashions change and that’s that. This is hardly an
attitude which is conducive to a principled development of
pedagogic thinking.
 The pragmatics of language teaching 41

One might say that in the light of the passage quoted on

page 39 there is no endorsement for the neglect of
grammar in Wilkins’ proposals either. The notional syllabus, as
he describes it, does take grammatical factors into account (or at
least, does not lose sight of them). However, it should be noted
that any method of presenting language will take grammatical
factors into account contingently in the sense that any linguistic
expression will of its very nature manifest grammatical cat¬
egories. The issue is not whether such factors will in some way be
represented, because they are bound to be, but whether they are
in pedagogic focus. The central question is how far and in what
ways does a syllabus encourage the teacher to direct learners to
an explicit awareness of grammatical factors. Much of the work
that claims to follow a notional/functional line of approach
leaves grammar as an implied presence: in sight, perhaps, but
not in mind. Whether this is a failure in the interpretation of the
evaluation of Wilkins’ ideas is an open question.
I shall discuss the role of grammar as a necessary communicat¬
ive resource in later chapters of this book. For now, the point I
wish to make is that in regard to the notional/functional
syllabus, although there is a change in the way of describing the
content units for language teaching, this change tends to be
assimilated into traditional and well-established ways of think¬
ing about the syllabus as a pedagogic construct. The fillers
change but they are assimilated into slots which remain the
same. So notions and functions are generally seen as replacing
linguistic structures as units of content, and a notional/functional
(and therefore communicative) orientation is seen to be incompat¬
ible with a concern for grammatical structure and meanings
intrinsic in form. There is little room for accommodation.
This is particularly paradoxical and revealing in the case of
notions. To begin with, the term notion is itself a cause of some
confusion since it has been interpreted in different ways in
relation to its companion function. Wilkins himself consistently
uses the term notion to refer to both kinds of category he talks
about, semantic and pragmatic: the term notion incorporates
functions. Conversely, Halliday uses the term function to refer to
the formal encoding of meaning within a grammar (see, for
example, Halliday 1985). In this semantico-grammatical sense
the term function incorporates notions. Let us clarify matters by
using the term notion to refer to what ^JO^ilkins calls semantico-
grammatical categories, and function to refer uniquely to what
42 The theory and practice of language teaching

Wilkins calls categories of communicative function. This would

seem to accord quite closely with common usage.
Now notions in this sense have, of course, always been
accounted for in conventional syllabus design. The lexical items
and the formal patterns they fill, which constitute the content
units of a structural syllabus, are precisely semantico-grammatical
categories. They are formal in character and have to do with
language at the level of the sentence and its constituents. The
notional/functional syllabus offers nothing of novelty here.
What is novel in notional/functional proposals is the compre¬
hensive inclusion, within syllabus specification, of what Wilkins
calls ‘categories of communicative function’. These have to do
not with formal but with pragmatic meaning. They cannot be
fixed in advance but are conditional on context because they
relate not to sentences but to utterances. The crucial question
that arises here, therefore, is how the given categories of
grammar which constitute the content units of a structural
syllabus are to be associated with the new categories of
communicative function: in other words, how (to use Halliday’s
terms this time) the meaning potential intrinsic in language form
gets realized in the pragmatic achievement of meaning.
The very raising of such a question makes it evident that
notional/functional syllabus proposals actually imply an exten¬
sion of previous practices, a development from the structural
approach to syllabus design with the semantico-grammatical
notions as the essential transition, common to both. But in fact
the notional/functional approach has generally been represented
as a radical departure, a complete break with the past. In
consequence, that aspect of it which links it with previous
practice has diminished almost at times to the point of
disappearance. Functions tend to dominate the scene with
notions appearing in a separate sub-plot in a relatively minor
role. As to the functions themselves, they are deprived of their
pragmatic identity and cast in a role in which they resemble the
grammatical units that they replace.
So we see ideas about the nature of communication inter¬
preted in a reductive way, assimilated into preconceived ways of
thinking to make them amenable to conventional pedagogic
treatment. They are not examined with a view to appraisal
which might require adjustments to concepts of syllabus design
to accommodate them.
In fact, a' close consideration of work done in discourse
 The pragmatics of language teaching 43

processing would raise the possibility of even more radical

accommodation. For example, there is evidence to suggest that a
good deal of communication is achieved not so much by the on¬
line assembly of analysed items but by the adaptation of
formulaic phrases (see, for example, Pawley and Syder 1983). It
would appear that such procedures, whereby pre-existing
patterns are adjusted by differential focusing, are a feature of
first language learning (see, for example, Peters 1983). So it
would seem reasonable to suggest that second language learners
might proceed in a similar way (cf. Hakuta 1974). This would
provide some explanation for the phenomenon of variation in
interlanguage systems (see Tarone 1983; Ellis 1985).
If one follows through the possible implications of this, one
comes to the thought that perhaps language should be presented
to learners primarily in lexical terms, setting conditions for the
gradual emergence of syntax as a focusing device—the very
converse of conventional practice.
This conception of syntax as a kind of lexical auxiliary is
explored in Chapter 5. I mention it here simply to illustrate the
implications that might arise from the more considered inter¬
pretation of theoretical ideas. There are of course cases where
the implications of ideas about communication have been
thought through and subjected to more accommodating inter¬
pretation. It would be wrong to suggest otherwise. But this kind
of critical thinking needs to be encouraged and sustained
throughout the language teaching profession as a whole, as,
indeed, a necessary condition on its professionalism. Otherwise
ideas which should instigate appraisal will continue to be made
into simple tokens for easy assimilation, converted into catch-
phrases or vague, fashionable buzz-words in vogue, and instead
of rational development, we shall continue to get change which
comes only with the vagaries of fashion.
All interpretation requires the reconciliation of the competing
claims of assimilation and accommodation. Too much accom¬
modation and one loses one’s bearings in the necessary con¬
ventions of thought. Too much assimilation and one is becalmed
in the doldrums and no progress is possible at all. The problem is
to know how to strike the right balance. What is needed as far as
language teachers are concerned is some way of making them
aware of the problem as it relates to their professional work and
of providing the means whereby they might arrive at interpreta¬
tions appropriate to themselves.
44 The theory and practice of language teaching

I have talked about interpretation which is accommodating so

that it yields implications for language teaching. But not all
implications are valid in reference to pedagogy. One can accept
that a certain theoretical position or set of research findings leads
to certain conclusions, but one can still question whether these
conclusions are relevant to one\s own case. The dividing line
here is a tenuous one if it exists at all. It is probably a matter of
degree of accommodation: the more you accommodate an idea
the more committed you become to an acceptance of its validity.
Accommodating interpretation implies a degree of evaluation,
which for the purposes of this chapter I would like to consider
separately.

2 Conceptual evaluation
There are two general assumptions which seem to be widely if
not universally accepted as self-evident. They are (to give them
convenient labels) the means/ends equation, and the efhciency of
natural learning. I do not think that in either case there has been
any failure in interpretation of the ideas proposed, but rather
that accommodation has been too readily accorded. In short I
believe the assumptions to be consistent with proposed ideas,
but of doubtful validity when subjected to evaluation.
The means!ends equation and the concept of authenticity. By
the means/ends equation, I mean the assumption that what the
learner has eventually to achieve by way of language ability
should determine what he does in the process of acquiring that
ability. Thus if learners are aiming to communicate naturally,
they need to be prepared for this by being involved in natural
communicative language use in the classroom. In short, the
language of the classroom, has to be authentic. The belief here is
that the language behaviour of natural use, which is the end of
learning, should be replicated as closely as possible in the
classroom as this language behaviour will also be conducive to
learning, to the means whereby communicative ability is achieved.
I have argued against this position elsewhere (Widdowson
1979, 1984a). This is not to say that the position is wrong but
only that it is arguable, or in other words, open to conceptual
evaluation.
Authenticity of language in the classroom is bound to be, to
some extent, an illusion. This is because it does not depend on
the source from which the language as an object is drawn but on
 The pragmatics of language teaching 45

the learners’ engagement with it. In actual language use, as the

work on discourse analysis and pragmatics makes abundantly
clear, meanings are achieved by human agency and are negoti¬
able: they are not contained in text. To the extent that language
learners, by definition, are deficient in competence they cannot
authenticate the language they deal with in the manner of the
native speaker. The language presented to them may be a
genuine record of native speaker behaviour, genuine, that is to
say, as textual data, but to the extent that it does not engage
native speaker response it cannot be realized as authentic
discourse.
Furthermore, if authenticity is to be defined as natural
language behaviour (and it is hard to see how else it might be
defined) there is also the difficulty that learners will naturally
incline to draw on their own language in any situation that calls
for uncontrived linguistic communication. So the situations
which are to stimulate the use of the language being learned will
have to be contrived in some way, and the learners will have to
co-operate in maintaining the illusion of reality. They will have
to be a party to the pretence and accept that the activities in class
are, to use a Goffman term, ‘framed’ as classroom events
(Goffman 1974).
Even when increased authenticity might be judged to be
desirable on motivational grounds, one is faced with the
problem that the process of learning depends on the recognition
of underlying regularities, on the identification of salient and
essential features from all the accidental complexities of actual
behaviour. It may be generally true that the natural language use
which constitutes the goal of learning is realized by a focus on
meaning rather than form, and is a matter of top-down rather
than bottom-up processing, but the process of arriving at that
goal, the development of the authenticating ability, calls for an
effective internalization of form and capability of analysis which
will allow for their use across a wide and unpredictable range of
different contexts. In other words, the very learning process
implies a focus on form as a necessary condition for the
subsequent focus on meaning.
The idea that learners should be discouraged from attend¬
ing to the formal properties of language is comparable to the
idea, prevalent in a previous era, that learners should be denied
all access to translation. Learners will attend to form and make
use of translation anyway because the learning process requires
46 The theory and practice of language teaching

them to do so. A pedagogy which denies this perversely creates

difficulties which hamper the learner in this task. The central
question is not what learners have to do to use lan¬
guage naturally, but what they have to do to learn to use
language naturally. In my view the authenticity argument is
invalid because it does not distinguish between the two
questions: it confuses ends and means and assumes that teach¬
ing language for communication is the same as teaching
language as communication (cf. Widdowson 1984a: Paper 16).
But no doubt these conclusions too are open to challenge. This is
why the issue needs to be evaluated for its pedagogic validity to
be established.
One way of proceeding would be to distinguish different
senses of the term authenticity. This is what Breen does. He
suggests that authenticity can be understood as relating not only
to the language selected to be taught but the task on which the
learner is engaged and the social setting which is created in the
classroom. He expresses these distinctions as four types of
authenticity:

1 Authenticity of the texts which we may use as input data

for our students.
2 Authenticity of the learners’ own interpretations of such
texts.
3 Authenticity of tasks conducive to language learning.
4 Authenticity of the actual social situation of the classroom
language. (1985:61) ''

This distinction enables Breen to say that ‘inauthentic language-

using behaviour might be authentic language-learning behavi¬
our’. The difficulty with such a conclusion is that one can claim
authenticity for anything that goes on in the classroom,
including mechanistic pattern practice and the recital of verb
paradigms, on the grounds that it may be conducive to learning
(type 3) and a feature of the conventional classroom situation
(type 4). This bears some resemblance to the definition of
communication referred to in Chapter 2 and seems to me to
generalize the meaning of the term authenticity to the point
where it ceases to have much significance. I would prefer to
retain the term to refer to the normal language behaviour of the
user in pursuit of a communicative outcome rather than the
language-like behaviour of the learner. Thus inauthentic language¬
using behaviour might well be effective language-learning
 The pragmatics of language teaching 47

behaviour, but to call the latter ‘authentic’ seems to me to

confuse the issue.
However, the kind of enquiry which Breen conducts here, the
interpretation and evaluation of ideas in reference to pedagogic
issues, is a very good example of the process of appraisal that I
am proposing. Its value lies in the way it opens up the issue for
reasoned discussion.
The efficiency of natural learning. Whereas the means/ends
assumption leads to the promotion of authentic, natural
communicative activity based on a belief in the determining
effect of eventual purpose, the assumption of the efficiency of
natural learning invokes the naturalness criterion in reference to
the learning process itself. The former assumption says that
language behaviour in the classroom has to be natural so as to
conform to the naturalness of language use: the latter assump¬
tion says that classroom behaviour has to be natural in
conformity with natural processes of language learning. Where¬
as the first assumption derives from ideas arising from descrip¬
tions of language use, the second derives from ideas emerging
from second language acquisition (SLA) research.
Such research is concerned with the identification and
explanation of developmental stages in the language acquisition
process, with the determinants of natural learning. The assump¬
tion is that if such research is successful, then pedagogy is a
matter of conforming to the revealed sequence of natural
acquisition and of setting up conditions in the classroom which
replicate those of its emergence. What this amounts to, in
Corder’s words, is ‘the accommodation of the structure of our
linguistic syllabuses and teaching materials to fit what is known
of the sequence of progressive complication of the aproxima-
tive systems of the free learner.’ (1981:77)
Note the term ‘accommodation’. Note, too, that such an idea
runs counter to the suggestion that syllabuses should be designed
along notional/functional lines. The implications for pedagogy
of SLA research on the one hand, and studies of communication
on the other, come into conflict and so stand in particular need
of conceptual evaluation. Corder’s proposal also makes the
equation between description and prescription which I question¬
ed earlier. Descriptions of ‘free’ learning sequences cannot
automatically determine the structure of contrived instruction
any more than can descriptions of attested language use. I shall
take this point up again in Chapters 9 and 10.
48 The theory and practice of language teaching

At all events SLA research, though of considerable theoretical

and descriptive interest, has yielded as yet nothing definitive
upon which one might confidently model an approach to formal
teaching which replicates the process of natural learning (see
Chapter 2). What we have are, in the words of T.S. Eliot, ‘only
hints and guesses / Hints follow^ by guesses’. These suggest the
importance of giving primacy to communicative fluency rather
than to formal accuracy (see Brumfit 1984a).
But even if definitive results were forthcoming, why should it
be assumed as self-evident that natural learning is necessarily the
most effective? Most of human progress seems to have come
about by making the contrary assumption that nature can be
improved upon by artifice of one kind or another. Social
institutions, including that of education, are set up to counteract
the shortcomings of nature, to control and exploit it and turn it
to human advantage. The very concept of pedagogy (whether
defined as art or science) presupposes invention and intervention
which will direct learners in ways they would not, left to their
own devices, have the opportunity or inclination to pursue. Now
clearly a teaching approach which goes against the grain of
natural disposition will create needless difficulties for the
learner, as I pointed out earlier in reference to translation and
the focusing on form, but it does not follow that pedagogy must
therefore simply accommodate that disposition.

3 Operation
Operation is that part of the mediation process which is most
readily recognized by teachers as their business. It is here that
techniques of various kinds are put into action to achieve
practical learning outcomes.
These techniques may be the conscious application of ideas
which have been subjected to previous appraisal and are
therefore the realization of principles, or they may simply be a
set of more or less formulaic activities sanctioned not by
appraisal but by the approval of authority. In the latter case,
operation may simply be a matter of conforming to standard
practices established by collective experience over time. Teacher
preparation would then be mainly an initiation, learning the
tricks of the trade from the old campaigners. Or perhaps
authority rqay come from some prestigious source external to
the domain of the practitioner, as when techniques are recom-
 The pragmatics of language teaching 49

mended on the grounds that they have the warrant of proven

theory. On the one hand we have an internal authority which
directs techniques to the maintenance of traditional practices,
and on the other we have an external authority which directs
techniques to the implementation of change. It is not surprising
that they come into conflict, with one side feeling threatened by
innovative theory and the other feeling frustrated by established
practice.
From the point of view of teacher activity as pragmatic
mediation, operational techniques are the enactment of ideas
previously subjected to appraisal which provides them with a
rationale. It may well be that there are good reasons for
following customary practices or for rejecting them in favour of
new ones, but teachers ought to know what these reasons are,
for it is this knowledge which provides for adaptability, the
alternative realization of principles through techniques appro¬
priate to different instructional settings. Particular activities in
the form of classroom tasks may appear to be very different and
to represent quite distinct views of pedagogy but on closer
consideration to be consistent with the same underlying prin¬
ciple of approach. Conversely, activities which seem on the
surface to be the same may turn out to realize different
underlying assumptions. A parallel might be drawn here with the
process of translation. The ability to recast an expression from
one language into the form of another depends on reference to
some underlying construct of meaning which may establish a
resemblance between expressions which appear very different
and a distinction between expressions which appear to resemble
each other. So it is with principles identified by evaluation and
the techniques which put them into operation. In both trans¬
lation and pedagogy, I suggest, effective mediation depends on
the recognition of the relationship between underlying ideas and
the surface forms which give them expression.
The conflicts that I referred to earlier commonly occur at the
level of technique when operation is implemented in dissociation
from appraisal, in the form, for example, of prescribed teaching
materials which have the appearance of novelty. This has the
effect of creating an opposition between established custom
within which teachers feel secure and new ways which they are
under pressure to adopt in the interests of progress. The result is
a crisis of confidence, often resolved by teachers returning to the
assurance of tradition which represents, after all, the more
50 The theory and practice of language teaching

influential authority of the culture in which they have been

nurtured.
But such conflicts can often be resolved and the oppositions
reconciled, at least to some degree, at the level of principle. As I
indicated earlier, the process of appraisal necessarily entails
reference to established ideas and'the balancing of the competing
forces of assimilation and accommodation. It follows that any
proposal for change would be referred to customary ways of
thinking, and this provides for the possibility of operational
techniques which realize n^ ideas being devised as an extension
of existing practices. Thus the damaging divide between what is
established and what is innovative is avoided, the present is seen
to be an evolution from the past and not an abrupt and
revolutionary change of paradigm which requires a radical shift
in allegiance.
Notice that appraisal, stimulated by new ideas, does not
necessarily undermine conventional assumptions and practices.
It may indeed confirm them. But it will do so by making them
explicit, so that in effect techniques become distinguishable
from the principles which inform them. In this way, con¬
ventional practices are rationalized and so made more adapt¬
able, as both Brumfit and Krashen point out.
It is quite common to hear teachers say that they do not
subscribe to any particular approach or method in their teaching
but are ‘eclectic’. They thereby avoid commitment to any current
fad that comes up on the whirligig of fashion. This might be
regarded as prudent common sense. But if by eclecticism is
meant the random and expedient use of whatever technique
comes most readily to hand, then it has no merit whatever. It is
indeed professionally irresponsible if it is claimed as a pedagogic
principle. ‘Haphazard’ rather than ‘eclectic’ would be the more
accurate term. If eclecticism is to be a matter of principle, it
cannot apply independently at the level of operational technique: it
must apply at the level of appraisal which techniques will make
variably operational in the manner previously discussed. Eclectic
techniques in this case will realize an underlying consistency.
This does not mean that they must directly and uniquely derive
from one particular model or method, excluding all others. This
is where I consider Krashen, for example, is mistaken. He has
harsh words to say about eclecticism. In so far as he means the
random use of techniques removed from all contact with
appraisal, then I think that he is right. But he appears to refect
 The pragmatics of language teaching 51

eclecticism as an evaluative principle as well and to argue that

pedagogic practice must be the application of a single theory
of language acquisition, namely his own. (See Chapter 2.) Indeed
Krashen’s work as a whole reveals very little in the way of
appraisal as I have defined the concept: ideas are hurried
through the processes of interpretation and evaluation and made
operational as rapidly as possible. Thus eclecticism is avoided at
the very stage in the process where it is most needed. The teacher
is persuaded into conformity, submits to authority and is
effectively discouraged from engaging in the process of prag¬
matic mediation which, I argue, defines the very profession of
pedagogy.

4 Empirical evaluation
I use the term ‘empirical evaluation’ to mean formative
evaluation in the sense of Scriven (1967). My own term is
intended to reflect the research orientation of this activity. It is to
be distinguished from summative evaluation, which I associate
with assessment, a periodic measurement rather than a continu¬
ous monitoring of process. My view of the function of
evaluation within the learning/teaching process is the same as
that of Breen and Candlin, who comment:

Evaluation within and of the curriculum can be a powerful

and guiding force. Judgements are a crucial part of know¬
ledge, learning and any educational process. By applying
judgements to the curriculum itself, evaluation by the users of
that curriculum can be brought into the classroom, evaluation
can be made to serve as a basis for new directions in the
process of teaching and learning ... it can shape and guide
learning and guide decisions within the curriculum process.
(1980:105-6)

Notice that both teacher and learner are seen as evaluators,

a view I subscribe to (see also Alderson 1985).
Here we are concerned with how the effects of teaching
techniques on learner behaviour are monitored and how the
techniques are modified accordingly. Empirical evaluation differs
from assessment, as the term is generally understood, in that it
focuses not on the measurement of learner attainment as
matched against norms or criteria of success but on the process
of learning and the factors which appear to influence it. It is
52 The theory and practice of language teaching

therefore a continuous enquiry into the relationship between

teaching and learning as it is enacted in particular classroom
contexts. As an integral part of the mediation model I have
proposed, it can, of course, only be carried out by the individual
teacher. This means that guidance for carrying out the task has
to be provided in programmes of teacher preparation. Just what
form such guidance might take is a question I shall return to in
Chapter 4.
Notice that empirical evaluation has implications both for the
operation and the conceptual evaluation phases of our scheme,
as is indicated by the arrows in the diagram on page 32. When
techniques are applied in the classroom and their effects are
monitored, the question arises as to whether and to what extent
these effects call for an adjustment of the techniques as
realizations of a particular principle, or for a reappraisal of the
principle itself. This (as I indicated earlier) is the pedagogic
version of the researcher’s dilemma in general: at what point
does a researcher decide that the hypothesis is untenable in its
present form and needs to be revised? The problem is one which
cannot ultimately be resolved in pedagogy but it needs to be
recognized, not in negative terms as a licence either for imposing
fixed ideas or for random expediency, but as a positive incentive
for continuing enquiry, which will be sensitive to the circum¬
stances of different learning/teaching situations.
Evaluation looks at the relationship between teaching and
learning: it therefore, logically enough, engages the participation
of both teachers and learners. This, again, distinguishes evalu¬
ation from assessment. With assessment learners simply provide
behavioural outputs: they are a source of data. They are not
invited to enquire into the process of their learning or how this is
affected by what teachers require them to do. With evaluation,
as I conceive it here, they take on a more active monitoring role
and, albeit under the ultimate direction of the teacher, become
partners in the pedagogic enterprise. Their attitudes to teaching
and their conceptions of the effects of classroom activities are
taken into account. The teacher’s evaluation of the effectiveness
of a technique will depend in part on the way the learners
evaluate its effects. It will not just be a matter of noting what
linguistic behaviour it provokes.
Evaluation, then, is to be distinguished from the measurement
of learner achievement. It is also to be distinguished from the
measurement of method. By this I mean schemes that have been
 The pragmatics of language teaching 53

devised in the past to establish the efficacy of a particular

approach to teaching. These necessarily seek to cancel out
particularities in the interest of identifying a set of determining
principles. The peculiarities of teachers and learners as interact¬
ing participants are eliminated as intervening variables, and the
evaluation is done by avoiding the mediation process which I
have been outlining. The evaluation I have in mind, as an
intrinsic element in pragmatic pedagogy, is itself a function of
teacher mediation and seeks to plot a course through the
variables of particular classroom encounters. Its purpose is not
to prove the efficacy of any particular method but to use a set of
principles as bearings for the development of different tech¬
niques.
My purpose in this chapter has been to propose a model of
language teaching and to show the interrelationship between its
different elements. It is the activation of this interrelationship
which defines teaching as a mediation process. But as I indicated
in Chapter 1, and is apparent from subsequent discussion, such a
definition represents teaching as essentially a research activity.
We need now to consider the question of the nature of pedagogic
research, how it differs from other kinds of research activity, and
what implications arise from such considerations for the
education of teachers. These matters are taken up in the next
chapter.
',.1 *' i -*

i • T- <

^ >1

( .V.' .•■■ -r

^ 1 . .i;»s ■ ;i. ? /rq

I ^ nji ft!
I . I , •»'Tj ^ •
.1 . i ■ '’•M
rT- ■■- j. ■■' • ' iC

*>• % ' ■ . rSK

/■ Ij

li - -

)lh« 1^' 41^ >»' »>Jj

«s*4 flMIHrtM: -r tfee
Y 'X t. V' *sr# ^ ^ mJ ir

<• k«9k4. r !•' •

O
4 Pedagogic research and teacher
education

Researchers and teachers

Research is commonly taken to be the specialist and reserved
occupation of theorists, an activity which is carried out in
detachment from the immediacy of actual events and requiring
knowledge and expertise of a kind which only academic
intellectuals can legitimately claim to have. The activity yields
findings which have the weight of authority and the stamp of
truth. This view of research in respect of language teaching has,
as I showed in Chapter 2, led to an unfortunate separation of
roles which has proved damaging to the pedagogic cause: the
researcher as the producer of truth on the one hand, and the
language teacher as a consumer of it on the other. The
paradoxical effect of this has been to put the researcher,
including the one who claims to be working within applied
linguistics, at a remove from the only contexts of application
which can provide substantiating evidence for the relevance of
the research. At the same time, the people who create these
contexts, the teaching practitioners, are often made to feel that
their own experience as pedagogic providers is not given
sufficient recognition but is, on the contrary, misprized as
lacking in rigour: wisdom which leads to understanding is
overridden by knowledge which leads to explanation. The
consequence of this division of roles is the kind of distrust and
absence of fruitful collaboration which marks any lack of
reciprocal relationship between cultures (of which more later).
What is needed, then, is a concept of research and teaching
which will bring them together into reciprocal dependence and
so reconcile the requirements of relevance and rigour. The model
of pragmatic mediation that I am proposing here is designed to
meet that need.
56 The theory and practice of language teaching

The nature of research

To clarify how the model outlined in the previous chapter

relates to, and draws upon, research as it is academically defined,
it will be helpful to start by considering the nature of research in
general terms. The range of activities carried out under its name
is very diverse and different in kind, from armchair theorizing to
the detailed accumulation and analysis of data, from meta¬
physical speculation to psychometric measurement. What they
all have in common is thafthey are all attempts to go beyond
appearances. They all seek, in their different ways, to discover
abstract categories and connections underlying familiar phe¬
nomena. The basic motivation of all research is the conviction
that things are not just as they appear to be, but are examples of
covert categories and relations of one sort or another. What
research does is to reformulate the familiar so that it assumes a
new significance. Once such a formulation is proposed then it
can of course be applied to other phenomena (other than those
originally used) and these are explained as additional instances
of the abstractions that have been established. Moreover, the
specification of new categories and relations provides the means
for manipulating phenomena by giving their underlying features
a realization which is not currently attested. Thus discovery
leads to invention. Research enables people to be aware of
different ways of conceiving of the familiar world and, if they act
on the new conception, to alter or extend their customary ideas
and practices. In this way research can be seen as a continuing
dialectical process: actuality is formulated as an abstraction
which in turn leads to a reformulation of actuality.
An obvious question arises from this: how do we know
whether the particular reformulations of a given piece of
research are valid or not? After all, one might dream up all kinds
of outlandish ideas, and make them look good by persuasive
argument decked out in the trappings of authority and the
careful selection of data. There are plenty of historical preced¬
ents for this, as I pointed out in Chapter 2. What they show is
that no matter how objective or impartial enquiries may seek or
claim to be they are always in some degree preconceived because
they are preconditioned by cultural assumptions of one sort or
another. The very conditions of systematic enquiry make this
unavoidable. For such enquiry is based on idealization, the
extraction of what is seen as essential from observable data, and
idealization obviously cannot be free of ideological influence: it
 Pedagogic research and teacher education 57

must be related to some pre-existing framework of belief. In

other words, research is bounded by the same sort of conditions
which bear on appraisal, as discussed above. Researchers
enquire selectively into what they believe to be essential
according to their conceptual bent, leaving the rest aside as
incidental. Thus they are naturally predisposed to find what they
are looking for. How, then, can validity be established?
Validity is a relative concept. Its relativity is a function of
degrees of cultural affinity, the extent to which the informing
disposition of the researcher is shared by others within the
community to which he belongs. Although a researcher in a
particular discipline, that is to say in a particular academic
culture, will work within conventions of enquiry defined by that
discipline, these will obviously in part be shared by other
disciplines and in part shared also by the wider community
within which the academic culture is located. The situation here
is a familiar one in anthropology: the recognition of degrees of
commonality of attitude and belief which provides for the
variable definition of what constitutes cultural boundaries.
So it is that we can apply to research of different disciplinary
provenance general criteria of appraisal approved by the wider
culture of intellectual enquiry. There are two kinds of general
criteria. One of these is conceptual and bears on the logical
coherence of theory. The other is empirical and bears on the way
the data of actuality are used to substantiate the theory. Thus in
relation to conceptual criteria, research might be criticized on
the grounds of inconsistency of reasoning or imprecision of
terminology or general intellectual ineptitude, using the stand¬
ards of argumentation approved by our culture (though not
necessarily by others). The employment of these general con¬
ceptual criteria is central to the process of interpretation, as
discussed in Chapter 3. As I indicated there, their employment is
often prevented, however, by too ready an assimilation of ideas.
If ideas sound good and meet the needs of a current state of
affairs, they will tend to be accepted in some simple assimilable
form without being subjected to scrutiny. With regard to
empirical criteria, research can be criticized on the grounds of
faulty methodology in the design of instruments for assembling
data or in the means used for measuring them. Whereas
conceptual criteria relate to how the abstraction has been
formulated empirical criteria relate to how the abstraction is
actualized in data.
58 The theory and practice of language teaching

These general criteria apply at the interpretation phase of the

pragmatic process I am proposing here. But evaluation calls for
more particular consideration and a closer look at the cultural
differences betw'een the domain of disciplines concerned with
language and learning and the domain of the subject of language
as it appears on the curriculum.
I have already discussed the more particular validation of
ideas as they bear upon issues in language pedagogy in terms of
the evaluation of principles in respect to transfer value.
Conceptual criteria, then, as they are more specifically invoked
to establish relevance across cultures or domains (from lin¬
guistics and other disciplines to language pedagogy), are used in
the process of appraisal. The more particular examination of
validity in respect of empirical criteria, however, has to do with
the evaluation of techniques since it is these which refer
abstractions back to actuality. Empirical criteria have to do with
the process of application. But the pedagogic methodology of
application is not the same kind of operation as the conventional
research methodology of selective observation and controlled
experiment.

The methodology of teaching and research

There are two ways in which a theoretical abstraction can be
matched against the actuality which it claims to account for.
One of them involves the observation of^events as they actually
occur, and the other involves experimentation which causes
events to take place. Both operations are exercises in data
control: the former by selection, the latter by manipulation. A
good deal of observational research has been done on what goes
on in classrooms, including language classrooms. Lessons have
been analysed according to descriptive schemes of different sorts
based on the abstraction of what the researchers have considered
to be the most salient or significant features of classroom
behaviour. In such classroom research, the teacher is part of the
data to be observed. Work of a manipulative kind has also been
done by using classes of learners as subject, but it is not the
teacher, in her normal teaching role, who conducts the enquiry
in these cases. If she is not data she is an intervening variable to
be eliminated. Either way she is not in charge.
But in th^ir professional role, teachers are, of course, in
charge: they are not intervening variables but intervening agents.
 Pedagogic research and teacher education 59

Their business is to manipulate behaviour. Unlike the experi¬

menter, however, they cannot, because of their prior respon¬
sibility to their pupils, simply eliminate or neutralize aspects of
the learning milieu which are inconvenient for the conduct of
experiment. Some of these aspects may well combine in complex
and inexplicable ways in the learning process and clearly the
teacher cannot take the chance of depriving learners of effective
learning conditions on the grounds that they cannot be explain¬
ed or controlled. The people in her charge are pupils and not
experimental subjects. The teacher is accountable in a way that
the experimenter is not: her purpose is to promote the
understanding of her pupils and not, except incidentally,
her own. Futhermore, as participants in the learning/teaching
operation, pupils have a role in the evaluation process, as I
pointed out earlier. They do not just generate data but enter into
collaboration in the continuous monitoring of classroom activity
and its effects.
There is, then, a difference between ‘outsider’ research
(whether observational or manipulative) which is carried out by
the external researcher with teachers as part of the data or the
object of enquiry, and ‘insider’ research which is carried out by
teachers themselves. We might refer to the former as ‘classroom-
oriented research’ (see Seliger and Long 1983) and the latter as
‘classroom-centred research’ (see Allwright 1983). How far
these two perspectives presuppose different procedures of
enquiry is a matter of considerable current debate (see, for
example, Mitchell 1985; van Tier 1988). One view is that insider
research calls for the free-ranging exploration of what goes on in
the classroom without the constraint of any preconceived
theory. Such a view would, of course, be in conflict with the
model I am proposing here, which involves the teacher in the
experimental application of ideas.
But at the same time, the classroom cannot simply be an
extension of the experimental domain. The empirical evaluation
of teaching/learning activities does not aim at the exact
specification of causes but at increased awareness of different
factors which bear upon the learning process in particular
classrooms. Certain generalities might well emerge, certain
principles may be borne out, and these can form the basis for
discussion with other teachers working in different circum¬
stances, and the basis for variable realization as different
techniques, as I indicated earlier. But classroom practice is not
60 The theory and practice of language teaching

put at the service of the principles, designed to test them out;

conversely, the principles serve classroom practices in that they
provide a way in which they may be more clearly understood
and more systematically carried out.
The point I am making is that pedagogic research calls for the
independent appraisal of ideas' as a precondition to their
application. That application cannot simply model itself on the
procedures of empirical research, however, since it has to
operate within contexts which preclude it. This does not mean
that these procedures cannot be adapted to classroom use.
Indeed, it may be that the best approach to the design of
guidelines for empirical evaluation is to begin with familiar and
well-tried procedures for the elicitation and observation of data
and see how far they can be adapted to serve a pedagogic
purpose. A comparative study of research procedures and
teaching techniques would be likely to provide a basis for
evaluation which combines the qualities of research rigour and
pedagogic relevance.
The value of using observational research procedures for the
purpose of evaluation is indicated by Long in the following
terms:

It should be stressed that the aim in providing teachers with

training in observational procedures is not to turn them into
classroom researchers . . . Rather, the idea is to equip them to
assess what they are doing when back in the field, and to
monitor any innovations they make in their teaching.
il983b:288)

Long gives a number of instances of such procedures. The

interesting question, in the present context, is how far these can
be co-ordinated with familiar pedagogic techniques, thereby
bringing research within the compass of normal teaching; or, in
more general terms, what implications does work on classroom
observation have for developments in teacher education (see
Allwright 1988).
One further observation might be made about pedagogic
research. The idea that teachers should adopt a research
perspective in their teaching has been current in educational
thinking (in Britain at least) for a good number of years,
associated in particular with the work of Stenhouse and Elliott
(for example, Stenhouse 1975; Elliott 1981). The term ‘action
research’ is used to refer to such research-oriented teaching. But
 Pedagogic research and teacher education 61

this is sometimes taken to mean the formation of groups of

teachers to generate their own ideas without reference to theory
and research of a more rigorous and academic character. ‘Be
your own expert’ is sometimes the cry and the do-it-yourself
attitude extends to educational thinking. This, it seems to me, is
symptomatic of a dangerous anti-intellectual trend in the
profession. I have been arguing here for what might be termed
appropriate research, analagous to the notion of appropriate
technology. But this does not mean that it has to be uninformed
and makeshift. The academic disciplines of linguistics, psycho¬
logy, sociology, and education provide the essential bearings for
the professional teacher both in the ideas they generate and in
their processes of enquiry, and it is the task of applied linguistics
(whose name belies its scope) to make the insights these
disciplines offer accessible for appraisal and application. The
kind of pragmatic approach to pedagogy that I have outlined
here depends on the recognition of the importance of intellectual
effort and puts theoretical rigour at the service of practical
relevance.
It is perhaps worth saying again that this argument for a
rational basis for pedagogy does not imply that everything about
good teaching can be reduced to what is rational. There is no
denial of the value of imagination and intuition, the instinctive
feel for what is effective. As Brumfit observes, ‘Being a good
teacher is never exactly the same thing as understanding
teaching; there are many things we can do well that we cannot
fully understand.’ (1987: 29) I would only add that the value of
what we cannot understand is only recognized by reference to
what we can. The principles of pedagogy have to do with the
craft of teaching. They cannot account for the artistry of the
individual practitioner. On the other hand, individual artistry is
likely to be enhanced by an increased consciousness of craft.
The research-oriented approach to teaching that I have been
proposing here obviously sets a high premium on teacher
preparation, and that in turn raises the issues of the education
and training of teachers.

The education and training of teachers

How can teachers be prepared to undertake the process of
pragmatic mediation that I have been proposing in these chapters?
The requirement for a research perspective which this necessarily
62 The theory and practice of language teaching

entails, the relating of abstraction to actuality, the use of

technique to realize principle, and indeed the whole process of
self-monitoring, presupposes attitudes and abilities which call
for the education and not just the training of teachers. Indeed, as
I have already indicated, it seems to me that the very claim of
pedagogy to professional status is based on the same supposi¬
tion. What then does teacher education involve and how does it
differ as a concept from teacher training?
In general terms, the distinction between education and
training can be formulated in the following way. Training is a
process of preparation towards the achievement of a range of
outcomes which are specified in advance. This involves the
acquisition of goal-oriented behaviour which is more or less
formulaic in character and whose capacity for accommodation
to novelty is, therefore, very limited. Training, in this view, is
directed at providing solutions to a set of predictable problems
and sets a premium on unreflecting expertise. It is dependent on
the stability of existing states of affairs since it assumes that
future situations will be predictable replicas of those in the past.
Education on the other hand is not predicated on predictability
in this way. It provides for situations which cannot be
accommodated into preconceived patterns of response but
which require a reformulation of ideas and the modification of
established formulae. It focuses, therefore, not on the application
of ready-made problem-solving techniques but on the critical
appraisal of the relationship between problem and solution as a
matter of continuing enquiry and of adaptable practice. For
further discussion see Peters 1967, 1973; Widdowson 1983, and
with particular reference to language teaching, Larsen-Freeman
1983.
It should be noted that this view of teacher education does not
give warrant to abstract speculation for its own sake in
dissociation from actuality. Rather it provides for the initiative
of invention whereby actuality can be variously interpreted and
changed. This does not mean that the educated teacher will
continually be in quest of innovation and reject conventional
practices out of hand. It means only that ideas both given and
new will be subjected to scrutiny and not simply accepted on
trust. The reason for the familiar bandwagon phenomenon is to
be found in educational failure.
Teacher education, then, provides for the appraisal of ideas in
order to make them more practically effective, because an
 Pedagogic research and teacher education 63

understanding of abstract concepts and their relationships

allows for adaptability in their realization. It follows that such
an educational perspective on teacher preparation does not deny
the importance of practical technique. On the contrary, it
proclaims its crucial role as effecting a renewal of connection
with the actuality of particular classrooms contexts. But
techniques are not goals in themselves, as they are in a training
perspective; they are the means for making ideas operational and
subject to modification in the light of evaluation. They do not
determine the activities of teaching and learning, but on the
contrary are themselves required to conform to the changing
perceptions and experiences of the teacher and learner.
The way techniques are to be considered in teacher education,
not as procedures to conform to but to exploit, bears on the
question of the use of existing teaching materials in the
preparation of teachers. These teaching materials are techniques
made manifest by the specific choice of language, designed and
arranged as a series of tasks or exercises for immediate
implementation. To the extent that such materials are realiza¬
tions of principles, instances of what exercises informed by
certain ideas might look like, then they are illustrations. They
can be assessed by reference to abstractions. To the extent that
such materials are designed for direct implementation with the
implicit claim for general pedagogic effectiveness, then they are
prescriptions. They presuppose that the particularities of dif¬
ferent classrooms are not determinants of teaching and learning
but are incidental. In this case, they can only be assessed by
reference to actuality.
As illustrations, therefore, teaching materials can be seen as
stimulants of enquiry calling for appraisal as a prerequisite for
application; as prescriptions, they call for application without
the requirement of appraisal, and so constrain the users into
conformity. A training perspective would obviously tend to¬
wards the prescription view of teaching materials, an educa¬
tional perspective would require the treatment of such materials
as illustrations.
This distinction between the illustrative and prescriptive
character of teaching materials, though obvious enough, does
not always seem to be recognized. The materials which I
designed myself as illustrative of a particular approach to
teaching (for example, Widdowson 1978; Allen and Widdowson
1973), have been criticized on the grounds that they have not
64 The theory and practice of language teaching

been subjected to evaluation and thereby given the seal of

practical effectiveness (Murphy 1985). But the point about the
evaluation of materials is that it cannot be carried out in
dissociation from the contexts of particular classrooms: it must
be part of the process of self-monitoring mediation that I am
advocating here. In this view there can be no possibility of any
global approval or rejection. It is interesting to note that
Murphy’s concept of evaluation in this respect is at variance
with that of Breen and Candlin which I referred to earlier.
I have talked so far about teacher training and education
without distinguishing between pre-service and in-service pro¬
grammes. It seems reasonable to suppose that there will be
differences in the kind of provision that needs to be made by
each of these stages of teacher formation. These differences are, I
believe, closely related to the training and education perspectives
outlined above.
Pre-service preparation initiates the prospective teacher into
the basics of professional activity. These are, in general, of two
kinds. There are those which relate to the craft of classroom
management and the use of routine procedures for organizing
class activity, the tricks of the pedagogic trade. Here what is
important for novices is the development of confidence when
confronted with a new and demanding, indeed threatening,
social situation, and for this they need to draw upon a set of
established and reliable techniques and learn to feel secure in the
straightforward business of actually putting them into practice,
whatever their validity in terms of learning effect might be.
Novice teachers clearly have to feel secure in their own role, and
establish their own identity, before indulging in experimentation
which could undermine their authority before they have actually
acquired it.
The second basic element of initiation is that which relates
newcomers to their fellow teachers, a process of acculturation
whereby they become members of the group. This too inclines
them to accept a set of conventional attitudes and practices. To
adopt too critical or enquiring a position at this stage would be
to run the risk of alienation. In their initiation into both the craft
and culture of pedagogy, it is in the interests of novice teachers
to conform in order to place their relations with pupils and
fellow teachers on a secure base and so to get established in their
role. In view of this, it seems clear that pre-service or initial
preparation needs to pay particular attention to training. This is
 Pedagogic research and teacher education 65

not to say that such courses would not also encourage an

awareness of wider theoretical implications or the kind of
appraisal I have associated with education, but this would be
more in the manner of a long-term investment rather than
something expected to yield immediate returns, something
which might influence attitude rather than instigate action.
The instigation of action, indeed of action research, is a matter
which must mainly be taken up by in-service provision. And it is
here, of course, that an educational orientation is required in
order that teachers might enact the role of mediator along the
lines I have been proposing. There is, of course, very extensive
provision already made in the field of in-service education for
language teachers, ranging from award-bearing year-long
courses in universities to the relatively informal meetings of
teacher groups on a self-help basis. With such programmes there
is, however, a persistent problem of renewal of connection with
the classroom. This is perhaps more evident in the case of longer
courses where teachers are displaced from their pedagogic
habitat for considerable periods of time, but it exists also in
shorter courses. What happens very often here is that particip¬
ants are inspired by the social and professional intensity of the
event but find that they have little to carry home with them
except a heady sense of general enlightenment which is often
quickly dispersed on its contact with reality. This is not to deny
the value of such courses; they provide, at the very least, a sense
of professional community and there is no doubt that some of
the inspiration they generate carries over into practice. But for
many participants what is needed is something more definite in
the way of a scheme of work of some kind which will direct and
maintain the momentum of the course into a continuing
programme of monitored activities in the classroom. It seems to
me that it should be the purpose of all in-service courses to
develop such schemes through the joint enterprise of the
participants. In this way, these courses prepare teachers for the
responsibility of their own continuing professional education.
They are provided with guidance which enables them to take the
initiative.
I would now like to make specific proposals as to how the
model of pragmatic mediation that I have proposed in these
chapters might be made operational within a programme of in-
service education for language teachers.
66 The theory and practice of language teaching

Model to module

I have presented the rationale for a pragmatic approach to

language pedagogy which casts the teacher in the role of
mediator between theory and practice. I have tried to show how
such an approach is necessary if language teaching is to draw
maximal benefits from developments in the informing disciplines
while retaining its own identity as a domain of enquiry which
has to meet its own conditions of relevance and accountability.
And I have indicated some of the negative consequences of
pragmatic failure.
The adoption of this model of pedagogy means that the
teacher is necessarily engaged in research as an aspect of
teaching and this research too has its own character and its own
criteria of relevance and success. It carries with it, too,
implications for teacher preparation. The conduct of pedagogic
research as I have defined it here presupposes attitudes and
approaches to techniques of teaching which are developed only
through an educational perspective and this in turn calls for a
continuous programme of in-service support. Such continuity
can only be assured by the design of schemes of work which
constitute guidelines for classroom action.
What I have in mind are portfolios or fascicules of data and
tasks representing modules of enquiry which deal with issues of
current pedagogic concern in a format which, being based on the
pragmatic model that I have proposed, draws the teacher into
the appraisal and application of ideas. The starting-point is not a
theoretical insight or a research finding or a recommended line
of approach but what the teacher finds problematic, what issues
are from her point of view in need of clarification and resolution.
Generally speaking, we know all too little about the teacher’s
perspective on pedagogy; we make generalizations which are not
based on anything in the way of reliable information. We are
very prone to pontificate about what teachers should think
about but we do not really know what they do think about: we
formulate issues on their behalf. What is needed is an investiga¬
tion into what language teachers in different classrooms in
different countries find problematic about conventional practice
or about proposals for innovation that have been put forward.
Modules could then be directed at issues which teachers
themselves see as of central concern.
In the cui'rent absence of information about problematic
 Pedagogic research and teacher education 67

matters from the teachers’ point of view, one can only make
assumptions based on impressions about the kind of issue that is
uppermost in their minds. The following might be suggested as
possibilities for the purpose of illustration.

Structure practice

Conventional wisdom supports the use of drills, substitution

tables, and other devices for pattern practice on the grounds that
they make a knowledge of language forms habitual. Recent ideas
about language use and learning insist on the primacy of
communicative activities in the classroom. Does this mean that
conventional ways are mistaken and should be abandoned? Is
there any place for practice exercises in a communicative
approach to language teaching?

Learner error

It has been suggested that learner errors can be seen as

evidence of learner achievement. Does this mean that they
should not be corrected under any circumstances? If there are
circumstances where error correction is legitimate, what are
they, and how does the teacher set about it without inhibiting
learner development?

Authentic materials

It has been traditionally supposed that the language presented

to learners should be simplified in some way for easy access and
acquisition. Nowadays there are recommendations that the
language presented should be authentic. How is it to be graded
so that it can be made accessible? Is simplification as a pedagogic
strategy inconsistent with the principles of a communicative
approach to language teaching?

The first stage in module design, then, is the identification of

problems from the teacher’s point of view. These may be
formulated in more or less specific terms and the rather general
issues that I have already outlined may be broken down into
more narrowly focused questions and be dealt with in more than
one module with provision made for explicit cross-reference. In
the case of structure practice, for example, more specific issues
might be:
Are meaningful drills possible?
What is the function of substitution tables?
68 The theory and practice of language teaching

How do language exercises differ from communicative tasks?

Modules might then be designed which deal separately with

each question but which have elements in common in respect to
underlying aspects of appraisal. In this way, the more specific
question would be related to a more comprehensive framework
of conceptual evaluation, and particular techniques seen to be
realizations of more general principles of teaching. The relative
narrowness of focus of particular modules, and the way this
corresponds with the need for cross-reference to make explicit
the common denominator of principle, are matters which cannot
be determined in advance but depend on what teachers conceive
their problems to be and on the actual experience of designing
and using such modules.
After identification of the problem to be addressed, the next
stage is to draw attention to ideas which bear upon it for inter¬
pretation and conceptual evaluation. This might best be done by
a judicious selection of quotations which teachers would be
guided to study in detail. The original problem could then be
reconsidered in the light of this appraisal and the next stage
would be to move into application, making the ideas operational
by the teacher’s devising activities in reference to given examples
drawn from existing teaching materials. These would sub¬
sequently be tried out and evaluated in the classroom, the
teacher having been given a set of possible procedures for doing
this. The following schematic sketch will give some indication of
what modules designed along these lines might look like.

Outline module: drills and meaning

Problem: To know whether the use of drills is consistent
with the teaching of meaning. Are meaningful
drills.a contradiction in terms?
Appraisal: Appropriate quotations from (for example):
Richards et <3/. 1984 . j r • • r
tor definition or terms
Rivers 1964
Dakin 1973 on the nature and use of
drills in language teaching
Palmer 1981
on the nature of meaning
Yule 1985

By consideration of points raised in the module, and matters

arising from them, teachers are made aware of the distinction
 Pedagogic research and teacher education 69

between drills promoting mechanistic performance and drills

promoting the unconscious internalization of mentalistic com¬
petence structures, and of the distinction between meaning as
semantic signification and meaning as pragmatic value in
context.
Conclusions emerging from conceptual evaluation: types of
drill can in principle be associated with types of meaning. They
can be used with a problem-solving element to develop a
knowledge of semantic meaning as signalled by lexis and syntax
within sentences and can be used in support of tasks which focus
on pragmatic meaning by providing a consciousness-raising
extension to communication-oriented activities.
Application: Sample drills provided, selected from existing
materials with reference to procedural guidance. Versions of
these to be used in class. Feedback in respect of learner attitude
and achievement to be recorded. Exercises adjusted accordingly.

I have given only a very rough and general indication of how

the model of teacher mediation I have presented and argued for
in this chapter might itself be made operational. The detailed
design of individual modules will need careful thought and will
need to respond to the problems which teachers themselves
recognize as important. Furthermore, since they are not direct¬
ives but guidelines they must be flexible enough to allow for
teacher participation and the exercise of initiative so that
teachers can follow their course in reference to the bearings
provided. These modules are like the data and suggestions
provided for project work and are therefore incomplete. In using
them in classrooms and in discussing them with colleagues,
teachers will change them, replace them, and develop their own,
more immediate to their own teaching circumstances. They
have, in this sense, an essentially catalytic character.^ Their

* These modules, not surprisingly, bear a resemblance to the elements of the

INSET scheme designed by C. N. Candlin and myself (see Anderson and Lynch
1988, Bygate 1987, Cook 1989, McCarthy 1990, Malamah-Thomas 1987,
Nunan 1988, Rea-Dickins and Germaine 1992, Wallace 1992, Wright 1987; all
titles in Language Teaching: A Scheme for Teacher Education.) Although they are
informed by the same rationale, these modules differ in that they are intended as
kits or worksheets for separate mini-projects, are more specifically focused on
particular topics and derive more directly from the perceptions of teachers. They
are more individualized and local, more subject to modification in use.
70 The theory and practice of language teaching

function is to stimulate a pragmatic approach to teaching and

teacher education.
In the chapters of this part of the book I have sought to provide a
rationale for such an approach. If appraisal now reveals that it has
some validity in principle, the next stage is application. And that is,
of course, beyond the scope of this discussion.
PART TWO

Aspects of language
 y
I ■ «

r,!

I
)
r
5 Preliminaries: approaches to
description

The definition of teaching as a pragmatic process of con¬

tinuous evaluation, as proposed in the previous chapter, does
not, of course, apply only to the teaching of languages but is in
principle applicable to the teaching of all subjects on the
curriculum. We move on now to aspects of teaching of more
particular concern to the language teacher, but as we do so we
should not lose sight of the more general educational context. As
will become apparent in subsequent chapters, many ideas
relating to the teaching of languages have a more common
currency across the curriculum—ideas like the creative class¬
room and task-based learning, for example, and the need to
resolve the competing claims of teacher authority and learner
autonomy. One might say, indeed, that recent developments in
communicative language teaching have brought language teachers
into closer alignment with their colleagues. But in this part of
the book my purpose is more specifically to consider aspects of
language and the extent to which they indicate how we are to
define our particular subject.

Language and linguistics: subject and discipline

It seems reasonable to suppose that the definition of a
language as a subject for learning should in some way be
informed by theories about the nature of language in general on
the one hand and by descriptions of languages in particular on
the other. After all, we would be surprised if the subjects of
physics or history were defined in disregard of theoretical ideas
and descriptive facts coming from the scholarly disciplines which
bear the same name. And so it seems self-evident that the subject
of language should also draw on a cognate disciplinary source
even though it happens in this case to have the different name of
linguistics.
As I have already indicated, there are broadly speaking two
74 Aspects of language

kinds of insight of potential relevance to language teaching

which we might expect linguistics to provide. One comes from
theory, and has to do with ideas about the nature of language in
general as a psychological and social phenomenon, internalized
in the mind, externalized in social life. The other comes from
description, and provides information about the properties of
particular languages as formal abstractions and as actualized in
use.
It seems to be commonly supposed that it is description rather
than theory as such which makes the most direct contribution to
language teaching. For example:

He [the language teacher] is not teaching linguistics. But he is

teaching something which is the object of study of linguistics,
and is described by linguistic methods. It is obviously desirable
that the underlying description should be as good as possible,
and this means that it should be based on sound linguistic
principles.
This is the main contribution that the linguistic sciences can
make to the teaching of languages: to provide good descrip¬
tions . . . There is no conflict between application and theory;
the methods most useful in application are to be found among
those that are most valid and powerful in theory.
(Halliday et al. 1964:166-67)

Similarly, Corder, though sounding a note of caution about

adopting too restrictive a model of description, expresses the
view that pedagogy draws selectively from descriptions of
language provided by linguistics. The difference is not qualitat¬
ive but quantitative, a matter of selection only:

Thus the descriptions of a language which are the output of a

first-order application of linguistic theory represent only an
inventory from which a selection must be made in order to
draw up the syllabus for any particular teaching operation.
(1973:143)

Both of these views hold that the influence of linguistics on

language teaching should be mediated through descriptions
which are informed by theory. These descriptions—the first-
order applications of theory, as Corder calls them—then serve as
a source of information, an inventory of items, from which the
language to, be taught can be directly derived. In this view,
teachers have no direct contact with theory, though it may be
dimly discerned through the description.
 Preliminaries: approaches to description 75

Type and token descriptions of language

We should note that the description of a language in the terms
of categories which theory provides may account either for types
of linguistic element in the abstract, or for tokens of linguistic
element as they actually occur in contexts of use. A description
which deals with abstract types will present linguistic forms and
their meanings as constituents of the conventional code. Where
these forms are syntactic combinations of words, the result is a
descriptive grammar like that of Quirk et al. (1972, 1985). With
such descriptions, actually occurring linguistic data, where they
are adduced at all, serve to exemplify category types. A
description which deals with actual tokens of language use
reveals the relative frequency of forms and their habitual co¬
occurrence in different contexts.
Thus a type description might present a comprehensive array
of structures and give each of them equal descriptive weight, but
a token description might well reveal that some of these were of
rare occurrence, or restricted to a realization through a limited
range of lexical items, almost exclusively confined to certain
contexts, or associated with certain meanings.
Token descriptions on a massive scale are now possible with
the development of the computer. The evidence they yield does
not just quantify the token occurrence of existing category types,
for the most part derived from intuition, but also suggests that
the types themselves stand in need of revision so that the
language as abstractly conceived by the linguist is brought into
closer correspondence with the language as actually realized by
the user.
The two kinds of description, then, take very different
perspectives on language. The type description considers lan¬
guage as abstract knowledge, the token description as actual
behaviour. The extent to which these approaches are comple¬
mentary has to do with the relationship between competence
and performance and is a matter of current debate. Sinclair, for
example, is in no doubt that recent developments in what I have
called token description require a radical revision of principles of
descriptive procedure in general, and have profound implica¬
tions for language pedagogy in particular:

For some years I tended to assume that the computers

would merely give us a better documented description of the
language, but I do not think that that position remains
tenable. Now that we have the means to observe samples of
76 Aspects of language

language which must be fairly close to representative samples,

the clear messages are:
a) 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 the new material.
b) The categories and methods we use to describe English
are not appropriate to the new material. We shall need to
overhaul our descriptive systems.
c) Since our view of the language will change profoundly,
we must expect substantial influence on the specification
of syllabuses, design of materials and choice of method.
(1985:252)

What Sinclair is saying is that type descriptions of language

knowledge based on intuitive impression and not grounded in
actually attested data do not provide an adequate source of
reference for language teaching. What is required is that the
language which is pedagogically presented should be a pro¬
jection of that which actually occurs as recorded by the
computer analysis of text.
I made the point earlier that descriptive facts about actual
usage do not necessarily determine what is to be included in a
language course. The abstractions of a type description may be
preferred on occasions in that they activate more effectively the
process of learning. Consider, for example, that type of syntactic
structure which Chomsky refers to as the kernel sentence, of
which the following are examples:

The birds are singing.

The farmer killed the duckling.

The likelihood is that such spare syntactic structures appear

very infrequently as independent forms in actual use. But as
Chomsky himself remarked when he abandoned them with some
reluctance as theoretical constructs, these kernel sentences
actually have considerable intuitive appeal. They seem to have a
certain psychological reality for native speakers as units of
knowledge in spite of their abnormality as units of behaviour.
One might surmise that this is because they represent elemental
features from which all other structures are compounded and so
constitute basic units of mental processing from which all
learning proceeds. Be that as it may, the intuitive reality of such
structures is borne out also by the fact that generations' of
language teachers have recognized that they have a leading role
 Preliminaries: approaches to description 77

to play in the initial stages of learning, because of what I have

called their valency or combining power. Now generations of
teachers may of course be wrong. But it will not do just to
dismiss the evidence of their experience out of hand.

The relevance of linguistic theory

The point is that whether descriptions are knowledge or

behaviour oriented, type or token, they cannot determine what
the teacher does. They have always to be referred to pedagogic
decision. But this decision will need to be informed by an
understanding of what theoretical assumptions underlie the
different descriptions. As we have seen, the usual belief is that
what teachers need is descriptions, with the implication that they
need not be bothered with the theoretical ideas upon which they
are based. I should like to put the contrary view that an
understanding of theory is of primary importance because this
provides a general perspective on the nature of language which
influences pedagogic principles and indicates how different
descriptions might be adapted for classroom use. In other words,
to refer back to the discussion in the previous section, an
understanding of theory is needed for evaluation.
Theories about language, its development as a mental con¬
struct, and its operation in social life, have had an enormous
influence on how teachers of second languages have conceived of
their subject. The basic principles which have been adopted in
these aspects of teaching which Sinclair mentions—‘specification
of syllabuses, design of materials and choice of method’—have
been drawn not from particular descriptions of the language but
from ideas, assumptions, and beliefs about language in general.
It seems to me to be entirely right that this should be so. For
although teachers will find descriptions in dictionaries and
grammars of immense value as sources of reference, ultimately it
will be up to them to decide on how theoretical ideas are to be
applied and evaluated in the pedagogic treatment of language.
But if teachers are to take such decisions they need to have
access to these ideas. This is in part what I hope to provide here
in respect of certain current theoretical perspectives and ped¬
agogic approaches. The two following chapters explore an issue
which has been prominent in the debate on language teaching in
recent years (and which is related to the distinction between type
and token descriptions which I discussed above), namely the
78 Aspects of language

relationship between meaning as formally encoded in a lan¬

guage, its grammar and lexis, and meaning which is achieved in
context by the exploitation of these formal properties.
It often seems to be supposed that a concern for grammar is
inconsistent with the principled of communicative language
teaching. This supposition is, I believe, based on an impover¬
ished concept of the nature of grammar, one which does not
account for the complementary functioning of lexis and syntax
as an essential resource for the negotiation of meaning in
context. In chapters 6 and 7 I would like to propose a
characterization of grammar and language use which shows
their interdependence.
6 Grammar, and nonsense, and
learning

Nonsensical sentences
The title of this chapter comes from a song in Oliver
Goldsmith’s play She Stoops to Conquer, sung by Tony
Lumpkin, a character who misspends much of his time in a
tavern called ‘The Three Pigeons’:
Let schoolmasters puzzle their brain
With grammar, and nonsense, and learning . . .

In Tony Lumpkin’s judgement, good liquor is much to be

preferred to all this. But we cannot afford to be so self-
indulgent and dismissive. Over recent years schoolteachers and
others concerned with the teaching of languages have ‘puzzled
their brains’ about the role of grammar in language learning, and
some maintain that students can manage without it. Others have
persisted in a more traditional view that language learning is
essentially the same as the learning of grammar. So what is the
relationship between the two?
Tony Lumpkin associates grammar with nonsense. He is not
alone. No less a person than Firth, not someone, one
supposes, who would have rejected all learning in favour of the
bottle, observes that grammarians ‘make regular use of non¬
sense’. He gives the example:
I have not seen your father’s pen, but I have read the book of
your uncle’s gardener. (1957:24)

This, though exemplifying the syntax of English, is in Firth’s

view nonsensical as an expression of meaning. So, he asserts, are
other rather less extreme examples, including the following, all
of which appear as instances of English in books by reputable
grammarians:

The farmer killed the duckling.

Pussy is beautiful.
The lion roars.
80 Aspects of language

Firth’s strictures would, of course, apply equally to a very

large number of sentences appearing in language teaching
textbooks and practised by pupils. In this case, nonsense is not
only associated with grammar but learning as well.
What Firth is pointing out is that sentences as artificial
constructs for exemplifying linguistic forms do not meet the
same conditions of making sense as do expressions naturally
used in the service of communication in context. They have no
‘implication of utterance’: whatever meaning potential they
might have is remote from any realization, since the contexts
which would provide the occasion for their use are of unlikely
occurrence. Of course, many people concerned with language
teaching have come to a similar conclusion. In consequence,
there has arisen a deep distrust of sentences and, by association,
of the grammar they exemplify. But grammar cannot be equated
with the devices used to exemplify its formal properties. There is
more to it than that.
Let us begin by considering Firth’s point more closely. It can
be related to Sinclair’s observation that examples of English are
sometimes given which the analysis of actual data reveals do not
in fact occur (see page 76). In both cases the objection is to the
presentation of expressions in English as if they were normal
uses of English. What is not clear, however, is the nature of the
abnormality which is being objected to. Is it to do with the
syntactic structure or to the way in which this structure is
lexically manifested? One may agree that the expression which
Firth dismisses as nonsense has a curious ring to it but this seems
to be due to the choice of lexis which results in a particularly
inconsequential proposition which it is hard to imagine ever
figuring in actual use. If yve alter the lexis, we can arrive at what
would seem to be an entirely normal expression:

I have not seen your father’s pen, but I have read the book of
your uncle’s gardener.
I have not seen your client’s proposals, but I have read the
report of your company’s accountant.

It may be that this relexicalized version has not been actually

attested and may never occur in the future but it acquires
normality because it is relatively easy to conceive of a context for
it. In this respect, the basic criterion for normality is not actual
occurrence but contextual plausibility. And this is something
which computer analysis of a corpus cannot of course determine.
 Grammar, and nonsense, and learning 81

Notice that the principle of contextual plausibility allows

legitimacy to expressions which arise in the contrived contexts of
the classroom. If it is allowed that such contrivance is ped-
agogically desirable as activating the process of learning, then it
sets its own conditions for normality. The crucial point is that
such expressions should be warranted by conceptual and
communicative purposes recognized as having point in class¬
room activity. These purposes do not, however, have to
correspond with those which are current or ‘authentic’ uses of
language in the world at large. This matter will be taken up
again in Part ^ of this book.
For the moment we need to notice that if normality is to be
defined not in terms of contextual plausibility but in terms of
frequency of attested occurrence, then in principle the teacher,
selecting language by this criterion, would be confined to a
presentation of expressions which realized a particular combina¬
tion of syntactic elements with particular lexical items. Such
combinations constitute formulaic patterns which are indeed of
very frequent occurrence in language use and need to be
accounted for (as I mentioned in Chapter 3) but they can hardly
be said to represent the total language to be taught.
Grammar is clearly central to the working of language. But it
is equally clear that its nature cannot be accounted for by
demonstrating its rules by a random use of any lexical items that
come to mind. I have suggested that it enters into some kind of
relationship with words and contexts. Grammar is not just a
collection of sentence patterns signifying nonsense, something
for the learner’s brain to puzzle over.
What is it then? And what is the nature of this relationship
between words and contexts I have referred to? It is obviously
important that we should have some clear idea about the nature
of the phenomenon as an aspect of language not just as a
preliminary but as a prerequisite for determining how it should
figure in pedagogy.
One obvious way of finding out what grammar is is to look up
the term in a dictionary. David Crystal has produced a work
which ought to be particularly well suited to our purpose. It is
called A First Dictionary of Linguistics and Phonetics (1980).
Two pages are devoted to the entry grammar. Unfortunately
they leave us little the wiser about its essential nature. It is, we
are told (among other things) ‘a systematic description of a
language’, ‘the way words, and their component parts, combine
82 Aspects of language

to form sentences’, ‘a device for generating a finite specification

of the sentences of a language’. So grammar is the name we give
to the knowledge of how words are adapted and arranged to
form sentences. And its operation can be exemplified by the kind
of nonsensical expressions that Firth complains about. But what
are sentences? One might demonstrate their formal properties
but what are they actually used for? Why do words have to be
subjected to adaptation and arrangement in this way?

Grammar, words, and context

There are after all occasions when words do very well on their
own. Consider the classic case of the surgeon performing an
operation and the utterances he addresses to his assistants:
‘Scalpel!’ ‘Swab!’ ‘Clamp!’ etc. No sign of grammar here: no
interrogative forms, modal verbs, question tags; no sentence at
all. Just words. The reason why communication is achieved here
by lexical means only is of course because the context of shared
knowledge makes it possible to use minimal cues. The con¬
ceptual or lexical meaning is sufficient for its indexical purpose
on this occasion.
The notion of indexical meaning (which I have expounded in
detail in Widdowson 1983, 1984a, and which I shall discuss
again in Chapter 7) is crucial here. By ‘indexical’ I mean the
function which is attributed to the linguistic sign by the language
user when it occurs in context. It is used as an indicator or
pointer to those features of the situation or existing knowledge
that need to be engaged to realize meaning. Indexical meaning,
achievable only pragmatically in reference to context, is con¬
trasted with symbolic meaning which inheres in the linguistic
sign as a stable semantic property. Thus the word scalpel has an
established signification as a linguistic symbol recorded in the
dictionary (‘a small, thin, very sharp knife used for surgical
dissection’). In the context of the operating theatre, the indexical
use of the sign allows for the identification of a particular
scalpel, and for the recognition that the surgeon is giving an
order. ‘Scalpel!’, then, takes on the indexical value in this
context of ‘Pass me that particular scalpel’.
In this surgical context, the words are themselves sufficient as
pointers to required meaning. Grammatical elaboration would
be redundant. Indeed it would be dangerous: it would make
communication less effective, the operation less efficient, and put
 Grammar, and nonsense, and learning 83

the patient in peril. By the time the surgeon had produced his
complete sentence, the patient might well have bled to death: a
victim of syntax.
In this case words alone are enough to indicate meaning
because of the high degree of contextual determinacy. On other
occasions, indeed on most occasions, we cannot count on the
context complementing words so closely, occasions when more
precision is needed to identify the contextual features which are
to be related to the conceptual meaning of the words to achieve
indexical meaning. And this is where grammar comes in. Let us
look again at the sentence that was cited earlier:

The farmer killed the duckling.

Stripped of its grammatical appendages and reduced to lexical

essentials this appears as three words:

farmer kill duckling

The very conceptual meaning of these three lexical items in

association allows us to infer a sort of unfocused proposition: a
process kill, two participants in the process, one an agent, farmer,
and the other a patient, duckling. Even if we were to change the
linear arrangement, the three words presented in association
would serve to indicate the same process and the same roles of
the participants:

farmer duckling kill

duckling farmer kill
kill duckling farmer

In all these cases, our knowledge of what these words mean in

English, and the very general context of our world knowledge
would lead us to suppose that in all cases the farmer is the agent
and the duckling at the receiving end of the action: the farmer
does the killing and the duckling gets killed.
But let us now alter the lexis somewhat and consider the
following three words in association:

hunter kill lion

We have the same process here, but now we cannot

distinguish between the participant roles of hunter and lion. In
the familiar world in which we live, ducklings are not known for
their propensity for attacking farmers; they are classed among
the victims of the killing process. But lions are a different matter.
84 Aspects of language

Hunters might seek to kill them, but they are quite capable of
turning the tables and acting out the agent role. There is nothing
outlandish in the idea of lions killing hunters. So here the
relationship between the lexical concepts has to be marked in
some way to make up for the inadequacy of the words to
indicate what part of the general context of knowledge is to be
engaged. There are two possible states of affairs here, not just
one. A common marking device used in English for such cases is
word order. Since the mere association of words will not
unambiguously point to meaning, the words need to be set down
in a particular arrangement. Thus the sequence hunter kill lion
signifies one thing, hunter agent, lion patient; lion kill hunter
signifies the opposite, the lion as agent (the killer), the hunter as
patient (the victim).
Word order is not the only conceivable grammatical device for
enhancing the indexical precision of lexical items. One might use
morphological rather than syntactic means. Many languages do.
We might propose, for example, that the participant roles in our
case might be marked by different suffixes: let us say o for the
agent role, om for the patient. Word order would then not be
needed for this particular purpose of role assignment. There
would be equivalence of meaning with' different word orders, as
in:

huntero kill lionom

lionom kill huntero

And conversely, of course, contrast of meaning with the same

word order:

huntero kill lionom

hunterom kill liono

Since word order is not now needed for signalling an increased

specificity of conceptual meaning it can be put to other purposes,
as we shall see presently.
But let us for the moment look at the word that signifies the
process or action itself: the word kill. Again although there will be
occasions when the bare lexical item will suffice to indicate mean¬
ing, when the context or the convergence of knowledge of those
concerned will provide the specificity required, it will generally
need to be supplemented by the addition of elements which give
the word a more precise conceptual focus. We need devices, for
locating the process in time and for indicating its own temporal
 Grammar, and nonsense, and learning 85

character. In other words, we need some way of marking tense

and aspect. In English this is done by a combination of addition
and alteration. The word is altered to signify present and past
time:

farmer kills duckling

farmer killed duckling

And auxiliary verbs, themselves altered as necessary, are

added to signify aspect:

farmer is killing duckling

farmer was killing duckling
farmer has killed duckling
farmer had killed duckling

In this way, the proposition is focused a little more clearly and

its dependence on contextual support decreases accordingly.
Again we should note that this increase in conceptual precision
can be achieved, and is achieved in other languages, without
recourse to grammatical devices of this kind. Lexical items might
be used instead so that the focusing effect is brought about by
extra words. We might propose something like the following for
English: then and now for past and present; be and have for
continuous and perfective:

then farmer kill duckling = farmer killed duckling

now farmer kill duckling = farmer kills duckling
then farmer be kill duckling = farmer was killing duckling
then farmer have kill duckling = farmer had killed duckling
now farmer be kill duckling = farmer is killing duckling
now farmer have kill duckling = farmer has killed duckling

Marking for tense and aspect, then, are other communicative

devices for getting features of context into focus, for providing a
sharper definition of what words mean in relation to the external
world. But we do not only report on events as they actually
occur, we also make judgements about them, and we can call
them into existence out of context. Again grammar is ready to
hand to provide the required refinement of the raw lexical
material:

farmer will kill duckling

farmer must kill duckling
farmer will have killed duckling
farmer must have killed duckling, etc.
86 Aspects of language

Devices are available too for giving variable prominence to

one or other of the participants in the process to identify it as the
topic. We noted earlier how word order can signal an assign¬
ment of participant role (agent, patient, and so on). It can also be
used to indicate what is to be presented by the speaker as the
topic. But then to avoid confusion we need some way of marking
the participant role distinction. In English the word by is used
for this purpose:

hunter kill lion (hunter topic and agent, lion patient)

hunter kill by lion (hunter topic and patient, lion agent)

Add tense and aspect specifications in the correct combinations

and we get:

hunter killed lion

hunter killed by lion
hunter was killing lion
hunter was killed by lion, etc.

The expressions we have arrived at so far still need further

refinement before they are presentable as sentences of standard
English. But they are gradually coming into grammatical focus.
And perhaps they suffice to demonstrate, in a rudimentary way,
how the arrangements and alterations of grammar provide
additional specification to lexical associations so that the words
can relate more precisely to features of context, including those
features which are incorporated into the knowledge of the
language users themselves. The greater the contribution of
context in the sense of shared knowledge and experience the less
need there is for grammar to augment the association of words.
The less effective the words are in identifying relevant features of
context in that sense, the more dependent they become on
grammatical modification of one sort or another. And of course
where there can be no possibility of shared contextual know¬
ledge, as in the case of unpredictable personal invention and
interpretation, grammar provides the guarantee of individual
conceptual freedom. Contrary to what Tony Lumpkin believes,
speaking for all those who have been subjected to the drudgery
of learning it in school, grammar is not a constraining
imposition but a liberating force: it frees us from a dependency
on context and the limitations of a purely lexical categorization
of reality.
 Grammar, and nonsense, and learning 87

The relationship between grammar and lexis

Grammar, then, can be seen as a resource for the adaptation of
lexis. But there is no absolute distinction between the two, only a
convenient distribution of semantic responsibility. Grammar is a
device for indicating the most common and recurrent aspects of
meaning which it would be tedious and inefficient to incorporate
into separate lexical items. Thus it might be possible to have
quite separate words for, say, kills, killed, is killing, and so on,
just as we have in English separate words for man (male human)
and woman (female human) or people (humans in the plural).
But then we would have to find separate words for every action
or event denoted by different lexical verbs: a mammoth and
unnecessary task. So grammar simply formalizes the most widely
applicable concepts, the highest common factors of experience:
it provides for communicative economy. Of course, as the
examples of man, woman, and people illustrate, one can
economize in the opposite direction as it were, from grammar to
lexis. Thus, in English, the grammatical structure the man who
brings the post can be lexically realized as postman-, an animal
which has been killed for consumption becomes meat. In this
respect, the dictionary and the descriptive grammar are comple¬
mentary. The dictionary shows how efficiency in the formulation
of meaning can be achieved by synthesis, the grammar shows
how it can be achieved by analysis. Each mutually supports the
other one as a compendium of conceptual and communicative
resource. Together they contribute the cultural means whereby a
society organizes and acts upon its environment by the establish¬
ment of communal categories of context. Such categories
naturally facilitate interaction within linguistic communities and
inhibit interaction between communities to the extent that they
differ in their conceptual and communicative economies. These
differences might have to do with the aspects of context which
are differentiated and generalized, or with the distribution of
responsibility for denoting these aspects within the formal
resources of the language, within its lexis, morphology, and
syntax.
The general issue arising here concerns the way in which lexis
and grammar act upon each other in the determination of
meaning. The traditional view in both linguistics and language
teaching is that grammar acts upon lexis. Thus, in the case being
considered here, syntax compensates for the imprecision of
88 Aspects of language

simple word association by setting tense and aspect co-ordinates,

assigning subject and object roles, and so on, thereby narrowing
down contextual possibilities. It is because grammar has the
effect of refining the relatively raw conceptual material of lexis
by systematizing it in this way that it is considered as the primary
determinant. Hence the custorrlary practice of establishing
structures and paradigms and then fitting words into them. But
we need to note that lexis can also act upon grammar. Consider,
for example, the case of the grammatical category of progressive
aspect. We are told (in Quirh et al. 1972: 93) that this indicates a
number of meanings, as for example:

Temporariness: activity at a particular moment, e.g. Joan is

singing well.
Limited duration: e.g. The professor is typing his own letters
(these days).
Characteristic activity: necessarily occurring with adverbs like
always and continually. It imparts a subjective, emotionally
coloured tone, e.g. John’s always coming late.

Now so long as one is careful in the selection of lexical

realization, then it is possible to distinguish the different
meanings supposedly signalled by the verb form. So, for
example, an expression like Joan is singing well could be
understood in either the temporariness or the limited duration
sense. But this is not the case with, for example, Joan is getting
dressed or John is having a shave. These lexical items act upon
the grammar to constrain the temporariness sense since as lexical
items they denote temporary activities. The professor is writing a
letter suggests a temporary activity, whereas The professor is
writing a book suggests an,activity of extended duration. But the
only difference between these two sentences lies in the choice of
the lexical item as object. And then there are cases which do not
seem to correspond with any of the given meanings. For
example:

Joan is going grey.

John is growing old.

These ‘activities’ are not temporary (more’s the pity) but they
can hardly be equated with the professor’s letter typing, which is
a periodic state of affairs and will not last. Nor is it appropriate
to call them ‘‘characteristic activities’. Going grey and growing
 Grammar, and nonsense, and learning 89

old are indeed processes rather than activities, irreversible and

permanent. We might compare:

The professor is typing his own letters for the time being.
"■Joan is growing old for the time being.

If we use the criterion invoked for the characteristic activity

sense of the progressive, namely that the admissibility of an
adverb signals a different sense, then we would have to increase
the subcategories of meaning associated with this particular
aspectual form of the verb. But then where do we stop? For it is
easy to think of other examples where the lexical choice will lead
to a further proliferation of senses. We have to conclude, I think,
that there is a very wide range of meanings which the progressive
can be associated with but that these are determined by the
particular choice of lexis and are not independently a property of
the grammatical category of aspect. The lexis, in this respect,
acts upon the grammar.
There is, then, a reciprocal relationship between grammar and
lexis: they act in concert in the discharge of their semantic duties.
But we should notice that there is really no hard and fast
semantic distinction between them, but only degrees of con¬
ceptual generality. Although it is common practice to distinguish
what Henry Sweet called ‘form-words’ as distinct from ‘full
words’ and to locate the former in grammar rather than lexis, it
has to be noted that these ‘form-words’ have emerged over time
from full-blown lexical sources, and for the most part still have
recognizable kin which show clear signs of the relationship. So it
is, for example, that the auxiliaries have and be co-exist with
verbs of the same form which denote possession and existence
and from which they derived. Similarly do occurs as a lexical
verb but also as a purely grammatical element whose function is
to carry tense and negation for the lexical verb. Compare, for
example:

Joan did her homework.

Joan did not do her homework.

The scale across which this gradual historical shift takes place,
a scale of increasing conceptual generality from lexical to
grammatical meaning, is evident too in the current state of the
language. We see lexical items of a high degree of generality
which have functions which are difficult to differentiate from
those of grammatical elements. Consider the following example:
90 Aspects of language

John has bought one of those new micro-computers. I don’t

know where we are going to put the thing/l don’t know where
we are going to put it.

Thing is a lexical item, it is a grammatical ‘function’ word, but

they are hardly to be distingpished in terms of degree of
generality.
There is, then, nothing contrary to the normal operation of
language in the proposition that lexical items of high generality
like then and now should be given grammatical status as in the
examples given earlier. This process has, after all, already taken
place in the signalling of future time with will. We might propose
the following paradigm:

The farmer then kill duckling—past time

The farmer now kill duckling—present time
The farmer will kill duckling—future time

It happens that only the last of these verb phrases is standard

English. But it is easy to imagine the development of the other
two in derived or dialectical variants. One can indeed see in
pidginized forms based on English this kind of assignment of
grammatical function to lexical items of high conceptual
generality. In Cameroon pidgin, for example, josnau (a deriva¬
tive from ‘just now’) is a marker of the present progressive and
nau nau a marker of the present perfect denoting the immediate
past as in:

Josnau a di chop—I am eating at the moment

A dong chop nau nau—I have only just eaten {Todd 1974:18)

Similarly, in Melanesian pidgin, the grammatical function of

possession is signalled by a derived form from the lexical item
belong as in:

pies bilong mi—my place

Another point that might be made about the relationship

between the ‘form-words’ of grammar and the ‘full words’ of
lexis is that it follows from the kind of auxiliary, context¬
complementary function of grammar being proposed here that
the proportional occurrence of these words will vary in different
kinds of discourse.
If, for example, language is used to establish a context of
shared knowledge rather than to identify aspects of a pre-
 Grammar, and nonsense, and learning 91

existing one, then there is likely to be a higher degree of explicit

lexical reference and so a higher proportion of full words. The
‘lexical density’ of such texts will be relatively high (for a full
discussion see Stubbs 1986b). Since a good deal of writing is
context-establishing in this way, it generally reveals higher levels
of lexical density than does, for example, spoken conversation,
which tends to be context-identifying, concerned with giving the
sharper indexical focus to shared lexical information which
form-words can provide. But one cannot simply associate high
lexical density with writing, and low lexical density with
speaking. What is at issue is the way different uses of language
realize the complementary relationship between linguistic re¬
sources and contextual factors.
In view of this complementary relationship of grammar and
lexis, it is not surprising to find that there are units of meaning
which are intermediary between lexical words and grammatical
structures, the existence of which again indicates that there is a
continuum between these levels of language. Such units are
sentence-like in that they are syntactically combined sequences
of words but they seem to be stored in the mind ready for use as
preformed unitary items, like words, already assembled for
immediate access. They are, therefore, formulaic in character
and although they may call for some adaptation for contextual
fit they are not composed on each occasion from constituent
parts. They are the result of recall and not of the composition of
components by the application of syntactic rules. These units
have been variously referred to as lexical phrases (Nattinger
1988), composites (Cowie 1988), and lexicalized sentence stems
(Pawley and Syder 1983). Pawley and Syder provide a number of
examples, among them the following:

You can’t be too careful.

Are you all right?
I see what you mean.
It just goes to show.
That’s easier said than done.
I don’t know and I don’t care.
It’s easy to talk.
I thought better of it.
It doesn’t bear thinking about.
Think twice before you do that.
I thought you’d never ask.
Think nothing of it.
92 Aspects of language

As Pawley and Syder point out, the number of such formulaic

or composite expressions runs into very many thousands and the
ability to apply them in use accounts for the fluency of the native
speaker. They are, therefore, a crucial component of competence
and they pose problems in both the description and the
pedagogy of language. '
With regard to description, computer analysis of text provides
the means for identifying these expressions since their normality
is a function of their occurrence as holistic units. So it becomes a
relatively straightforward matter to list them as an inventory.
But it is less easy to see how they might be described in reference
to grammar on the one hand and lexis on the other. They vary in
their tolerance of adjustment: some are quite flexible and some
can be adjusted quite radically and these we may think of as
more grammatical. Others are more fixed and allow little room
for adjustment, and these we might think of as more lexical in
character. But they are ranged on a continuum and a satisfactory
account really depends on a theory of language which encom¬
passes both grammar and lexis within the same descriptive
scheme and makes explicit the relationship between them. At
present they tend to be treated apart, so that these formulaic
lexico-grammatical units are left rather in limbo.

Semantics and pragmatics

I have suggested that it is the function of grammar to reduce
the range of meaning signalled by words so as to make them
more effective in the identification of features of context, thereby
providing for the increased indexical potential of lexis. But of
course grammatical modification cannot account for the particu¬
lars of meaning which are signalled on particular occasions.
Grammar can only denote degrees of generality. It cannot refer
to individual cases. Now, as will have been noticed, in all the
examples I have given, one very crucial element has been missing
to make the expressions grammatical as instances of standard
English—we need determiners of some sort. For example:

The farmer kills a duckling.

A hunter killed the lion.

And so we arrive at last at the fully focused sentence. These

articles, definite and indefinite, now increase the specificity of the
lexical meaning of the nouns. The farmer, for example, narro'ws
 Grammar, and nonsense, and learning 93

attention to one who is known to both speaker and hearer, a

farmer indicates one who is not. But the hearer still has to act on
this indication and find which particular farmer is thereby being
referred to. All the grammar does is to signal that there is one. As
with the other grammatical devices we have been considering,
the function of the article is to set contextual co-ordinates in a
way which narrows down the range of inference.
We come here to the question of the relationship between
lexis/grammar on the one hand and context on the other,
between semantics and pragmatics. This is explored further in
Chapter 7, but we might note in passing here that the extent to
which these contextual co-ordinates are encoded within gram¬
mar varies across languages: what is semantically signalled by
grammatical means in one language is left for pragmatic
inference in another. To the extent that the grammar formalizes
the same principles as those of preferred pragmatic inference one
can think of grammar, as Levinson has suggested, as ‘frozen
pragmatics’ (Levinson 1987). Such a suggestion is, of course,
entirely consistent with the ideas about the function of grammar
as outlined in this chapter.
To illustrate this distribution of work between semantic
encoding and pragmatic inference, that is to say, the extent to
which grammar fixes the contextual co-ordinates, consider the
following English expressions:

1 The president wanted to resign.

2 The president wanted him to resign.
3 The president said he wanted to resign.
4 The president assumed he wanted to resign.

In sentence 1 the identity of the person who wants to resign is

grammatically established as the president and in sentence 2 it is
established as some male person who is not the president: the
grammar itself specifies disjunctive reference for him. In sen¬
tences 3 and 4, the reference value of he may be either the
president or some other male person. Although one would
probably incline pragmatically to the former interpretation in
sentence 3 and to the latter interpretation in sentence 4, the
context could easily override these preferences—most obviously,
perhaps, in the case of the president being a woman. On the
other hand, sentences 1 and 2 are impervious to such contextual
effects. But of course, even here there is pragmatic work to do to
establish the identity of the people concerned. The use of definite
94 Aspects of language

reference and third person pronoun signal that shared know¬

ledge is to be engaged, that is to say, that contextual information
is needed which is not abstracted in the codified forms of
grammar. In all these cases, what needs to be inferred is the
relationship between what has become generally conceptualized
in grammatical form and what is particular in context.
Communication, then, can only be achieved by relating
language with context: grammar simply makes it easier to
establish the relationship by setting, as it were, more exact co¬
ordinates. But the language user is still left with the problem of
engaging the particular features of actuality which are relevant
on a specific occasion. Knowing these devices for narrowing
down contextual possibilities does not imply that one can judge
how best to act upon such knowledge, how much can be left to
be inferred from context, how much needs to be made
grammatically explicit. Something must always be left unsaid.
But how much? An example from history might serve to
illustrate the point: another instance of the possible disasters
attendant on an ineffective use of grammar.
The scene is the Battle of Balaclava. On high ground, at a
customary safe distance from the action, the British general Lord
Raglan is directing troop movements by sending his orders by
messengers on horseback. From his vantage point he sees in one
part of the field the enemy trying to retreat with their artillery
and he sends a message to his brigade of light Cavalry. It reads as
follows:

‘Lord Raglan wishes the Cavalry to advance rapidly to the

front, follow the enemy and try to prevent the enemy carrying
away the guns.’

The Cavalry commanders. Lord Cardigan and Lord Lucan,

receiving this message, recognize that the definite articles signal a
particular front of battle and particular guns which both they
and Lord Raglan are supposed to know about. But they are in
the valley. Lord Raglan is upon the heights. The commanders
cannot actually see what Lord Raglan intends to refer to. They
do not in fact share the same context. The general has made an
unwarranted assumption about shared knowledge. The only
front that the Cavalry commanders can see is right at the end of
the valley where the main Russian army is massively assembled,
secure behind their heavy guns. For them this front and these
 Grammar, and nonsense, and learning 95

guns are the only possible ones indicated by the definite articles.
So, since theirs is not to reason why, they attack, with disastrous
consequences. And that is how the Charge of the Light Brigade,
the most celebrated and glorious calamity in British military
history came about—all because of a failure in the effective use
of grammar to make an appropriate connection with context.
Not all such failures, of course, are as historically momentous.
But they are of very common occurrence.

Grammar and learning

But let us return to Tony Lumpkin once more: ‘Grammar, and
nonsense, and learning’. What about learning? I have presented
grammar as a device for mediating between words and contexts.
The device itself is very complex, and its complexity cannot be
explained only by invoking communicative function (see New-
meyer 1983). It is subject to other informing influences: the
general and essential parameters of universal grammar; the
particular and accidental developments of its own social history.
But for language learners to learn only the intricacies of the
device without knowing how to put it to use is rather like
learning about the delicate mechanisms of a clock without
knowing how to tell the time. What is crucial for learners to
know is how grammar functions in alliance with words and
contexts for the achievement of meaning.
The teaching of grammar, as traditionally practised, does not
promote such an alliance. On the contrary, it is the formal
properties of the device which are commonly given prominence.
Words come in only as convenient for purposes of illustration. In
other words, lexis is put to the service of grammar. But as I have
shown, the function of grammar depends upon its being
subservient to lexis. Teaching which gives primacy to form and
uses words simply as a means of exemplification actually denies
the nature of grammar as a construct for the mediation of
meaning. I would suggest that a more natural and more effective
approach would be to reverse this traditional pedagogic depend¬
ency, begin with lexical items and show how they need to be
grammatically modified to be communicatively effective.
Within the category of lexical items, I include the formulaic
patterns I referred to earlier. If they do figure so prominently in
competence, it does not seem reasonable just to disregard their
96 Aspects of language

existence and leave their learning to chance. Studies in first

language (see Peters 1983) and second language acquisition (see
Gleason 1982; Vihman 1982) suggest that the way learners
proceed is to begin with these units as lexical complexes
associated with certain contexts and then pick them apart
analytically as the need arises. Some are dismantled entirely into
separate components for generative reassembly in reference to
grammatical rule, others are partly dismantled but are left as
adaptable formulaic frameworks to be adjusted to circum¬
stances; some again remain as holistically fixed, essentially large-
scale lexical items.
In view of this one might consider presenting language as
lexical units, both as single words and as complex packages, and
then creating contexts which constrain the gradual elaboration
of the first, the gradual analysis of the second. In this way
grammar would not be presented as primary but as a con¬
sequence of the achievement of meaning through the modifica¬
tion of lexical items. This would not imply that all of the many
thousands of formulaic expressions would be expressly taught.
The object would be to use them to develop learning procedures
which would provide the basis for learners subsequently to
process language and acquire the packaged units for themselves.
Such an approach to language presentation of course means
that contexts have to be contrived to motivate this lexical
modification and to guide the learner in the discriminating and
differential use of grammatical analysis. Again, traditional
teaching has tended to dissociate grammar from context and to
deal in isolated sentences. A pedagogy which aimed at teaching
the functional potential of grammar along the lines I have
described, would have to get learners to engage in problem¬
solving tasks which required a gradual elaboration of grammar
to service an increasing precision in the identification of relevant
features of context. In this way, learners would realize the
communicative value of grammar in the very achievement of
meaning.
I use the term ‘realize’ here in a deliberate double sense. On
the one hand the approach I am proposing would lead the
learners to realize (in the sense of actualize) grammatical
potential in contexts of use, that is to say that it would lead to
effective behaviour. But on the other hand, the approach would
also make, learners realize (in the sense of recognize) the
 Grammar, and nonsense, and learning 97

significance of grammar, and raise their consciousness of its

relevance.
A distinction is commonly made between teaching language
and teaching about language and the belief is expressed that the
first has to be done by providing direct experience and avoiding
the explicit knowledge promoted by the second. Conscious
awareness and reflection on one’s own experience is thought to
have an inhibiting effect. A version of this belief is incorporated
into Krashen’s distinction between (unconscious) acquisition
and (conscious) learning, already discussed in Chapter 2. Now
although there will obviously be occasions when this belief is
warranted, when learners are of an age, for example, at which
they would not have the capacity or disposition for analytic self¬
reflection, there seems no good reason for supposing that the
belief is universally valid. On the contrary it seems on the face of
it to be likely that with some learners a conscious awareness of
how language works and the subjection of their experience to
analysis would suit their cognitive style, increase motivation by
giving added point to their activities, and so enhance learning. It
would enable them to make comparisons between the language
they are learning and their own language, and engage in the kind
of rational enquiry which is encouraged in other subjects on the
curriculum. In this way the language to be learned could be
associated with a wider experience of language and education.
There are, then, arguments in favour of such consciousness-
raising (see also Sharwood-Smith 1981; Rutherford 1987). The
question, as always, is how they apply in particular cases.
It seems sometimes to be supposed that what is commendable
about a communicative approach to language teaching is that it
does not, as a structural approach does, have to get learners to
puzzle their heads with grammar. If we are looking for nonsense,
this suggestion is a prime example. For if this were really the
case, a communicative approach would have little or nothing to
commend it. For language learning is essentially learning how
grammar functions in the achievement of meaning and it is a
mistake to suppose otherwise. The question is how should
grammar be learned so that its intrinsic communicative char¬
acter is understood and acted upon. This cannot be done by
restricting attention to its formal properties, the relations and
regularities which make up the internal mechanism of the device.
No matter how legitimate it might be to define the scope of
98 Aspects of language

linguistics in this way (and this is currently a controversial

matter), it will not do for language pedagogy. Learners need to
realize the function of the device as a way of mediating between
words and contexts, as a powerful resource for the purposeful
achievement of meaning. A communicative approach, properly
conceived, does not involve the 'rejection of grammar. On the
contrary, it involves a recognition of its central mediating role in
the use and learning of language.
7 The negotiation of meaning

Sentence meaning and utterance meaning

Grammar, as I pointed out in the preceding chapter, can only
go so far. When it has done its work there still remains the task
of relating it to aspects of the context which are particular and
cannot of their nature be accounted for in advance. In this
chapter I shall consider in more detail what is involved in this
task, how context acts upon grammar so that the specific
meanings of particular expressions are realized and commun¬
icative outcomes brought about. So we move from semantics to
pragmatics, from virtual to actual meaning.
What I am seeking to do is to outline a model of language use.
But we must also bear in mind, particularly in view of the points
raised in Chapter 5, that such a model needs to be evaluated in
terms of its pedagogic relevance. How far is such a model
serviceable for language learning? How far is it congruent, or in
competition with, the model of grammar outlined in the last
chapter? But before these questions are addressed we need to
consider the model of how meanings are realized in context.
We can begin with the crucial, if rather obvious, point that
understanding what people mean by what they say is not the
same as understanding the linguistic expressions they use in
saying it. This is simple enough to demonstrate. Consider an
expression in English. The following will do:

The letter is in the drawer.

Considered as a sentence this poses no problem for under¬

standing. But as a use of language, as an utterance, presented like
this in isolation, it is quite incomprehensible, because we cannot
attach any meaning to it. The idiom is significant: we attach
meanings to linguistic expressions, and we do this by invoking
some pre-existing knowledge or other, or some co-existing
feature of the situation of utterance. Anybody actually pro-
100 Aspects of language

ducing this expression with the intention of being meaningful

would suppose that the addressee can make an attachment, can
relate the language to some shared conception or perception of
the world and so achieve the intended meaning. The letter (the
one we have just been talking about, the one that arrived by post
this morning, the one containing Aunt Kitty’s cheque ...) is in
the drawer (the one in the desk, the one in the dressing table
upstairs, the one in the kitchen where we always put the
post. . .). Every linguistic expression contains the potential for a
multiplicity of meanings and which one is realized on a
particular occasion is determined by non-linguistic factors of
context.
A sentence has only one invariant meaning, or if it has more
than one, as in the case of structural or lexical ambiguity, its
meanings can be exactly specified. Utterances, on the other hand,
are protean in character. Their meanings change continually to
suit the circumstances in which they are used.
We must be careful, however, not to be too carried away by
this idea. The multiplicity of utterance meanings does not mean
that any linguistic expression can mean anything at all in
complete disregard of what it means as a sentence. I cannot say
that the letter is in the drawer and mean that the cake is in the
cupboard, or that the letter was or will be in the drawer. The
conventional meaning of linguistic signs, and their combinations
in sentences, constitutes types of conceptualization codified as
linguistic knowledge and the tokens of particular and actualized
instances must clearly be set in correspondence with them. The
letter I am referring to now at the moment is a particular token
instance of letter as a lexical item, a general conceptual type, a
codified abstraction. It is simply that the type is more or less
stable, established by convention, whereas the token is not since
it is conditioned by context. And language use must always be a
matter of actualizing tokens as appropriate.
I say ‘more or less stable’ because, of course, conventions of
meaning change over time. Hence the need to revise dictionaries
and grammars as certain particularities of use become general¬
ized as usage. And these conventions are not of common
currency throughout a speech community. A dictionary, for
example, does not record meanings which are within the
competence of all language users. As Johnson-Laird has pointed
out we customarily understand and use words in context whose
meanings we do not know in any exact sense:
 The negotiation of meaning 101

Discourse rarely depends upon speakers having complete and

identical representations of the meanings of the words they
use. It is perfectly possible to communicate with little or no
such similarity or else children would never learn their native
tongue. Adults, too, can communicate successfully with an
incomplete knowledge of meaning. When, for example, you
read the sentence:

After a hearty dish of spaghetti, Bernini cast a bronze of a

mastiff searching for truffles.

you may understand it perfectly well even though, on

reflection, you may not be entirely sure exactly what alloy
bronze is, or what sort of dog a mastiff is, or what truffles
are.” (1983:225)

So the type of meaning that is known may be very general

{bronze is a kind of metal, mastiff a kind of dog). All that is
required is that it can be realized as a token which is sufficient
for the purpose of the communication concerned. There will, of
course, be occasions when the purpose calls for increased
specificity of type: we would need to be more knowledgeable
about the precise meaning of bronze, for example, in the context
of a textbook on metallurgy. But then the argument would apply
to other words in that context which had less bearing on the
purpose in reading.
Communication, then, is a matter of the mutual accom¬
modation of type and token as appropriate to purpose. Our
concepts of meaning provide us with bearings on what words
mean in context and the context in turn provides us with
evidence for extending our conceptual representation of these
meanings.

Symbol and index

It will be convenient to mark the distinctions I have been
drawing by means of certain semiotic terms. I introduced them
briefly in Chapter 6 and it is now time to consider them more
closely. The linguistic sign as type we can call the symbol. A
knowledge of language will enable us to decipher strings of
symbols as sentences and it is this knowledge, generally referred
to as linguistic competence, which it is the traditional business of
linguists to account for, and the traditional business of language
102 Aspects of language

teachers to teach. Comprehension in the sense of understanding

sentences is then a semantic matter of deciphering symbolic
meanings. But this knowledge will not alone enable us to
understand language in use for this is always a matter of realizing
the particular token meanings of signs in association with the
context of utterance. This is a pragmatic matter of achieving
meaning by using linguistic signs as evidence. The sign in the
utterance, therefore, does not function as a symbol but as an index:
it indicates where we must look in the world we know or can
perceive in order to discover meaning. It directs our attention away
from the language itself. In the case of the letter and the drawer, the
signs fail in their indexical function because, appearing as they do
in an isolated expression, they simply direct us into a void. They
can be deciphered semantically as symbols without difficulty, but
they cannot be interpreted pragmatically as indices (for further
discussion see Widdowson 1983,1984a: Paper 9).
Whereas symbolic meanings inhere in the signs themselves,
indexical meanings must be achieved by the language user
associating symbols with some relevant aspect of the world
outside language, in the situation or in the mind. This associ¬
ation may be fairly straightforward, the range of possible
reference fairly narrow, and then the interpretation of what the
addresser intends is easy. People who have particular knowledge
and experience in common, whose contextual realities, so to
speak, are closely congruent, will manage to communicate by
engaging relevant aspects of contexts with only sparing use of
the linguistic resources at their disposal. Conversely, of course,
those who have little in common have to place greater reliance
on the language.
But we do not only communicate with people with whom we
share our personal lives. We need also to participate in wider
networks of interaction which extend from the individual into a
complexity of connections with the groups and institutions that
constitute the society we live in. And here we cannot rely on
particular instances of shared knowledge and experience. We
need to refer to more general and conventional assumptions and
beliefs which define what is accepted as normal or typical in
respect of the way reality is structured and to the conduct of
social life. This common knowledge of shared experience and
conventionally sanctioned reality can be called schematic know¬
ledge: it is the knowledge which is acquired as a condition of
entry into a particular culture or sub-culture.
 The negotiation of meaning 103

Schematic knowledge, then, is a necessary source of reference

in use whereby linguistic symbols are converted into indices in
the process of interpretation. But we should note, too, that
language development itself, the acquisition of knowledge of
symbolic meanings, is activated by the need to extend schematic
knowledge so as to cope more effectively with the social
environment. We learn language in order to manage our affairs
in the world we find ourselves in. Language is the means of
initiation into the conventions of conceptualization and com¬
munication which define particular cultures. It has a crucial
socializing purpose. Formal education extends this process by
providing guidance into different ways of conceiving of the
world and different ways of conveying these concepts in modes
of description, argument, and so on. Particular subjects on the
curriculum can be seen as different sub-cultures in which reality
is variously reformulated.
But if it is the case that language learning is activated by the
socio-cultural purpose of schematic extension, that we learn
language in order to get a better grasp of the world so that we
can turn it to our advantage, then it would seem to follow that a
central problem in the teaching of a foreign language lies in the
provision of some comparable activating purpose. In other
words, we need to identify areas of schematic knowledge which
the learners will accept as independently relevant and worth
acquiring so that the learning of the language is seen as the
necessary means to a desired end. If this is so, the first question
to be asked in designing a language course should not be ‘What
language do we need to teach?’ but rather ‘What do we need to
teach that will stimulate the learning of language?’ The defining
of the subject English is in this view primarily not a matter of
language but of non-linguistic content. One might argue then
that the effectiveness of language teaching will depend on what
is being taught, other than the language, that will be recognized
by the learners as a purposeful and relevant extension of their
schematic horizons.
This argument can be invoked to support the programme of
language immersion in Canada, referred to in Chapter 2, and
other proposals to associate the foreign language more closely
with other subjects on the curriculum (see, for example,
Widdowson 1968, 1978). The idea of task-based learning in
general also bears on this issue. The success of such tasks as
activators of language use for learning will depend on the extent
104 Aspects of language

to which they engage the learner in conceptual and communicat¬

ive activities which they feel are worthwhile in their own right
(see Prabhu 1987).
There are matters concerning the teaching of languages for
specific purposes (LSP) which come up for consideration here as
well. It would seem by its very nartie that this enterprise is indeed
concerned with associating language with purpose. But the
purpose has not always been taken as primary in the pedagogic
approach adopted. Often specific language associated with the
eventual purpose has been identified and then activities devised
which lead to no schematic extension, either because the
specialist content is familiar or because the tasks are not
designed to engage appropriate modes of thinking. Hence
although the language is related to purpose this is not realized in
the process of learning (for further discussion see Widdowson
1983: Chapter 3).
We may say, then, that the achievement of indexical meaning
is commonly a matter of making a connection berA^een the
linguistic sign and the relevant aspect of schematic knowledge. If
we refer to linguistic knowledge, the internalization of the
symbolic function of signs, systemic knowledge, then we can
think of the realization of meaning in actual language use as a
matter of taking bearings on two points of reference: systemic
knowledge on the one hand and schematic knowledge on the
other.
Now, one can distinguish two kinds of schematic knowledge.
On the one hand there is knowledge of conceptual content or
topic area. This I have referred to elsewhere as ‘ideational’
(Widdowson 1983). Carrell (1983, 1987), who has been much
occupied in determining the effects of schematic knowledge on
reading comprehension, refers to such knowledge constructs as
‘content schemata’. These she defines as ‘background knowledge
about the content area of a text—for example, a text about
washing clothes, celebrating New Year’s Eve in Hawaii, building
a canoe, or about the economy of Mexico, the history of
Canada, problems of nuclear breeder reactors, etc.’ (Carrell
1983: 84). The other kind of schematic knowledge has to do
with mode of communication. This I have referred to elsewhere
as ‘interpersonal’ (Widdowson 1983). Carrell’s term for it is
‘formal’ and she defines it as ‘background knowledge about the
formal, rhetorical, organizational structures of different kinds of
texts’ (1983: 83-4).
 The negotiation of meaning 105

Negotiating procedures

This taking of bearings on systemic and schematic knowledge

is the procedural activity which converts type to token, symbol
to index, and so actualizes particular meanings. It is the
continuous process of plotting a position and steering an
interpretative course by adjustment and prediction. It is in this
sense that language use can be regarded as essentially a matter of
the negotiation of meaning.
It will be clear that on any particular occasion of meaning
negotiation the more familiar the schematic content or mode of
communication, the less reliance needs to be placed on systemic
knowledge, and vice versa. If, for example, I am engaged in
reading a text on a subject in which I am well versed (where the
ideational or content schema is familiar), which has been written
in a manner conventionally associated with writing on this
subject (where the interpersonal or formal schema is familiar),
then I shall only need to pay attention to the linguistic signs to
the extent that they key in this schematic knowledge and indicate
how it is to be extended. Many of the specific meanings which
can be deciphered out of the symbol are not activated because
they are not indexically required. This, of course, relates to the
point made by Johnson-Laird earlier; that an effective (i.e.
indexical) use of language does not depend on knowing precise
(i.e. symbolically complete) meanings.
On the other hand, if I encounter a text which deals with an
unfamiliar content area and does so in accordance with
communicative or rhetorical conventions which are new to me,
then I obviously have to look much more closely at the language
itself as a source of information as to what might be meant. I
have to count on the symbol itself as providing the main
evidence for meaning and, of course, as providing the means for
creating schematic knowledge which I do not have in advance. In
brief, if schematic knowledge is in short supply on a particular
occasion, then the more we need to invoke systemic knowledge
as a means of compensating for the deficiency and if we are
thereby able to convert symbol to index, then the act of meaning
negotiation itself has the effect of extending or altering the
schematic knowledge we started with. This is the process we call
learning.
As an illustration of this learning process, consider the
following passage:
106 Aspects of language

Oil-bound paints may sometimes fail when used on certain

building materials. When an acid and alkali react together the
result is a salt and water. If the acid is fatty such as linseed oil,
then the result is a soap and water. This is known as
saponification. Many building materials such as lime mortars
and plasters, Portland cement and asbestos cement develop
alkalis. If such surfaces are coated with an oil-bound paint,
particularly in the presence of even small quantities of
moisture, they will cause saponification. The paint may blister
in a mild attack or show yellow soapy runs in a severe attack.

If one is already in possession of schematic (content) know¬

ledge about oil-bound paints and building materials, then one
will already know that the former contains acids like the fatty
acid in linseed oil and the other contains alkalis, that soap
contains salt, and so on. Interpretation of the passage is thereby
facilitated and one can use it selectively to pick up additional
information to extend the schema—that the reaction referred to
is known as saponification, for example. But without such
knowledge, the reader is required to focus on the language and
make considerable efforts of inference. Thus one has to infer that
the second statement here functions as some kind of explanation
of the first statement in the passage, so that the failure of oil-
bound paints on certain building materials is caused by the
reaction of an acid and an alkali. So acids, it would seem, are
associated with the paints, alkalis with,,the building materials.
This is confirmed later when examples of these ‘certain’ building
materials are provided (lime mortars and plasters, etc.). The
relationship between the second and third statement is problem¬
atic for the unknowledgeable reader. At first sight it would seem
that there are two possible results from acid -I- alkali reaction.
Closer consideration, however, reveals that this cannot be the
case. ‘When an acid and alkali react together’ means when (all)
acids and (all) alkalis react together: the structure here signals a
general statement. It follows that the next statement, which
provides a particular instance of the general category {the
acid . . . such as linseed oil), must be an illustration, so a soap is
a particular manifestation of a salt, and saponification a
particular example of the general process of acid -I- alkali
reaction.
So it is that close attention to the language itself and reference
to systemic knowledge allows us to negotiate meaning and
 The negotiation of meaning 107

acquire the kind of information which for the reader in the

schematic know is provided in advance.
If one’s purpose is to develop the ability to negotiate meaning
in this way, then one would choose to present students with
passages whose content was unfamiliar. But it would be
necessary to ensure that the learning which resulted in respect of
such content was accepted by students as worth the processing
effort. In the present case, one would suppose that information
about the effect of paint on building materials is of fairly limited
appeal.

Reciprocal and non-reciprocal negotiation

Of course, if occasions arise when the symbols resist indexical
conversion, for all our procedural efforts, then interpretation
will fail. In this case, the addresser, the message sender, has made
unwarranted assumptions about the extent to which the same
schematic knowledge is shared by the receiver, and leaves too
many gaps, or gaps too broad to be bridged. When this happens
in reciprocal discourse, in face to face conversation, where the
discourse develops by the exchange of speaker role, then the
situation can be remedied (given the impulse to co-operate and
to accommodate the reality of the interlocutor’s world) by an
overtly interactive negotiation whereby intention and interpreta¬
tion are brought into an approximate convergence as required
by the purpose of the interaction. Two people who speak the
same language can always work towards an understanding
about what each of them is talking about and what each of them
wishes to achieve by what they say. They can therefore make
their meanings accessible by talk. Of course, they may not accept
the beliefs, attitudes, or intentions that are expressed; one can
always object to what somebody else says, but it does not make
much sense to do this unless one understands its meaning.
In reciprocal discourse, then, interlocutors can always estab¬
lish, by the turn-taking of talk, the necessary grounds of shared
knowledge, and so arrive at a mutually satisfactory schematic
convergence. But the case is different of course in modes of
language use which are non-reciprocal, of which what I am
doing at this moment is an obvious example. As I write, I make
assumptions about what my reader already knows about
language and language learning, about the common ground of
knowledge, and what he or she cannot be expected to know
108 Aspects of language

since it concerns the idiosyncrasies of my own view of these

matters. So I have to make decisions as I go along as to which
points need to be spelled out, which of them can be stated
sparely, or which of them do not need to be mentioned at all but
can be left to be inferred. No negotiation by active and
reciprocal participation is possiblb: the reader and writer cannot,
as conversationalists can, work together in directing the course
of the communication. This does not mean that no negotiation
takes place. The writer is engaged in a kind of vicarious
interaction with a presumed reader and anticipates and provides
for likely reactions. The reader for his part is drawn into the
discourse role that the writer has cast him in (for further
discussion see Widdowson 1979: Paper 13; 1984a: Section 2).
But of course the writer’s assumptions may be mistaken: the
reader may know more or less than has been assumed, in which
case the negotiation will falter. If the reader has more knowledge
than the writer has supposed, then he will tend to disregard the
discourse that has been plotted in the act of writing and simply
take from the text whatever best suits his purpose. If the reader
knows less of the writer’s world than supposed, he will have to
draw on systemic knowledge to furnish the necessary clues and if
this strategy fails, then the meanings remain inaccessible, the text
non-negotiable.
I have talked of the negotiation of meaning as a function of the
convergence of schematic knowledge, achieved by the con¬
version of symbol to index. But it should be noted that this
convergence need not be complete. It is unlikely that we ever
achieve an exact match between intention and interpretation,
and we probably would not know it if we did. We arrive at the
degree of convergence necessary to the purpose of interaction
and no more. Comprehension is never complete: it is always
only approximate, and relative to purpose.
So far in this chapter I have focused on meaning negotiation as
a matter of transaction. But we need to recognize its inter¬
actional character as well (see Brown and Yule 1983). I
mentioned earlier that a particular meaning might be made
accessible but that it does not necessarily follow that it will be
acceptable. The effective transaction of communicative business
requires the parties concerned to enter into a sort of social
contract involving adherence to what has been called the ‘Co¬
operative Principle’ (see Grice 1975). Language cannot symbol¬
ically signal its own meaning but has to be contextually
 The negotiation of meaning 109

connected to yield indexical meaning, so it follows that some

mutually agreed ground rules for co-operation have to be
assumed. We suppose, for instance, that what somebody says to
us will be relevant to the occasion or to what has just been said,
that people do not just issue utterances at random. We suppose
that when somebody says something to us, it is meant to be
informative, has some warrant in fact and is not a deliberate
obfuscation or falsehood (for further detail and discussion see
Grice 1975; Levinson 1983; Sperber and Wilson 1986).
Now if these maxims of co-operation are not honoured by a
speaker on a particular occasion and the hearer has reason to
suppose that this is intentional, the hearer will look for meanings
other than those which are explicitly expressed. If it is obvious to
both of them that what the speaker has just said is false, or so
obvious as to need no comment at all, the hearer will look for
implications, that is to say what is implied other than what is
expressed.
But why should these maxims be flouted in this way? Why
should people deny the co-operative principle by being deviant,
making their meanings less accessible, their transactions there¬
fore less effective?
The answer is that there is an affective as well as effective
consideration in the negotiation of meaning. To communicate
we need to co-operate. But co-operation involves risks. It
requires us to move into somebody else’s territory of self and to
leave our own vulnerable to intrusion. We therefore tend to be
circumspect and protective, particularly of course when our
dealings involve some threat to the fragile security of the other
person’s composure or self-esteem, or our own (for further
discussion see Brown and Levinson 1978; Widdowson 1984a:
Paper 7). The negotiation of meaning which is both accessible
and acceptable, therefore, involves the reconciliation of two
potentially opposing forces: the co-operative imperative which
acts in the interests of the effective conveyance of messages, and
the territorial imperative which acts in the interests of the
affective wellbeing of self. One imperative might impel the
speaker to be direct, to get the message across; the other might
impel the speaker to be deviant, to avoid offence or embarrass¬
ment. Where reciprocal interaction is concerned in particular,
the achievement of communicative purpose also involves, indeed
depends upon, the maintenance of rapport between the inter¬
locutors. Those participating in conversational encounters have
110 Aspects of language

to have a care for the preservation of good relations hy

promoting the other’s positive self-image, by avoiding offence,
encouraging comity, and so on. The negotiation of meaning is
also a negotiation of social relations. And indeed in some cases
of conversation, the maintenance of rapport, the sharing of
affective territory, the achievemeht of mutually acceptable states
of mind is not the means towards an effective communicative
transaction but the very object of the interaction itself. People
often talk to each other in order to lower their territorial
barriers, to indulge in the pleasure of rapport, the sharing of
common ground. There may be very little transactional business
enacted, or effectively none at all (see Aston 1986).

The negotiation of meaning and language learning

So much for the general account for how (as I see it) meaning
is achieved in the natural use of language. I turn now to the
question of the relevance of such an account to language
learning, in particular as that process is carried out in classrooms
under the direction, or at least the surveillance, of the teacher.
Perhaps the first point that needs to be made is that the
relationship between use and learning differs in respect to first
and second language situations. In natural first language
acquisition, the child, growing up through involvement in
naturally recurrent events, learns about the world through
language and concomitantly learns language through an engage¬
ment with the world. The two processes are, so to speak,
symbiotically related; they are the mutually reinforcing deter¬
minants of development. Thus systemic and schematic know¬
ledge develop concurrently, each supportive of the other. This
experience cannot be replicated in second language acquisition.
Here learners have already been socialized into the schematic
knowledge associated with their mother tongue: they are
initiated into their culture in the very process of language
learning. When they confront uses of the foreign language they
are learning, their natural inclination is to interpret them in
reference to this established association, and rely on the foreign
language as sparingly as possible. They will invoke as much
systemic knowledge of this language as is indexically necessary
and no more, using both their first language and the foreign
language tactically as a source of clues to meaning, while taking
bearings, as usual, on their schematic knowledge.
 The negotiation of meaning 111

The nature of learner error comes up for consideration here.

When learners are called upon to use the language being learned
for some communicative purpose, a purpose other than language
practice, then they vv^ill be naturally disposed to draw upon the
systemic resources which have proved serviceable in the past for
the achievement of indexical meaning. These, of course, have
been predominantly those of the mother tongue. In this respect
learner errors which reveal first language influence are the
natural reflex of procedures of meaning negotiation.
Errors have generally been attributed to cognitive causes,
evidence of the learner’s psychological process of rule formation.
But they can also be seen as communicatively motivated, the
realization of available resources to get a message across. Thus
although it may be the case that communicative demands result
in error because the learner does not have enough time to access
second language systemic knowledge, equally they may do so
because such demands quite naturally lead learners to call upon
whatever resources they have at their disposal, some of which
will be drawn from their own language. It is also likely, in
reference to the matters discussed in Chapter 6, that under
communicative pressure learners will place more reliance on
lexical means than on the intuitive assumption that context can
compensate for an absence of refinement in grammatical
signalling. In this respect, the errors of learners have some
resemblance to other uses of language where the exigencies of
the communicative situation license the disregard of grammat¬
ical niceties: telegrams, for example, or the laconic instructions
of the surgeon who appeared in the preceding chapter. In short,
then, errors can be referred to social/communicative as well as
psychological/cognitive causes.
In view of all this, one might characterize second language
pedagogy as a set of activities designed to bring about the
gradual shift of reliance from one systemic resource to another
for the achievement of indexical purposes.
The essential point is that meaning negotiation, which, in
normal circumstances, is always a matter of achieving an
objective by the most economical means, will be carried out by
taking whatever short cuts are available. It does not in itself
provide conditions for the acquisition of a systemic knowledge
of the foreign language, in spite of what those may say who
speak of ‘comprehensible input’ as the determining factor in the
acquisition process (cf. Krashen 1982 and the discussion in
112 Aspects of language

Chapter 2). The internalization of the system as a communicat¬

ive resource is only likely to happen when there is a concentra¬
tion on symbol to index conversion, when the potential
value of symbols is actualized indexically in the process of
discovering new meaning; that is to say, when there is a
recurrent association of new schematic knowledge with new
systemic knowledge. Such a state of affairs is normal in first
language acquisition, where there is a concurrent discovery of
language and the world, as I suggested earlier. But this focusing
on form, as a condition for comprehension, will usually have to
be artificially induced by some contrivance or other in a foreign
language situation.
Now it might be objected that I am talking here about
language knowledge and not communicative behaviour, about
how grammar might be learned and not about the development
of a language using ability. The point here, I think, is that if the
latter is to be an ability, and not just a performance repertoire,
then it has to be based on the internalization of systemic
knowledge as a communicative resource. There is a tendency
these days, and in some places, to suppose that systemic
knowledge will be acquired as a natural corollary to commun¬
icative activities. This supposition is, I think, based on the
simplistic idea that the natural conditions of language learning
through use that obtain in a first language setting and in
naturalistic contexts for second language acquisition can be
directly replicated in foreign language classrooms.
And yet our eventual objective must be to prepare learners to
cope with the natural conditions of language use. We come here
to what seems to me to be the central dilemma in second
language pedagogy: the conditions appropriate for acquiring
communicative resources are different from the conditions of
their use.
Consider in this connection the point I made earlier that
comprehension is necessarily incomplete and dependent on
purpose. Now if the intention is to elicit authentic behaviour
corresponding to that of the natural language user, then learners
need to be primed with purpose before they are required to give
evidence of understanding, and indeed the evidence itself has to
be of the kind which would emerge as a natural outcome. The
reading of the passage, or the listening to an extract of spoken
language, has to be such as to be dependent, a part of some
activity of broader significance which provides it with a cause
 The negotiation of meaning 113

and a consequence which have independent point. If one rejects

this authenticity condition, then one has to find ways of
presenting comprehension tasks so that the learner’s interest is
engaged in spite of the artificiality, so that they are induced into
co-operating with the contrivance. In the first case, with a
dependent text, one has to allow for the exercise of schematic
knowledge which will to some degree lead to the bypassing of
language and the avoidance of inference. It will deflect attention
from the language itself: acting as an authentic user, the learner
denies himself access to the necessary data for learning. In the
second case, with a detached text, more reliance will have to be
placed on inference and the use of systemic knowledge. But if we
seek to direct attention to this text as data to stimulate the
acquisition of systemic knowledge, we run the risk of making the
learner press for precision of meaning beyond any real purpose,
thereby misrepresenting the very nature of language use. In the
first case we have the simulation of user behaviour, in the second
case the stimulation of learning behaviour. Both have their
pedagogic uses but they will logically refer to different criteria
for the selection of texts and the design of tasks.
The question then is, how can we continue to induce both the
internalization of language as a resource and the ability to use it
appropriately? I shall examine this question in more detail in the
chapters in Part 3, but meanwhile we might consider briefly
how different approaches to language pedagogy in general can
be characterized in reference to this question.
Simplifying somewhat, we might say that a structural approach
to language teaching lays emphasis on systemic knowledge and
makes the assumption that once this is acquired the learners will
discover for themselves how it is put to use in communication.
Classroom activities will tend to be those focusing attention on
deciphering rather than on interpretation by indexical inference.
Language difficulty will generally be measured in terms of
decipherability, the problem of which can be eased by reducing
the symbolic complexity of the text.
A notional/functional approach essentially seeks to establish
correlations between systemic and schematic elements. It associ¬
ates concepts and communicative acts with their most common
or ‘standard’ expressions in the foreign language. In this respect,
it focuses on lexical/grammatical co-occurrences in formulaic
phrases of the kind discussed in Chapter 6. Classroom activities
here would prepare the learners to recognize the relevant co-
114 Aspects of language

occurrences and correlations as they occur in actual use.

Language difficulty will be seen in terms of nonconformity to
standard or normal ways of expressing notions and functions.
Neither approach takes as its central concern the exercise of
procedures for meaning negotiation, which require the relating
and mutual adjustment of systemic and schematic knowledge for
the realization of indexical value, after the manner previously
described, and which can provide the learner with the opportun¬
ity to learn the language through using it.
But an approach which did promote a negotiation of meaning
in a natural way, seeking to cast the learner into the role of user,
would itself run into problems, as I have indicated. For it would
encourage a reliance on schematic knowledge and a correspond¬
ing avoidance of an engagement with the systemic features of the
foreign language, or at best a tactical use of them which would
not lead to their internalization as a more general strategic
resource.
What seems to be needed is an approach which recognizes the
necessary contrivance of pedagogy and seeks to guide learners
through graded negotiating tasks. These would require them to
take bearings on both systemic and schematic knowledge and
would shift the focus of procedural work in a controlled way.
There seems to be no obvious reason why such tasks should not
also allow learners to refer to the systemic and schematic
knowledge of their own language and culture. This would take
pedagogic advantage of the learners’ own experience, and would
help to ensure that the tasks were independently purposeful—a
crucial design feature if the tasks are to induce language
processing consistent with natural use.
The purpose of such an approach would be to demonstrate
that the second language has the same potential for use as the
first language, encourage learners to draw on their own
experience of language by applying familiar procedures to the
interpretation of second language use, and so to teach the second
language system not as an end in itself but as a resource for the
achievement of meaning. But such an approach raises a number
of issues in pedagogy about course design, methodology, and the
roles of teacher and learner. These issues are taken up and
considered in more detail in Part 3.
PART THREE

Aspects of teaching
 ^*f1-
4

>/r.
‘I »»J lr««
8 General perspectives on
pedagogy

In the preceding part of this book (Chapters 5-7), I outlined

two approaches to the description of language, showed how they
are related, and indicated certain implications for teaching that
seemed to follow from them. In this part I want to pursue
these implications in more detail, focusing on pedagogy.
The two approaches to language description can be called the
semantic and the pragmatic. The semantic approach provides an
account of how the language contains within itself, within its
grammar and lexis, the essential resources for meaning. The
pragmatic approach on the other hand focuses on how these
resources have to be exploited for language users to achieve
meaning. On the one hand, then, we have meaning seen in terms
of a potential contained within linguistic forms. On the other
hand, we have meaning seen in terms of the procedures and
contextual conditions that come into play in order for this
potential to be realized. These two approaches to language
description are, as I have indicated, complementary. They
inform different approaches to language pedagogy. These too
can be seen as complementary, although they are commonly
presented in opposition.

Communicative language teaching

One way of characterizing the pedagogy of language teaching
in vogue over the past ten years or so is to say that it is, in a
word, communicative. But this is not very illuminating. The term
has been bandied about so freely, has been so liberally used as a
general marker of approbation, that its descriptive value has all
but vanished.
For one thing, we need to be clear whether the term is meant
to refer to the purpose or to the process of learning. It is often
supposed that the structural approach (what Prabhu (1987) has
recently dubbed as S.O.S.—structural-oral-situational) denied
118 Aspects of teaching

the communicative purpose of language teaching and proposed

in its stead that of structural manipulation. But this is not so.
The emphasis on structures was associated with the process of
learning, as the means towards an end, and this was not at all
intended to preclude a communicative purpose. On the contrary,
it was intended to promote it. This is clear from the quotation
from Lado which I cited in Chapter 2 (see pages 11-12). Here
structural manipulation through pattern practice is proposed as a
process which Hcilitates a communicative purpose. Lado argues
that such practice, far from confining the learner to mechanistic
performance, creates conditions for effective communicative use.
It leads to the internalization of patterns at a subconscious level,
and thereby leaves the mind ‘free to dwell on the message
conveyed through the language’. These patterns, says Lado, are
reduced to habit, where they belong, ‘so that the mind and
personality may be freed to dwell in their proper realm, that is
on the meaning of the communication rather than the mechanics
of grammar’ (Lado 1957).
So Lado quite clearly recognizes the ability to communicate as
the primary objective of language learning and conceives of
structural practice only as a means to that end. The essential
difference between his position and that of those who advocate
communicative language teaching would seem to lie not in any
disagreement about the centrality of communicative purpose but
in the concept of communication itself. And this difference
necessarily leads to different proposals as to how the ability to
communicate should be taught.
Lado talks of ‘the message conveyed through the language’,
and of ‘the meaning of the communication’ as linguistically
encoded. It would appear from this that he conceives of meaning
as intrinsic to language itself, a property signalled through the
medium of language. This medium concept, which defines
meaning as a function of the linguistic sign as formal symbol,
can be contrasted with a concept of meaning as significance
which is mediated by human agency. In this latter view, meaning
is not a semantic matter of encoding and decoding messages by
reference to linguistic knowledge but a pragmatic matter of
negotiating an indexical relationship between linguistic signs
and features of the context. It is not transmitted through the
semantic medium of language, it is achieved by the pragmatic
mediation of language users. So the question is not what
linguistic expressions communicate but how do people com-
 General perspectives on pedagogy 119

municate by using linguistic expressions. The medium account

of meaning is therefore associated with the semantics of sentence
grammar, the mediation account with the pragmatics of lan¬
guage use, as discussed in Part 2.

Pedagogic approaches: medium and mediation

What we have seen in language teaching, as in the study of
language, over recent years is a shift of emphasis from the
medium to the mediation view. An approach to pedagogy
informed by the medium view will focus attention on the
syntactic and semantic properties of the language itself and look
for ways of manipulating them for the purposes of transmission.
Learner activity will be directed at increasing receptivity. They
will be involved in activities which are designed to facilitate the
internalization of units of meaning so that they are put in store,
so to speak, ready for use when required. Such activities will
typically be exercises for the provision of practice. An approach
informed by the mediation view will focus attention on creating
conditions for negotiation. The learners will be engaged in
activities designed to achieve purposeful outcomes by means of
language. The activities here will be typically tasks for problem
solving.
These two perspectives on meaning can, then, be quite
naturally associated with different approaches to pedagogy,
different ways of defining the language to be learned as a subject
on the curriculum. Although these approaches may not deter¬
mine practice in any absolute way, they can nevertheless be
characterized in terms of certain interrelated tendencies.

Comparison of approaches
The medium perspective will tend to see the syllabus as
primary since it is here that linguistic units are specified and
ordered. These units may be given a formal or notional or
functional definition but they are in all cases presented as
packages of meaning. The assumption is that once they are taken
into psychic store they are ready to be issued as ready-made
tokens of communication. Methodology is essentially sub¬
servient, directed at facilitating this internalization process. In
this respect, structural and notional/functional syllabuses are
both informed by a medium view of communication, which is
120 Aspects of teaching

why it is so easy to rewrite one in terms of the other. With a

mediation view, on the other hand, the methodology becomes
primary. This is bound to be so since meaning can only be
achieved by action and is not simply a reflex of knowledge. It is
what learners actually do which results in meaning: communica¬
tion is a function of activity which engages language for the
achievement of purposeful outcomes. It is not surprising,
therefore, that as the mediation view has taken hold, so the
relevance of the syllabus has been redefined or called into
question, even to the extent of denying the need for it altogether
as a projected plan of work (see, for example, Candlin 1984).
The two perspectives will tend to see the organization of
language for learning in very different terms. All pedagogy is
involved in some kind of grading or ordering to allow for the
gradual development of learning. The medium view will
naturally provide for the gradual accumulation of units of
meaning: it will be disposed to adopt the principle of text
control, seeking to direct development by degrees. Since meaning
is confined within language it has to be transmitted according to
the channel capacity of the learner for the internalization of new
linguistic forms. But in the mediation view meaning, of course, is
not so confined. What this means is that if learners are prompted
by a particular purpose, if they are oriented to some outcome or
other, and are informed by relevant schematic knowledge, then
they can be exposed to more language than they can process as
sentences. The mediation view allows^ for ordering on the
principle of task control, whereby language is regulated not by
input but by intake. The channel capacity of the learners is
increased by context. Meanings which are beyond the learners’
linguistic competence to decipher by analysis are interpreted by
inference, which is activated by communicative purpose and
achieved by cross-reference to schematic knowledge. So the
mediation perspective will favour a pedagogy of discovery
whereby language is learned as a contingent consequence of
carrying out activities which engage the language with the
learners’ knowledge and experience of other things. Whereas the
text control of the medium view is a language filtering process,
with the learner as recipient, the task control of the mediation
view is a language focusing process, with the learner as agent.
Text control necessarily entails the contrivance of language
teaching units, a pedagogic fabrication which shapes the raw
materials of actually occurring language so as to make it simpler
 General perspectives on pedagogy 121

and more accessible. Task control, on the other hand, will tend
to favour the presentation of language in its natural state and
advocate the cause of authenticity.
These two perspectives on meaning will be inclined to
conceive of the roles of teacher and learner in different terms. In
the medium view, the learner is dependent on the teacher as a
source of information. The teacher’s task is to transmit, the
learner’s to receive what is transmitted. The process of trans¬
mission is seen as the same as the process of learning. Since
meaning is encoded and decoded in the foreign language which
the teacher knows and the learner does not, the learners’
previous experience of meaning in the mother tongue is
irrelevant. Indeed it is a nuisance since it is likely to act as a
disruptive influence on learning. But if one takes a mediation
view, the matter is altogether different. Learners already know
how to use language to mediate meaning. They can draw on
their own experience and employ the same sort of procedures to
achieve meaning in the new language as those they use to achieve
meaning in the mother tongue. The learner experience is not
now an interference to be eliminated but a resource to be
exploited. Accordingly, the learner assumes a more positive role.
Learning is not now seen as conformity to the conditions of
transmission controlled by the teacher but as a self-generating
process by the learners themselves.
This difference of role relates closely to a fourth difference of
principle between the two views. If meaning is thought of as
signalled uniquely through the medium of language, then any
departure from the rules of the language will result in a
distortion of meaning, and communication will accordingly be
impaired. So a medium view necessarily sets a high premium on
correctness, and learner errors are seen as failures to internalize
the devices necessary for the proper formulation of meaning.
They are defective sentences. The mediation view, on the
contrary, sees these nonconformist features of learner behaviour
as positive signs of successful learning since they show the
learner employing procedures for using whatever linguistic
resources they have to hand to mediate meaning. These so-called
errors may be defective sentences from the medium point of
view, but effective utterances from the mediation point of view.
In summary, there are two ways of conceiving the nature of
meaning, and these can be seen as informing different perspect¬
ives on language teaching, two pedagogic paradigms if you will.
122 Aspects of teaching

One focuses on meaning as transmitted through the medium of

language. It will concentrate on the devising of syllabuses of
preplanned schemes of work based on text control whereby
learners are directed by the teacher into the gradual reception of
units of meaning. If the learners conform to these conditions of
pedagogic transmission then they will learn the language as the
code for the transmission of meaning. Nonconformity is
negatively evaluated as error. The other perspective focuses on
methodology, the instigation of classroom activity, allows for
more exposure to language by guiding the learner by means of
task control, provides for the exploitation of previous experience
and for the exercise of initiative on the part of the learner.
Nonconformity is positively evaluated as the achievement of an
interim interlanguage.
Two paradigms, then, two sets of pedagogic principles
identifiable by a consistency with a particular view of the nature
of communication: either as the transmission of meaning by the
semantic medium of language or as the achievement of meaning
by the pragmatic mediation of the language user.
It is, of course, the mediation view which has become
prominent over the past fifteen years. There are a number of
reasons for this. For one thing, language teaching has been
sensitive to developments in the study of language: the extension
of the scope of linguistic description beyond the sentence, the
study of actually occurring language in context, the interest in
speech acts and pragmatics. At the same time, work in second
language acquisition has suggested that it is the creative
exploitation of language to achieve purposeful outcomes which
generates the learning process itself. Both language use and
language learning, it would appear, are to be characterized in
terms of mediation.
There is another influence at work as well. The medium view
is associated with authority. With its emphasis on transmission
and conformity, it promotes the conservation of established
social values and is consistent with a concept of education as the
means of maintaining conventions and persuading people into
their acceptance. Such an ideology is not well suited to the spirit
of the age, at least as this is perceived in some parts of the world.
It has been called into question on the grounds that it
perpetuates the rule of privilege and denies the rights of self-
determination and dissent. The mediation view is obviously
more attuned to more liberal ideas, allowing as it does for
 General perspectives on pedagogy 123

discovery and self-expression. It emphasizes initiative rather

than initiation, the autonomy of learning rather than the
authority of teaching. Its consistency with what would seem to
be a more enlightened social and educational ideology might
seem reason enough to accept it as self-evidently preferable
without further question. And some people, it would appear,
have accepted it on these grounds, inspired by the humanistic
fervour of the times.
Is there then nothing to be said for the medium perspective on
pedagogy? Is it to be dismissed as entirely discredited? We may
accept that communication is not semantic, a matter of
deciphering, of encoding and decoding meaning in sentences;
but we should note that this deciphering does demonstrate a
knowledge of linguistic resources which are indispensable in the
achievement of meaning by mediation. The language constitutes,
in Halliday’s terms, a meaning potential, and this can be
manifested through sentences and so internalized. But the
potential also needs to be realized as use, related to context,
made actual, externalized as a purposeful outcome by medi¬
ation. It is not enough that the learner knows linguistic resources
as an internalized potential, he must also know how to access
this knowledge and realize it as a resource. Knowledge of
language is a necessary condition for communication but it is
not, as Lado seemed to imply, a sufficient condition. Language is
a medium for the demonstration of meaning potential but this
can only be realized by mediation.
It follows from this that the two perspectives (like the two
approaches to language description outlined in Chapters 5-7, to
which they relate) can be seen as being not in conflict but
complementary so long as they do not make (as they sometimes
seem to do) exclusive claims to truth. The meaning which is
present as potential within linguistic forms serves as a set of
bearings for the language user. It sets the guidelines, the
parameters with reference to which meanings can be achieved by
mediation. Language users cannot spin meanings out of nothing
any more than they can achieve communicative purposes simply
by reciting sentences.

Complementary approaches
Now what we have here I think is a very general principle of
mutual dependency which we can apply to the other aspects of
124 Aspects of teaching

pedagogy I have referred to as associated with the two different

perspectives on meaning. Consider, for example, the relationship
between syllabus and methodology. One needs to allow meth¬
odological flexibility so that the opportunities for learning can
be exploited as they emerge. But learn is a transitive verb: one
has to learn something and this logically entails a measure of
conformity to some scheme of things prefigured in advance. The
most effective way of formulating a scheme of things, the object
of learning, is to design a syllabus. This then serves as the
necessary frame of reference. As far as language teaching is
concerned, it seems to me that the syllabus is a device for
specifying meaning potential as incorporated within linguistic
units and it is the methodology which realizes this potential by
mediating activities. It makes no more sense to attempt to
account for mediation within a syllabus than it does to make
methodology simply the means for manifesting the units of the
language that the syllabus specifies. What one needs to do is to
establish the proper concern of each aspect of the pedagogic
process so that they can be effectively complementary. This
matter is dealt with in some detail in Chapter 9.
Or consider the matter of allowing for the exercise of learner
initiative. There is no doubt that the imposition of correctness,
which is consistent with a medium view of meaning, has the
effect of inhibiting the learners’ engagement of relevant pro¬
cedures for mediation acquired through an experience of their
own language. There is no doubt that so-called errors are indeed
evidence of the learners’ success in realizing meaning potential.
But they are also evidence of limitations in the learners’
knowledge of meaning potential as encoded in the language they
are learning. Learners therefore need to be provided with
guidance in order to extend the range of the knowledge they can
draw upon. This does not mean that they have to be subjected to
direct correction, for this, being a shift from realization to
manifestation, has the effect of undermining efforts at medi¬
ation, but it is perfectly possible to refer learners to correct forms
and to set tasks which constrain their use without compromising
the achievement of purposeful outcomes. Again, the question is
how to create conditions for more effective mediation by
referring the learners to the meaning potential they must exploit
which is encoded within the language itself. One way in which
this might be done is discussed in Chapter 10.
I have suggested that the pedagogic movement of the seventies
 General perspectives on pedagogy 125

and eighties that goes under the banner of communicative

language teaching can be characterized not by a novel concern
for communication but by a different conception of the nature of
communication from that which prevailed in the preceding era.
The tendency has been to suppose that since meaning is not
uniquely signalled by language itself but is achieved by medi¬
ation, then all that is necessary is for learners to be involved in
problem-solving tasks without explicit reference to the formal
resources for meaning intrinsic in the language. My argument is
that there needs to be explicit reference, that it is the purpose of
pedagogy to actualize the mutual dependency between linguis¬
tically coded potential and its realization.
To do this is to recognize that the learners are not by the very
nature of their role free agents: their agency is bound to be
delimited. They cannot be charged with the sole responsibility
for their own learning. It is to recognize, too, that there is no
point in a pedagogy which does not intervene, does not exercise
authority, does not set the limits which define the subject. Such a
pedagogy, indeed, would be a contradiction in terms. The
central issue is how to apply these constraints or limitations so
that they constitute enabling conditions which facilitate the
learning process rather than restrictions which inhibit it. What
we need to consider, then, is how the roles of teacher and learner
can be made effectively complementary so that effective learning
comes about as a consequence of their interaction. This is an
issue that is taken up in Chapter 11.
 • - f-. '
‘ ^

I tr VI

' i <••

• h'A '

J-s ifl'
»i I '

-i* \i

ir«*
ftlr ^ ji'U Wt
» - ' . I »t»Cy C«

a><
^3^ il •• < <
<■ .'I^f r ^ - ! --•

... A -
'M
,i'/T .«
9 The problems and principles of
syllabus design

Pedagogic and educational aspects of the syllabus

In order to avoid confusion (or at least to reduce it) I had
better begin by making clear w^hat I mean by the concept I intend
to talk about in this chapter. I shall take a syllabus to mean the
specification of a teaching programme or pedagogic agenda
which defines a particular subject for a particular group of
learners. Such a specification not only provides a character¬
ization of content, the formalization in pedagogic terms of an
area of knowledge or behaviour, but also arranges this content
in a succession of interim objectives. A syllabus specification,
then, is concerned with both the selection and the ordering of
what is to be taught (cf. Halliday, McIntosh, and Strevens 1964;
Mackey 1965). Conceived of in this way, a syllabus is an
idealized schematic construct which serves as reference for
teaching.
My main concern in this chapter is with the syllabus as the
formulation of pedagogic goals. But it is important to recognize
that the syllabus is also an instrument of educational policy. The
goals are formulated not only in reference to pedagogic
effectiveness but also in accordance with ideological positions
concerning the nature of education in general, what Clark refers
to as ‘educational value systems’ (1987).
Formal education can be defined as a superimposed second-
order culture which consists of schemes of conceptual organiza¬
tion and behaviour designed to supplement the first-order
processes of the primary socialization of family upbringing. As
such its purpose is to give ideas, attitudes, actions, beliefs, and so
on, a shape which they would not otherwise have, so as to
prepare learners to participate in areas of social life beyond their
immediate environment and to extend the range of their
individual experience.
The relative weighting which is given to each of these general
128 Aspects of teaching

educational provisions, for future social role on the one hand

and for personal development on the other, will depend on how
different policies interpret the general purposes of education.
Some, for example, will focus attention on future social role,
define this in specific occupational terms, and seek to direct the
curriculum towards the satisfaction of projected manpower
needs. In this case, syllabuses will be designed to be accountable
to measures of utility and will be, to use terms borrowed from
Bernstein, ‘position-oriented’. Syllabuses to meet the needs of
English (or any other language) for specific purposes would be of
this kind. Other policies will be more favourably disposed
towards the individual, allow for a greater degree of divergence
and self-realization and incline to ‘person-orientation’ (see
Bernstein 1971).
The syllabus realizes educational policy and at the same time
is realized by pedagogic methodology. What happens when the
two are incompatible? Consider, for example, a case where
educational policy is designed to prepare pupils to fit into an
established social order. Such a position-oriented policy would
set a high premium on conformity, and this would be incompat¬
ible with a methodology which was person-oriented and
encouraged independent initiatives in learning. Conversely, one
might have good reason to suppose that a person-oriented
methodology with an emphasis on task-based discovery is to be
favoured on the grounds that it is more effective in the
promotion of learning. But this might run counter to a prevailing
position-oriented educational policy which called for conformity
and submission to teacher control.
As I argue later in this chapter, the teacher is not bound to
interpret the syllabus in. line with its intentions: methodology
can always find some room for manoeuvre. Nevertheless the
influence of policy is powerful and cannot be disregarded,
particularly, of course, because examinations assess the syllabus
which embodies it, and not the methodology. Furthermore, to
the extent that policy is informed by an established educational
orthodoxy, it is likely to reflect prevailing expectations and
attitudes on the part of both teachers and learners, not only in
respect of the language subject but also in respect of other
subjects on the curriculum. One cannot expect that learners will
take readily to modes of behaviour in the language class which
are at variance with those which are promoted in their other
lessons.
 The problems and principles of syllabus design 129

The general point to be made relates to the discussion in

the second chapter of this book: it is that pedagogic proposals
have always to be referred to socio-cultural factors in par¬
ticular educational settings. Whatever good reasons there may
be for adopting a pedagogy of person-orientation, one has to
recognize that there are societies, and educational systems
which serve them, for which this is a dangerous doctrine. The
design of a syllabus and its implementation by means of
methodology can never be simply a pedagogic matter. Having
noted this point, we can now go on to consider this matter of
implementation.

Syllabus and methodology

The task for teacher is to realize the syllabus as a course of
action by whatever methodological means seem most appropri¬
ate for the activation of learning. The syllabus itself is an inert
abstract object. It has much the same relationship to learner
activities as does any abstract model of knowledge to the
actuality of behaviour. Its function is not to provide a prospectus
of everything that the learner has to do. It is, so to speak, a set of
bearings for teacher action and not a set of instructions for
learner activity. What learners do is not directly determined by
the syllabus but is a consequence of how the syllabus is methodo¬
logically mediated by the teacher in the pursuit of his own course
of instruction.
It follows from this that changes in syllabus as such need have
no effect on learning whatever. They will only do so if they
inspire the teacher to introduce methodological innovations in
the planning and execution of activities in the classroom which
are consistent in some way with the conception of content and
the principle of ordering proposed in the new syllabus. The main
purpose of syllabus reform, in this view, is therefore to alter the
perspectives of teachers, shift their customary points of reference
and so provide them with a different set of guidelines. Unless
teachers clearly understand what principles inform these guide¬
lines and how they can be acted upon by means of an
appropriate methodology which promotes learning activities,
then new syllabus proposals are likely to remain in the region of
wishful thinking and pious hope, interesting ideas for academic
debate but ineffectual in the domain of pedagogic practice. The
effect of a syllabus on pedagogic practice depends on the
130 Aspects of teaching

effective appraisal and application of the ideas that inform it (see

Chapter 3).
Failure to recognize that the syllabus as a source of teacher
reference can only effect learning through methodological
mediation has led, I think, to the mistaken (but quite common)
assumption that a notional/fun(itional syllabus is synonymous
with a communicative approach to language teaching and that,
more generally, communicative pedagogy is primarily a matter
of syllabus design. But a notional/functional syllabus is of itself
no more communicative tlian is a ‘structural’ one. Communica¬
tion is what may or not be achieved through classroom activity;
it cannot be embodied in an abstract specification. Of course, as
I have already said, teachers may be inspired to realize the
implications of a particular way of specifying content and its
arrangement. But equally, they can ignore such implications, or
not recognize them, and proceed to teach in reference to a notional/
functional syllabus in just the same way as they previously
taught in reference to a structural one.
Conversely, it is perfectly possible to adopt a communicative
methodology in the realization of a syllabus designed along
structural lines. So I do not see that there is anything paradoxical
or perverse in saying that there is no such thing as a
communicative syllabus (see Stern 1984). The assumption that
there is, or can be, has, it seems to me, tended to deflect attention
from the real issues of communicative pedagogy which are
essentially methodological in character, and which therefore
relate crucially to the mediating role of the teacher.
Ways of characterizing and ordering the content units of a
language syllabus do not, then, determine classroom activity (for
a discussion see Johnson 1979; Breen 1987). But they may carry
implications about what a:ctivities might effectively be promoted
as consistent with the syllabus rationale. It is of interest to
enquire, therefore, what principles appear to inform different
approaches to syllabus design, and what implications might be
drawn from them for methodological practice.

The specification of syllabus content

In the approach to the definition of the language subject which
was orthodox until recent times, the characterization of content
was derived from formal models of linguistic description: the
units for teaching were assumed to be the same as the units' of
 The problems and principles of syllabus design 131

grammar. Latterly, in accordance with a different orthodoxy,

characterization has been in reference to the concepts and
actions (notions and functions) which these formal elements
most commonly realize when language knowledge is put to use.
What motivates this change of perspective? Mainly, it would
seem, considerations not of language learning but of language
use. The reason for defining language content in terms of
notional/functional rather than formal structural units is that
these are seen as being more immediately relevant to what
learners will need eventually to do with the language once they
have learned it. The emphasis is on the objectives and not the
procedures of language learning, on purpose not process. It is a
needs or goal-oriented rationale, expressed by Wilkins in the
following way:

The process of deciding what to teach is based on consider¬

ation of what the learner should most usefully be able to
communicate in the foreign language. (1976:19)

It would seem to follow that such a characterization is also

based on the assumption that learning behaviour must be some
sort of pre-scripted rehearsal for actual use. The implication for
teaching is that methodology should ensure that classroom
language is as close an approximation to authentic language
behaviour as possible. In this way, the learner accumulates a
performance repertoire for subsequent enactment when required.
This goal-oriented rehearsal assumption runs directly counter
to that which informs the characterization of syllabus content in
structural terms. It is not that the structural syllabus denies the
eventual communicative purpose of learning but that it implies a
different means to its achievement. It is often suggested that the
designers of such syllabuses supposed that language was of its
nature entirely reducible to elements of formal grammar and
failed to recognize the reality of use. But this, as I have pointed
out in Chapter 8, is a misrepresentation. Such syllabuses were
proposed as a means towards achieving language performance
through the skills of speaking, listening, reading, and writing.
That is to say, they were directed towards a communicative goal
and were intended, no less than the notional/functional syllabus,
as a preparation for use. The difference lies in the conception of
the means to this end. ‘Structural’ syllabuses are designed on the
assumption that it is the internalization of grammar coupled
with the exercise of linguistic skills in motor-perceptive manip-
132 Aspects of teaching

ulations (usage) which affords the most effective preparation for

the reality of communicative encounters (use). The content of
such syllabuses, therefore, is not conceived of as units of
performance, bits of pre-scripted communicative behaviour, to be
accumulated and issued like tokens when future occasions arise.
The content comprises units of linguistic competence, made
manifest as usage in spoken and written sentences but making no
claim to the status of natural utterance. Their learning con¬
stitutes an investment in competence (and not an accumulation
of performance capital) to be variously realized to meet
unpredictable requirements in the future.
In general terms the underlying implications behind the two
approaches to syllabus design might be formulated as follows.
The notional/functional syllabus implies that the subject lan¬
guage is to be taught as units of communicative performance for
accumulation. The structural syllabus implies that the subject is
to be taught as units of linguistic competence for investment.
It is interesting to note, in this connection, that in the first
volume of English Language Teaching published in 1946, the
editor A.S. Hornby provides examples of language teaching
techniques and makes it quite clear that these are not designed to
have any communicative validity: ‘None of the material given
above can be described as “conversation”. It is oral drill, and
even the expanded answers in the later stages are mechanical.
Real conversation cannot be taught in the artificial atmosphere
of the classroom and we may as well recognize the fact.’
But Hornby would of course have claimed that this mechan¬
istic process contributes to the eventual purpose of achieving the
ability to communicate. His position is the same as that of Lado,
referred to in Chapter 8. Teaching language for communication
as an end is not the same as teaching language as communication
as a means (see Widdowson 1984a: Paper 16).
Now it will be evident that although these two perspectives on
the language subject are commonly represented as in opposition,
they are really complementary, each compensating for the
limitations of the other. We need the communicative orientation
of one combined with the competence orientation of the other.
Communicative ability must clearly incorporate competence of
some kind and cannot just be a performance repertoire. The
syllabus has to allow for the investment of knowledge for future
and unpredictable realization. With regard to the notional/
functional syllabus, therefore, we need to dissociate communica-
 The problems and principles of syllabus design 133

tion from accumulation. The structural syllabus provides us with

the investment principle we want, but this is associated with the
formal properties of the language and, as is evident from past
experience, this sort of investment is not readily realized as
communicative behaviour. With regard to the structural sylla¬
bus, therefore, we need to dissociate investment from linguistic
forms as such, and apply it to communicative function. We
might summarize these general points in the following way:

learning learning
syllabus type content unit principle consequence

structural formal investment linguistic

competence

communicative
N/F functional accumulation
performance

Figure 2

An interesting question arises here with respect to these

principles of investment and accumulation, namely the extent to
which an adherence to the former principle necessarily entails an
analytic concept of content, and an adherence to the latter
principle a holistic one. Certainly the units of a structural
syllabus are discrete items of analysed knowledge, as Bialystok
defines it, and she seems to associate this with investment with
regard to language proficiency:

The assumption is that if knowledge is analysed, then certain

uses may be made of that knowledge which cannot be made of
knowledge which is unanalysed. (1982: 183)

Of course, as I noted in Chapter 6, a good deal of the native

speaker’s knowledge of language remains unanalysed as holistic
conglomerate units or formulas which resist analysis by rule
application (see Bolinger 1976; Pawley and Syder 1983) and
there is a good deal of compelling evidence that first and second
language learners invest in such formulas in the process of
learning (see Peters 1983; Vihman 1982). We should be careful,
therefore, to allow for the acquisition of unanalysed knowledge
as an aspect of competence and not assume, as the structural
134 Aspects of teaching

approach tended to do, that investment necessarily entails

analysis.
Interestingly, Wilkins associates analysis with the notional/
functional syllabus and refers to the structural syllabus as
synthetic (Wilkins 1976). But he uses these terms, it would seem,
not to characterize the nature of the content itself but what
teachers and learners do with it. Thus, according to Wilkins, the
teacher who works from a structural syllabus is constrained to
an incremental presentation of the discretely ordered items so as
to synthesize them in the learner’s mind as knowledge, whereas
working from a notional/functional syllabus will require the
presentation of language as holistic chunks of behaviour, leaving
the learner to do the necessary analysis.
So a synthetic syllabus is one which presents language as
analysed units to be synthesized in the process of learning
whereas an analytic syllabus presents language as synthesized
units to be analysed in the process of learning. We have, then, in
principle, a clear distinction: two competing approaches to
pedagogy, informed by diametrically opposed theories of learn¬
ing. What is not clear, however, is how this distinction is
exemplified by the structural syllabus on the one hand and the
notional/functional syllabus on the other, as Wilkins describes
them.
It is not obvious, for example, why the sentences which serve
as the basic units of a structural syllabus are any more analytic in
character than the speech acts or communicative expressions
which are the basic units of the notional/functional syllabus. The
two kinds of unit would appear to be exactly isomorphic, the
one being a functional variant of the other. Consider an
example:

Shall we go to the zoo?

This, says Wilkins, is an example of the category of com¬

municative function of suasion; more specifically its use is to
carry out the act of suggestion. Exactly the same combination of
words might appear in a structural syllabus. But here it would be
presented not as an act of suggestion but as one type of
interrogative sentence. But in both cases we have a collection of
words synthesized into a larger unit. They differ only in the way
the unit is conceived: formally on the one hand, functionally on
the other. And in both cases there is need for analysis if the
expression is to be taken as an example of how the language
 The problems and principles of syllabus design 135

works rather than as a fixed phrase without any generative

potential. That is to say, formally the learners will need to devise
other interrogative sentences with different lexical items and
syntactic constituents, and functionally to make other sugges¬
tions for carrying out different activities.
Of course, the kind of analysis will be very different. With
regard to the interrogative form, it is a matter of identifying the
device of subject/auxiliary verb reversal in English, of inferring
the contrast:

Shall we go to the zoo?

We shall go to the zoo.

This might then lead on to further analysis of the verb, the

recognition of the need to isolate tense from its lexical
connections and provide it with its own form in interrogative
sentences of this sort. Thus, we might infer the following
contrasts:

We went to the zoo.

"^Went we to the zoo?
Did we go to the zoo?

With regard to the expression of suggestion, analysis would be

a matter of identifying this modal auxiliary in association with
the first person plural pronoun as the realization of this
particular function, so that the following, for example, do not
have the same functional potential:

Will we go to the zoo?

Shall I go to the zoo?

So whether this combination of words is being presented as a

formal or a functional unit of meaning, the learner needs to
analyse out the first two words as a separate component which
has transferable signifying potential. The difference lies not in
the degree but in the kind of analysis which is applied.
It is possible to conceive of syllabuses which are designed in
accordance with the opposing principles that Wilkins suggests,
but which would be different from the structural and notional/
functional types that he describes. In reference to the broad
distinction which was made in Part 2 between semantic and
pragmatic approaches to language description, we might pro¬
pose, for example, that a synthetic syllabus is one which works
semantically from the inside out, so to speak, whereas the
136 Aspects of teaching

analytic syllabus is one which adopts the reverse, pragmatic

procedure and works from the outside in.
Thus a synthetic syllabus would define its units in lexical terms
and, along the lines suggested in Chapter 6, would gradually
introduce elements of grammar as needed to modify words in
association to make them more contextually effective. The
synthesizing of words into syntactic structures would be
motivated by communicative requirement. In this respect it
would not just deal in formal manipulations of the kind usually
associated with the structural syllabus, but in the realizations of
meaning which are usually claimed to characterize the notional/
functional syllabuses. But it would be thoroughly synthetic in the
Wilkins sense.
An analytic syllabus, on the other hand, would not be bottom-
up but top-down. That is to say, it would present language in
the form of larger textual units and set tasks of different kinds
which would direct the learners’ attention to specific features,
formal or functional, of the language they were exposed to.
Analysis would then be induced by means of controlled
procedural work.
It is possible too to conceive of syllabuses which reconcile the
opposing principles of synthesis and analysis. Crombie, for
example, has put forward proposals for what she calls a
relational approach to syllabus design (Crombie 1985). The
basic unit here is the propositional relation (see also Widdowson
1979: Paper 20). This can operate within a sentence where it
connects different constituents together, or it can operate across
sentence-like elements of larger textual units. The propositional
relation, therefore, operates both in the narrow domain favour¬
ed by the synthetic syllabus and the broader domain favoured by
the analytic syllabus. For example:

The cancellation of the match was on account of the rain.

This propositional content can be unpacked, so to speak, and

expressed in terms of a main and subordinate clause:

The match was cancelled because it was raining.

But the same cause and effect relationship between these

events can be expressed by a variety of ways across separate
elements of texts which are more or less contiguous, more or less
explicitly connected. For example:
 The problems and principles of syllabus design 137

The match was cancelled.It was raining.

The match was cancelled.This was because of the
rain.
It was raining. The match was cancelled.
It was raining . Because of this, the match was
cancelled.

Where there is an absence of explicit signalling and the

presence of intervening, potentially distracting, text (indicated
by the dotted lines) there will be an increased dependence on the
kind of interpretative procedures discussed in Chapter 7.
A syllabus based on the different realization of propositional
relations could either proceed synthetically by projecting out¬
wards from sentence to text, or analytically by focusing inwards
from text to sentence.
What all this would seem to show is, as I indicated earlier, that
syllabus type does not determine its mode of implementation in
any particular methodology. I do not think it follows that
learners who are practised in the use of formulaic units will
necessarily analyse them so as to have their constituent parts
accessible for other combinations and applicable in a general
range of contexts. But now that methodological issues have
come into the discussion, we need to consider them more closely.

Methodological implications of content specification

What kind of methodological practices would be consistent
with the ideas about language and learning which are implied by
these different approaches to syllabus design? A teacher using a
notional/functional syllabus as reference, if she is sensitive to
what is implied, will tend to promote activities in the classroom
which have as much communicative verisimilitude as possible,
so that they function effectively as rehearsal. She will favour
materials which are ‘authentic’ as instances of use. Her metho¬
dology, in brief, will be behaviour-oriented. A teacher acting on
the implications of a structural syllabus, on the other hand, will
tend to promote activities which serve to internalize the formal
properties of language and will accept the need for contrivance.
Her methodology will be knowledge-oriented. The danger of the
former methodology is that learners may accumulate perform¬
ance units without being able to refer them to any underlying
competence which alone can provide the sort of adaptable
138 Aspects of teaching

resource required for actual communication and for further

learning. The danger of the latter methodology is that the
learners may find that having made their speculative investment
in form, they cannot gain effective access to it: they cannot act
on their knowledge. ^
It seems clear that what is wanted is a methodology which will
(in reference to the diagram on page 133) provide for communicat¬
ive competence by functional investment. I have indicated else¬
where what such a methodology might involve (e.g. Widdowson
1983, 1984a and Chapter 10 in the present book). Very generally,
it would engage the learners in problem-solving tasks as purpose¬
ful activities but without the rehearsal requirement that they
should be realistic or ‘authentic’ as natural social behaviour. The
process of solving such problems would involve a conscious and
repeated reference to the formal properties of the language, not
in the abstract, dissociated from use, but as a necessary resource
for the achievement of communicative outcomes. Such a
methodology makes a virtue out of the necessity of classroom
contrivance and sees it not as a debilitating constraint but as a
facilitating condition for the development of communicative
competence in the language.
I will return to the matter of a problem-solving methodology
later in relation to issues arising from the ordering of syllabus
content. But for the present we need to note that such a
methodology is not determined by a commitment to either of the
syllabus types that have been discussed but depends on realizing
a complementary relationship between them. It can either
develop as a communicative complement to the principle of
investment or as an investment complement to the idea of
functional content. It does not matter which syllabus serves as
the point of reference so long as the teacher can realize it
appropriately in this complementary fashion in actual classroom
activities. This is why so much discussion about the design of
‘communicative’ syllabuses is misdirected. Which kind of syllabus
a teacher has to work with is relatively unimportant. This is
fortunate since she very often has no choice in the matter
anyway. What is important is that teachers should understand
the principles underlying the characterization of content in a
particular syllabus so that they might adopt or adapt these
effectively in the area where they do have room for independent
action, namely in the mediating activities of classroom
methodology.
 The problems and principles of syllabus design 139

The ordering of syllabus content

Discussion so far has centred on the different ways that course
content is characterized as defining what aspects of language
might have pedagogic primacy in getting learners to acquire
relevant proficiency as an eventual objective. We have yet to
consider the question of how such content might be ordered so
as to serve effectively the process of learning. I referred to
ordering earlier as the specification of interim objectives. These
can be conceived of in two ways. One is in terms of their
surrender value or communicative pay-off in respect of practical
use. This would yield a series of stages each representing a level
in proficiency for use. The criteria for such staging relate to what
is to be learned by what point in the way of realizable profit, as it
were, and are determined by considerations of investment and so
on which we have already dealt with in the discussion of the
characterization of content. The second way of conceiving of
ordering, however, has to do with language development as a
function of the learning process itself. This is a very different
matter.
The distinction I am making here between the different
criteria for ordering corresponds, in my mind, with that familiar
in testing between achievement and proficiency. Tests of
achievement measure progress in learning. Since this progress is
not an accumulation of standard forms but the development of a
non-standard interlanguage, there is, or should be, no require¬
ment in achievement tests that performance should match up to
the norms of correctness or appropriateness associated with
native speaker behaviour. Tests of proficiency, on the other
hand, measure the ability to access and to act upon what has
been learnt to realize effective communicative behaviour. Here
learner performance clearly does have to be set against the
norms of native speakers.
A central question in regard to ordering as a way of
facilitating achievement (as distinct from ordering as a way of
providing for levels of proficiency) is the extent to which content
can be ordered so as to conform to, or at least be congruent with,
the sequences which mark the natural process of learning. In a
paper first published in 1979 (though reprinted since), Corder
made the following remarks:
Efficient language teaching must work with, rather than
against, natural processes, facilitate and expedite rather than
140 Aspects of teaching

impede learning. Teachers and teaching materials must adapt

to the learner rather than vice versa. The study of interlan¬
guage is the study of the natural processes of language
learning. What has been discovered so far suggests that the
nearer we can approximate language teaching to the learning
of a second language in an informal setting the more successful
we shall be.
And he goes on to derive from this the following practical
proposal:

. . .the accommodation of the structure of our linguistic

syllabuses and teaching materials to fit what is known of the
sequence of progressive complication of the approximate
systems of the free learner. (1981: 77)
If we accept that such accommodation is desirable (and this, as I
have suggested in Chapter 3, is not as self-evident as it might seem
to be) then the question arises as to whether (nearly ten years on)
we now know enough about second language acquisition and
natural emergence of interlanguage to be able to sequence
content accordingly, thereby adapting the teacher’s syllabus to con¬
form to the ‘built-in’ syllabus of the learner, as Corder would wish.
The last ten years have seen a great proliferation of research
into second language acquisition (SLA). The collection of papers
published to honour Corder himself (Davies, Criper, and
Howatt 1984) gives a good indication of its scope and subtlety,
as does the synthesis provided by Ellis (1985). This research has
yielded an abundance of fascinating facts and speculations about
the language learning process and its possible determinants.
What it has not yielded is anything so specific as a definitive
linear succession of items that defines interlanguage develop¬
ment, and which could directly inform content order in a
syllabus.
Indeed there would appear to be good evidence in support of
the intuitive conviction that natural learning does not proceed
strictly in this linear fashion anyway (see Long 1985: 79 and the
references cited there). The indications are of an emergence of
certain salient features of language, resembling general prin¬
ciples, rather than the precise formation of rules, which follow a
development sequence of an implicational kind with one feature
following dependently on the previous appearance of another.
But at no regular intervals. This is because in between these
appearances, other features may intervene in no fixed order
 The problems and principles of syllabus design 141

playing out a supporting role. To account for this variable

emergence, Ellis has proposed a distinction between order and
sequence of acquisition. The former is the succession without
fixed intervals of the more general organizing principles or
salient features of grammar, and the latter is the succession of
these together with other features appearing without predeter¬
mined order in linear arrangement (Ellis 1985: 64). This
proposal incidently has interesting parallels with the structural
analysis of folktales and myths where the objective is to identify
the underlying (and perhaps universal) order of elements
through the varying sequences of particular events which appear
in different narratives.
We should note, therefore (to return to the relatively non-
fictional world of second language acquisition research), that if
any stable and generalizable findings do eventually emerge from
such research, they can, if this current hypothesis is sustained,
only relate to order (in Ellis’s terms) and not to sequence. So they
could not provide criteria for any precise specification but only
for a more general staging of content, motivated though by
learning needs and a concern for achievement rather than by the
requirement for a usable communicative return and a concern
for proficiency. Syllabus design based on such findings would
presumably consist of a series of general parameters representing
the established implicational sequence of general grammatical
principles, subsequently given more specific realization by
methodological means, involving the incidental inclusion of
other features.
Two further points might be made in reference to this question
as to whether content arrangement in a syllabus might be
informed by findings relating to the natural growth of language
in the learner’s mind. The first has to do with the observation
made earlier that there is no necessary correspondence between
achievement in learning and proficiency in use. Achievement
progresses through the approximative systems of interlanguage
and, as Pienemann points out, to model a syllabus on this
progression would involve the inclusion of interlanguage forms
which, by definition, do not conform to the rules of the standard
language (see Pienemann 1985). The assumption that a syllabus
should only include correct instances of language runs counter to
the notion that it should be fashioned to correspond with the in¬
built syllabus of the natural learner. Bearing in mind that
syllabuses are designed as teaching specifications and generally
142 Aspects of teaching

prescribed by educational authorities, it is highly unlikely, even

if it were deemed to be pedagogically desirable, that a syllabus
which specified deviant forms as teaching items would find much
institutional favour.
A second point has to do with how linguistic knowledge is to
be defined. Studies in second language acquisition have generally
defined it in terms of the internalization of morpho-syntactic
rules of the language system (see Ellis 1985: 288—9). What is not
clear, though, is how freely available such rules are in respect to
different lexical realization and contextual function. A learner
may know certain rules but only within a limited range of
application. She may know them as they are associated with a
certain set of lexical items but not be able to generalize them to
others; she may know them as they are used to realize certain
communicative functions but not be able to extend them into
other domains of use. In short, the rules may only be known in
relative degrees of analysability and accessibility (see again
Bialystok 1982).
Considerations such as these lend support to Selinker’s belief
in the need to bring together the fields of second language
acquisition and language for special purposes (ESP) (see, for
example, Selinker 1984; Selinker and Douglas 1985). In ESP we
are concerned (among other things) with how the discourse
conventions which characterize different domains of use are
textualized by particular choices of the lexis and grammar of the
language in question. Thus it might be demonstrated that a
certain rhetorical function in one discourse domain is typically
realized by means of a particular tense (see Trimble 1985). It
would be the purpose of an ESP course to establish the
relationship between tense and function in this domain. But
there is no guarantee that the learner would be able to dissociate
this tense from the lexical and contextual connections which
define this function so that it was freely transferable to other
functions and other domains. It could be learned only as the
element of a relatively restricted formula. In other domains, the
learner might make use of other forms in his interlanguage
repertoire. This particular tense would, then, be only partially
acquired as a contextual variant. But an ESP syllabus which was
designed to provide only for the use of language in this particular
domain would not of its nature make provision for any
extension of this restricted range.
It is worth pointing out, however, that variable competence is
 The problems and principles of syllabus design 143

not only to be associated with the interlanguage of learners. It is

a fact of life for native speakers also. They too may be able to
access certain areas of knowledge in recognition for receptive
purposes but fail to access them in recall for production. They
may know certain words as they appear in certain phrases and as
they relate to certain contexts but be unable to use them in free
transference. All language knowledge is only relatively analysed
and accessible: a good deal of it, as I indicated earlier, takes the
form of conglomerate units and is contextually tagged.
All languages reveal the co-existence of lexical and syntactic
variants which are associated with different functions, forms
which are semantically equivalent but pragmatically distinguish¬
able, appropriate in some contexts but not in others. So when
the language learners develop context related variants in their
interlanguage, they are following their natural instincts as
language users in a way we would wish to encourage as evidence
of communicative competence. The only problem is that the
learners’ variants are deviations from the standard norm, and in
this respect are evidence of linguistic incompetence.
It is clear from all this that if SLA research is to yield anything
of relevance to syllabus design which is directed at the
development of communicative competence, then it will need to
define its object of enquiry so as to take into account the natural
language phenomenon of contextual variability (see Ellis 1985:
Chapter 4; Tarone 1988).
What we need to know, then, is how learners extend the range
of application of the rules they internalize, how they develop in
making use of ‘procedures for accessing the knowledge’ (Bialy-
stok and Sharwood-Smith 1985). This would clearly have a
bearing on the arrangement of syllabus content. Instead of
arranging items in some absolute sequence, assuming that once
the item was learned in one lexical and contextual environment,
it would transfer automatically to all others, we could instead
order content cyclically in terms of an increasing range of
application, requiring of methodology that it should provide
activities which develop the necessary accessing procedures.

Methodological implications of content ordering

Work on second language acquisition gives us (as yet at least)
little in the way of reliable guidance as far as syllabus design is
concerned. So what pedagogic relevance does it have? The
144 Aspects of teaching

question brings us back once more to the primacy of methodo¬

logical considerations. The factors which bear upon natural
language acquisition are clearly many and complex, but they can
be grouped under two main headings. There are those which
constitute essential conditions. These are of a cognitive character
and they set the psychic co-ordinates, as it were, of natural
language development. The second sort of factors are of a
communicative character and these trigger off or activate the
developmental process, and provide the enabling conditions for
growth by providing an experience of language. Now if we
cannot fashion a syllabus to incorporate the essential conditions
of the natural learning process what we can do is create these
enabling conditions in the classroom by means of a methodology
which allows the learner to acquire language as a function of
natural language behaviour.
It is this kind of thinking that informs the ‘natural approach’
(Krashen and Terrell 1983) and, in a different and more rigorous
way, the work of Prabhu in South India (Prabhu 1985, 1987;
Johnson 1982; Brumfit 1984a). For Krashen, as we have seen
(see Chapter 2), natural language behaviour seems to involve
little more than the reception of comprehensible input, which he
confidently proclaims to be, 'the [sic] fundamental pedagogical
principle in second language teaching’ (Krashen 1981). This
principle, the Input Hypothesis, is formulated as follows:

1 We acquire by understanding input language that contains

structures a bit beyond our current level of competence.
2 Speech is a result of acquisition, not a cause.
3 If input is understood, and there is enough of it, the
necessary grammar is automatically provided. {1983:159)

Prabhu also believes that grammar is acquired as a function of

meaning realization, but accepts the need for more active
involvement on the part of the learner whereby meaning is
achieved in the process of problem solving. For Krashen, it
would appear, the input acts upon the mind, and this serves as a
sufficient catalyst. How the act of comprehension is actually
achieved is not considered. Nor, for that matter is the issue of
analysability and the possibility of a restricted range of applica¬
tion as discussed earlier. For Prabhu, on the other hand, it is the
intellectual effort of reasoning which sets the internal cognitive
process of acquisition into motion. The mind acts upon the
input. Although in both cases the assumption is that involvement
 The problems and principles of syllabus design 145

with meaningful uses of language will be effective in developing

grammatical competence, Prabhu’s approach provides for ana¬
lysis as a condition of conscious learning.
The kind of task which Prabhu found to be most effective was
one which involved what he calls ‘reasoning-gap activities’. Such
an activity ‘involves deriving some new information from given
information through processes of inference, deduction, practical
reasoning, or a perception of relationships or patterns’ (1987:
47). Such an activity is more likely of its very nature to lead to an
analytic abstraction of intrinsic meanings in language than is the
passive exposure to input that Krashen would seem to advocate,
which relies on the activation of the language acquisition device
without the learners’ rational involvement. Another reason for
Prabhu’s preference for reasoning-gap activities is that it allows
for teacher intervention and the control of learning:

When a reasoning-gap activity proves difficult for learners, the

teacher is able to guide their efforts step by step, making the
reasoning explicit or breaking it down into smaller steps, or
offering parallel instances to particular steps. (1987:48)

Interestingly enough, this bears a marked resemblance to the

methodology of what Prabhu calls the structural-oral-situational
(S.O.S.) approach and which his procedural pedagogy is quite
explicitly intended to replace. The difference is, of course, that in
Prabhu’s proposals the methodology is not applied to the
language itself but to problem-solving activities which are
designed to induce language contingently. Even so, such explicit
focusing on the elements of the problem is, one would think,
bound to direct attention to the elements of language in which
they are expressed. One might draw a parallel with the kind of
problems which are set in the teaching of mathematics (and
indeed the task which Prabhu describes involving the use of
railway timetables could be readily adapted to such a purpose).
In this case learners would need to apply mathematical rather
than linguistic procedures as a means to the solution. But they
would at the same time be learning the use of algebraic equations
and other formal devices for calculation. It seems reasonable to
suggest that the use of the formal devices of language would be
learned in a comparable way.
At all events, it is clear that the methodology of the Prabhu
approach is not one which simply allows natural acquisition to
take its course by undirected discovery or the instinctive reaction
146 Aspects of teaching

to input. Monitoring is much in evidence. It is for this reason

that he describes it as learnmg- but not learner-centered (see
Prabhu 1985).
Although there are differences between them, the approaches
adopted by Krashen and Prabhu are both based on an
assumption which runs directly qounter to that which informs
the design of a structural syllabus. Rather than supposing that a
grammatical investment (demonstration and practice) will yield
a communicative return, the assumption here is that a com¬
municative investment (input or interaction) will yield a gram¬
matical return. In reference to the distinction made earlier the
rationale for communicative methodology relates here not to the
goal but to the process of learning: it is based not on the
accumulation but on the investment principle.
The view here is that the activation of natural acquisition will
be sufficient, without the deliberate teacher inducement of
conscious learning by directing the attention of learners to form.
It is a view that not everybody feels disposed to share. Brumfit,
for one, does not.

A consequence of the view that learning does not assist

acquisition is the dismissal of several traditional learning
activities as peripheral. Examples are exercises and drills and
grammatical explanation procedures . . . and an authoritative
rejection of such procedures needs to be based on firmer
evidence than has been forthcoming. (1984a: 320)

The role of the syllabus

These are not the only pedagogic traditions which are called
into question by the transfer of power, so to speak, from teacher
to learner, by the reliance placed on natural acquisition rather
than directed learning. The teacher’s authority for direction, the
syllabus itself, becomes vulnerable. Since it cannot be fashioned
to model the acquisition process, it can only be an instrument of
teacher imposition (no matter how benevolent) and so, it might
be argued, it can and indeed should be dispensed with, at least as
it is traditionally conceived. Instead one might retain the term
but redefine it to mean not the projection of a scheme of work
but a retrospective record of classroom activity, a product of the
methodological process rather than a preconceived framework
within which such a process can operate (see Candlin 1984).
 The problems and principles of syllabus design 147

One way of allowing for the initiative of the learner to plot his
own course by following his natural flair is effectively to
abandon altogether the traditional notion of syllabus. Another
less extreme strategy is to redefine content not in terms of the
forms or functions of language as such but in terms of problems
of a conceptual or communicative character which require the
use of language for their solution. This is the strategy favoured
by Prabhu in his proposals for a so-called procedural syllabus.
This can be seen (if this does not seem immodest) as a
modification of the proposal I first put forward nearly twenty
years ago (Widdowson 1968) and which is currently finding
some belated favour, that language learning is most effectively
promoted by the contingent use of the language in the study of
other subjects on the curriculum.
It is important to note that this proposal is based on a process-
oriented investment rationale rather than a goal-oriented accu¬
mulation one. This has not always been appreciated. Cook, for
example, comments that ‘Many language teachers would deny
that their main purpose in teaching English was to enable their
students to understand physics or geography in English.’ (1983:
231)
But the point is that the subject is meant to create conditions
for purposeful activity whereby the process of language learning
is engaged: it does not constitute the unique goal of learning.
The argument is that if learners have learned the language
purposefully, then they will have invested in an ability which can
be put to purposes beyond those which originally served to
develop it. This in turn invests language teaching with an
educational rather than a training purpose (see Widdowson
1983).
The modification of this subject-based proposal, as put
forward by Prabhu, retains the investment principle but dis¬
sociates it from established curriculum subjects. What it essen¬
tially involves, therefore, is the devising of a subject (logical
reasoning, for example) which applies activities associated with
other subjects to neutral topics. This kind of problem solving, as
a methodological procedure, was mentioned earlier in this
chapter as a way of reconciling the apparently competing
implications of structural and notional/functional syllabuses.
But the use of problem-solving tasks as units of syllabus content,
as distinct from activities for syllabus realization, encounters a
number of difficulties. By what criteria, for example, are such
148 Aspects of teaching

tasks to be sequentially arranged? If the sequence is to be in

accord with natural learning, as it would be only consistent to
require, then reliable information is needed about cognitive
development at different stages of maturation, about the
conditions, psychological and social, which attend the emer¬
gence in the mind of general' problem-solving capabilities.
Armed with such information, we could perhaps relate these
capabilities to certain task types, analysed into their constituent
features and then given token realizations and ranged in an order
of increasing complexity.^ Such information is not, to my
knowledge, currently available.
Suggestive findings relating to this issue have, how¬
ever, emerged from general psychological research on child
development. It seems clear, for example, that tasks increase in
difficulty the more remote they are from the schematic know¬
ledge of the children. Margaret Donaldson (1978) takes issue
with Piaget on the question of the child’s reasoning ability at
about the age of five or six. Piaget’s claim is that children at this
age are incapable of reasoning out the relationship between
classes and sub-classes of objects and offers the following
evidence in support of his claim. When children are presented
with two obvious sub-classes of a more general class, say four
red flowers and two white flowers, they will fail to answer
correctly a question of the kind:

Are there more red flowers or more flowers?

t*

Children will regularly reply that there are more red flowers.
But the reason for this may be that children are interpreting what
they suppose to be a likely utterance rather than the actual
words which are said. The question is an odd one which is
unlikely to occur in the context of ordinary discourse. A more
normal question would be:

Are there more red flowers or more white flowers?

In other words the children may be reforming the language to

a model of their familiar world and this schematic knowledge
overrides the specific sense of the words—a common phenom¬
enon in language use, as we have seen from the discussion earlier
in Chapter 7.
Donaldson points out that experiments show that when tasks
are related rnore closely to the child’s experience, when the
reasoning is associated with familiar schemata and embedded in
 The problems and principles of syllabus design 149

other and more purposeful activities, then the child reveals no

such deductive deficiency.
We return here to issues concerning the negotiation of
meaning raised in Part 2. There it was pointed out that the
less schematic knowledge can be engaged in interpretation (itself
a kind of problem solving of course), the more reliance will need
to be placed on systemic knowledge, or knowledge of the
language. Now it would seem that there is good reason to
suppose that schooling in general through all subjects seeks to
develop in children the capability for dealing with meanings in
language which are not schema-derived, that is to say which are
not dependent on the immediate context of experience. In
reference to research carried out in Canada, for example, David
Olson observes:

All of these studies suggest that schooling is responsible for an

important reorientation to language, the differentiation of
what the sentence means, its direct or ‘sentence meaning’. It is
my conjecture that schooled competence, including com¬
petence in intelligence tests, is based on the ability to respond
to sentence meaning, the meaning in the syntactic and lexical
structures of the sentence, or as I prefer to state it, to treat
language as if the meaning were ‘in the text’. The focus on
language, on mastering and explicating the meanings in the
language, takes many forms: formulating definitions of words,
devising verbal rules, and drawing valid inferences. These are
of course among the very intellectual effects associated with
schooling. (1981:385)

If Olson is right (and he acknowledges that what he says is

conjecture) then the development of the problem-solving ability
or conceptual competence in schoolchildren may not be distinct
from the development of competence in the language. This
would suggest in turn that task grading might be done in
reference to degrees of schematic familiarity and dependence on
meanings intrinsic to text. It would also suggest in reference to
the points made earlier about Prabhu’s methodology that a
focusing on the elements of the reasoning-gap task might well, in
practice, involve directing attention to the elements of language
itself as a corollary.
At all events, we have (at the moment) no definite guidance on
how tasks might be sequenced in terms of conceptual complexity
or implicational dependence. So the proposal for a task-based
150 Aspects of teaching

syllabus, if it is to be concerned with the principled arrangement

as well as the specification of content, runs into precisely the
same problems as beset the proposal to model a syllabus on the
sequence of natural language acquisition. Against this, it might
be suggested that there is no need for such precision in task
gradation, that a rough and intuitive calculation of complexity
will do. But then how does one guard against intuition referring
to linguistic criteria for such a calculation? And if the task
ordering does not correspond with a natural order, then it can be
argued that it will inhibit natural cognitive initiative in just the
same way as an uninformed linguistic ordering constrains
natural language development. Either way, the result is just the
sort of imposition and interference in learning processes which it
is the very purpose of a procedural syllabus to avoid.
The attempt to specify and order content units of a syllabus so
as to correspond with natural learning, then, runs into problems
for which we have (at present at least) no solutions, whether
these units are defined in linguistic or non-linguistic terms. So if
all syllabuses constitute an unnatural confinement which im¬
pedes the intuitive process of learning, why not accept the view
previously mentioned and dispense with them altogether, letting
courses develop out of classroom activity under their own
momentum as a function of communicative interaction? But
there are problems here too.

The limitations of intuitive learning

The belief that learning is best achieved by the unconstrained
exercise of natural intuition would appear to be based on two
related assumptions which, if not false, are certainly unfounded.
I have already referred to these assumptions in Part 1 of this
book. The first is that the operation of this natural intuition is
necessarily made less effective by being constrained. The second
assumption is that natural learning is necessarily more effective
than learning that is induced by artifice. On closer examination
both of these assumptions turn out to be of doubtful validity.
Against the first assumption one can argue that, unless one
takes an extreme deterministic position, the very process of
learning depends on the presence of constraints of one sort or
another. Indeed, learning would seem to be of its nature a matter
of accommodating intuitions to patterns of belief and behaviour
established by convention. Learning, one might argue, is
 The problems and principles of syllabus design 151

essentially an implied act of social conformity. This does not

mean, of course, that there is no room for individual manoeuvre,
for creativity, for the working of the imagination beyond the
bounds of convention, but all of this presupposes the existence
of such bounds. No freedom is possible unless there are limits to
define it. So the constraints on learning are in effect the essential
conditions of its existence. The child, for example, learns its
mother tongue in the process of socialization, in the process, that
is to say, of recognizing the limits to individual initiative defined
by the conventional structures of its society. The child is not a
free agent. If it were, it would learn nothing. So to speak of
unconstrained learning, is really a contradiction in terms. The
question is not how to set the learner free from constraints but
what kind of constraints are likely to promote learning most
effectively, and most likely, therefore, to provide eventual
freedom of action. In this view it makes no sense to talk about
dispensing with the syllabus, or of conceiving of it as a
retrospective record because this simply denies the very nature of
learning. Far from removing an impediment, it actually imposes
one. We cannot avoid the responsibility of specifying and
ordering content in advance in a syllabus. The question is: how
can its design be made (in computer parlance) learner-friendly?
The second assumption underlying the belief in intuitive
learning is that what is natural is self-evidently to be favoured as
intrinsically good and desirable. This is a notion that has always
had a powerful sentimental appeal. It would seem to arise from a
longing for primitive innocence. But civilizations are based on a
denial of its truth. They seek to improve on nature by artifice,
and they are created on the contrary assumption that nature can
be controlled, crafted, adapted, and fashioned into more
manageable and convenient form. The assumption here is that
what is natural is inchoate and inefficient compared with what
can be contrived by the conscious application of human reason.
It seems to me that education, one of civilization’s essential
services, is similarly based on a belief in contrivance. Its business
is not to defer to natural intuition but to direct it by ingenuity so
that it is effectively applied in the acquisition of knowledge and
ability. This does not mean that learners are thereby committed
for life to strict confinement within the conventional constraints
that serve the process of learning. Once knowledge and ability
are acquired they can be turned against the conventions which
brought them into being. But unless the learner first acquires
152 Aspects of teaching

these capabilities by accepting such conventions he is powerless

to challenge them. Education in this way can provide for equal
opportunity in a way that nature most assuredly does not.
Teaching, in this view, seeks to remedy by artifice the
deficiencies of natural processes, and induce development which
would otherwise be left to the vagaries of chance. In the case of
second language learning, the process is inefficient in the normal
circumstances of untutored acquisition, so it seems to be a
perversion of pedagogy to seek to replicate these in the
classroom. This does not, of course, mean that the particular
characteristics of learners are to be deliberately disregarded. The
artifice has to be such as to engage their interest and be
consistent with their cognitive and affective dispositions. Arti¬
fice, like art itself, has to carry conviction.
The rejection of a syllabus as a projected programme of work
for the teacher to follow in favour of allowing learning to
happen by natural instinct is, in effect, to deny learners the
necessary conditions for learning, and this is all the more
pernicious when such denial is imposed by invoking the principle
of individual freedom.
Apart from (but related to) the point that there needs to be
some framework for learning, there is a second positive effect of
the delimitation provided by a syllabus. This has to do not with
the process of learning but with the well-being of the learner. A
syllabus offers security. We return here to the suggestion I made
at the beginning of this chapter that the general purpose of
education is to initiate people into a secondary culture. This
means that they are drawn away from a sole reliance on the
patterns of familiar experience, and the more removed the
secondary culture is from the primary, the greater the danger of
alienation, and the loss of confidence and self-esteem. In such a
situation, it is natural for learners to seek the solidarity of their
peers and set up a counter-culture to that of the classroom. But if
they are to be induced into formal learning they are likely to
need some definite direction which will indicate the paths they
are to follow, something known in advance, some frame of
reference. This is what the syllabus can provide. No doubt there
are some people who need no such guidance, who can plot their
own course without worrying about losing their way. But many,
it would appear, because of cultural or personal disposition,
need the suppqrt which a projected plan of work can provide.
So the very nature of the educational enterprise, the very
 The problems and principles of syllabus design 153

purpose of the promotion of learning, would seem to require the

design of an informing framework within which classroom
activities can operate and in terms of which they take on
significance. The syllabus provides the teacher with the basis for
such a framework. How it is to be made operational as a
particular course of instruction, and how instruction is itself to
be defined, are methodological matters which are the concern of
individual teachers. But to dispense with the syllabus is an act of
self-inflicted disablement.

Criteria for syllabus design

If we are to accept the need of a syllabus, then by reference to
what criteria is it to be designed? Criteria relating to the natural
learning process, as might be revealed by research on second
language acquisition, are not self-evidently relevant, even if they
were available. It is of interest to note that in a recent paper one
of the most authoritative figures in the SLA field, in considering
the relevance of such research for syllabus design, can only
conclude by putting forward proposals for task-based teaching
(Long 1985). This, one has to say, is rather a small pedagogic
return on such an extensive investment of research effort.
Enquiries are now increasingly being directed at the effect of
classroom instruction on what is assumed to be the natural
acquisition process (see Ellis 1985: Chapter 9; Long 1983c).
This would seem logically to involve a convergence of SLA
research with that which has been carried out on classroom
interaction (see Seliger and Long 1983; Mitchell 1985) E^rch
and Kasper 1985; Allwright 1988; van Lier 1988), although
curiously, Ellis deals with these areas of research in two quite
separate chapters (1985: Chapters 6 and 9).
Insights may well emerge from such a convergence (as they
may from the relationship mentioned earlier between SLA and
LSP), but the variables involved in any consideration of different
pedagogic practices have proved in the past to be so diverse and
elusive that it is not easy to imagine at the moment what reliable
guidance such insights might afford. The key notion of in¬
struction itself has not so far in SLA research been subjected to
much conceptual analysis and tends to be dealt with (as do other
key concepts like communication and comprehension) in a
quantitative and rather simplistic way (how much instruction
rather than what kind). It is not easy to see how the shift to a
154 Aspects of teaching

more valid qualitative enquiry can retain reliability and so avoid

the sort of difficulties encountered in, for example, the Penn¬
sylvania Project (Smith 1970) in seeking to evaluate the effects of
different kinds of classroom encounters; or the sort of rather
inconsequential proliferation of learner characteristics coming
out of research on the ‘Good Language Learner’ (see Naiman et
al. 1978; Rubin 1975, 1981; Stern 1983). It is not easy to see
either, in the light of the observations made earlier, how any
findings (no matter how suggestive for methodology) would
have any direct bearing on syllabus design as such.
Meanwhile, what are we to do? On the one hand we need the
syllabus to give shape to instruction (however conceived) but on
the other hand we do not seem, to have any definite or
authoritative guidance as to how to set about its design. Neither
goal-oriented criteria which relate to proficiency nor process-
oriented criteria which relate to achievement seem to be
satisfactory, whether these are applied to the specification of
syllabus content or to the manner of its arrangement. And I
think it is unlikely that any research at present or in the future
will provide us with anything very definite to resolve these
difficulties.
But then I am not sure that this matters very much. For I am
not convinced that the issue of just how the content of a syllabus
should be specified and arranged is actually a very crucial one.
To my mind, what is crucial is that the principles upon which the
syllabus has been designed, whether in terms of structures,
notions, functions, topics, or tasks, should be made quite explicit
so that teachers can submit them to appraisal and application. In
this way they can make use of the syllabus as a set of bearings for
the plotting of their own course in a lesson sequence, and for the
realization of aspects of language and learning which the
syllabus of its nature cannot account for. Conceived of in this
fashion, a syllabus is a rationalized construct whose principles of
design are made plain so that teachers can refer to them, and not
just to the design itself, to make their own methodological plans
to suit the circumstances of their particular classroom. In this
explicit, explanatory form, the syllabus becomes an important
element in the continuing education of teachers as they experi¬
ment with its variable realization in the process of actual
teaching (see Chapter 3). And such a conception of the syllabus
acknowledges that although learners must ultimately accept
individual responsibility for their own learning, it is the teacher
 The problems and principles of syllabus design 155

who must, as social agent, prepare them for this responsibility by

setting the limits which condition learning. And she cannot do
this unless she herself has limits within which she can define her
own enterprise and it is these which it is the essential purpose of
the syllabus to provide.
 I' ^ ' I' ‘t Cj:w V* .

./ w rf(fwJ4 u

■' 1

^ ’ ir» lui
•'»

... ■■■..'• .1
-■. . ■-•'■■ . ' :

«i ,a.' ■ • ^ . . v'-.v 1. 104;

■-It : ■■♦••)•

■ I -in

■ . . ^ . i< •»■

^■K^i mv^t '* •*

K 11
.1 *!
; «* . ..► * V
' . 'Si
M9E. . . ..'».Vi
tttmr |k(^. CM|
it

# • ' ■'.

■ . •,v1
•* rr 1
10 Methodology for the teaching
of meaning

Structural and communicative approaches: knovs^ing and

doing
It follows from the characterization of language proposed in
the preceding part of this book that learning a language as a
natural human accomplishment involves getting to know some¬
thing, and being able to do something with that knowledge.
Language learning has two sides to it: knowing and doing
(competence and performance), the first associated with a
medium and the second with a mediation perspective on
meaning, as these are defined in the introductory chapter of this
section. And as I pointed out there, different approaches to
language teaching have tended to emphasize one rather than,
and often at the expense of, the other.
Thus the approach to language teaching in fashion until the
recent past, the so-called ‘structural approach’, focuses attention
on knowing. Here items of language, words and sentences, are
presented and practised in a way which is intended best to help
the learners to internalize them as forms containing meaning
within themselves, as semantic capsules, so to speak. The
assumption is that once learners have achieved this semantic
knowledge, then they will be able to use it pragmatically to do
things: to converse, to read, to write; to engage, in short, in
communicative activity of the same kind (if not to the same
degree) as that associated with their mother tongue. The
assumption is, generally speaking, that the primary task of
teaching is to impart knowledge and that the learners can be left
to find out how to do things with it for themselves.
This does not mean that classrooms where this approach is
followed are devoid of activity. Typically there is a great deal of
doing: learners speaking in pairs and in groups, reading
passages, composing sentences, busy practising the four skills.
158 Aspects of teaching

But in this approach, these activities are seen essentially as a

means to the internalization of knowledge and not as ends which
are achieved by the use of knowledge. Their purpose is to
stimulate participation in the use of skills to help in learning the
language system as a medium for meaning. They differ therefore
from the kind of doings which we would associate with normal
uses of language, where the purpose is to achieve some outcome
outside language learning and where linguistic knowledge is
drawn upon as a means and not an end, as a frame of reference
for the mediation of meaning.
Let me dwell on this point for a moment, because it seems to
me to be an important one. In the approach I am talking about—
the structural approach—the activities which call for the
exercise of different skills are designed to help the learners to
consolidate their knowledge of language. They are devices which
are designed to service language learning. The doing is sub¬
servient to knowing. But of course in the usual business of
language using, you do not go around speaking sentences in
order to practise them, or reading passages of prose in order to
get more exposure to particular structural patterns. In normal
language use, in normal activity with language, we always have
some purpose which language is there to serve. The knowing is
subservient to doing.
It has been pointed out often enough in recent years that the
disadvantage of this structural approach is that it does not allow
the learners to use language in a natural \yay. They tend to fixate
on form for its own sake, internalize the language system as a
separate body of knowledge and fail to learn for themselves how
to use it. Having been trained to direct what they do in the
service of knowing, they have difficulty in reversing the
dependency and so to direct what they know in the service of
doing. This is not surprising. For the structural approach
requires the learners to conceive of the foreign language as
something very different from their mother tongue, something
designed as a subject with its own rules for learning which seem
to have very little in common with the learner’s own experience
of language. For their own experience will have been in using
language in the mediation of meaning for some purposeful
outcome and not in the contemplation of the formal and
semantic properties of the medium itself.
The principles of structural language teaching, then, seem to
be at variance with the natural use of language which it is the
 Methodology for the teaching of meaning 159

purpose of such teaching to promote as the eventual aim of

learning. The structural means of teaching would appear to be
inconsistent with the communicative ends of learning. This
paradox can be resolved, in practice, by the dominance of
teaching so that in effect learners do not learn how to do
communicative things with what they have acquired as know¬
ledge, but simply learn how to display their knowledge accord¬
ing to conventions established by teaching and to meet the
requirements of the examinations based on the same con¬
ventions. These are traditionally designed to test a knowledge of
the subject itself as defined by the language teachers, and not what
learners have learned to do by means of the subject. Thus such
examinations are based on the norm of what has been taught,
not the criterion of what needs to be learnt.
One way of resolving the paradox, then, is to retain the
structural approach principle and make practice conform to it.
Another way is to adopt a different principle, indeed a directly
opposite principle, and make the means of teaching conform to
the communicative ends of learning. We adopt, in other words, a
communicative approach to language teaching and concentrate
on doing.
The communicative approach reverses the emphasis of the
structural. It concentrates on getting learners to do things with
language, to express concepts and to carry out communicative
acts of various kinds. The content of a language course is now
defined not in terms of forms, words and sentence patterns, but
in terms of the concepts, or notions, which such forms are used
to express, and the communicative functions which they are used
to perform. Hence the notional/functional syllabus. But, as I
argued in Chapter 9, this definition of course content is not
enough to ensure that there is an emphasis on doing in the
language classroom. There also needs to be a methodology
which will implement this course content in such a way that
learners will be activated to realize the notional and functional
character of the course specification. It is perfectly possible to
treat notions and functions as items to be learned in the same
way as structures, as pieces of knowledge to be put in store
without any necessary implication for actual use as natural
behaviour. If a methodology, a set of classroom techniques,
is focused on teaching knowledge in the abstract, and directs
all learner activity to that end, then it will fail to realize
the communicative possibilities within a notional/functional
160 Aspects of teaching

syllabus. It will indeed convert such a syllabus into what to all

intents and purposes is a structural one. The notional/functional
syllabus only becomes ‘communicative’ when it is implemented
by appropriate methodology.
A crucial element of the communicative approach to language
teaching is the adoption of a methodology which will encourage
learners to do things with the language they are learning, the
kind of things they will recognize as purposeful and commun¬
icative and have some resemblance to what they use their own
language to do. The assumption is that if the learners use the
language in this way, then they will learn it contingently, as a
natural consequence, that knowing will emerge from doing. In
other words, what teaching is concerned with is setting up
conditions for effective performance with the language on the
assumption that in learning how to perform pragmatically
learners will somehow be able to acquire knowledge of the
language itself inferentially by themselves.
So we have two approaches characterized by different
emphases. The principles, put simply, amount to this. The
structural approach is based on the belief that language learning
comes about by teaching learners to know the forms of the
language as a medium and the meaning they incorporate; that
they will learn how to do things with this knowledge on their
own. The communicative approach is based on the contrary
belief that language learning comes about when the teacher gets
learners to use the language pragmatically to mediate meanings
for a purpose, to do things which resemble in some measure
what they do with their own language. They will learn a
knowledge of the language itself, the formal and semantic
properties of the medium, as they go along, without the teacher
having to draw explicit attention to it.
It is of course the communicative approach which is in current
fashion. It is not difficult to see why its principles should be so
appealing. They bring the means of learning into alignment with
its eventual ends—the achievement of an ability to use language
to communicative effect. Furthermore, at the same time, they
represent the language to be learned as the same sort of natural
phenomenon as the language the learners already know, and so
allow them to draw on their own experience in the process of
learning. This in turn means that the focus of attention shifts to
the learner, \vho becomes the dominant partner in the pedagogic
enterprise so that instead of having the assertive teacher
 Methodology for the teaching of meaning 161

dictating to the submissive learner, we have the teacher

submissive to the requirements that assert themselves as neces¬
sary for successful learning. I shall take up this question of the
roles of teacher and learner in more detail in Chapter 11.
So the communicative approach seems not only to be more
natural and less contrived, but also seems to have the added
advantage of providing for the human rights of learners. So on
both pedagogic and democratic grounds it might seem self-
evidently preferable to the structural approach. A happy
coincidence of good causes, one might think: the learners
achieve their language learning objectives as a function of
increased independence, under the benevolent guidance or
counsel of the teacher, who is no longer the autocratic controller
of their destiny.

The problems of a communicative approach

Such an approach has its obvious attractions. Unfortunately,
it also has its problems, as I indicated earlier in this book (see
particularly Chapters 3 and 8). Two might be mentioned of
specific relevance to the present discussion. One of them has to
do with natural language learning and the other with natural
language use.
The natural learning problem is this. It turns out that learners
do not very readily infer knowledge of the language system from
their communicative activities. The grammar, which they must
obviously acquire somehow as a necessary resource for use,
proves elusive. So quite often the situation arises where learners
acquire a fairly patchy and imperfect repertoire of performance
which is not supported by an underlying competence. Their
doing does not seem to lead naturally to knowing, as has been
optimistically assumed.
The reason for this can be inferred from the characterization of
language use presented in the chapters in Part 2. As I observed
there, in using language we quite naturally follow the principle of
communicative economy whereby we pay attention to linguistic
features only to the extent needed to make connection with context
and so achieve the indexical meaning suited to our purpose. This
process of mediation does not depend on linguistic analysis. On the
contrary, it will often involve the matching up and adjustment to
context of pre-assembled patterns of language of the sort
discussed in Chapter 6. Some adjustment might be needed, but
162 Aspects of teaching

not analysis. Indeed, analysis would in many instances be out of

place and disruptive. So the process of communication which
calls for external synthesis with context would run counter to
the process needed to develop competence, which calls for the
internal analysis of the language itself. Thus the conditions
favourable for establishing external relations necessary for
effective use are different from the conditions favourable for
establishing internal relations necessary for effective learning.
But it might be objected that these conditions are not in fact
different in the natural development of first language acquisition.
This is true, but irrelevant. The natural language acquisition
process is a long and rather inefficient business. For the child
acquiring its mother tongue, knowing a language and internal¬
izing its formal rules is something that happens gradually
through trial and error; its competence is the result of a recurrent
refocusing as required by increasing performance demands over
a period of extensive exposure to and experience of language.
Courses for the teaching of other languages in schools just
cannot provide such conditions for natural learning.
Nor does it make any sense to try. For the whole point of
pedagogy is that it is a way of short-circuiting the slow process
of natural discovery and can make arrangements for learning to
happen more easily and more efficiently than it does in ‘natural
surroundings’. This is what schools are for, whatever subject we
are dealing with. Pedagogy is bound to be a contrivance: that is
precisely its purpose. If what went on in classrooms exactly
replicated the conditions of the world outside, there would be no
point in pedagogy at all. And, in respect of the present argument,
the advantage of pedagogy is denied if it just leaves learners to
learn by doing without quite deliberately contriving ways of
assisting them in getting to know the language system at the
same time, as the essential resource for their doings.
Put simply, the problem connected with natural language use
is this. As I pointed out in Chapter 7, when using language
indexically either to convey our intended meaning, or to
interpret the meanings of other people, we adopt strategies of
differential access and analysis in the interests of effective
communication. We do not derive meaning from a close
consideration or analysis of the language itself, we bypass it if we
can, treating it rather cavalierly as a set of clues which we can
follow in order to discover meanings elsewhere. When language
is used it becomes a set of pointers which indicate what area'of
 Methodology for the teaching of meaning 163

knowledge of the world must be brought to mind for inter¬

pretation to be achieved. In other words, we do not find meaning
in the medium, we achieve it by mediation.
For example, when we read the newspaper in the ordinary
course of everyday life, we do not submit the language to close
scrutiny, carefully deciphering every word and sentence to
extract its meaning. The language provides us with access to
what we already know, it indicates to us what already existing
knowledge we need to engage as a frame of reference into which
any new information might be fitted. It is only when we cannot
engage a frame of reference that we focus attention on the
language itself. So when we come across a word we do not
know, we do not immediately reach for the dictionary, we pass
over it, assuming that we will be able to fit it into the framework
subsequently by inference. Only when we fail to make the fit,
when a word does not activate our knowledge, do we begin to
enquire into its specific meaning.
To recapitulate the arguments of Chapters 5—7, natural
language use in the expressing and interpretation of meanings is
only in part a matter of language. The process involves taking
bearings, as it were, on two kinds of knowledge. One of these is
indeed knowledge of the formal properties of language, its
semantics and syntax, the meanings of words and their combina¬
tion in sentences, a knowledge of the properties of the medium,
systemic knowledge. The other kind of knowledge is that which
we have of the particular world we live in, our beliefs, ideas,
experiences, cultural values, and so on, schematic knowledge.
The meanings we achieve through the use of language, whether
we are sending or receiving them, are a function of a dual
reference to these two kinds of knowledge.
But in natural language use, as we have seen, systemic
knowledge is subservient to the schematic. It is a means to the
achievement of meaning and not an end in itself. And when we
are using language to deal with what is familiar to us as we
normally do, we pay as little attention to language as possible.
But when we do encounter unfamiliar content, or when we want
to express ideas which are novel and which do not fit
conventional schematic patterns, we, as native speakers, have
systemic knowledge to fall back on. We can use this as a
communicative resource.
Here then is the essential problem about natural language use
for language learning. We do not want our learners to bypass
164 Aspects of teaching

language when they use it, as it is natural for native speakers to

do, because they do not have the systemic knowledge as a back¬
up resource to rely on. This is precisely what we want them to
acquire and it is the purpose of pedagogy to assist them in
acquiring it.
Let me summarize the points I Kave made against a too-ready
acceptance of the primacy of doing. To try to replicate the
conditions of natural communicative use of language in class¬
rooms is mistaken for two basic reasons. First, to do so is to deny
the whole purpose of pedagogy, which is to contrive economical
and more effective means for language learning than is provided
by natural exposure and experience. Second, natural language
use typically deflects attention from language itself and pre¬
supposes a knowledge of the language system as a basic resource
which learners have, by definition, not yet acquired.

Reconciling the oppositions: an illustration

It would appear that a strict adherence to neither structural
nor communicative principles of approach, as I have outlined
them here, will provide satisfactorily for the interplay between
knowing and doing upon which (I have argued) effective
language use and language learning depend. We need to
reconcile these contraries by methodological procedures which
draw on both approaches and realize the necessary interdepend¬
ence of knowledge and behaviour.
This is not, of course, a new discovery. Many teachers
recognize this need for reconciliation well enough. The profes¬
sion of language teaching is not neatly divided into groups of
teachers who adhere strictly to one approach or the other. All
distinctions of the kind I am making here are based on an
idealization, an abstraction of salient features, and are con¬
founded in reality. The two approaches I have outlined are
modified and combined in different ways under the influence of
varying circumstances. In other words, they are expediently
applied. So there are already methodological compromises to be
found in existing teaching materials and in the practices of
individual teachers which can be said to bring about the
reconciliation that I refer to. What I want to do now is to
illustrate this process of reconciliation by reference to a
particular exercise in materials design. My purpose is to
demonstrate how the specific techniques proposed derive from a
 Methodology for the teaching of meaning 165

consideration of general principles. The demonstration, there¬

fore, is an example of the pragmatics of language pedagogy
discussed in the first part of this book: having subjected ideas
to critical appraisal, we now move on to make them operational.
The illustration is a set of communicative grammar materials
originally designed to complement a course which had been
written along notional/functional lines for the Arab World. ^
This course was based essentially on the assumption which has
already been referred to—that knowing will be derived from
doing by natural inference without direct intervention in the
form of explicit demonstration or conscious practice.
Grammar, then, was to be assimilated incidentally as a
function of communicative activity. One difficulty with this
approach to methodology relates to points I mentioned at the
beginning of Chapter 9: pedagogically enlightened though it
may seem to be, such an approach is not closely congruent with
the educational context. Both teachers and learners were
accustomed to an approach in which explicit grammatical
directions were provided. The new course required learners to
shift from a referential to an inferential mode of learning (from
referring to knowledge in doing things to inferring knowledge
from doing things) and this led to some degree of disorientation.
And it turned out that the learners could not always discover
their own grammatical bearings by generalizing from particular
instances of behaviour. Grammatical knowledge did not always
follow as a necessary corollary of communication. The learners,
in short, and the teachers too, needed some sort of chart which
marked out the grammatical features of the learning terrain to
help them to find their way. This is what these communicative
grammar materials set out to provide.

^ The course which these materials were designed to complement is the

Crescent English course. The materials themselves were published under the title
of Communicative Grammar, the first four books of which were written by Ann
Brumfit and Scott Windeatt and the last two books by Christopher Hyde. Both
the main Crescent course and Communicative Grammar appear under the
imprint of ELTA/OUP (English Language Teaching for the Arab World/Oxford
University Press) Beirut.
It should be noted that the Crescent course has been subject to recurrent
revision over the years in the light of teachers’ experience in using it and in
consequence there has been a more focused attention to the formal properties of
language in later editions. The development of this course from its early
conception to its most recent revision is an interesting case history of the
relationship between pedagogic ideas and the educational context which sets
conditions for their implementation (see Chapter 8).
166 Aspects of teaching

What they did was to reformulate the grammar which

appeared informally and contingently in the main course. This
involved the selection of those elements of grammar which
seemed to call for focal treatment (leaving the rest to be inferred
from peripheral use) and the design of units of work which
would draw attention to the aspSects of conceptual and com¬
municative meaning which these elements encoded as semantic-
ally intrinsic in their forms (that is to say, their denotation), and
also to the relationship these elements contracted with others as
terms within the linguistic systems of English (that is to say, their
sense relations).
But simply to take language originally presented as a kind of
communicative scatter and reformulate it explicitly as grammar,
even though tagged with notional/functional values, would
clearly have run the risk of presenting it as an area of knowledge
in detachment from its communicative realization. It would not
have provided the necessary transitional dependency between
knowing and doing. Grammar had to be internalized not just as
a formal system but as a resource for use. This meant that
activities needed to be devised which would require learners to
access grammatical knowledge for some purpose other than
practice and so to realize for themselves the latent capability
within grammatical knowledge for the achievement of meaning.
In other words, the activities would need to establish the
relevance of systemic knowledge and the credentials of the
medium as a crucial factor in the realisation of meaning by
mediation.
The procedure was first to demonstrate how grammatical
forms typically correlate with areas of meaning and then to
involve learners in problem-solving tasks whereby they could
put these forms to purposeful use. An example of the demon¬
stration stage appears in Figure 3.
Two features of this demonstration call for comment. First,
the little diagrammatic device at the beginning is intended to
represent the conceptual meaning of the particular grammatical
form. This ‘grammargraph’ (as it was called) is a mnemonic
device which indicates by non-verbal symbolism both the
standard or core denotation of the form and, by contrast with
other grammargraphs, the sense relations it contracts with other
forms as terms within the grammatical systems concerned. Thus,
this particular^ grammargraph clearly contrasts with others, shown
 Methodology for the teaching of meaning 167

Figure 3

PAST TIME
Past Now Future
—►

I was turning left at the time of the accident.

7 Whose fault was it?

7.1 Which car is Fuad's, and which is Ahmed's?
Read the sentences and look at the map to Find out.
Then write the names in the correct boxes.

This morning Fuad was driving along South

Street while Ahmed was driving along West
Street. At 10.01 they had an accident at the
crossroads.

7.2 Ahmed and Fuad told the police about the accident. They
both told the truth. What were they doing at the time of the
accident? Were they turning left or right? Was one of them
driving carelessly? Read the passages to find out, and then
tick the correct boxes for the sentences below.

Ahmed Fuad
'At 10.00 I was driving along West Street. I was 'At 10.00 I was driving along South Street with
going to the bank. I was watching the traffic my friend, Ali. We were going to the market
lights and while I was driving past the market, and we were not going very fast. We were
they changed to green. The other car was talking when the accident happened.'
turning left out of South Street when it hit my
car. I was not going very fast.'
True Untrue

1 Fuad was turning left.

2 Ahmed was turning right.

3 Fuad and Ali were talking.

4 Ahmed and Fuad were not going very fast.

(continued)
 168 Aspects of teaching

PAST TIME

7.3 So whose fault was it? Who was driving carelessly when the
accident happened? Complete the passage. ^

Ahmed and Fuad.very fast, but.was driving carelessly. The

accident was.'s fault. He.. left into.Street when

the traffic lights in.Street were red.

7.4 Complete the tables.

he
she
UJD/& .the traffic lights.
it
At the time
1
of the
accident,
we
you .to the market.
they

he
she
.carelessly?
it
At the time
1
of the
accident, we
you .left or right?
they
 Methodology for the teaching of meaning 169

in Figure 4, which represent the simple past, the present

continuous, and the present perfect. How far the kind of symbolic
representation exemplified by these grammargraphs is effective in
teaching conceptual meaning is an open question. Some areas of
meaning clearly lend themselves more naturally to such representa¬
tion than do others.

PAST SIMPLE
Past Now Future
1

PRESENT CONTINUOUS
Past Now Future

PRESENT PERFECT
Past Now Future
1

Figure 4

Spatial prepositions, for example, are straightforward enough.

Figure 5 (overleaf) shows a set of grammargraphs, which represent
motion and position in the second book of Communicative
Grammar (see also Quirk etal. 1972: 307).
Things get more difficult, however, the more remote meanings
become from immediate perceptual experience. Here one relies
on the learner acquiring the conventions which are established
by the devices themselves. But there can be problems even with
concepts which seem simple because they are more or less
directly derivable from the perceived world. Take the notion of
170 Aspects of teaching

MOTION AND POSITION

□ •

relative proximity {near/far, herelthere, this/that, etc.). An

obvious vv^ay of symbolizing this would be to show two figures,
one in the foreground, the other in the background. But a correct
interpretation depends on a familiarity with the convention of
perspective drawing. Otherwise the figures could be read as
representing relative size. It has to be remembered, then, that
 Methodology for the teaching of meaning 171

pictorial and diagrammatical devices are based on cultural

convention just as linguistic devices are.
Nevertheless, for all this, it seems worth pursuing the
possibilities of grammargraphs of this kind. For one thing, they
can be particularly useful in demonstrating the abstract sense
relations that hold between grammatical elements, the way in
which they contrast with each other as terms within a system.
They can also be used as frames of reference for establishing the
differences between the mother tongue and the language being
learned in respect to the manner in which reality is encoded.
The second feature of the demonstration to be noted is that the
learners are not cast in the role of passive recipients of information.
They are drawn into participation through the solving of a simple
problem which requires the repetitive use of the form in question.
This leads to the completion of a substitution table of the
conventional kind which formalizes the knowledge they have
derived from the preceding activity and serves as a source of
reference for subsequent use. There is provision for doing, then, in
this demonstration stage, but it is doing at the service of knowing, a
variant of conventional form-focused activity.
The demonstration stage seeks to establish the standard or
canonical notional/functional valency of the linguistic forms,
their prototypical value as it were. In the tasks which follow,
learners are required to act on this information through activities
which provide an opportunity for repeated use of these forms,
but as motivated by a problem-solving purpose. Here the
dependency of language and task is reversed: knowing is put at
the service of doing. The outcome is a solution and not a
substitution table.
An example of a problem-solving activity which follows on from
the given demonstration appears as Figure 6 (pages 174-5).
The relationship between the demonstration stage and the
problem solving that follows it is in some respects similar to the
relationship between pre-task and task in Prabhu’s procedural
approach to teaching. In both cases, the first activity primes the
learner for the second. The difference is that the priming in
Prabhu’s case is not explicitly linguistic:

The pre-task is a context in which any difficulties which

learners may have in understanding the nature of the activity
_seeing what information is given, what needs to be done,
and what constraints apply—are revealed and the teacher is
172 Aspects of teaching

able to provide appropriate assistance, perhaps by para¬

phrasing or glossing expressions, by employing parallel
situations or diagrams, or by re-organizing information. In
this sense, the pre-task is preparation for the task, since
learners are less likely, vv^hile engaged later in a similar activity
on their own, to fail to see whdt is given and what needs to be
done. (1987:54)

I do not myself see why explicit reference to language should

be so stringently avoided in preparation for the task. Linguistic
information and linguistic constraints are just as much a part of
the nature of the activity as anything else. And in practice, the
learners are indeed given linguistic instruction in the pre-task by
means of ‘paraphrasing and glossing expressions’.
Apart from providing learners with linguistic priming, the
demonstration stage also, like Prabhu’s pre-task, prepares the
learners to engage in a particular kind of problem-solving
activity—so it seeks to establish a close relationship between
language and reasoning. In the demonstration stage it is the
latter which is the dependent partner and in the stage which
follows, the dependency is reversed. In Prabhu’s scheme, the
reasoning is always given primary emphasis.
Of course, the crucial condition on this reversal of dependency
is that learners will engage with the tasks as purposeful problem¬
solving activities and not as linguistic exercises. The tasks have
to be such as to seem worthwhile inr their own right and
independently motivating. And they of course also have to be
within the conceptual scope of the learners. Here we encounter a
general difficulty with reasoning-gap activities, one that was
touched upon in Chapter 9, namely that we have no reliable
measure of their relative complexity. Complexity, furthermore,
may be culturally variable. What is a familiar mode of reasoning
in one culture may not be familiar at all in another. It should also
be noted that too great a reliance on reasoning might preclude
the learner from engaging more affectively with the language,
might narrow the scope for imaginative and emotional involve¬
ment. Prabhu has countered this argument by making a
distinction between learning and teaching. He acknowledges
that his tasks are factual and rational in meaning-content and
that they require no procedures for increasing emotional
involvement,,but adds: ‘This does not imply any denial of value
to emotional involvement in learning. What it implies is a
 Methodology for the teaching of meaning 173

recognition of the much greater suitability of rational activity for

language teaching.' (1987: 52)
In other words, emotional involvement is seen as an incidental
and contingent effect of teacher-directed reasoning. In this
respect it seems to resemble the acquisition of grammatical
competence itself in Prabhu’s scheme of things. But letting
emotional involvement happen (by connivance almost) is very
different from encouraging it actively as a matter of positive
pedagogic policy as in humanistic approaches to language
teaching (for example, Asher 1982; Moskowitz 1978; Rogers
1969). In such approaches, indeed, such a distinction between
learning and teaching is hardly tenable.
The fact remains that reasoning-gap activities in Prabhu’s
approach, and in the communicative grammar materials present¬
ed in this chapter are, in the terms I used earlier (Chapter 2),
designed for cognitive regulation, and they do restrict the
engagement of the learners’ personality in a way which
proponents of humanistic learning, who believe in affective
regulation, would find unacceptable. It should be noted,
however, that these communicative grammar tasks are designed
to be complementary, so that they do allow for the possibility of
learner personality to be more extensively and affectively
engaged elsewhere in other parts of the programme. I will return
to this point later in the chapter.
It should be stressed that the problems which are posed in
these tasks are not language problems but problems which
require a use of language for their solution. The learners do not
just manipulate language as an end in itself, but realize its
potential as a means for achieving outcomes which have
independent point. The design of the problems, therefore, seems
to reconcile two features which are commonly associated with
two different approaches to the teaching of language: linguistic
repetition, with its necessary focus on form, and non-linguistic
purpose, with its necessary focus on meaning. Linguistic
repetition is a feature of a structural orientation to teaching with
its emphasis on knowing: learners are required to practise
particular structures so as to facilitate unconscious assimilation.
Non-linguistic purpose is a feature of a communicative orienta¬
tion with its emphasis on doing: here learners are engaged in
activities which deflect attention away from the linguistic forms
being used. But in these communicative grammar materials, the
tasks are so designed that their solution depends on the repeated
 174 Aspects of teaching

PAST TIME

8 What were they doing at 11.00?

At 11.00 ye.sterday morning, a thief took a box
of watches from Salim's watch shop. A '
policeman arrived, and stopped seven people
near the shop. He asked each of them: 'What
were you doing at 11.00?'
The policeman wrote down their answers in
shorthand, because he wanted to write quickly.

8.1 Look at the policeman's notebook. What did he write?

Use the key for the shorthand words and the map below to find out.

POLICEMAN’S NOTEBOOK
Majed:
Nedal
Jassem; I
Ibrahim
and Hamad
Ahmed: I
Omar
Nedal: I
Omar

Ibrahim We
and Hamad; Jassem
Omar; I
Majed

KEY FOR THE SHORTHAND

WORDS

Figure 6
 Methodology for the teaching of meaning 175

PAST TIME

8.2 What were the seven people doing at 11.00? Use the
policeman's notebook to complete everybody's answers.
The thief told two lies, and everybody else told the truth.
So if two people said the same thing, it was true.

Majed:

Jassem:

Ahmed: I. in the ...

Omar. ... in the

Nedal: I. in the...

Omar. in the

Ibrahim
& Hamad: We. in the

Jassem.in the

Omar: I.in the.

Majed.in the .

8J Who told the lies? Read the sentences in 8.2 again to find out.
Then write everybody's names in tbe boxes round tbe map,
and draw lines to show where everybody was.
Wbo was in tbe watch-shop?

.told two lies, so he was the thief.

176 Aspects of teaching

use of the language items concerned: repetition, therefore, is a

function of purpose. The learner practises language in the
process of solving the non-linguistic problem. There is no
conflict between form and meaning: the one complements the
other.
This accommodation of thd two cardinal principles of
repetition and purpose, which in combination provide for the
internalization of grammar as a resource for use, places
constraints of course on the design of the tasks themselves. In
particular, they operate at a remove from verisimilitude of the
simulation of the real world. An attempt to use scenes and events
from everyday life (favoured as typically communicative in other
teaching materials) would have resulted in distortion after the
manner of a structural approach in order to provide for the
required repetition of linguistic forms, and this would inevitably
have involved a corresponding reduction in communicative
point. Neither the problem to be solved nor the language needed
to solve it could have been so sharply focused. The procedure,
therefore, was to devise problems whose very purpose lay in
contrivance, which sought to create their own contexts of
significance and which in this respect bear a closer resemblance
to what learners are required to do in other subjects on the
curriculum than to naturalistic uses of language in the social
world outside the classroom.
There are two further points of principle implicit in this design
feature of the tasks. First, the approach to language teaching is
brought into closer alignment with activities associated with
other areas of school work and allows for the possibility of
integration within the curriculum as a whole. Secondly, there is
the implication that what the learners are acquiring through
these problem-solving tasks is not primarily a repertoire of
behaviour ready for direct deployment but a capacity for actual
use based on the internalization of grammar as a communicative
resource. This point relates to the distinction which I made in
Chapter 9 between accumulation and investment. What these
materials aim to bring about is not a rehearsal of ‘authentic’
communication in the sense of naturalistic social behaviour but
an investment in grammar which has been authenticated as
meaning potential by having been realized through these tasks in
the achievement of purposeful outcomes.
 Methodology for the teaching of meaning 177

Further developments: text sequences

I should make it clear that I am not suggesting that
methodology should consist entirely of an unrelieved diet of this
sort of communicative grammar. These tasks should be seen as
complementary to activities which are less specifically focused
on the medium, which allow for less controlled mediation of
meaning and for the contingent learning of language. They could
be complementary in two ways. They could provide a retro¬
spective reformulation of elements of grammar which had been
informally introduced in communicative guise in previous
teaching. This was the purpose for which these particular
materials were originally designed. But tasks of this kind can
also be complementary in a prospective way: they can prepare
the ground for problem solving as the negotiating process which
(as I indicated in Chapter 7) defines language use in general.
Thus one can envisage, for example, a development from these
tasks in the form of work on texts which gradually shifted
dependence from systemic to schematic knowledge and which
required learners to negotiate interpretation by the discriminat¬
ing reliance on both. In this way they would gradually
approximate to ‘natural’ language behaviour by taking short
cuts and adjusting their level of attention to language according
to purpose. But they would previously have been provided with
the basic systemic knowledge which is necessary as a back-up
resource to be referred to when required.
Such a sequence of tasks would make increasing demands on
procedures for negotiating meaning. Whereas the problem
solving involved in the tasks of these communicative grammar
materials has primary focus on the medium, and so relates more
specifically to the matters discussed in Chapter 6, later texts
would emphasize the procedures of mediation which are
discussed in Chapter 7. Such a development would make
operational the necessary interdependency of these two per¬
spectives on methodology for the teaching of meaning.
One source of texts which are of particular interest here is
literature. As I have argued elsewhere (for example, Widdowson
1984a: Section 4), in literature we find not reference to reality
conventionally conceived but representation of alternative con¬
structs of reality, not actual but possible worlds, existing in a
different dimension. These cannot of their nature simply be
recognized by calling up relevant schematic knowledge: they
178 Aspects of teaching

have to be realized through the language which creates them.

Even when literature makes mention of objects, places, and
events which are familiar, they have to be located in the created
context of an imagined world which cannot be familiar, so their
significance is never just a matter of recognition. The problem
posed by a literary text is how to Use systemic knowledge to gain
access to a self-enclosed world that has no existence outside the
text itself, and is not subject to normal standards of truth or
rationality. The interpretation of literature, then, involves acts of
schematic realignment and these can only be achieved by a
particularly intensive exploitation of the language medium. And
of course the literary writer encourages the reader in such
exploitation. Consider, for example, the following opening of a
short story by Somerset Maugham:

The Unconquered
He came back into the kitchen. The man was still on the
floor, lying where he hit him, and his face was bloody. He was
moaning. The woman had backed against the wall and was
staring with terrified eyes at Willi, his friend, and when he
came in she gave a gasp and broke into loud sobbing.

The use of the personal pronoun would normally signal

shared knowledge. The pronoun would refer to some identifiable
person. But here of course there is no person that the reader can
recognize as referent for the pronoun that appears at the
beginning of the passage. The reader is therefore drawn into the
text and projected forward in quest of information as to who
this person might be. And then another unidentified person
appears: ‘the woman’. Which woman? Who is she? Why is she
so terrified? Why has the second man been hit by the first?
Where has the first man come from? What has been happening
here? The only way to find out is to pay close attention to the
text: interpretative procedures are put, as it were, on special
alert. And we read on, creating the fictional world by exploiting
the language as we go along (see Widdowson 1985).
It is often suggested that learners should be primed to read
texts by establishing a purpose for reading beforehand, by
preparing them by means of a pre-task which is in effect a
pretext. Obviously this can often be an effective strategy. But
some texts, notably (though not exclusively) literary ones, fire
designed to attract attention and so provide their own priming.
 Methodology for the teaching of meaning 179

Here contextual discovery can be encouraged in preference to

pretextual preparation.
The point about literary uses of language is that communicat¬
ive significance is realized as a function of a close attention to the
language itself. As a further illustration, consider the following
simple little poem by Emily Dickinson:

A word is dead
When it is said,
Some say,

I say it just
Begins to live
That day.

As before, we have no context that we can refer these

expressions to. The words appear on the page as having some
sort of self-enclosed significance—a significance which does not
carry over in a prose paraphrase:

Some people say that once a word is said it is dead, but I say
that it is only then that it begins to live.

This simply sounds banal and inconsequential. So there must

be something in the particular choice of language arranged in
this particular pattern which is meaningful. Without contextual
connections, the only clues to significance must be contained
within the language of the poem, inherent in those features of
form which do not carry over into a conventional prose
paraphrase which presupposes context. So we are drawn into the
poem. What, then, might we infer from it?
Consider, for example, the way in which the propositions
have been verbally fashioned. The predicate in the first line is
stative: ‘A word is dead'. That in the second line might be
interpreted as a passive, but its parallel syntactic position and its
phonological and metrical equivalence gives it stative force in
association with the preceding line. A word is dead = a word is
said. The two lines, we may say, thus indicate stasis, absence of
movement. They do not only express the proposition ‘a word is
dead when it is said’, they actually represent, in the very syntax,
the identical state of ‘deadness’ and ‘saidness’. In the second part
of the poem, however, the verb in the predicate is dynamic, being
active in voice and inceptive in aspect, so that the expression
‘begins to live’ can be said both to refer to birth and at the same
time to represent it. This effect is furthermore enhanced by the
180 Aspects of teaching

simple and uninterrupted continuity of the syntactic pattern

which carries the reader over line boundaries as if impelled by
the elemental life-force itself.
The sense of the finality, the abrupt closure of death, in the
first verse of the poem and the contrasting active and initiatory
movement in the second verse are represented by syntactic and
prosodic structuring and the meanings intrinsic to the grammar
of the verb phrase. The death and life of words are in contrast,
not only as they are expressed explicitly in the propositions, by
virtue of lexical meaning, but also as they are implicitly
represented in the form. What is said about words becomes
indistinguishable from what is done with them. Their death is
denied in their very use: they come alive as they are said as is
evident from the interpretations they provoke.
Access to both the fictional world of character and event in the
short story opening, and to the expression of experience in the
poem depend, in different ways, on paying close attention to
language. In both cases we find a kind of problem solving
directed towards the realization of communicative significance.
The meaning achieved is very different from that which comes
from the cognitive tasks illustrated by the communicative
grammar materials demonstrated in this chapter. For one thing,
literary texts encourage a more affective engagement with
language and so increase scope for personality involvement. But
different though the meaning is, the point is that it too depends
on a focus on language itself. Now, clearly, careful thought
needs to be given to how literary texts should be selected and
presented so that they are pedagogically effective. But it is not
difficult to see how, in principle, such texts can be used, in
association with others of a referential rather than representa¬
tional kind, where the communicative purpose and reference to
schematic knowledge call for less conscious concern with the
systemic meanings intrinsic in language. Further discussion of
the ways in which literature and language study can complement
each other is to be found in Brumfit and Carter (1986) and
Widdowson (1975,1984a: Section 4,1985).
So we can envisage a programme consisting of texts selected
and presented in such a way as to require a differential and
discriminating attention to systemic and schematic knowledge.
The sequence would in effect gradually shift dependency on
these two kinds of knowledge in the procedures for negotiating
meaning.
11 The roles of teacher and learner

This book began with a consideration of the relationship

between the roles of teacher and researcher. It ends with a
consideration of the relationship between teacher and learner.
Some aspects of this relationship have already been touched on
in the preceding chapters. In particular I pointed out how a
mediation-oriented methodology allows for the greater exercise
of learner initiative, with the consequent modifications of
directive control on the part of the teacher. But although one
might commend this redefinition of relationship, whereby
learners are able to participate more fully and freely in the
development of their own learning, there are certain constrain¬
ing factors in force. One has to do with the fact that all learning
must be delimited in some way as a consideration of its
happening at all (I discussed this matter in Chapter 1). Other
factors have to do with the nature of teacher and learner roles in
general, with what might be called the sociology of the
classroom. These are the factors I shall be concerned with in this
chapter.

Types of role
The term ‘role’ is a familiar one and frequent use has worn
away its foreign circumflex. It can be defined, generally, as a part
people play in the performance of social life. Roles, whether
achieved or ascribed, are kinds of conventional script, or pre¬
script, which constrain the individual person to assume a
persona in conformity to normal and expected patterns of
behaviour. A role, to put it in more precise sociological terms, is
‘a set of norms and expectations applied to the incumbents of a
particular position’ (Banton 1965: 29). The incumbents we are
concerned with here are those who take up positions in the
language class—teachers and pupils, or learners. The questions I
now want to discuss are, what are the norms and expectations
182 Aspects of teaching

associated with these particular roles? And what particular

positions do the incumbents occupy?
The classroom provides the context for the enactment of these
roles: but the classroom should not just be perceived as physical
surroundings but also conceived as social space. The difference is
important and can be marked by a terminological distinction:
setting for the physical context, scene for the socio-psychological
one. These are terms and distinctions proposed by Hymes:
Setting refers to the time and place of a speech act and, in
general, to the physical circumstances. Scene, which is distinct
from setting, designates the ‘psychological setting’, or the
cultural definition of an occasion as a certain type of scene . . .
In daily life the same persons in the same setting may redefine
their interaction as a changed type of scene, say, from formal
to informal, serious to festive, or the like. (1972: 60)
Hymes identifies setting and scene as factors in the speech
event, the situated communicative occurrence of language use.
Since the classroom lesson is a type of speech event, it should be
possible to characterize it in terms of these and the other factors
which Hymes proposes.
With reference to setting, we might consider the physical
features of the classroom which facilitate or constrain certain
interactive procedures. If the walls are thin, this fact sets limits
on activities which might provoke too much noise; if the desks
are ranged in lines facing the teacher and^fixed to the floor, this
fact inhibits the setting up of small group discussion. So setting
factors will have an influence on the kind of scene which the
teacher wishes to create. One might compare the effect of seating
arrangements in such places as courts of law or parliamentary
assemblies.
But it is the socio-psychologically defined context, the scene,
which is of particular relevance in recognizing how roles are
assumed by classroom incumbents. The protagonists have to be
in position both socially and physically before the scene is set
and the play can begin. The teacher comes into the room. There
is a lull in the hubbub, a transitional phase of settling down.
Then: ‘Right. Quiet please. Sit down.’ The tumult and the
shouting die. The scene is set. The classroom is constituted as a
kind of social scene and the lesson starts. The participants, or
the players, assume their normal and expected positions.
But what are these positions exactly? We have names for them
 The roles of teacher and learner 183

in English: ‘(school)master’, ‘(school)mistress’, ‘teacher’, on the

one hand; ‘pupil’, ‘student’, ‘learner’, on the other. But the terms
on each side are not in free variation: they seem to suggest
different things, conjure up different images. And this gives some
indication, I think, that the roles of the classroom protagonists
are not as straightforward as they might at first appear.
Let us look first at the terms ‘pupil’ and ‘student’. These
denote stable and socially established roles at different stages of
life (studenthood in Britain seems to coincide roughly with the
right to vote) and they are institutionalized and a part of the
incumbent’s identity. So the terms can be used to mark social
categories and to specify occupation. Someone might say, in a
conversation, ‘Rupert, you know, is a student at Oxford’, or we
might read as a newspaper item ‘Fiona Chetwynd, a pupil at The
Priory School, is the new junior East Basingstoke badminton
champion.’ But people are not categorized or identified as
learners. Being a learner is not an occupation but an incidental
activity. The term denotes a role of temporary engagement
which does not attach to the incumbent as a continuing
characteristic: it is not a means of identification. We cannot say
that Rupert is a learner at Oxford or Fiona a learner at The
Priory School. And, by the same token, we have a National
Union of Students but not a National Union of Learners.
Learners are not kinds of people. Students and pupils are. You
can be a learner whenever you like; you do not have to formalize
your position.
In Britain, though, we do have a National Union of Teachers.
This, together with the fact that this is the same sort of
organization as the National Association of Schoolmasters and
Women Teachers, indicates that the term ‘teacher’ is ambiguous.
It can be used to refer to an identifying and categorizing role, in
which case it is synonymous with ‘(school)master’ and ‘(school)
mistress’, and corresponds to the terms ‘pupil’ and ‘student’. But
it can also be used in reference to a temporary and incidental
role, which is engaged as an activity as and when occasion
requires, and in this case it corresponds to the term ‘learner’.
What I am suggesting, then, is that we find two really quite
different kinds of role enacted in the classroom. One has to do
with occupation and is identifying (pupil, student; master,
mistress) and the other kind has to do with activity and is
incidental (learner), with the term ‘teacher’ ambiguous, able to
denote both. Now it happens that, as with other things, they
184 Aspects of teaching

order this matter better in France. TTie terms professeur, ecolier and
etudiant(e) denote the identifying occupational roles, the
terms enseignant(e) and apprenant(e) the incidental activity
ones.

Two kinds of classroom engagement

Well, this may or may not be of interest, but how is it
relevant? It would seem to me that what this distinction suggests
is that the classroom encounter is compounded of two kinds of
engagement. These sometimes converge and sometimes conflict,
and it seems to me that it is of central concern for classroom
methodology to get these two kinds of engagement to synchron¬
ize effectively.
One kind of engagement involves the identifying roles: the
teacher as professeur (master or mistress) in some sort of social
interaction with the pupil. This let us call an interactional
engagement. The norms and expectations defining appropriate
behaviour here will be very much a matter of social attitude and
educational ideology. This kind of engagement is a microcosmic
school version of the macrocosm of social life and reflects the way
educationists believe pupils should be socialized. It services the
hidden curriculum of acculturation and the promotion of accepted
values. It is the mode of interaction itself which is intended to have
the principal educational effect.
But the classroom also has what we might call a transactional
purpose—that is to say, it is meant to meet certain explicit
learning objectives, to instigate activities directed at achieving
specified goals. Here what I have called the incidental roles are
brought into play: teacher as teaching person [enseignant] on the
one hand, learner (apprenant) on the other. The norms and
expectations here relate to pedagogic purpose: ways of defining
roles which are likely to be the most effective for dealing with a
particular subject on the overt curriculum, for developing
specified knowledge and skills, for meeting the demands of the
examination.
The two kinds of engagement and their different role
realizations can of course key in with each other in a convenient
and complementary way, with a particular kind of interaction
facilitating a particular kind of transaction, educational and
pedagogic perspectives nicely converging into a single focus. But
 The roles of teacher and learner 185

it can also happen that they are at variance. Proposals for change
in one kind of engagement can conflict with the conventions
which characterize the other. A rationale for one kind of
engagement can be mistakenly supposed to apply with equal
validity to the other.
With this in mind, consider the following example of a type of
classroom interaction. The teacher is accorded high status and
commands deference. This might be symbolized by dress: formal
clothing (a jacket and a tie for a man) or perhaps even an
academic gown. There is protocol to be observed: pupils fall
silent when the teacher enters the room, they stand up and chant
a choral greeting. One of their number cleans the board. There
are ritualistic practices to be strictly adhered to. Pupils are
addressed by their surnames, the teacher by title (Mrs A, Mr B).
The interaction itself is tightly controlled. Only the teacher has
the right to initiate exchanges. Pupils can only contribute when
they make a bid by raising the hand and when this is
acknowledged and ratified as a claim for a speaking turn. Only
one pupil speaks at once. The teacher only asks questions to
which he or she already knows the answers. The rights and
obligations associated with the teacher and pupil roles are clear,
fixed and non-negotiable. They are established by ‘norms of
interaction’, another of Hymes’s factors, which he defines
as ‘specific behaviours and properties that attach to speaking
—that one must not interrupt, for example, or that one may
freely do so . . . that turns in speaking are to be allocated in a
certain way’. (1972: 63—4)
Now one may think that this kind of interactional engagement
is just a quaint and rather Dickensian remnant of the past, to be
dismissed out of hand and clearly to be condemned as bad
practice. But on what grounds?
Well, we might invoke a more enlightened ideology and say
that the kind of education promoted by this type of interaction is
one which forces the individual into conformity with existing
patterns of power, schools the pupils into maintaining an
iniquitious social structure which favours a self-appointed elite
and effectively acts as an instrument of disenfranchisement. The
pupils are put in this position in the classroom so that they can
be more effectively kept in their place in social life. We may
indeed, as many others have done, challenge the idea that there
needs to be any clear definition of role at all. We may wish to
think of the classroom engagement as being not a position-
186 Aspects of teaching

oriented but a person-oriented interaction (to use Bernstein’s

distinction, see Chapter 9) and so get rid of all this cumbersome
ritual. No sartorial signs of office, no gown, no suit or tie—just a
sweatshirt and sneakers. No position and no imposition. ‘Don’t
call me Sir. Don’t call me Mr Brown. Call me Dave.’ ‘Don’t call
me Miss. Don’t call me Miss Brown. Call me Liz.’ Equal
opportunities and human rights . . .
So we might object to the traditional interaction I have
presented on the ideological grounds that it runs counter to
enlightened educational thinking. It is fundamentally undemo¬
cratic. But we could also object to it on more expedient and
practical grounds by saying that, whatever the moral or political
or social objections might be, such an interaction is to be
proscribed because it is inconsistent with the kind of pedagogic
transaction needed to facilitate language learning. It just does not
create the right sort of enabling conditions. Now this is a very
different sort of objection and needs different arguments to
sustain it.
It might be pointed out, for example, that such a rigid
definition of role impedes the natural learning process since it
does not allow for learner initiatives: it does not give the learner
scope to draw on the available resources of intuition and
inventiveness, or to engage freely the procedures for learning
which he or she has acquired through a previous experience of
language. Nor does such a role definition allow for the provision
of group work (see Brumfit 1984a: Chapter 5; Wri^t 1987:
Section 2.3). A teaching role which determines learner activity
by directing it radially, so to speak, from the pedagogic podium
is, of course, based on the assumption that learners can only
learn from approved and ,appointed teachers and not from each
other. Thus the collective potential that learners bring to class as
a resource for learning is left unexploited. We might adduce
evidence from research on the ‘Good Language Learner’ (see, for
example, Naiman et al. 1978) and on natural second language
acquisition (see Ellis 1985) in support of the contention that
such a close adherence to fixed prescripted roles in the classroom
is detrimental to effective (and affective) learning.
Now these are points which bear on pedagogic and not
educational issues. They concern the optimal conditions for
learning. The interaction in this case is evaluated not in its own
social terms hut in reference to its degree of congruence with a
preferred type of pedagogic transaction.
 The roles of teacher and learner 187

It is not always easy to keep these two kinds of classroom

engagement distinct. One is tempted to suppose that if a
particular role-relationship between teacher and learners is
transactionally effective in one set of circumstances then it will
transfer (and should be transferred) to others. But the effective¬
ness may depend on a particular interactional role-relationship
between teacher and pupil which is simply not sanctioned as
educationally desirable in a different social situation. A human¬
istic, group therapy approach to pedagogy may be highly
effective (and affectively highly enjoyable) in places which
favour person-oriented education, but impossible to implement
in places where different educational ideology calls for a very
different kind of interactional engagement in class, one based on
clear positional definition established by tradition. Again, it is
tempting to believe that if a particular concept of interaction has
an especially appealing ideological ring to it, then it must needs
be transactionally effective; that setting pupils free of their
traditionally ascribed roles, for example, will as a corollary
make them better learners. But this, of course, does not follow at
all. If there is a causal relationship, then it needs to be
demonstrated. It cannot just be taken on trust, or on faith, or
confirmed by the fiat of ideological commitment.

Teacher authority and learner autonomy

This dual functioning of the classroom encounter that I have
briefly outlined has a bearing on two related issues which are
prominent in present debate in our profession, and which make
their appearance, in various guises, in many conference pro¬
grammes. One of these is teacher authority. The other is learner
autonomy.
Challenging authority in general is a popular activity at
present in those societies where the populace has the liberty to
do so. In education, the teacher as professeur has come under
some suspicion as a possible agent of authority which seeks to
maintain the power of privilege, schooling pupils into obedient
compliance. In pedagogy (or in our branch of it at least) the view
in vogue among those who claim expertise seems increasingly to
be that expressed nearly two thousand years ago by Cicero:
‘Most commonly the authority of them that teach hinders them
that would learn.’
But I think it is important to realize that the exercise of
188 Aspects of teaching

authority in interaction is different from the exercise of authority

in transaction, as I have defined those terms.
In interaction, the teacher—as professeur—claims a superior
and dominant position by virtue of a role which has been
socially ascribed to him or her: ‘I am your teacher. By the
authority vested in me I have the right to ask you to behave in a
certain way, whether you like it or not. And you, in your role,
have the obligation to obey.’ So the exercise of authority in
interaction is more or less authoritarian.
But the teacher as enseignant exercises authority in trans¬
action by virtue of the achieved role of expert. His or her
authority is based on professional qualification. Dominance
derives from the claim to be able to teach, to make the
transaction successful in respect of its specified objectives. In this
case there is no assertion of right but a claim to knowledge: not
‘Do this because I tell you and I am the teacher’ but ‘Do this
because I am the teacher and I know what’s best for you.’
Transactional exercise of authority, therefore, is more or less
authoritative.
This difference between being authoritarian and authoritative
is nicely illustrated in a recently published paper by Barry
Taylor. He argues for the need to foster ‘self-investment’ and
whole-person goal accomplishment as a condition for effective
learning through engagement with communicative tasks. Such
an approach, he says,

points at the need to maintain a non-authoritarian presence

throughout this process so that students can feel secure and
non-defensive to enable them to learn not because the teacher
demands it of them, but because they need to in order to
accomplish their own goals.

But then he goes on in the next paragraph:

This approach stresses that sharing the responsibility for

structuring learning with the students does not require that
teachers abdicate their fundamental authority to guide and
structure their classes. (1987:58 [my italics])

How we view the exercise of authority in interaction will

depend on our attitude to education and the society it serves.
The ideological trend in the western world is towards a less
authoritarian position and perhaps this is indeed necessary for
the maintenance of an open society. Be that as it may, this is a
 The roles of teacher and learner 189

quite different issue from the exercise of authority in transaction.

For no matter how we view pedagogy, no matter how much
initiative we believe should be allowed to the learner, the teacher
as enseignant must surely retain an undiminished authority. The
increase in learner-centred activity and collaborative work in the
classroom does not mean that the teacher becomes less authorit¬
ative. He or she still has to contrive the required enabling
conditions for learning, still has to monitor and guide progress.
And all this presupposes an expertise, applied perhaps with
more subtlety and consideration and discretion than before, but
applied none the less. I see no future whatever for any pedagogy
which undermines the authority of the teacher in his or her role
as enseignant, as ultimately responsible for the management of
classroom transactions. Indeed, if one does not allow the
legitimacy of this authority, then I do not see any point in talking
about pedagogy at all. It seems to me that it is because these
differences in the exercise of authority have not been properly
recognized that the authoritative actions of the teacher have at
times been discredited quite improperly as authoritarian im¬
positions of power.
Finally, a word about learner autonomy. This too has its
transactional and interactional aspects. As far as the trans¬
actional aspect is concerned, the learner {apprenant) really only
exercises autonomy within the limits set by teacher authority.
The learner is never really independent, it is the kind of
dependency which changes.
A distinction is commonly made between the natural contexts
of learning which allow for the exercise of independence and the
contrived contexts of instruction which enforce constraints. The
first are associated with the primary socialization of
family and friends, and the second with the secondary social¬
ization of the school. But the distinction can be too starkly
drawn. In upbringing the child is subject to all manner of
controls and is constrained to conform to established patterns of
behaviour whether these are determined by members of the
family or peer group. Those in contact with the child contrive to
shape its behaviour and there can be penalties for nonconformity
at least as disagreeable as those devised by the school. Indeed
one might define the very concept of learning in general as the
recognition of appropriate authority.
So the idea of ‘natural’ learning as being a process of
untrammelled discovery or self-directed quest seems to me to be
190 Aspects of teaching

misleading. The process is always directed. The issue is not

whether or not it should be subjected to direction, because it
always is, but what kind of direction is most ideologically
desirable and most pedagogically effective. This is not to deny
the possibility that there are certain parameters of language
development which are innately programmed in the mind.
However, the fixing of these parameters in the learning of
particular languages is, I suggest, always dependent on the
action of external control of one kind or another, applied with
varying degrees of indirectness and subtlety.
Learner autonomy in the transaction of language learning is
therefore necessarily restricted. And it is interesting to note that
when interactional autonomy is given the opportunity to assert
itself, it can have consequences which run counter to trans¬
actional purposes. Consider again, for example, the case of
group work.
The teacher may arrange for the class to work, let us say, on
problem-solving tasks in groups. From a transactional point of
view, this has the advantage of increasing participation in the use
and practice of language (see Long and Porter 1985). But if the
groups are to operate by exercising interactional autonomy in an
independent and unconstrained way then they will quite
naturally develop their own norms and expectations and these
will apply not to the role of pupil at all but to the role of peer
group member.
The crucial point here is that there are in fact two separate
interaction groups in a classroom. One of them is overt and
consists of the teacher as professeur and the collectivity of
pupils. The other is covert for most of the time and consists of
the children themselves within their own peer group, where the
criteria for pupilhood no longer apply. It is this other group
which is in interactive operation in the hubbub before the
teacher comes in to set the scene and start the lesson. This group
then, as it were, goes underground, only surfacing from time to
time, often with disruptive effects. Now this group of course
has its own social norms and these have little correspondence
with those of the pupil. It has its own standards of appropriate
behaviour and these are not likely to match up with what
the teacher expects from group activity. Thus the group
may see it as appropriate in reference to their norms of
behaviour to withdraw commitment, to sabotage the activity, to
ridicule attempts to use the foreign language and to give high
 The roles of teacher and learner 191

prestige to comic incorrectness and impropriety. There are, then,

obvious dangers in allowing groups to exercise a genuine
interactional autonomy. This should not be taken as a recom¬
mendation that all interactional autonomy should be sup¬
pressed, but only that it should be subject to careful consideration
lest an ideological zeal for democratic gesture should undermine
the whole pedagogic enterprise.
I believe that the success of this enterprise, to which we are all
presumably committed, depends on our recognizing and re¬
solving the difficulties inherent in the dual functioning of roles in
the classroom encounter that I have outlined, and on a
reconciliation of the claims of authority of the teacher on the one
hand, and the claims of autonomy of learner and pupil on the
other. As in other areas of social life, success in transaction and
interaction in the classroom depends on our knowing the parts
we have to play and how they relate with those that others enact
in the encounters in which we are engaged.
 - hj.' vUa»»'^<u.o;f'.^fS;iq
jT'f V' 1*. ‘ m >n>v70.1w fUi <'’ 0V'if?
t ii'*^**^ ffjrt '-iw, <A^w.*Jv*r- •
viocitontJ^ ^ 'f. I- r ^•w,,-.. f/;; U H? . ^<'^7 '.’jl 111

<tA3l4» VJ ■’ •S''* 'J'i .\'>f..-..<i

—Mu.bUjmhtitrmx \ .. .^vt^<r/3t.i nc '-.x
' vt-., :rrf *>f If wit
^ *^a.:irr rhir^ .-j? *’ r :..'j'tii'I
<j- Ll!( ■ VX.K- • »Mitul iU»-r7r|
)' 'I - .«•
u tilTf- •..:' ^nnk. ;• '» . ; • f* li!

. tf,'. lu ' .: - ' I

-K , %ftt, 7 »ff I' •l,»^".. .'1'.!^

iJii C.' fv ' ■ ill fj 'l-'llij
' v.i : » «(i u
r. . f .*)J '••■<1 . 1 ■ } y r.i{ ovf

► *Vl*i 4 ;)
■• / -tf
, ♦ ■
' ' * -4, ,.t -• .V '

> L 1.
'I *1 *'* f

* jwe
ft
■ ifcsfc^r*

‘,..1 ',• .'•

• j r s ^*1

•‘'i /j -Jp
t«r IB ’
rv' .: ' ’ jC' ihc jciiv . ro
1
1 ifXli/, . n/w -ji* 4-id 1
Conclusion

Chapter 11 ended with a consideration of the relationship

between teacher authority and learner autonomy. The central
question this poses is: how can learner behaviour be delimited in
such a way as to direct initiative towards the achievement of
effective learning? We have come full circle to the issue to which
I referred in Chapter 1, namely the need to acknowledge that
individual enterprise can only be meaningful if it is bounded.
Creativity is only possible in reference to established convention.
Freedom presupposes restriction. All the chapters in this book
can be read as variations on this basic theme. The activities of
language users, language teachers, and language learners can all
be characterized as ways of achieving particular purposes by
reference to a general conceptual framework which at the same
time delimits these activities and defines their significance.
Put simply, the proposition is that all new experience is
necessarily circumscribed by what is given. An absolute exercise
of initiative, to the extent that this is possible at all, would result
in confusion. In reference to language use, one cannot just use
words and structures at random in the hope that context will
somehow provide the necessary co-ordinates of meaning. Mean¬
ings are not freely negotiable. They are constrained by estab¬
lished rules and conventions: the rules of the language system,
the conventions of their communicative use. But although they
are constrained, they are not determined. There is always room
for manoeuvre. Competence in language must be a matter of
knowing these constraints and using them in association with
context to narrow down the possibilities of meaning on actual
occasions. These constraints, these rules and conventions are
general variables which have to be given particular values.
These variables and values correspond to the principles and
techniques of language teaching. Techniques are activities which
are designed to meet the needs of a particular context, like
linguistic utterances. But they also realize more general prin-
194 Conclusion

ciples, just as the utterance realizes more general categories of

linguistic and social knowledge no matter how closely it may be
keyed in with the needs of the context. The ability to adjust
linguistic behaviour contingently to meet particular commun¬
icative requirements presupposes a knowledge of general rules
and conventions. In the sam'e way, the ability to adjust
techniques to account for the demands of different classroom
contexts presupposes a knowledge of more general pedagogic
principles. In other words, the pragmatics of language teaching
discussed in Part 1 correspond with the pragmatics of language use
discussed in Part 2.
What then of language learners? They, too, need some given
conceptual framework within which to operate if their activities
are to have any point. The central task of pedagogy is to find the
framework which is most effective for learning. This will
necessarily involve taking into account the attitudes, interests,
and predispositions of the learners themselves. How far this calls
for open consultation or negotiation with learners in the design
of courses, how far this design is established as a projected plan
or as a continuous process, are matters which can only be
determined in reference to particular teaching/learning situ¬
ations. But whatever process is most expedient, the end result
must be a framework of some kind, some set of bearings to
enable learners to find their way. It should also be borne in mind
that whatever initiative learners are allowed, the destination
will, in most cases, be determined by .decisions beyond their
control. This is one aspect of language learner behaviour which
distinguishes it from that of normal language use: it is
necessarily prescribed. The contexts in which it occurs are
contrived for its inducement. One can disguise the contrivance,
stimulate learner incentive and exploit learner initiative in all
manner of ways. These are the usual tactics of teaching. But
ultimately they serve the basic strategy of control. They are
manoeuvres within confinement. We come here to another
distinguishing feature of language learner behaviour. It is only
valid to the extent that it induces knowledge, that is to say, that it
leads the learner to extrapolate, to generalize beyond the
immediate and particular circumstances of its occurrence. As we
have seen, using a language already acquired is a matter of
referring the particular to the general, but to acquire a language
one needs to infer the general from the particular.
All this points to the conclusion that we should not assume
 Conclusion 195

that language using behaviour is necessarily effective as language

learning behaviour, or that natural learning potential once
released from the inhibiting confinement of teacher control will
lead learners to home in on their objectives. One does not solve
the complex problems of language pedagogy by simply invoking
the concepts of authenticity of language on the one hand, and
the autonomy of learners on the other. There needs to be a
continuing process of principled pragmatic enquiry. I offer this
book as a contribution to this process—and as such, it can have
no conclusion.
 jkr'f.fr-

fr-f'rwk? i • ' . f- • ' **• r ^

^ y43»-. , 3.
■kf' '; ■■-:■' ■♦" * ',
«
1 - J ■ '/>•»,->(/>^f t ■-w*■''■■■ ..*">
\
• ' ' ifr i- ■• •• • 'rt: ^ '}'.Jr>:i
■ •* V' yh* r'- •' - • ■

tr/'
■ *••.■..' f^'Vr ' ^V'l'■ ■•■; 7 ? T (*' 1■ f. 'T* rfr^C'i
• -;■ ■- ■ . •) 'y, -.'r

*4 Jf t-r- * T-.f^ , '■ -;

'■ -. ‘ .1 ♦ »• K*' * .i ■

i •»'i> * ■•*■■?> M • ' •( *• H

'it > t- »/i ' ' >{.-'•• • .■ •, • • 1 •: IS . 1

• . •■•I i .T. . .lit - - li.

,1 r»i ' . ' '• .r.rn. t ‘31 <i . i>I>
L' ■ 'J ■ .* 1- . :i riK Je*. • 1,

' . -1 1 • . “."tJ ninf(

.«>•» he
‘ ^ ^ " 1 O'. . ' ■ ' . M ,
• ‘ .> ■ k- ■- . : - - ' fc^t
'-■ %1 • '«nu«^ (0
' r«» HZ’ m mtnj
■ 't * • . r . .^ViA De.'i.'r*Kiun

> "orci.^ l}»C«''

_ JUI wfui
i: Hr ‘ifcC: >r t;
•'..HU <»rair^ 4ti
« !..,(’rrr tti
^\i- U Ul 4l|
* \-.cv .*u* Ttjc.nitii; ‘^1
I'.t ^ IWir iti
itcTP lu «m4M
' c!- bf^f»w^v^ I: «
» MA^.. . ft. i* f > rii I

''I; V* , ' * i ’ •
• l> vO»irf ihi
4b ^ • tl. J *r»tr, A.<
’ ■< Jc‘‘

, t, ( •

»: .'Hi. •*
I Uic I
Bibliography

Alatis, J. E. (ed.). 1978. International Dimensions of Bilingual

Education. Washington D.C.: Georgetown University Press.
Alatis, J. E., H. H. Stern, and D. Strevens (eds.). 1983. Applied
Linguistics and the Preparation of Second Language Teachers:
Toward a Rationale. Washington D. C.: Georgetown Univer¬
sity Press.
Alderson, J. C. (ed.). 1985. ‘Evaluation.’ Lancaster Papers in
English Language Education 6. Oxford; Pergamon.
Alderson, J. C. and A. H. Urquhart (eds.). 1984. Reading in a
Poreign Language. London: Longman.
Allen, J. P. B. and P. van Buren (eds.). 1971. Chomsky: Selected
Readings. London: Oxford University Press.
Allen, J. P. B. and H. G. Widdowson. 1973. English in Physical
Science. London: Oxford University Press.
Allwright, R. L. 1983. ‘Classroom centred research in language
teaching and learning: A brief historical review.’ TESOL
Quarterly 17/2.
Allwright, R. L. 1988. Observation in the Language Classroom.
London: Longman.
Anderson, A. and T. Lynch. 1988. Listening. Oxford: Oxford
University Press.
Asher, J. 1982. Learning Another Language Through Actions:
The Complete Teacher’s Guide Book (2nd edition). Los
Gatos, CA; Sky Oaks Publications.
Aston, G. 1986. ‘Trouble-shooting in interaction with learners:
the more the merrier?’ Applied Linguistics 7/2.

Bachman, L. 1990. Pundamental Considerations in Language

Testing. Oxford: Oxford University Press.
Banton, M. 1965. Roles: An Introduction to the Study of Social
Relations. London: Tavistock Publications.
Bernstein, B. B. 1971. Class, Codes and Control 1. London:
Routledge and Kegan Paul.
198 Bibliography

Bialystok, E. 1982. ‘On the relationship between knowing and

using linguistic forms.’ Applied Linguistics 3/3.
Bialystok, E. and M. Sharwood-Smith. 1985. ‘Interlanguage is
not a state of mind: an evaluation of the construct for second
language acquisition.’ Applied Linguistics 6/2.
Bolinger, D. 1976. ‘Meaning and memory.’ Forum Linguisticum
1/1.
Breen, M. P. 1985. ‘Authenticity in the language classroom.’
Applied Linguistics 6/1.^
Breen, M. P. 1987. ‘Contemporary paradigms in syllabus
design.’ Language Teaching 1013,4.
Breen, M. P. and C. N. Candlin. 1980. ‘The essentials of a
communicative curriculum in language teaching.’ Applied
Linguistics 1/2.
Breen, M. P., C. N. Candlin, L. Dam, and G. Gabrielson. 1988.
‘The evolution of a teacher training programme’ in Johnson,
(ed.) 1988.
Brown, G. and G. Yule. 1983. Discourse Analysis. Cambridge:
Cambridge University Press.
Brown, P. and S. Levinson. 1978. ‘Universals in language usage:
politeness phenomena’ in Goody (ed.) 1978.
Brumfit, A. and S. Windeatt. 1984. Communicative Grammar 2
and 3. Beirut/Oxford: English Language Teaching for the
Arab World/Oxford University Press SARL.
Brumfit, C. J. 1983. ‘The integration of theory and practice’ in
Alatis etal. (eds.) 1983.
Brumfit, C. J. 1984a. Communicative Methodology in Language
Teaching: the Roles of Fluency and Accuracy. Cambridge:
Cambridge University Press.
Brumfit, C. J. 1984b. ‘The Bangalore Procedural Syllabus.’
English Language Teaching Journal 38/4.
Brumfit, C. J. 1984c. ‘Theoretical implications of Interlanguage
studies for language teaching’ in Davies etal. (eds.) 1984.
Brumfit, C. J. (ed.). 1984d. General English Syllabus Design.
Oxford: Pergamon.
Brumfit, C. J. 1985. Language and Literature Teaching: Erom
Practice to Principle. Oxford: Pergamon.
Brumfit, C. J. (ed.). 1986. The Practice of Communicative
Language Teaching. Oxford: Pergamon.
Brumfit, C. J. 1987. ‘Concepts and categories in language
teaching methodology’ in Kuhlwein (ed.) 1987.
 Bibliography 199

Brumfit, C. J. and R. A. Carter (eds.). 1986. Literature and

Language Teaching. Oxford: Oxford University Press.
Brumfit, C. J. and K. Johnson (eds.). 1979. The Communicative
Approach to Language Teaching. Oxford: Oxford University
Press.
Burt, M., H. Dulay, and E. Hernandez. 1973. Bilingual Syntax
Measure. New York: Harcourt Brace Jovanovich.
Bygate, M. 1987. Speaking. Oxford: Oxford University Press.

Canale, M. and M. Swain. 1980. ‘Theoretical bases of com¬

municative approaches to language teaching and testing.’
Applied Linguistics 1/1.
Candlin, C. N. 1983. ‘Principles and problems for INSET in
implementing a communicative curriculum’ in Alatis et al.
(eds.) 1983.
Candlin, C. N. 1984 ‘Syllabus design as critical process’ in
Brumfit (ed.) 1984d.
Candlin, C. N. 1985. ‘Teacher-centred training: costing the
process’ in Quirk and Widdowson (eds.) 1985.
Candlin, C. N. and H. G. Widdowson (eds.). 1987. Language
Teaching: A Scheme for Teacher Education. Oxford: Oxford
University Press.
Carrell, P. 1983. ‘Some issues in the role of schemata, or
background knowledge, in second language comprehension.’
Reading in a Foreign Language 1/2.
CarreU, P. 1987. ‘Content and formal schemata in ESL reading.’
TESOL Quarterly 11/3.
Carter R. and M. McCarthy (eds.), 1988. Vocabulary and
Language Teaching. London: Longman.
Chomsky, N. 1959. Review of Verbal Behaviour by B. F.
Skinner. Language 35.
Chomsky, N. 1965. Aspects of the Theory of Syntax. Cambridge,
Mass.: MIT Press.
Chomsky, N. 1980. Rules and Representations. Oxford: Black-
well.
Clark, J. L. 1987. Curriculum Renewal in School Foreign
Language Teaching. Oxford: Oxford University Press.
Coates, J. 1982. The Semantics of the Modal Auxiliaries.
London: Croom Helm.
Cole, P. and J. Morgan (eds.). 1975. Syntax and Semantics 3:
Speech Acts. New York: Academic Press.
200 Bibliography

Collins COBUILD English Language Dictionary. 1987. Lon¬

don and Glasgow: Collins.
Cook, G. 1989. Discourse. Oxford: Oxford University Press.
Cook, V. J. 1983. ‘What should language teaching be about?’
English Language Teaching Journal 37/3.
Corder, S. P. 1973. Introducing Applied Linguistics. Harmonds-
worth: Penguin Books.
Corder, S. P. 1981. Error Analysis and Inter language. Oxford:
Oxford University Press.
Coulmas, F. 1981. Festschift for Native Speaker. The Hague:
Mouton.
Coulthard, R. M. (ed.). 1986. Talking about Text. English
Language Research: University of Birmingham.
Cowie, A. P. 1988. ‘Stable and creative aspects of vocabulary
use’ in Carter and McCarthy (eds.) 1988.
Crombie, W. 1985. Discourse and Language Learning: A
Relational Approach to Syllabus Design. Oxford: Oxford
University Press.
Crystal, D, 1980. A First Dictionary of Linguistics and Phon¬
etics. London: Andre Deutsch.

Dakin, J. 1973. The Language Laboratory and Language

Learning. London: Longman.
Dakin, J., B. Tiffen, and H, G. Widdowson. 1968. Language in
Education. London: Oxford University Press.
Davies, A., C. Griper, and A. P. R.> Howatt (eds.). 1984.
Interlanguage. Edinburgh: Edinburgh University Press.
Donaldson, M. 1978. Children’s Minds. Glasgow: Collins/
Eontana.

Edelhoff, C. 1983. ‘Putting the communicative curriculum into

practice: the organization of teacher in-service education and
training in Hesse, Germany’ in Alatis et al. (eds.) 1983.
Edelhoff, C. 1985. ‘A view from teacher in-service education and
training’ in Quirk and Widdowson (eds.) 1985.
Elliott, G. 1981. Self-Evaluation and the Teacher. London:
Schools Council.
Ellis, R. 1985. Understanding Second Language Acquisition.
Oxford: Oxford University Press.

Faerch C. and G, Kasper (eds.). 1985. ‘Foreign language learning

 Bibliography 201

under classroom conditions.’ Studies in Second Language

Acquisition (Special Issue) 7/2.
Firth, J. R. 1957. Papers in Linguistics 1934-1951. London:
Oxford University Press.

Gleason, J. B. 1982. ‘Converging evidence for linguistic theory

from the study of aphasia and child language’ in Obler and
Menn (eds.) 1982.
Goffman, E, 1974. Frame Analysis. New York: Harper and
Row.
Goody, E. (ed.). 1978. Questions and Politeness. Strategies in
Social Interaction. Cambridge: Cambridge University Press.
Gregg, K. 1984. ‘Krashen’s monitor and Occam’s razor.’
Applied Linguistics 5/2.
Grice, H. P. 1975. ‘Logic and conversation’ in Cole and Morgan
(eds.) 1975.

Hakuta, K. 1974. ‘Emergence of structure in second language

acquisition.’ Language Learning 24.
Halliday, M. A. K. 1985. An Introduction to Functional
Grammar. London: Edward Arnold.
Halliday, M. A. K., A. McIntosh, and P. Strevens. 1964. The
Linguistic Sciences and Language Teaching. London: Longman.
Harley, B. and M. Swain. 1984. ‘The interlanguage of immer¬
sion students and its implications for second language
learning’ in Davies et al. (eds.) 1984.
Hartford, B., A. Valdman, and C. R. Foster (eds.). 1982. Issues
in International Bilingual Education. New York: Plenum
Press.
Hatch, E. 1979. ‘Apply with caution.’ Studies in Second
Language Acquisition 2/1.
Hornby, A. S. (ed.) 1946. English Language Teaching 1/1.
Hudson, R. A. 1980. Sociolinguistics. Cambridge: Cambridge
University Press.
Hymes, D. 1972. ‘On communicative competence’ in Pride and
Holmes (eds.) 1972.

Johnson, K. 1979. ‘Communicative approaches and com¬

municative processes’ in Brumfit and Johnson (eds.) 1979.
Johnson, K. 1982. Communicative Syllabus Design and Meth¬
odology. Oxford: Pergamon.
202 Bibliography

Johnson, K. (ed.). 1988. Programme Design and Evaluation in

Language Teaching. Cambridge: Cambridge University Press.
Johnson-Laird, P. N. 1983. Mental Models. Cambridge: Cam¬
bridge University Press.

Krashen, S. D. 1981. Second Language Acquisition and Second

Language Learning. Oxford: Pergamon.
Krashen, S. D. 1982. ‘The fundamental pedagogical principle in
second language teaching.’ Studia Linguistica 35/1, 2.
Krashen, S. D. 1983. ‘Second language acquisition theory and
the preparation of teachers: towards a rationale’ in Alatis et
al. (eds.) 1983.
Krashen, S. D. and T. Terrell. 1983. The Natural Approach.
Oxford: Pergamon.
Kuhlwein, W. (ed.). 1987. ‘Linguistics in Applied Linguistics.’
AIL A Review 4.

Labov, W. 1984. ‘Intensity’ in Schiffrin (ed.) 1984.

Lado, R. and C. C. Fries. 1957. An Intensive Course in English.
Michigan: University of Michigan Press.
Larsen-Freeman, D. 1983. ‘Training teachers or educating a
teacher?’ in Alatis et al. (eds.) 1983.
Leech, G. N. 1983. Principles of Pragmatics. London: Longman.
Levinson, S. N. 1983. Pragmatics. Cambridge: Cambridge
University Press.
Levinson, S. N. 1987. ‘Pragmatics and the grammar of anaphora:
a partial pragmatic reduction of binding and control pheno¬
mena.’/owrwa/ of Linguistics 23/2.
Long, M. H. 1983a. ‘Native speaker/non-native speaker con¬
versation and the negotiation of comprehensible input.’
Applied Linguistics 411.
Long, M. H. 1983b. ‘Training the language teacher as researcher’
in Alatis etal. (eds.) 1983.
Long, M. H. 1983c. ‘Inside the “Black Box”: methodological
issues in classroom research on language learning’ in Seliger
and Long (eds.) 1983.
Long, M. H. 1985. ‘A role for instruction in second language
acquisition’ in Pienemann and Hyltenstam (eds.) 1985.
Long, M. H. and P. A. Porter. 1985. ‘Group work, interlanguage
talk, and second language acquisition.’ TESOL Quarterly
19/2.
 Bibliography 203

Long, M. H. and J. C. Richards (eds.). 1987. Methodology in

TESOL: A Book of Readings. New York: Newbury House.

Mackey, W. F. 1965. Language Teaching Analysis. London:

Longman.
McCarthy, M. 1990. Vocabulary. Oxford: Oxford University
Press.
McLaughlin, B. 1978. ‘The Monitor Model: some methodo¬
logical considerations.’ Language Learning 1913.
McLaughlin, B. 1987. Theories in Second Language Learning.
London: Edward Arnold.
Malamah-Thomas, A. 1987. Classroom Interaction. Oxford:
Oxford University Press.
Mitchell, R. 1985. ‘Process research in second language class¬
rooms.’ Language Teaching 18/4.
Morrison, D. M. and G. Low. 1983. ‘Monitoring and the second
language learner’ in Richards and Schmidt (eds.) 1983.
Moskowitz, G. 1978. Caring and Sharing in the Loreign
Language Classroom. Rowley, Mass.: Newbury House.
Murphy, D. F. 1985. ‘Evaluation in language teaching: assess¬
ment, accountability and awareness’ in Alderson (ed.) 1985.

Naiman, N,, M. Frohlich, and H. H. Stem. 1978. The Good

Language Learner. Toronto: OISE.
Nattinger, J. 1988. ‘Some current trends in vocabulary teaching’
in Carter and McCarthy (eds.) 1988.
Newmeyer, F. J. 1983. Grammatical Theory: Its Limitations
and Possibilities. Chicago: University of Chicago Press.
Nunan, D. 1988. Syllabus Design. Oxford: Oxford University
Press.

Obler, L. K. and L. Menn (eds.). 1982. Exceptional Language

and Linguistics. New York: Academic Press.
Olson, D. R. 1981. ‘The literate native speaker: some intellectual
consequences of the language of schooling’ in Coulmas 1981.

Palmer, F. R. 1981. Semantics. Cambridge: Cambridge Univer¬

sity Press.
Pawley, A. and F. Syder. 1983. ‘Two puzzles for linguistic
theory: native-like selection and native-like fluency’ in
Richards and Schmidt (eds.) 1983.
204 Bibliography

Peters, A. M. 1983. The Units of Language Acquisition.

Cambridge: Cambridge University Press.
Peters, R. S. 1967. The Concept of Education. London:
Routledge and Kegan Paul.
Peters, R. S, (ed.). 1983. The Philosophy of Education. London:
Oxford University Press.
Piaget, J. 1955. The Child’s Construction of Reality. London:
Routledge and Kegan Paul.
Pienemann, M. 1985. ‘Learnability and syllabus construction’ in
Pienemann and Hyltenstam (eds.) 1985.
Pienemann, M. and K. Hyltenstam (eds.). 1985. Modelling and
Assessing Second Language Acquisition. Clevedon: Multi¬
lingual Matters.
Prabhu, N. S. 1985. ‘Coping with the unknown in language
pedagogy’ in Quirk and Widdowson (eds.) 1985.
Prabhu, N. S. 1987. Second Language Pedagogy. Oxford:
Oxford University Press.
Pride, J, B. and J. Holmes (eds.). 1972. Sociolinguistics: Selected
Readings. Harmondsworth: Penguin Books.

Quirk, R., S. Greenbaum, G. N. Leech, and J. Svartvik. 1972. A

Grammar of Contemporary English. London: Longman.
Quirk, R., S. Greenbaum, G. N. Leech, and J. Svartvik. 1985. A
Comprehensive Grammar of the English Language. London:
Longman.
Quirk, R. and H. G. ’Widdowson (eds."^). 1985. English in the
World. Cambridge: Cambridge University Press.

Rea-Dickins, P. and K. Germaine, 1992. Evaluation. Oxford:

Oxford University Press.
Richards, J, C, and R. W. Schmidt (eds.). 1983. Language and
Communication. London: Longman.
Richards, J. C., H. Weber, and J. Platt (eds.). 1984. The
Longman Dictionary of Applied Linguistics. London: Long¬
man.
Rivers, W, 1964. The Psychologist and the Foreign Language
Teacher. Chicago: University of Chicago Press.
Rogers, C. 1969. Freedom to Learn. Ohio: Charles E. Merrill.
Rubin, J. 1975. ‘What the “Good Language Learner” can teach
us.’ TESOL Quarterly 9/1.
Rubin, J. 1981. ‘Study of cognitive processes in language
learning.’ Applied Linguistics 2/2.
 Bibliography 205

Rutherford, W. 1987. Second Language Grammar: Learning

and Teaching. London: Longman.
Rutherford, W. and M. Sharwood-Smith. 1988. Grammar and
Second Language Teaching. New York: Newbury House.

Sampson, G. 1980. Schools of Linguistics. London: Hutchinson.

Schiffrin, D. (ed.). 1984. Meaning, Form, and Use in Context:
Linguistic Applications. Washington D. C.: Georgetown
University Press.
Scriven, M. 1967. ‘The methodology of evaluation’ in Tyler et
al. 1967.
Seliger, H. W. and M. H. Long (eds.). 1983. Classroom Oriented
Research in Second Language Acquisition. Rowley, Mass.:
Newbury House.
Selinker, L. 1984. ‘The current state of Interlanguage studies: an
attempted critical summary’ in Davies et al. (eds.) 1984.
Selinker, L. and D. Douglas. 1985. ‘Wrestling with “context” in
Interlanguage theory.’ Applied Linguistics 6/2.
Sharwood-Smith, M. 1981. ‘Consciousness-raising and the
second language learner.’ Applied Linguistics 2/2.
Sinclair, J. M. 1985. ‘Selected Issues’ in Quirk and Widdowson
(eds.) 1985.
Sinclair, J. M. and R. M. Coulthard. 1975. Towards an Analysis
of Discourse. Oxford: Oxford University Press
Smith, P. 1970. A Comparison of the Cognitive and Audiolingual
Approaches to Foreign Language Instruction: The Penn¬
sylvania Foreign Language Project. Philadelphia: Center for
Curriculum Development.
Spada, N. 1986. ‘Some effects of the interaction between type of
contact and instruction on the L2 proficiency of adult
learners.’ Studies in Second Language Acquisition %ll.
Spada, N. 1987. ‘Relationships between instructional differences
and learning outcomes: A process-product study of com¬
municative language teaching.’ Applied Linguistics 8/2.
Sperber, D. and D, Wilson. 1986. Relevance: Communication
and Cognition. Oxford: Basil Blackwell.
Stenhouse, L. 1975. An Introduction to Curriculum Research
and Development. London: Heinemann Educational.
Stern, H. H. 1978. ‘French immersion in Canada: achievements
and directions.’ Canadian Modern Language Review 34.
Stem, H. H. 1983. Fundamental Concepts of Language Teaching.
Oxford: Oxford University Press.
206 Bibliography

Stern, H, H. 1984. ‘Review and discussion’ in Brumfit (ed.)

1984d.
Stubbs, M. 1986a. ‘A matter of prolonged field work: notes
towards a modal grammar of English.’ Applied Linguistics 7/
1.
Stubbs, M. 1986b. ‘Lexical density: a computational technique
and some findings’ in Coulthard (ed.) 1986.
Swain, M. 1978. ‘Bilingual education for the English-speaking
Canadian’ in Alatis (ed.) ,1978.
Swain, M. 1982. ‘Immersion education applicability for non¬
vernacular teaching to vernacular speakers’ in Hartford et al.
(eds.) 1982.
Swain, M. and S. Lapkin. 1981. Bilingual Education: A Decade
of Research. Toronto: Ontario Ministry of Education.

Tarone, E. 1983. ‘On the variability of Interlanguage systems.’

Applied Linguistics 4/2.
Tarone, E. 1988. Variation in Interlanguage. London: Edward
Arnold.
Taylor, B. 1987. ‘Teaching ESL: incorporating a communicat¬
ive, student-centred component’ in Long and Richards (eds.)
1987.
Tickoo, M. L. (ed.). 1987. Language Syllabuses: The State of the
Art. Singapore: RELC.
Todd, L. 1974. Pidgins and Creoles. London: Routledge and
Kegan Paul.
Trim, J. L. M., R. Richterich, J. A, van Ek, and D. A. Wilkins.
1980. Systems Development in Adult Language Learning: A
European UnitICredit System for Modern Language Learning
by Adults. The Council of Europe/Oxford: Pergamon.
Trimble, L. 1985. English for Science and Technology. Cam¬
bridge: Cambridge University Press.
Tyler, R,, R. Gagne, and M. Scriven. 1967. Perspectives of
Curriculum Evaluation. Chicago: Rand McNally,
van Lier, L. 1988. The Classroom and the Language Learner.
London: Longman.
Vihman, M. M. 1982. ‘Formulas in first and second language
acquisition’ in Obler and Menn (eds.) 1982.
Wallace, C. 1992. Reading. Oxford: Oxford University Press.
White, L. 1987. ‘Against comprehensible input: the input
hypothesis' and the development of second language compet¬
ence.’ Applied Linguistics 8/2.
 Bibliography 207

Widdowson, H. G. 1968. ‘The teaching of English through

Science’ in Dakin et al. 1968.
Widdowson, H. G. 1975. Stylistics and the Teaching of
Literature. London: Longman.
Widdowson, H. G, 1978. Teaching Language as Communica¬
tion. Oxford: Oxford University Press.
Widdowson, H. G. 1979. Explorations in Applied Linguistics.
Oxford: Oxford University Press.
Widdowson, H. G. 1983. Learning Purpose and Language Use.
Oxford: Oxford University Press.
Widdowson, H. G. 1984a. Explorations in Applied Linguistics
2. Oxford: Oxford University Press.
Widdowson, H. G. 1984b. ‘The incentive value of theory in
teacher education.’ English Language Teaching Journal 39/2.
Widdowson, H. G. 1985. ‘The teaching, learning and study of
literature’ in Quirk and Widdowson (eds.) 1985.
Widdowson, H. G. 1989. ‘Knowledge of language and ability
for use.’ Applied Linguistics 10/2.
Wilkins, D. A. 1976. Notional Syllabuses. London: Oxford
University Press.
Wilkins, D, A. 1981. ‘Notional syllabuses revisited.’ Applied
Linguistics 2/1.
Wright, T, 1987. The Roles of Teacher and Learner. Oxford:
Oxford University Press.

Yule, G. 1985. The Study of Language: An Introduction.

Cambridge: Cambridge University Press.
"la

\r' t'il'in3 .iJ H

. ' ■
. I s:!». Ura / ' * v (ibi«6

■' - . V ' V -

' • •) ’ ■.-•,•;■-••■ i ; tr- ■ '.

•>- • r . iV4 ' . . , ■ ><.*>!',' '. >1' .

.H > I : ) H .'W. '•

..••/•; , ■ ■ '

Xiic4 • T' /L J M fj
,■. • • •I ' t. j;(i. > • j

r'firrwij^ .'••*'■ : M

’•» ■’f: I*,*® . .j K W

r., - .

/ . • • /■ 'i .-.su.sl J/
'n ■ .'■ f ' *' .|Qb*||3M; ■.•■■’'■ 'V-.fjyi'.-i.i
. it».'rt'< .ta >il V
t • ^ f *. % I j 1
.■■,1 • .. , - . \'R v*V5 ~A''k .)

t ,• Jf-f •' • ''■ *;i‘ - . ■ V J '■ ' rirtivjt ,

~ \ 1
■ V

. - >• r

_ f
"•t;, . ■

Ttii^
•‘1. »Tie btifui
pA«fC4«< /
Index

accessibiliry of language knowledge Canada, immersion programme

18.142 14-16
accommodation, vs. assimilation Candlin, C.N. 51
38-9,43 Carrell, P. 104
acculturation, hidden curriculum 184 categories, grammatical 41-2
accumulation, learning as 132-4 Chomsky, N. 9-10,11,17-18,76
accuracy, vs. acquisition 17-19 Clark, J.L. 127
achievement, vs. proficiency 139,141 classroom
acquisition context and activities 3,137-8,
order and sequence 141 157-8,182
vs. accuracy 17—19 engagements 184-7
vs. learning 19—22 instruction, effect on language
action research 60—1 acquisition 153—5
affective regulation 14 sociology 181
analysability of language knowledge classroom-centered vs. classroom-
133.142 oriented research 59
analytic syllabuses 133—7 cognitive regulation 14
application 31—3 communication 21—5, 42—3
applied linguistics 5—6, 8—10, 32—3 communicative approach 117—25,
appraisal 31—3, 49, 50 157-80
apprenant 184,189 communicative competence 39^0
articles 92-5 communicative function, categories 42
aspects 85, 86, 88—9 communicative grammar materials
assessment 51, 52 165-80
assimilation, vs. accommodation communicative syllabuses 130
38-9,43 competence
authenticity 44—7 communicative 39—40
authentic materials 67 vs. performance 17-18, 40,157—61
authoritarian vs. authoritative 188 composites 91-2
authority comprehensible input 22—5
association with medium view 122 comprehension 108,112-13
of teacher 187—91 computer analysis of language 35—6,
autonomy of learner 187—91 75,92
conceptual evaluation 31, 32, 35,
backsliding 18 44-8
Banton, M. 181 conformity, vs. nonconformity 122,
behaviourist psychology 11-12 128
Bernstein, B.B. 128 consciousness-raising 97
Bialystok, E. 18,133,143 content schemata 104
bilingualism, through immersion context 82—7,93-5,99-114
14-16 contextual plausibility 80-1
Bilingual Syntax Measure 16 contextual variability 143
Breen, M.P. 46—7, 51 convergence 107-8
Brumfit, C.J. 33—4, 61,146 Cook, V.J. 147
210 Index

co-operative imperative 109 form-words 89-91

Co-operative Principle 108-9 free learning 47
Corder, S.P. 47, 74, 139-40 frozen pragmatics 93
correctness, association with medium full words 89, 90—1
view 121, 124 functions
Council of Europe 12 vs. notions 41-2
coverage 36 see also notional/functional
Crescent English course 165 n. syllabuses
Crombie, W. 136 ‘Fundamental Pedagogical Principle’
cross-cultural accommodation 6 19,22,23
Crystal, D. 81—2
curriculum, hidden and overt 184 goals
student 188
decipherability 113 in syllabus 39,127, 131
definite article 92—5 Goffman, E. 45
denotation 166 ‘Good Language Learner’ 154,186
description 35, 74—8 grammar 40—2, 79—98, 165—76
determiners 92—5 grammargraphs 166, 169—71
direct method 24 group work 186
discourse 45,101, 107-10,142
discovery, association with mediation habit formation 11-12
view 120,122-3 Halliday, M.A.K. 41, 74,123
doing, vs. knowing 157—61 Harley, B. 15
Donaldson, M. 148 Hatch, E. 26
drills 67, 68-9 hidden curriculum 184
Hornby, A.S. 132
eclecticism 50—1 humanistic learning 13-14
education Hymes, D. 182,185
formal 127-8
of teachers 55-70 ideational knowledge 104
educational policy, relation to syllabus immersion 14—16
127-8 indefinite articles 92-5
educational value systems 127 index, vs. symbol 101—5,112
efficiency of natural learning 47—8 indexical meaning 82-3,102,104
Elliott, G. 60 inference'93—4,165
Ellis, R. 141 initiative, learner 4—5,124,147
emotional involvement in learning input hypothesis 144
172-3 in-service teacher education 65-70
empirical evaluation 31, 32, 51-3 INSET scheme 69n
enabling conditions for language instruction, effect on language
acquisition 144 acquisition 153-5
enseignant 184, 188, 189 interaction, in meaning negotiation
errors, learner 67, 111, 121,124 108
essential conditions for language interactional autonomy of learner
acquisition 144 190,191
evaluation interactional engagements 184,185
conceptual 31, 32, 35, 44-8 interactional exercise of authority 188
empirical 31, 32, 51-3 interlanguage 43, 139,140,141, 143
internalization 112, 158
Firth, J.R. 79-80 interpersonal knowledge 104
focus of form 20-1 interpretation 31, 32, 35, 38—44
formal education 127-8 intuitive learning 150—3
formal knowledge 104 investment, learning as 132^, 147
formative evaluation 51
formulaic expressions 43, 81,91-2, Johnson, Samuel 7-8
95-6 Johnson-Laird, P.N. 100-1
 Index 211

kernel sentence 76 meaning

knowing, vs. doing 157—61 indexical 82-7,102, 104
knowledge, 104 lexical 37, 88-9
vs. learning 5 methodology for teaching 157-80
Krashen, S.D. 15,19, 20, 21, 22, 23, negotiation 99-114
34, 50, 97, 144 symbolic 81, 102
meaning potential 123
Lado, R. 11-12,118 means/ends equation 44-7
language acquisition see acquisition mediation approach 30-8, 118-23
language description see description medium approach 118-23
language learning see learning methodology
language prescription see prescription medium and mediation approaches
language for specific purposes (LSP) 119-20
104,142 relation to syllabus 124,129-30,
language teachers see teachers 137-8,143-6
language use see use for teaching of meaning 157-80
learner autonomy 187-91 of teaching and research 58—61
learner errors 67, 111, 121,124 modules 66, 67—9
learner initiative 4—5,124,147 monitoring 22-3
learner’s role 121,181—91 Monitor theory 19-25
learning morpho-syntactic rules 142
free 47 Morrison, D.M. 22—3
and grammar 95—8 Moskowitz, G. 13
humanistic 13-14
natural approach 144
intuitive 150—3
natural language use 161—4
as investment and accumulation
natural learning, 47—8, 140—1,161^,
133-4,147
189-90
natural 47-8,140-1,161-4,
needs-oriented rationale 131
189-90
negotiation of meaning 99—114
as process 36 117—18
nonconformity, vs. conformity 122,
vs. acquisition 19—22
128
vs. knowledge 5
non-linguistic purpose 173,176
learning-centred approach 146
non-reciprocal negotiation 107—10
learning process 105—7,186—7
nonsense 79-82
Levinson, S.N. 93
norms of interaction 185
lexical density 91
notional/functional approach 113
lexical items 36, 83—6, 88—90,
notional/functional syllabuses 12,
95-6
39^2,130-7,159-60
lexicalized sentence stems 91-2
notions
lexical meaning 37, 88—9
vs. functions 41—2
lexical phrases 91—2
see also notional/functional
lexis, relationship with grammar
syllabuses
87-92
linguistic repetition 173,176 observational research 58, 59, 60
linguistics 5—6, 8-10, 32—3 Olson, D. 149
linguistic theory 77—8 operation 31, 32,48-51
literature, source of texts for teaching operational research, teaching as 2-4
of meaning 177—80 order of acquisition 141
Long, M.H. 23-4, 60 ordering of syllabus content 139^6
Low, G. 22—3 overt curriculum 184
LSP (language for specific purposes)
104,142 Pawley, A. 91-2
pedagogic research 55—70
Mackey, W.F. 36 pedagogy
manipulative research 59 general perspectives 117-25
materials, teaching 63^, 165-76 pragmatics 29—53
212 Index

principles 1—6, 19, 22, 23 Richards, J.C. 22

purpose 162 roles, teacher’s and learner’s 32-3,
relevance of research 26 121,181-91
Pennsylvania Project 154
performance, vs. competence 17-18, Sampson, G. 10
40,157-61 scene 182
person-orientation 128, 129, 186 schemata, content 104
Piaget,]. 38,148 schematic knowledge, vs. systemic
pidgin 90 knowledge 102—4,106—7,110—14,
Pienemann, M. 141 163-4
policy (educational), relation to Schmidt, R.W. 22
syllabus 127-8 Scriven, M. 51
position-orientation 128, 185-6 ^ second language acquisition (SLA)
Prabhu, N.S. 117,144-6, 147,171-3 learning 110—11
practice research 23, 47-8,140,142,186
exercises 67 self-investment 188
vs. theory 1—70 self-regulation 14
pragmatics Selinker, L. 142
frozen 93 semantico-grammatical categories 41
language teaching 29-53 semantics, vs. pragmatics 92-5,117
pedagogy 29-53 sense relations 166
vs. semantics 92—5,117 sentence meaning, vs. utterance
prescription 35 meaning 99-101
pre-service teacher preparation 64—5 sequence of acquisition 141
principles 1—2 setting 182
problems, relation to solutions 7-27 Sharwood-Smith, M. 18,143
problem-solving 138,147-8,171, Sinclair, J.M. 35, 36, 75-6, 77, 80
177, 180 Skinner, B.F. 11
procedural syllabuses 147 SLA see second language acquisition
process of learning 36,117-18 SLOPE test 21-2
professeur 184,187—8,190 socio-cultural factors 129
proficiency, vs. achievement 139,141 sociology, of classroom 181
progressive aspect 88-9 solutions, relation to problems 7-27
propositional relations 136-7 S.O.S (structural-oral-situational)
psychology, behaviourist 11-12 approach 117—19, 145
pupil 183 Spada, N. 15
purpose, vs. repetition 173,176 speech acts 182
purpose of learning 36,117—18 speech events 182
Stenhouse, L. 60
Quirk, R. 75, 88 Stern, H.H. 15
structural approach 113,117-19,
reasoning 148—9 157-61
reasoning-gap activities 145,172,173 structural-oral-situational (S.O.S.)
reciprocal negotiation 107-10 approach 117-19,145
reference, to reality 177,179 structural syllabuses 131—7
regulation 14 structure practice 67, 68-9
relational syllabuses 136—7 student 183
relevance 25—7, 29 summative evaluation 51
repetition, vs. purpose 173,176 Swain, M. 15
representation, of reality 177 Syder, F. 91-2
research syllabus design 127—55
pedagogic 55-70 syllabuses
relativity and relevance 25-7, 29 effect on communicative
second language acquisition (SLA) competence 39—40
23,47-8,140,142 medium and mediation approaches
teaching as 2—4 119-20
 Index 213

relation to methodology 124 vs. practice 1-70

symbol, vs. index 101-5, 112 Threshold Level 12-13
symbolic meaning 81, 102 tokens of linguistic element 75-7,
syntax 43, 75, 88 100-1
synthetic syllabuses 134-7 training of teachers 55-70
systemic knowledge, vs schematic transaction 108, 186
knowledge 102-4,106-7, 110-14, transactional exercise of authority 188
163-4 transactional purpose of classroom
184
task-based learning 103-4,170, 171, types of linguistic element 75-7,
173-7 100-1
task-based syllabuses 149—50
task control 120-1 use 131,158,161-4
Taylor, B. 188 utterance meaning, vs sentence
teacher authority 187-91 meaning 99—101
teachers
education and training 55-70 valency 36, 37,171
role 32-3, 121, 181-91 validity 57
teaching materials 63^, 165-76 value systems, educational 127
techniques, relation to principles 2 variants 143
telegraphic speech 17
tenses 85, 86 whole-person approaches
territorial imperative 109 (humanistic) 13—14
tests of achievement and proficiency whole-person goal accomplishment
139 188
text control 120—1 Wilkins, D.A. 39,41-2,131,134
text sequences, interpretation 177-80 word frequency studies 36
theory words 82—7
linguistic 77-8

ESOLCENTitg

■ 4 MAR 2004
OXFORD
APPLIED
LINGUISTICS

Aspects of Language Teaching

\

In this book Professor Widdowson provides a

timely critical review of the most prominent issues
in language teaching today. Part 1 presents the case
for the relevance of theoretical enquiry to classroom
practice. Parts 2 and 3 serve to demonstrate this
relevance by showing how different perspectives on
linguistic description lead to different approaches to
the teaching of language. At a time of wide-ranging
debate about, language teaching, particularly in the
broader contexts of educational principle and
policy, the force and authority of Professor
Widdowson’s arguments will be welcomed by all
those concerned to uphold professional standards in
language teaching and teacher education.

H. G. Widdowson is Professor of English Linguistics at

the University of Vienna.