-
J,
/,\
DESCRIBING ENGLISH LANGUAGE
SERIES ED.ITORS
]ohn Sinclair . Ronald Carter
A Grammar of Speechr.
David Brazil.;.,.
"o
..~'.
"o'
"~':t;"
~~Lr,.~".~~Ir;e
Oxford University Press 1995
-
Oxford University PressWalton Street, Oxford OX2 6DP
Oxford New York TorontoDelhi Bombay Calcurta Madras
KarachiPetaling jaya, Singapore Hong Kong TokyoNairobi Dar es
Salaam Cape TownMelbourne Auckland
and associated companies in Berlin Ibadan
Oxford and Ox(ord English are trade marks of Oxford University
Press
ISBN O 19437193 X
David Brazil 1995
All rights reserved. No pan of this publication may be
reproduced, stored in a retrievalsystem, or transmitted, in any
form or by any means, electronic, mecha nical,
photocopying,recording, or otherwise, without the prior permission
of Oxford University Press.
This book is sold subject co the condition that it shall nor, by
way of trade or otherwise, belent, re-sold, hired out or otherwise
circulated without the publisher's prior consent in anyform of
binding or cover other than that in which it is published and
without a similarcondition including this condition being imposed
on the subsequent purchaser.
Typeset by Wyvern Typesetting Ltd, BristolPrinted in Hong
Kong
ftif.rl>t,Yt~t::;~.~t:;V;,'
~-~
ht',,;;}.:'
~;
~jf,
f:'.i";.::lrs
{{;-!
1~.';-
';~.'';''';-!-~:
i~+t.~~
1~~:
)::t;~"'"'"J:1.':-:.
't:;~,
f,~
~~'t'
1~;~,~;
ii
t"~tif'Zl
~\-
i
ContentsAcknowledgements
The author and series editors
Foreword
Xl
Xll
Xlll
Transcription notations xv
Introduction
An exploratory grammarStarting assumptionsWhy do we want a
linear grammar?Discourse analysisWhat is a sentence grammar?Product
and processWhy speech?Who is it for?
12447
101112
1 The argument and organization of the bookCommunicating in
timeImmediate constituent grammarsFinite state grammarsSummary of
the argumentDevelopment of the description
1517202122
2 Used language
Sample of dataUsed speech is purposeful
Going through the motionsInteractionWhat can be told or
asked?Communicative need
Participants co-operateDealing with mismatchesProjecting a
need
Existential values
24262829303131323233
v
-
1,
Transcription notation
//p+// proclaiming tone (dominant)+ reduplicationo element with
zero realization# end of increment& and or so
XVI
.,
Introduction
An exploratQry grammarGrammars are of many kinds, and it is
important to say somethingat the outset about what kind of grammar
this is. There are manyways of describing the differences between
them, but one particularway will be useful for present purposes:
they may aspire to bedefinitive or they may be avowedly
exploratory. A definitivegrammar is helpful if one is in search of
what passes for a fact aboutthe way a language works: for instance,
if one needs guidance on aquestion of grammatical correctness or an
explanation of somethingmore theoretical=- for instance, why
certain sentences seem to becapable of two or more quite different
interpretations. Such agrammar starts from the assumption that
canons of correctness havesomehow been established, and/or that
well-founded explanationsare actually there to be found. Existing
definitive grammars do norexplain everything, however.An
exploratory grammar is useful if one is seeking possible
explanations of some of the many still unaccounted-for
observationsone may make about the way language works. Ir accepts
uncertaintyas a fact of the linguist's life. Its starting-point can
be captured in thephrase 'Let us assume that .. .' and it proceeds
in the awareness thatany assertions it makes are based on nothing
more thanassurnptions; the aim is to test these assumptions against
observablefacts. .The distinction just made is not, of course, so
clear-cut in practice,
or so well recognized as we may have seemed to suggest. Ir is,
forinstance, a hazard of the grammarian's business that users
frequentlyaccord definitive status to grammars which their writers
haveintended to be exploratory: carefully formulated hypotheses are
alltoo likely to be misinterpreted as expressions of God-given
truth. Itis partly for this reason that the present writer's
intentions areemphasized at this point. The book will be of the
'Let us assume .. .'kind.
1
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
-
Introduction
Starting assumptionsThis book differs from most of the current
work in the field ofgrammar in the rather different set of
assumptions ir invites readersto entertain. As its title indicates,
moreover, ir does so specifically inrelation to English speech. And
for reasons that will be given later,its main focus is upon speech
that is produced as monologue.
A cornmon assumption that underpins much comtemporarygrammar can
be crudely expressed like this:
Let us as sume that the mechanisms whereby words are assembledto
make larger units will be revealed to us if we begin by thinkingof
speakers as aiming, in everything that rhey do linguisticalIy,
atrhe production of objects which we cal! 'sentences'.
The alternative that this book explores can then be equally
crudelyexpressed like this:
Let us assume that the mechanisms whereby words are assembledto
make larger units will be revealed to us if we begin by thinkingof
speakers as pursuing some useftilcommunicative purpose andas
aiming, at any one time, at the successful accomplishment ofthat
purpose.
~iIf we say we are seeking to develop a grammar of communication
!'
to put aJongside existing grammars of sentence-making, we shall
.~have a working description of our intention which wilI do service
~until we are able to refine it. j~'~,"'~"
The notion of the sentence is so much part and parcel of what we
\:take grammar to be about that it is not immediately obvious how
we!can dispense with it as a key part of our conceptual apparatus.
It '.,may, indeed, seem perverse to try to do so. At this point it
is Possible!to do no more than give two close!y related reasons for
making so'rrfundamental a break with tradition. ~;
The first reason is that the notion of sentencehood brings with
it t.the idea of a self-contained object. We think we know where a
~:,sentence begins and ends; it has a special kind of completeness
that :~makes it seem to be in some way separable, both from any
other ~.speech that may precede or follow it and from the unique
:~background of speaker-hearer understanding against which every
',;sample of purposeful speech operates. To a very large extent,
the \c-'\ ;~,sentence which has no special context, either of a
linguisric or of a 1< J ~
2 t;'f\tt
I
lntroduction /
'real world' kind, has been taken to be the proper object of
thegrammarian's attentions. But it is not characteristic of
language usersto produce such free-standing objects; they produce
pieces oflanguage in the performance of some communicative activity
whichis meaningful in the situation they presently occupy. The
concern ofthis book is with the language that such users do
produce, so it isprobably better not to put ourselves in the way of
being tempted toentertain the idea of a sentence - that is, a
potentialIy free-standingand unused object-at al!.
L The second, and in some way more important, reason arises
fromthe first. It is very widely agreed, among those who have
approachedthe task of formulating grammars, that in order to
explain how theparts of a sentence are related to each other, a
structural mode! of arather special kind is necessary. Ir is the
kind of mode! that enablesus to show how small constituents combine
with others to makelarger constituents, which then combine with
others to make evenlarger ones, and so on - the complete sentence
is viewed as a kind ofChinese box assembly. We are not concerned at
this point with themany different ways in which this apparently
fundamental factabout linguistic structure has been represented,
but with the receivedbelief that no grammar of the sentence can do
without it. We willenter once more into the realm of
supposition:
Let us suppose that the need for this particular kind of
structuraldescription arises directly from the common preoccupation
withthe potentially free-standing object.
It may be only because we want to be able to abstract away from
allthe conditions surrounding an act of purposeful speech that
suchdescriptive apparatus is necessary. But language users, engaged
asthey are with some particular communicative task, can reasonably
beexpected to take into account what grammarians ha ve so
oftene1ected systematically to ignore. Seen from the user's point
of view,the notion of an unused and context-free object makes
little sense.We may reasonably ask whether the form that the
solution takes,that is to say the whole nature of the resulting
grammaticaldescription, is simply an inevitable consequence of the
peculiarnature of the problem we set out to solve.
Let us therefore further suppose, then, that by taking into
account
3
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
-
lntroduction
the conditions and circumstances that are accessible to the
user,we shall be able to manage without invoking the
'consituent-within-constituent' view of grammar. ,
Why do we want a liner grarnrnar~Ir is essentially the above
supposition that this book sets out toexplore. It develops one
tentative view of how a purpose-drivengrammar-as opposed to a
sentence-oriented one-might beconstructed; and, crucially, it does
so without having recourse to anynotion of constituency of me
hierarchicalIy organized kind.
The reasons for wanting to make such an exploration arediscussed
at greater length throughout the book. A brief statementwill be
sufficient here.
Speech is an activity that takes place in time: speakers
necessarilysay one word, follow it with another and then with
another, and soon. There are obvious and well-recognized
difficu!ties in reconcilingthis increment-by-increment presentation
of speech with a hier-archical constituent-within-constituent
account of how language isorganized. But many aspects of verbal
communication can scarcelybe investigated at all except in the
context of a view of speech as anongoing event, and there is no
need to labour the point that anaccount of grarnrnar which
dispensed with the Chinese box imagewould be more readily adaptable
for use in such investigations.
Users of the book will, therefore, need to be willing to read it
inthe same exploratory spirit as that in which it was written. It
is notdirectly concerned with specific applications, either to the
solution oftheoretical problems or to those practicaI matters in
which languageteachers and others might be interested but its
relevance to suchproblems and to such pedagogical matters will, it
is hoped, becomeapparent to all who are engaged with them. Its
reason for existing isthat, given the present sta te of linguistic
knowledge, a seriousatternpt to take into account the fact that
speech proceeds lineariyspatently desirable.
Discourse analysisAmong the many reasons for this, we might note
some of thequestions that engage the attention of scholars working
in fields thatwe can broadly describe as discourse analysis and
conversational
4
'i~~re:;:,"j
:$.~::{1:~~:.~
tit.::.0-.,(
)~
-;,~i
\!'""J',g."J(1',
,~
fl;'1'."!o'~'(~;~.:,gl'
. ,~~,rI ~\~
l!f.':jIi.,'!l.r,:
.,~~~-.ti::).".~- ...!,'t
J~
)J:i:r~.'.,:~
",t~~.'l:lf.~~
,~
{r.'
~*/:..~'0 ~~~.
~';~.tit'
lntroduction
analysis. Very much of the work that has been done in these,
fle,ldshas been described as analysis 'above the sentence', a
descriptionwhich not only takes sentence grammar for granted, as
someth~ngalready described, but also often implies the exist~nce of
some kl~ldof barrier or discontinuity. Practitioners on both sides
of the barnertend to t'ake little note of what those on the other
side are doing. Yetit is self-evident that those configurations of
words that thegrammarian calls sentences are, in som~ inescapable
sense, part ofthe communicative activity that the discourse analyst
commonlyattends to.
One influential approach to discourse analysis , mclair and
~-Coulthar 975), has used a hierarchicaIly conceived
grarnmaticalmodel as an analogue for 'discourse structure': the
various units ofdiscourseare visualized as being in a relationship
whiclrparallels theunits that a Hallidayan model posits for the
relationship among theconstituent parts of the sentence (Halliday
1961). This aP.J?ro~ch hasthe advantage of recognizing that the two
kinds of orgaruzanon arelikely to be essentially of the same kind -
that is, language users areunlikely to be operating two distinct
organizational system~ at thesame time. It does not, however, avoid
the problem of assurrung thatthey are discontinuous. The units of
one mode of analysis can onlybe described as having a tendency to
coincide with those of the other:
The units of the lowest rank of discourse are acts and
correspo~most nearly to the grammatical unit clause, but when we
describean item as an act we are doing something very different
from whenwe describe it as a clause. Grammar is concerned with the
formalproperties of an item, discourse with the functional
properties, Iwith what the speaker is using the item foro
.-J(Sinclair and Coulthard 1975: 27)
The particular kind of discourse that Sinclair and Coultharcl
had inmind when devising this model was that which teachers and
pupilsjointly construct in the pursuance of a certain kind of
classroomlesson. Modifications of it have been applied to other
kinds ofdiscourse. A criticism sometimes voiced is summarized by
tubb: .....not al! discoitse is as highly structured as that of the
classroom; suchan approach lis primarily applicabIe to relatively
formal_situ~tions inwhich a central aim is to formulate and
transmit preces ofinformation' (Stubbs 1983: 146); it is less
obviously applicable to
5
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
-
Introduction
'casual conversation between social equals, where the
generalfunction of much of the discourse may be phatic and social'
(ibid.:146).We might add that it is less obviously applicable, too,
to the
analysis of spoken monologue, since the use made of the
centralconcept of the exchange depends upon there being frequent
changesof role as between speaker and listener(s). As Coulthard
points out,even within the c!assroom this presents problems:
... at the simplest leve! there is the question of whether a
two-minute lecture by the teacher is one inform or a
series.(Coulthard 1985: 134)
Although some kinds of discourse can undoubtedly be analysed
inaccordance with a constituent-within-constituent model, it is
lessclear that such a mode! is necessary, in the way that one is
consideredto be necessary to deal with sentence structure. The
constituent partsof events of many kinds can, of course, be
perceived to partake inthis kind of relationship; but at least some
of those events might beequally amenable to description along
increment-by-increment.lines.When we turn to researchers who have
taken conversation, whethercasual or otherwise, as their
starting-point, we find greater re!ianceupon a far simpler kind of
organization. The concept of theadjacency pair which we owe to
Sacks (1967-71) says little aboutthe relationship between certain
utterances that is not covered by theobservation that one follows
the other in time: that answers followquestions, greetings follow
greetings, and so on. A particular interestof those who have
followed the conversational analysis line has beenthe efficient
management of speaker change. This would seemnecessarily to demand
attention to what can properly happen atsuccessive moments in the
temporal deve!opment of theconversation.
This is not the place to speculate about how much of what
Sinclairand Coulthard have presented in the context of a
hierarchicalapproach could be re-presented, without serious loss,
in terms of asimple linear one. There is good reason to suppose,
however, thatmany of the questions we want to ask about spoken
discourse couldbe sensibly addressed on the basis of the rather
unremarkable factthat the events that comprise such discourse occur
one after the other.
The task of reconciling two seemingly incompatible accounts
of
6
Introduction
how language is organized at a sub-sentential leve!, and
howdiscourse is organized at a supra-sentential one can, In fact,
bevisualized in a way which reverses Sinclair and Coulthard's
solution.If discourse can be described in terms of a purely linear
apparatus,can grammar-not rhe grammar of the sentence, but the
grammar ofthe functional increments of which discourse is composed
- bedescribed in a similar way? In asking this question we are, of
course,questioning the need for the distinction between formal
properti~sand functional properties which Sinclair and Coulthard
find itnecessary to make. What we are saying is that it is the
entityoperating as a functional increment in the discourse
whosegrammatical organization we want to describe. Can we show
thatword-like objects are arranged in sequences which perform
dis-coursal functions without first showing that they comprise
non-functional sentence-Iike objects? In so doing, can we show how
suchsequences operate as contributions-which are themselves
sequen-tially arranged-to a discourse? Can we, in other words,
dispensewith the notion that language users rely upon one set of
principieswhen they are organizing the essential building blocks of
theirspeech - their sentences - but make use of another set in
arrangingthem to meet perceived communicative needs?
hat i a sentence grammar:?Before seeking to embark upon an
answer to the questions we havejust po sed, we must look briefly at
what the term 'sentence grammar'implies. Anyone familiar with
contemporary linguistic scholarshipwill know that no account is
likely to satisfy everyone who isinvolved in it. Having noted that
the term 'grammar', or rather itsGreek equivalent, was once used to
make very wide r~ference to 'theart of writing', Lyons goes on to
say:
The history of western linguistic theory until recent times is
verylarge!y a history of what scholars at different times held to
fallwithin the scope of 'grammar' taken in this wider sense.(Lyons
1968: 133)
What has happened in recent times is that the terrn has come to
havemore restricted application:
7
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
-
Introduction
Grammar, we will say, gives rules for combining words to
formsentences. It thus excludes, on the one hand, the
phonological
, description of words and sentences, and on the other, an
accountof the meaning that particular words and sentences bear.
yons 1968: 133)
Lyons is at pains to underline the provisional nature of what
hesays by pointing out that our notions of 'sentence' and 'word'
are notunproblematical: but, in order to proceed we rnust, for the
timebeing at any rate, take these notions for granted.
If we continue in the spirit of Lyons' tentative formulation,
wehave also to recognize, as he does, that the mechanisms
wherebywords are presumed to combine to make sentences can
berepresented in various ways. Contemporary linguists tend to work
inthe light of differing views of those mechanisrns, views which
derivefrom differing sets of tenets: we have a number of schools
oflinguistics. We are likely to describe what we practise variously
astransformational, systemic, stratificational, functional grammar,
andso on. Debate occurs, of course, across the boundaries that
thesedifferences crea te; but it has, to a very much greater
exrent, occurredbetween members of one or other of the schools.
That is to say,discussion has been carried on within certain
assumptions whichmembers of the school have, to use Lyons' words,
agreed to 'take forgranted'.
We have already said that a fundamental difference that
separatesthe work of some linguists from that of others is the
status theyconfer upon the mechanism, or formalism, that they see
thernselvesas seeking to perfecto A dom~ant tradition, which ow
itsinspiration and many of its basic tenets to the work of
lI..::.homs.!sy,aspires to what we may think of as a definitive
account of naturallanguage. The search is for a formalism which
will relate theobservable evidence provided by all such language to
certainuniversal features. These features will then be seen to be
related toinvariable attributes of the huma~ mind. In this view,
the goal oflinguistics is to be seen as an advance in our
understanding of theproperties and functions of mind:
At the leve! of universal grammar, he [the linguist] is trying
toestablish certain general properties of human intelligence.
8
l:1i:r:(~:~~'-.;tlt:"'n:J,~+,'.1~',.~,
,,~,'lf':fJ
"'1;?.s-~,');
[1s;-~,t!,t'~
;~
t~':'~T'~
'f~'f\~~j;~:1.
IK'~rt. !~
',~,,
'.i"
':f,~rf.
't.~;~,~~
:~1.!,~;"~tl~
:fil
lntroduction .'
LingCs'tics so characterised is simply the subfield of
psychologythat deals with these aspects of the mind.(Chomsky 1968:
24).We may note in passing that, in so far as such a grammar
makes
the sentence its starting-point, any deductions one may make
aboutpsychological processes will depend for their ~alidity on
u.sersactually using sentences. This is something that this book
calls 1I1toquestion. The claim that Chomsky makes can be expressed
roughlythus:
Knowledge of one's language comprises one's knowledge of howall
the sentences of the language are recognized as sentences, andof
how to make and interpret them as such.
Alongside this, we might propose an alternative claim: ~
~~nKnowledge of one's language comprises one's ability to engage in
Ithose communicative events with which one is from time to
timeconfronted. ..-J
We can properly speculate as to whether it is only because of
the waythe description is set up that the first of these seems to
be a logicalprerequisite for the second.
Chomsky acknowledges that other aims may be pursued:
1 do not, by any means, intend to imply that these are the
onlyaspects of linguistic competence that deserve serious
study.(Chomsky 1962: 530)
There seems to be little doubt in his mind, however,
thatgrarnmars which do not address the central psychological issue,
as heformulates it, will be intrinsically less interesting than
those whichdo. _
Halli ay takes a different position. He sees a grammar asessen
la ly a tool for the examination of a piece of language with
aparticular analytical purpose in mind; the nature of the
grammarmay be due then, in part at least, to the nature of the
task: _
We should not, perhaps, take it for granted that a description
interms of a formalised model, which has certain properties
lackingin those derived from models of another kind, will
necessarily bethe best description for all of the very diverse
purposes for which
9
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
-
I .
Introduction
the descriptions of languages are needed. In assessing the value
ofa description, it is reasonable to ask whether it has proved
usefulfor the purposes for which it was intended.(Halliday 1964:
13) ,
The steps toward a description that are presented in this book
areoffe~ed .in t~e spirit of Halliday's pluralistic approach. Its
primarymotrvation is, as has been said, the belief that there are
many taskscur.rently facing the student of language, whether within
linguisticsor In one of the neighbouring disciplines, that might
well be tackledmore productively with the aid of a grammatical
descriptiondiffering quite radically from any of those that are
presentIy on offer.
Product and procesA hint as to the nature of such a description
is provided by Hallidayhimself:
V Traditionally, grammar has always been a grammar of
writtenlanguage: and it has always been aproduct grammar.
['Product' ishere used as one term of the Hjelmslevian pair
process/product.] Aprocess/product distinction is a relevant one
for linguists beca use'it corresponds to that between our
experience of speech and ourex erience of writing: writing exists
whereas speech happens.
alliday)1985: xxiii)
~alliday makes the distinction in order to define more clearly
theaims of the text-explicating description that he presents. Since
'text'is self-evidently a product, he can properly exclude from
con-sideration those factors which a process-oriented approach
wouldneed to take into account. But, there are many other reasons
forwhich we might need to have recourse to a grammatical
description;and for some of them, at least, a view of language as
somethinghappening might be a more promising srarting-point than
the moreco~mon one - view of it as something which exists as an
objectavailable for ost ha xamination. .
Halliday pro~~r-r1 sys that such an undertaking involves
'goingback to the beginning'. Many of the most basic concepts upon
whichproduct grammars have been erected will require
re-examination,This will sometimes require comparisons: we may want
to comparethe way a new kind of formalism might deal with a
certain
10
)lntroduction
phenomenon to how a previous model would deal with it.
Readerstrained in the use of one or other of the existing grammars
willdoubtless find themselves making such comparisons. When
thishappens some important considerations should be kept in mind.
Oneis that what is said should not be interpreted as a claim that
one wayof doing the analysis is better than another in any absolute
sense.The only way in which 'better' can be used is to say that one
solutionto a problem is more consistent than another with the
principiesupon which the present sryle of analysis is based.
Furthermore, thepurpose for examining certain instances in this
book is simplyto demonstrate that the principies underlying a
purpose-drivengrammar can be applied consistently, and provide
worthwhileinsights into how users proceed when assembling
speech.
Why speech?Part of the answer to this question has been
anticipated in previousparagraphs. Q.yr 'experience of speaking',
to use Halliday's term, is
J of something that begins, continues, and ends in time: it
happens. As'""1 spea ers,we know that causing it to happen is not
always without
its problems: our ability to put together what we want to say
may\ not always be equal to the pressure to keep up with ourselves,
so to1 speak, in the delivery of our message. As listeners, too, we
frequentlyfeel ourselves under similar pressure. The fact that time
is passingmakes it imperative to decode what we hear promptly so as
not tomiss what comes next.. Our experience of writing on the other
hand is often of somethingthat has been worked out in detail
beforehand: the act of carefulcomposition, at least, consists in
visualizing more or less clearlywhat we want to say, assembling the
language we need in order tosay it, and only then embarking upon
the mechanical business ofputting it down.
Such a comparison of two distinct modes of proceeding is,
ofcourse, far too simple to represent the truth. Much spoken
languageis pre-planned in considerable detail, and can therefore be
expectedto exhibit 'product' characteristics. Much writing is done
in the kindof unprepared way that we have associated with speech,
and leftunedited: it can thus be viewed as 'process'. It is no part
of the aimof this book to describe how and why speech and writing
differ in
r
11
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
-
Introduction
t~e way that HaIliday (1985) does -though having available a
real-time grarnmar might weIl be found advantageous by
anyoneattemptmg such a task. Speech has been chosen
forexamination~rincipaIly b.ecause its general nature as 'an
activity happening intime makes it a more transparently suitable
object for a 'process'approa~h. To this we might add two further
reasons. Firstly, as we~ave said, the nee? for ~ real-time grammar
seems to be most urgentm work upon vanous k.mds of spoken data.
Secondly, using spokendata enables us to take rnto account from the
outset a feature whichis absent from written data-intonation.
This b?ok is, then, a. 'grammar of speech' in that it is a
grarnmarwhich grves due consideration to those circumstances in
whichspeech is characteristically produced. There is no
intendedimplication in the title thatthe resulting formalism will
be any lesscapable of accommodating written English than spoken
English.It is rather that, by striving to remain aware of the
circumstancesin which people usuaIly speak, we can more easily
escape, so tospeak, from some of the suppositions that underlie the
variouskinds of product grammar available. Itis an arternpt to
describe theword-by-word assembly of speech in away which takes
into accountsome aspects of what we can, perhaps, agree is the
'reality ofspeaking'.
As we ha ve made c.lear, t!ls book is concerned with some very
basicand general considerations rhat affect our view of how
thegrammatical. orga~ization ?f speech should be perceived. It
stopsshort of discussing detailed applications. It amounts to
anexamin~tion. o.f the starting~P?int at which much of our
thinkingabout linguistic matters origtnates, and might be
considered asdealing with matters that relate to the interests of a
number ofdifferent ki ds oE reader.------ -
Firstly, ~here a~e those whose interest is in the analysis of
spokendata~ and m particular those who are concerned in some way
withrelating recent and current investigations into what happens
'abovethe sentence' to detailed descriptions of how language is
organized'below' this hypothetical watershed.
Secondly, there are wide-ranging implications in what is
proposed
12
lntroducon
here for those who are engaged in teaching English, whether as
themother tongue or as second or foreign language.
In the latter connection, the apparent difference between
learninga language and learning about it has often made for
difficulties indeciding upon appropriate classroom practices. The
conflictingdemands of learning to cornrnunicate and learning
grammar have ledto a great deal of debate, not aIl of it well
informed. In the rnother-tongue situation, the seeming paradox that
schoolbook grammarwith which students are presented can often seem
so difficult tomaster, even though they have a perfectly adequate
workingcommand of their native language, is one of the reasons for
a similarlongstanding dispute. Opinions vary widely about the
usefulness ofteaching explicit grammar to such students at al!. In
both cases,dogmatism has had an unfortunate tendency to take the
place ofenquiry into why these apparent conflicts existo The
possibility, asthis book suggests, that the rule systems purposeful
speakershabitually work with are of a quite different kind fram
those thatthey find in sentence grammars, might be one which is at
least worthentertaining. One mght reasonably ask whether
difficulties arsebecause learners of both kinds are being asked to
regard as a productsomething whch, from their perspectve as users,
most obviouslypartakes of the nature of a real-time process.
Thrdly, the attempt to adopt a user's perspective might make
thebook nteresting to another kind of reader. The question of
thepsychological reality of a grammar of a Chomskyan knd has
beencentral to the arguments that followers of that tradition
advance.The present book avoids involvement in this kind of
argumento Itdoes, however, seek to represent speaker's progress
through anutterance - hence, by implication, the listener's
interpretation of theutterance, in ways which might reasonably be
said to accord withhow they would report what they think they are
doing; they might,for instance, say that they were telling or
asking someone somethingrather than generating sentences of a
declarative or interrogativekind. The terminology is selected, as
far as possible, to reflect this.The exposition is couched largely
in terms of rnornent-by-momentdecisions. Although there is, of
course, no suggestion that these arealways, even often, conscious
decisons, the method reIies upon theconvention that they are in
some sense real to the users. Whether thisconvention s anything
more than a rnetaphor, or whether users
13
if,/l.
1!}
"
;
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
usuarioHighlight
-
lntroduction
really are psychologically engaged with the mechanisrns in
anythinglike the way suggested, is necessarily an open question. lt
is aquestion to which psycholinguists might find it .
worthwhileaddressing themselves.Finally, the question of the
feasibility of a linear, real-time
description of syntax is one which currently occupies the
attention ofthose who are engaged in manlmachine interaction. If
suchinteraction is to replicate purposefullanguage, it would seem
to beimportant to get a clear view, not only of how a real-time
grammarof such language might be represented, but also of the
extent towhich participants in a person-to-person interaction are
dependentupon pre-existent understandings to appreciate what is
going on.
14
usuarioHighlight
usuarioHighlight
usuarioHighlight
