-INGLES-

FONETICA Y FONO LOG I A 3

INGLESA III

PAG. IV A IX
PAG. 1 A 160 (166 CO PIAS)LC

Paul Tench
Cassell
Wellington House, 125 Strand, London WC2R OBB
127 West 24th Street, New York, NY 10011
Paul Tench 1996
All rights reserved. No part of this publication may be reproduced or
transmitted in any form or by any means, electronic or mechanical, including
photocopying, recording or any information storage or retrieval system, without
permission in writing from the publishers.
Paul Tench is hereby identified as the author of this work as provided under
Section 77 of the Copyright, Designs and Patents Act 1988.
First published 1996
British Library Cataloguing in Publication Data
A catalogue record for this book is available from the British Library.
ISBN 0-304-33690-4 (hardback)
0-304-33691-2 (paperback)
Library of Congress Cataloging-in-Publication Data
Tehch, Paul.
The intonation systems of English / Paul Tench,
p. cm.
Includes bibliographical references and index.
ISBN 0-304-33690-4. - ISBN 0-304-33691-2 (pbk.)
1. English languageIntonation. I. Title.
PE1139.5.T46 1996
421'.6dc20 95-5267=
CII
Typeset by Ben Cracknell Studios
Printed and bound in Great Britain by Biddles Ltd, Guildford and Kings Lynn
 Contents

Preface vii
Acknowledgements ix
C h a p te r 1 Intonation: what it is and what it does 1
1. Definition 2
2. System 7
3. Structure 12
4. Functions 16
4.1 The organization of information 16
4.2 The realization of communicative functions 18
4.3 The expression of attitude 20
4.4 Syntactic structure 21
4.5 Textual structure 23
4.6 The identification of speech styles 26
3. Conclusion 29
C h a p te r 2 Tonality: the units o f intonation 31
1. Neutral tonality 31
2. Marked tonality 34
3. Lists 36
4. Marked them e 37
5. Adjuncts 37
6. Tags 38
7. Tonality contrasts in grammar 39
7.1 Defining and non-defining items 40
7.2 Apposition 41
7.3 Verb phrases 44
7.4 Negative domain 45
7.5 Tteport clauses 46 ,
7.6 Clause complements 48
8. Intonation unit boundaries 49
C h a p te r 3 Tonicity: the focal point o f intonation 53
1. Tonic syllables 53
2. Neutral tonicity 56
3. Broad focus 57
4. Narrow focus 59
5. Marked tonicity 6l
6. The tonicity o f final adjuncts 64
 The Intonation Systems o f English

7. Tonicity by default 68
8. Tonicity and grammar 70
C h a p te r 4 Tones: the tunes o f intonation 73
1. Prim ary and second ary tones 73
2. Primary tones: falls, rises and fall-rises 74
2.1 Falls 75
2.2 Rises 77
2.3 Fall-rises 79
3. Tones and the status o f inform ation 80
3.1 Falls and rises 80
3.2 Fall-rises 83
4. Tones and the com m unicative functions 86
5. D om inance and deference in com m unicative functions 91
5.1 Information, reality and belief 92
5.2 Suasion 98
5.3 Social exchanges 101
5.4 Conclusion 105
C h a p te r 5 Tone variations 107
1. Attitudinal m eanings 107
2. Intonational lexicons 108
2.1 OConnor and Arnold 109
2.2 Pike 115
2.3 Halliday and key 119
2.4 Crystal 121
2.5 Brown 123
3 Intonational resources for attitudinal m eaning 124
3.1 Variations in tones 125
3.2 Heads 128
3.2.1 Low and high heads 129
3.2.2 Wide descending and ascending heads 131
3.2.3 Stepping heads 132
3.2.4 Glissando heads 133
3.3 Pre-heads 134
4. Sum mary 136
4.1 Tones 136
4.2 Heads and pre-heads 137
C h a p te r 6 Intonation in a m odel o f com m unication 138
1. M odels o f com m unication 138
2. Intonation in the m odel 148
3. The intonation systems 154

R eferences 155
In d ex 158

VI

mb \
 Preface

1995 m arked - it seem s incredible - the fiftieth anniversary of

K enneth Pikes publication o f The In to n a tio n o f A m er ic a n E n g lish .lt
is a classic; it is systematic and com prehensive to an extent not
rem otely m atch ed by any su b seq u en t sch olar o f A m erican
intonation.
1995 m arked - and this seem s alm ost as incredible - the twenty-
fifth anniversary o f M. A. K. Hallidays A C ou rse in S p o k en E nglish:
In to n a tio n . This too provides a thorough coverage, but o f British RP
intonation.
In their different ways, Pike and Halliday have b een the inspiration
behind this book. Pikes work, in my estimation, is seriously under
rated as a com prehensive descriptive statem ent o f an essential
feature o f language. To Halliday, I owe the debt o f an insight into the
place that intonation holds in an integrative view ofTanguage. My
own aim has b e e n to p rod u ce a co m p reh en siv e descriptive
statem ent o f British RP intonation w ithin an integrative view o f
spoken discourse.
I owe debts elsew here: to David Crystals P r o s o d ic System s a n d
In to n a tio n in E nglish, w hich is still the m ost useful account o f the
phonetics o f English intonation, and to other writers on intonation
like Brazil, Brown, Cruttenden and Ladd. I owe a debt to Robin
Fawcett and Barrie W ynn for their encouragem ents in my work. And
last, but certainly not least, Lowe debts to Julia Bullough, Ann Parry,
Cristina Rita and Je a n V errier for th eir patient w ork on the
m anuscript.

vii
 Acknowledgements

The author and publishers are grateful for permission to reproduce the
following:
Extracts from D. Crystal and D. Davy, A dvanced Conversational English
(Longman, 1975): by permission of Professor David Crystal.
Figure on p. 49, from G. Brown, K.L. Currie and J. Kenworthy, Questions o f
Intonation (Croom Helm, 1980): by permission of Routledge.
Figure 3.1, from David L.E. Watt, An instrumental analysis of English
nuclear tones in P. Tench (ed.), Studies in Systemic Phonology (Pinter, 1992):
by permission of Dr David Watt.
Figure 4.1, from Louis Alexander, First Things First (Longman, 1967): by
permission of Longman Group Limited.
Figure 5.2, from M. Liberman, The In ton ation al System o f English
(Garland, 1979).

IX
 Intonation
WHAT IT IS AND WHAT IT DOES

Intonation refers to the rise and fall of the pitch of the voice in spoken
language. When you say something, you cannot say it without some kind of
intonation - even a monotone can be classed as a kind of intonation.
Intonation is inevitable in speech.
It is also important: we have all made an observation like It is not what
they said, but the way they said it. The way they said it is a rough guide to
what intonation is.
Language students and teachers will also find intonation of interest as it is
part of the structure of any particular language. People can hear that the
intonation of English is different from the intonation of French, or German,
or Russian or any other European language, and is decidedly very different
from the intonation of Japanese, Hindi, Swahili, Quechua and all the other
non-European languages of the world. Furthermore, the intonation of
English varies within the range of national and regional accents that we
recognize: the so-called drawl of Southern USA intonation is quite distinct
from the melody of Welsh English intonation; and the intonation of Scottish,
Irish and Liverpudlian English seems to have rises where others have falls,
and vice versa. People are thus aware of intonation in a very general way,
even though the details of it are not properly appreciated.
Intonation has traditionally not had the same kind of attention in the past
that has been accorded to the study of consonants, vowels and word stress.
Rhythm has also been the focus of attention to a greater extent than
intonation because of its crucial role in poetry. But in the last two decades,
linguists have been turning to intonation in a much more systematic fashion
as a result of the upsurge of interest in discourse studies, and as a result,
much more is now known.
People have always been very much aware of consonants and vowels
because, in the written form, they are the units of pronunciation that are
most readily identified - even though there is an enormous discrepancy
between spelling and pronunciation (as is the case, for instance, in English,
French and Irish). And because we write in units of words, people also seem
The intonation systems of English

to be aware of word stress: people know the normal stress pattern of words:
return has the second syllable stressed, but hospital the first and cigarette the
last. Many people are aware of alternative stress patterns in modern English
in words like controversy, contribute and harass. But people are much less
able to talk confidently about intonation and rhythm because these are
features of language in use rather than of language in units (like words).
Although people can recognize the significance of the way something was
said, there isnt the same certainty in talking about intonation as there is
about talking about words and their pronunciation.
This book is intended to introduce to language students and teachers a
description of English intonation that will take into account the actual nature
of intonation and its functions in spoken language. Intonation is in fact part
and parcel of the English language, as it is for every language of the world.
Intonation is inevitable whenever a language is spoken; it is important,
because we eventually realize that it carries meaning and will often be the
most important part of a message (consider Oh they did, did they spoken in
a menacing way); and it is integral to the study of any language, for it links
up not only with meaning, but also with grammar, pronunciation and
spoken discourse at large.

1. Definition
A firmer definition than the one we gave on the first line of this chapter is
that intonation is the linguistic use of pitch in utterances. By saying
linguistic, we hope to avoid reference to other uses of pitch such as singing,
and to subjective, aesthetic evaluations about how nice and pleasant (or
ugly and unpleasant) an accents intonation is. (Some people say that they
cannot stand a certain accent and its intonation; this is obviously not a
linguistic judgement, since the allegedly unpleasant accent still has all the
forms and functions of intonation that it needs in order to carry meaning.)
We also specify pitch as the essence of intonation, and in this way we
make intonation distinct from the broader concept of paralanguage.
Intonation does in fact have both a linguistic and a paralinguistic dimension.
The liriguistic dimension concerns the message itself: how many pieces of
information there are; what information is new; whether the message is
complete or incomplete; whether the speaker is telling you something or
asking you, or whether the speaker is turning to a new topic or finishing off
an old one. The paralinguistic dimension concerns the messenger rather
than the message: the speakers state of mind, their degree of politeness and
their effort to associate or dissociate from you. But paralanguage - and this
is the point - involves not only pitch, but also volume, tempo and voice
quality, indeed all the vocal effects that are available within a given language

2
 Intonation: what it is and what it does

community, effects like giggling, sobbing, tut-tutting, husky voice, speaking

through the teeth and many, many more. Intonation itself is confined to the
use of pitch alone. Whereas we will attempt to cover both the linguistic and
the paralinguistic dimensions of intonation in this book, we will not be
attempting to cover the whole of paralanguage.
Our definition further specifies that intonation is concerned with
utterances. This- is important to note, because pitch variation is used to help
signify words in many languages. If a language uses pitch variation to
differentiate between words, we call that language a tone language. There
are many examples of tone languages in the world; indeed maybe up to 50
per cent of the worlds languages are tone languages. Chinese is a well-
known example: if you take a word like m a, it could mean different things
depending on whether it was spoken on a low pitch, a high pitch, a mid
pitch, a falling pitch or a rising pitch. Here is an example from another tone
language, Thai: these are five different words which could all be spelt kha, as
far as consonants and vowels are concerned. However, if kha is pronounced
with a low pitch, it means spice-, if it is pronounced with a high pitch, it means
trade-, with a mid pitch, herb-, with a falling pitch, kill-, and with a rising pitch,
leg-.
(1.1) _ kha = spice
- kha = trade
- kha = herb
\ kha = kill
/ kha = leg
It may be tempting to seek some kind of semantic link between the
meanings (e.g. spice and herb), but that would be quite fortuitous, as trade,
kill and leg testify.
These five different pitch variations are in contrast with each other and
help to differentiate word meanings; linguistically, these contrasting pitch
variations are called tones, and because their function is to identify words,
they can be more specifically called lexical tones. Their contrastive function
can be compared to the contrastive function of phonemes, cf. the contrast
between /l/ and /r/ in English, in lam b and ram. In fact, their contrastive
function is more akin to the contrastive function of word stress in a language
like English, of insight /'insait/ and incite /in'sait/, and words like invalid,
entrance and all the noun/verb pairs like import, survey, accent, etc. A well-
known word-stress contrast is found in Spanish: termino (= end (noun)),
term ino (= I finish) and termino (= he/she/it finished). A tone language
does not usually have word-stress contrasts as well, and languages that
employ word-stress as a means of lexical contrast do not usually have a
comprehensive system of lexical tone contrasts.
Our definition of intonation has thus to make clear the difference

3
The intonation systems o f English

between pitch being used for lexical tone, and pitch being used for
intonation purposes. We shall now illustrate the latter. If you say
(1.2) Johns going out, isnt he

with a falling pitch on isnt he, it will sound as if you are pretty sure of your
facts. However, if you say the very same utterance with a rising pitch on isnt
he, you will sound as if you are not so sure of your facts. Say (1.2) in these
two ways and check your impressions: going down on isnt he should give
you an impression of certainty, going up an impression of uncertainty, even
of questioning.
Notice that the words are the same in both cases and that the difference
in meaning is solely a matter of pitch variation. This is parallel to the Thai
example in (1.1) where the consonants and the vowel are constant and the
difference in meaning is solely a matter of the pitch variation. The distinction
between the Thai example and the English example is obvious: whereas the
Thai example concerns differences in words (i.e. lexical tone), the English
example concerns differences in utterances (i.e. intonation). Intonation is
thus the linguistic use of pitch in utterances.
Three more points need to be made, to prevent possible confusion. The
first is this: very often, utterances consist of single words like Yes, No, Well,
Right, Excellent, Bother! and so on. These single-word utterances will take a
range of intonations: Yes with a falling pitch will sound like a definite
statement of agreement in response to a question; Yes with a rising pitch will
'sound like agreement with what somebody else is saying, without
interrupting them; and Yes with a fall and then a rise all within the single
word will sound as if you only half-agree. Say these three variations to
yourself; no doubt you will be able to think of many more variations. Try to
work out their differences in meaning. However, note this; the basic
meaning of Yes has not changed; it retains its basic meaning of agreement
throughout. It is not like the case of Thai kh a which has radical differences
of meaning. English Yes retains its basic meaning; what changes is its effect
on the dialogue. If someone asks you a question and you reply Yes with a
falling pitch, that signals to the other person that you are in agreement with
them, and because there is assured common ground between you, the other
person can go on to a new point. If, however, you reply by saying Yes with a
falling-rising pitch, that signals that while you do agree to a certain extent,
you do not do so whole-heartedly; this will probably have the effect on the
other person of feeling the need to backtrack, or modify their argument, or
think of new ways of trying to convince you. Your intonation of Yes will have
an effect on the way the conversation proceeds.
The point being made here is that although the differences of pitch can
fall on a single word in English, it is a distinctly different matter from the
lexical tone of a language like Thai. It just so happens that in these cases the

4
 Intonation: what it is and what it does

utterance consists of a single word, but because that one word has
constituted a whole utterance, it is subject to the inevitability, i.e. the ever
present nature, of intonation.
If, instead of replying Yes, you had replied I agree, the same intonation
patterns could have applied. You could have said I agree with the same kind
of falling pitch to indicate a definite statement, or with the same kind of
falling-rising pitch on -gree to indicate half-agreement, or with the same kind
of rising pitch which indicates that you do not wish to interrupt. This
utterance consists of two words, and we are less tempted to think that the
variation of pitch affects the meanings of the words: it clearly affects the
meaning of the utterance.
You could experiment with these three patterns on other single-word
utterances and check that the basic meaning of the word does not change
and that the kind of meaning associated with falls and fall-rises is roughly
parallel in each case; the case of rises is a little more complicated, because
rises can sometimes mean I am going to continue or I am letting you
continue on the one hand, or I am asking you or appealing to you on the
other. But the fact that you can separate off the meaning of the intonation
from the basic meaning of the word is proof that we are handling intonation
in these cases and not lexical tone. (If you think about it, you cannot separate
off the meaning of the pitch variation in Thai words, because it is part of the
word itself - just like word-stress in English.)
The second point is that intonation is tied to utterances rather than to
sentences. Spoken language is not as neat and tidy as written language. If by
sentence, you are looking for neat, well-formed structures of recognized
syntactic patterns, you will not always find them in spoken language. Here is
an example as recorded in Brown, Currie and Kenworthy (1980):
(1.3) I regret + putting the people out of the out of the South Side and
central Edinburgh you know ++ I dont think ++ especially after
the war you know after the ++ war when they started the ++
redevelopment and the ++ well the authority more or less made it
that everybody was to go outside you know ++ the gardens and
houses but ++...
Why is it so confusingly unstructured? This is, nevertheless, typical of1
informal spontaneous conversation; it would not be typical of reading a
passage aloud where all the decisions of what to say have already been
taken. Brown, Currie and Kenworthy (1980: 47) explain:
In producing spontaneous speech the speaker has to decide on a topic, select
the staging procedures for presenting his (sic) topic..., determine what he (sic)
must introduce as new and what he can take as given, sort out the appropriate
syntactic structures, select lexical items, check that his listener is following what
he is saying and agreeing with it, make clear that he wishes to continue with or
to give away his turn, quite apart from speaking. In spontaneous speech,

5
The intonation systems of English

especially of a quite unrehearsed kind, where a speaker is more or less painfully

working out what he wants to say as he goes along, many speakers will produce
non-fluent speech.

So, it is no good expecting well-formed, precisely executed sentences in all

forms of spoken language. Sometimes, the sentences are so long and
complicated with abandoned or half-abandoned elements, and cluttered
with asides and appeals, that many, many units of intonation are employed
to manage them. On the other hand, two consecutive sentences may be so
short as to be contained within a single unit of intonation; it should not be
too difficult to think of a situation in which the following is said quickly in
one go:
(1.4) He did. I saw it.
A number of intonation studies do misleadingly refer to sentence
intonation or sentence prosody. Such studies usually confine themselves to
the melodies of single sentences spoken in isolation or in a simple two-part
dialogue. This is only a fraction of the reality, since spoken language is
overwhelmingly longer than single sentences or two lines of dialogue;
spoken language involves discourse.
The third, and final, point to be made in preventing confusion is a
relatively minor one. Occasionally, a dichotomy is established between tone
languages and intonation languages.1 This is a complete misunderstanding
about the nature of intonation. Some languages use pitch variation for
identifying words, i.e. lexical tone; other languages, like English, dont.
However, all languages use pitch variations for intonation purposes, even
including all the tone languages; intonation relates to utterances and
discourse, and of course, speakers of all languages (tone and non-tone
languages) use their languages for utterances and discourse. You might
wonder how a tone language manages to use pitch variation for lexical tone
and intonation: what happens is that the intonation patterns are
superimposed upon lexical tone, so that if the intonation pattern falls, the
lexical tones - whatever they are: high, mid, low, etc. - all fall relatively lower,
and so on. What is noticeable is that a tone language often has a simpler
intonation system than a non-tone language, and that it will employ
alternative linguistic devices - in grammar, usually - to compensate. This
seems a reasonable enough assumption, so that any languages use of pitch
is not overloaded.

6
 Intonation: what it is and what it does

2. System
Intonation is integral to languages, and therefore to language study and to
language learning and teaching. .Neglect of intonation in the past is now no
excuse for neglect in the present. Traditionally, pronunciation manuals
concentrated on consonants, vowels and word stress, and lexical tone in
tone languages: this area of interest is often labelled word phonology. Word
phonology is readily accessible to anyone with an interest in language,
because words and their pronunciation are easily identifiable; their
representation, through spelling, is also easily recognized in the written
form of languages. (It does not matter if the language uses an alphabetic
script, a syllabary, or pictographs; it is nevertheless words - as basic terms of
any message - that get represented.)
Rhythm and intonation are less easy to talk about because they accompany
whole messages. Rhythm is readily identifiable in poetry, but the role of
intonation in the recitation of poetry is less widely acknowledged. This does
not mean that no one had ever investigated the role of intonation until
recently, but simply that it did not receive equivalent coverage and attention.
Indeed, some eminent names in acting have drawn attention to intonation
and rhythm, namely David Garrick and Joshua Steele in the eighteenth
century; and some eminent linguists in the first half of this century have
published interesting studies, for instance, Armstrong and Ward, Jones,
H.E. Palmer and Pike. However, it was not until the availability of tape
recordings that linguists were able to investigate intonation more thoroughly.
The acquisition of tape recorders produced a new dimension to intonational
research in the 1950s and 1960s. Although Jassem, Kingdon and OConnor
and Arnold bridge the gap between the older studies and the newer, it was
Halliday and Crystal who set the pace by analysing long stretches of tape-
recorded spoken discourse. Not only were long stretches of spoken
discourse available for listening and re-listening, but recordings could be
subjected to a wide range of acoustic experimentation. The amount of
information that is now available is exhaustive; theories and descriptions of
intonation can now be based and verified on much more objective grounds.
Crystal (1969) is still, possibly, the most comprehensive discussion on the
phonetic nature of British English intonation, but Halliday (1967) sought to
present a more linguistic orientation and emphasized the phonological
nature of intonation. A most important part of this emphasis was to draw
attention to intonations role in the speakers organization of information. In
previous studies, the emphasis had been primarily on the attitudinal role of
intonation, i.e. how a persons feelings were expressed. Halliday, rather,
drew attention to the informational role of intonation, i.e. how a persons
meanings were expressed. Earlier studies had largely overlooked this aspect,
H.E. Palmer being the notable exception.

7
The intonation systems o f English

Halliday introduced the notion of a trio of systems operating in English

intonation: tonality is the system by which a stretch of spoken text is
segmented into a series of discrete units of intonation which correspond to
the speakers perception of pieces (or chunks) of information; tonicity is
the system by which an individual, discrete, unit of intonation is shown to
have a prominent word which indicates the focus of information; and tone
is the system of contrasting pitch movements in each unit of intonation,
which, among other roles, identifies the status of the information, e.g. major,
minor or incomplete.
We should look at this concept of system more closely. In linguistic terms,
a system means that there is a choice between one possibility and another,
resulting in a difference of meaning. When we looked at (1.2) above, we
noted that the wording could be identical for different renderings, one with
a falling pitch and the other with a rising pitch. The choice of pitch
represents a system. In the case of our discussion of Yes and I agree, we
noted that there were at least three possible renderings, with a different
meaning attached to each possibility. If there is a choice of three
alternatives with a meaning attached to each of them, we could call that
choice a three-way system - in this case, a three-way system of pitch
variations (or tones).
Just as there is a system for tones, there are also systems for tonicity and
tonality. How are you most likely to say this question?
(1.5) What are you going to do tonight?
At a good guess, the word do would be the most prominent word: it is
pitched quite high, but the voice drops to a low pitch which is subsequently
heard in the following word. Do is thus the focus of information, and in any
case, is likely to be the main point of interest in the question.
But say (1.5) again, with the emphasis on tonight. What change of
meaning takes place? The words remain the same, but the main point of
interest is no longer on what the person is going to do. It is as if the speaker
is indicating that he or she knows what the person might be doing in the
afternoon or tomorrow evening, etc., or what the person had done the night
before, etc., but in contrast to these points of time, the interest of the
question now lies in what the person plans to do tonight.
Now, you could make the word y ou the focal point, and again you could
consider what change of meaning takes place. It is as if the speaker is
alluding to the fact that they know what they are going to do, but now wants
to know what you, i.e. the addressee, in contrast to everybody else, plan to
do.
Can you think of changes of meaning that are caused by making going or
a re prominent (or even, possibly, what) ? This illustrates the tonicity system:
if you change the prominent word, you change the focus of information, and

8
 Intonation: what it is and what it does

thus create a different meaning. The tonicity system relates to the choice of
prominent word.
Tonality is also subject to a system. If you change the number of
intonation units, you change the number of pieces of information. You
could say (1.6) as one piece of information:
(1.6) Im going into town this morning

but you could present it as two pieces of information by separating off this
m orning with a little pause and saying:
(1.7) Im going into town |this morning
This sounds as if you have given one piece of information, and then,
possibly even as an afterthought, added an extra piece of information to
indicate when you were off to town. This would entail making town
prominent, and then, secondly, m orning prominent as well. The difference
between (1.6) and (1.7) is simply the speakers perception of the message as
either one piece of information or two. (You might argue that (1.7) would
have a comma after town; but remember we are talking about spoken
discourse; commas belong to written discourse.)
The tonality system also links up with certain syntactic choices in English.
There is a difference in meaning and in grammar between (1.8) and (1.9).
(1.8) My brother who lives in Nairobi...
(1.9) My brother |who lives in Nairobi...

In (1.8) there is no pause before who, and the relative clause tells you which
brother the speaker is talking about, i.e. not the brother who lives anywhere
else; the relative clause defines which brother; it restricts the referent. In
(1.9) , with a pause before who, the relative clause does not define or restrict;
it adds an extra piece of information. The difference in tonality - whether
there is a single unit of intonation in (1.8) or two in (1.9) - matches a
difference in grammar, and also a difference in meaning; (1.8) implies that
there are other brothers, (1.9) does not.
This book will devote a considerable amount of space to describing these
three systems of English intonation. The description of the intonation in
terms of systems shows that intonation is as systematic as other parts of the
phonology. And because it is systematic, it can be presented in terms of
differences of meaning, and that in turn implies that intonation can be
taught and learned. There is no need to think of intonation as a nebulous
phenomenon that can only be appreciated in subjective, emotional terms,
and is so personal that it defies careful analysis. On the contrary, although
there is a personal, subjective, emotional element to it, intonation is mainly
conventional. If it was not conventional, we could never know .what
meanings it conveys; in order to interpret an intonations meaning, we have

9
The intonation systems o f English

to assume that the intonation pattern that one person uses means the same
thing when somebody else uses it. To take a simple example, how do we
know that an intonation pattern is a menacing one if we dont have an
agreement (or convention) that that is what that particular pattern means?
It is because intonation is conventional and we know what different patterns
mean that we can make the comment about not liking the way something
was said. And because there is conventionality about intonation, we can
analyse it reasonably objectively and describe it reasonably succinctly.
Before we leave this concept of system there is one other major point to
be made. We shall be concentrating on a description of English intonation in
this book and on the tonality, tonicity and tone systems in particular. But just
as each language has its own system of consonants and vowels and word
stress (and lexical tones), each language has its own system of intonation.
This is partly why the intonations of different languages sound different;
although there are other features that affect the overall sound of the
language, like voice quality, the patterns of intonation are distinctively
different too.
Each language has its own system of intonation. It seems inevitable that
each language will manifest tonality, tonicity and tones, but the way they do
so will vary considerably. (We say will manifest and will vary, because only
relatively few languages have ever had their intonation system investigated
in any detail, and those that have, suggest that the kind of variety displayed
will be applicable to all languages.) It seems inevitable, also, that each
language will manifest a system of tone contrasts, but not necessarily a
system in tonicity and tonality. In Hausa, a language of West Africa, for
example, the prominent word in an intonation unit is always the last one; in
other words, there is no possibility of variation as in English, and thus no
possibility of choice: no choice, no system. It is conceivable, too, that a
language might have a fixed structure in tonality, too, in which case again the
same principle applies: no choice, no system. What this means, simply, is that
the kinds of meaning that are conveyed in English by choices in tonicity and
tonality are conveyed in other languages by other means - usually
grammatical, but by lexical means as well.
The next point to make in this comparison of the intonation systems in
different languages is that similar sounding patterns in two languages might
possibly convey different meanings, just as similar meanings might be
conveyed by different patterns. I remember overhearing a conversation
between a German lady and a British lady who had moved to Germany. It
sounded as if they were going for each other hammer and tongs, and I even
wondered whether I should intervene. However, I had no need to worry
because the door opened and the British lady beamed: Schwester A. is such
a darling! The mistake I had made was to interpret German intonation in

10
 Intonation: what it is and what it does

terms of my English intonation, and I had come to the wrong conclusion.

In the same way that we can misinterpret the intonation of another
language, we can also mis-produce it when we are speaking in another
language. When you are still in the early stages of learning to speak in a new
language, it is often difficult to reproduce the new intonation and sound
right. I remember learning Welsh but having difficulty in adjusting from an
English intonation to a Welsh one. This is a difficulty that most learners of a
foreign language have; if they are conscious of it, there is a good chance of
correct adjustments; if they are not aware of it, then of course intonation will
remain a problem.
But is faulty intonation a problem? It often is, but not always. What is very
noticeable is the tolerance that native speakers show when a learner fumbles
over consonants, vowels and word stress (and even lexical tone in tone
languages). This tolerance derives from the native speakers awareness of'
word phonology and possibly from their familiarity with other learners
efforts (or indeed their own) of the problems of pronouncing the words of
another language. But such tolerance does not always extend to intonation,
for two reasons. Firstly, native speakers are not usually so aware of
intonation, a point that we have made before. But secondly, and possibly
more alarmingly, a mistaken intonation still means something; it might well
be that you intended to convey one particular meaning but they interpreted
it quite differently. This is how misunderstandings of national character can
get perpetuated; for instance, the normal information-giving intonation of
one language might happen to coincide with the intonation of arrogance in
another. This could, quite clearly, lead people into thinking that the others
actually are arrogant, because they sound arrogant. Misunderstandings can
happen in other areas too: people who are giving information might sound
as if they are always checking up to see if you have understood them, and so
sound patronizing. Others who do not adjust the words of prominence in
their own intonation units will have difficulty in doing so when they speak
English, with the effect that a British person might misinterpret the real focus
of information and the ensuing discourse get miscued.
The problem is that mistaken intonation patterns still mean something,
but obviously not the intended meaning, and so misunderstandings can
easily follow, it seems important, therefore, for learners to adjust to the
system of their target languages, and for native speakers to become aware of
the system that they themselves use, albeit unconsciously, in their own
language.

11
The intonation systems of English

3. Structure
The intonation of English not only displays system, it also displays structure.
Indeed, structures of some kind will be found in all intonation systems in the
worlds languages. Think again of the question in
(1.5) What are you going to do tonight?

or the statement in
(1.6) Im going into town this morning

each being said in one go. Each is considered as a unit of intonation, each
handling one piece of information (whether asking about it or stating it).
Then, think again of the statement in
(1.7) Im going into town 1this morning

being said in two goes. In this case, each part is a unit of intonation: one
piece of information is given in the first unit, a second piece in the second.
In the same way, the part of the utterance given in (1.8) is a single unit of
intonation, whereas the same wording in (1.9) is a sequence of two units: the
first unit of (1.9) simply gives the topic (My brother) and the second provides
an extra piece of information about the topic.
Each intonation unit has a structure; we will illustrate the intonation unit
structure first of all with a very familiar saying:
(1.10) A dog is a mans best friend

Imagine, perhaps, that someone says this just as the conversation turns to
the topic of dogs. It is most likely to be said as a single unit of intonation,
with the word fr ie n d being most prominent, and the pitch of the voice
falling to a low level on that word. (Notice we have then specified all three
systems: tonality (segmentation of the discourse into units of intonation) =
one unit; tonicity (identification of the most prominent word within the
unit) = friend-, tone (specification of the contrastive pitch movement) =
falling.) In (1.10), the word fr ie n d is said to be the nucleus or the tonic
syllable. These two terms represent two traditions in the description of
intonation; in this context they mean the same thing, as shown in Table 1.1.
The part of the utterance up to best is called the pre-tonic segment, which
in turn can be divided into the pre-head and the head. The head is the part
which begins with the first stressed syllable, known as the onset syllable, in
this case the word dog-, the word a before dog is unstressed and precedes the
head, hence its label pre-head. The structure is thus:
pre-tonic segm ent tonlc/nucleus
pre-head head
A 'd og is a 'm a n s 'best 'friend
TABLE 1.1

12
 Intonation: what it is and what it does

The stressed syllables are traditionally marked by the tonic or nucleus is

marked either in bold, or in capitals or, as we shall do, by underlining.
The tonic (or nucleus) is obligatory, because it is that part of the
intonation unit that bears the contrastive pitch movement, the tone; it is also
always stressed not only because it bears the tone but also because of its
essential prominence. The other parts of the unit are optional, in the sense
that they may or may not happen to be present.
The sentiment of (1.10) could easily be rendered as
(1.11) 'Dogs are 'mens 'best 'friends
in which case there is no pre-head, because there happens to be no
unstressed syllable before the onset syllable of the head (Dogs).
Now suppose the topic of the conversation had been different, say,
horses, and somebody claimed that in fact horses were mens best friends.
Another could retort:
(1.12) 'Dogs are 'mens 'best 'friends
with contrastive emphasis on Dogs right at the beginning of the intonation
unit. In (1.12), Dogs is the most prominent; it is, therefore, the tonic (nucleus)
and is not preceded by anything; therefore, in the case of (1.12), there is no
pre-tonic segment, no head, and no pre-head.
What about the words that follow Dogs in (1.12)? The tonic (nucleus) and
all following words in the same unit are collectively known as the tonic
segment (obviously, in contrast to the pre-tonic segment). The tonic
segment can then be subdivided into the tonic and the tail. The tail thus
refers to all the words subsequent to the tonic syllable. (The use of the term
tail and the choice of topics in (1.10), (1.11) and (1.12) are purely co
incidental!)
You might ask why it is necessary to identify so many points in the
structure of the intonation unit. The answer is that different pitch
movements can contrast with each other at each point of structure, and thus
produce changes of meaning. The pitch of the pre-head may be either
higher or lower than normal, and these variations have a significance; the
pitch movement in the head is capable of many variations and these, too,
have their significance; the pitch movement at the tonic provides the basis of
the tone system, and the tone has an effect on the pitch of the tail. The sole
purpose of this elaborate dissection of the intonation unit is to indicate the
points or areas where intonation can vary and produce differences of
meaning. It is the explication of the choices in meaning that determines the
essence of linguistic structures.
In response to (1.12) - the contrastive, assertion that it is dogs who are
mens best friends - someone might agree by saying
(1.13) yes |they are |arent they

13
The intonation systems o f English

There are three units here (three goes).

(1.13a) yes

stands all by itself; it consists of a tonic, but no head, pre-head, or tail.

(1.13b) they are

consists of a tonic and a prediead. It may be thought strange to have a

structure that contains a pre-head, but no head, but it is perfectly possible. In
(1.13b), there is no onset syllable preceding the tone; they is unstressed and
therefore does not constitute an onset syllable. Without an onset syllable,
there is no head. What this means is that the variations of meaning
associated with the head are simply not available to (1.13b). (They were not
available to (1.13a) either.) (1.13b) also lacks a tail.
(1.13c) aren't they

has the obligatory tonic, has also a tail, but no head, and no pre-head. It will
be seen how the tonic is obligatory and the head, pre-head and tail are
optional. If the latter happen to be present in an intonation unit, they
constitute points or areas where choices of intonation (and, therefore,
choices of meaning) are potential. The choice of tone (and meaning) at the
tonic is always realized.
The formal structure of the intonation unit is displayed in Table 1.2.
pre-tonic se g m e n t to n ic se g m e n t

pre-head head to n ic / n u cle u s tall

1.10 A 'dog is a 'm an's 'best friend

1.11 'D o gs are 'm en's 'best frie n d s

1.12 bogs are 'm e n s 'best 'friends
1.13a yes
1.13b they are
1.13c aren 't they
TABLE 1.2

One final, small theoretical point. The structure of the intonation unit has
been presented mainly in terms of words so far; the tonic is the most
prominent word, the tail consists of the words following the tonic, and the
pre-tonic segment refers to the words preceding the tonic. We have also had
recourse to refer to syllables, stressed and unstressed. Which should we use:
words or syllables? It seems inevitable that when we think of the content of
the message contained in an intonation unit, we shall refer to words.
However, strictly speaking, an .intonation unit is a structure of sound and
belongs to phonology.
Halliday (and others2) show how intonation belongs to phonology. There
is a hierarchy of phonological structures and units. It can be described in the

14
 Intonation: what it is and what it does

following way: an intonation unit has a structure (is made up) of one or
more rhythmic units, or feet; each foot has a structure of syllables; and each
syllable has a structure of phonemes. Thus there are four ranks of
phonological structure: at the lowest level, phonemes, then syllables, feet
and intonation units.
It can also be shown that there is a relationship of function, as well as
structure, between the ranks. Certain phonemes have certain functions in
the syllable: vowels, in the main, function as the nuclei of syllables,
consonants as the margins. Then, certain types of syllable have their
functions in feet (rhythmic units): stressed syllables function as the nuclei of
feet, unstressed syllables as margins. And certain types of feet have their
function, in intonation units: the foot containing the tonic acts as the nucleus,
the others contribute to the head and the tail. The nucleus at each rank is
obligatory; the rest of the structure may or may not be present. It seems
inevitable, therefore, that when we think of the formal structure of the
intonation unit, we shall use phonological terms like feet and their structure
of stressed and unstressed syllables. The labels and examples of Table 1.2
can thus be interpreted in terms of words or syllables, depending on
whether we are dealing with meaning or form, respectively.
But what if the tonic consists of a word with more than one syllable, as in
. (1.14)?
(1.14) 'Snakes are 'mens 'most 'dangerous 'enemies
The most prominent word, en em ies, is the tonic segment; the stressed initial
syllable is the tonic syllable and the subsequent unstressed syllables are the
tail. In (1.15), the prominent word gets divided even further:
(1.15) But 'thats 'quite ri'diculous

The prominent word is ridiculous-, the stressed syllable is -die--, the following
syllables constitute the tail, but the initial, unstressed, syllable belongs
intonationally to the head.
We have now considered all parts of the structure of an intonation unit.
The tonic (or nucleus) is obligatory; without it we cannot identify a complete
unit. The head, pre-head and tail are optional in the sense that they may or
may not be present. It should also be remembered that we identify these
components solely on the grounds that they are points, or areas, where
meaning can be affected by variations in an intonation pattern. It should
now be possible for you to identify the four components in each of the
examples listed from (1.5) to (1.9); remember that (1.7) and (1.9) comprise
two separate units each.3
The fact that intonation units display structure (as well as system) should
help to dispel the notion that intonation is so nebulous and subjective a
phenomenon that it defies analysis and description. Because the structure of

15
The intonation systems o f English

intonation units can be identified and described, it should now be possible

to view intonation as a phenomenon that can not only be analysed, but
taught, understood and learnt. Furthermore, the intonation system and
structure of one language can be compared and contrasted with those of
another.

4. Functions
We turn now to the uses we make of intonation. System and structure relate
to the nature of intonation: what intonation is like. We turn now to what
intonation does: its functions.
Some of its functions have already been alluded to, and here we will
briefly introduce and illustrate its six major functions. i

4 .1 Th e organization of inform ation

Perhaps the most important, and least appreciated, function of intonation is I

to present the management of information as the speaker perceives it. In a !
typical conversation a person knows the information they want to present,
but then has to present it in pieces that are manageable not only to
themselves but also to the people they are addressing. We saw a snippet of
conversation in (1.3) above. It was simply too much information to present
in a single unit, and the speaker had to divide it up and present it a piece at
a time. The main topic is presented first:4
(1.3a) I regret + putting the people out of the out of the South Side and j
central Edinburgh
l
and is followed up immediately with an appeal to his addressee to believe
him:
(1.3b) you know
i
He then starts, but abandons, a new idea; he is obviously still trying to sort
out in his mind how he might best present his ideas:
(1.3c) + + I dont think

He now restarts, having decided to set out first of all a reference to the |
past:
(1.3d) ++ especially after the war you know

(Incidentally, the y ou know of (1.3d) is. not made as prominent as that of

(1.3b) was; hence it does not constitute a separate intonation unit in (1.3d).)
As he fumbles in his mind for the words he needs, he repeats the time
reference and proceeds:
(1.3e) after the ++ war when they started the ++ redevelopment

16
 Intonation: what it is and what it does

The pause before redevelopm ent suggests a difficulty in deciding upon the
word, but he cannot think of the next word and so abandons the next unit:
(1.3f) and the
(1.3g) ++ wel' the authority more or less made it
and follows up with the next point:
(1.3h) that everybody was to go outside

and an appeal as in (1.3b)

(1.3i) you know
and another new point:
(1.3j) ++ the gardens and houses
The organization of information involves decisions about the division of
information into manageable pieces and their staging - what comes first,
what follows, what precedes, and so on. It also involves grading the pieces
of information into major and minor and tying them up into coherent
sequences. The division and staging are handled by tonality, with
boundaries of intonation units; and the grading is handled by tone: rises,
falls, and fall-rises.
There is also another important aspect to the organization of information,
and that is the decision as to what should be made prominent in any piece
of information. This is usually discussed in terms of new and given (or
old) information, and is neatly illustrated in (1.3). You will notice that in
(1.3a) the word p eo p le is more prominent than either South Side or central
Edinburgh. What is left non-prominent, especially after the tonic, is
considered by the speaker as given information, i.e. it has been mentioned
or alluded to before, or is being treated as common knowledge or as obvious
in the current situation. Thus what is new in (1.3a) is people, and we must
assume that the locations have been mentioned before - either by the
speaker or someone else participating in the conversation.
In (1.3d) the word w ar is prominent; it is new. But in (1.3e), it is not
prominent and is treated, understandably, as given.
This information structure within the intonation unit is handled by
tonicity: the location of the tonic.
Not all spoken discourse is as unfluent as the snippet of conversation that
we have been analysing. Spoken discourse which has been rehearsed, e.g.
newsreading, story-telling, joke-telling,, teaching, preaching, etc. is likely to be
much more fluent, but whether the discourse is executed fluently or not, it
will still be subject to the speakers management of the information. The
type (or genre) of spoken discourse is immaterial in this connection:
decisions must be made by the speaker in every discourse with respect to

17
T he intonation systems of English

the division and staging of information (tonality), the structure of new and
given information within each unit (tonicity) and the grading of one piece
of information against another (tone).
This function of intonation, the organization of information, seems to be
a basic function, so basic, in fact, that people are hardly conscious of it.

4 .2 Th e realization of com m unicative functions

Another basic function of intonation is to present the speakers purpose in

saying something; whether the speaker is telling you something, asking you,
ordering you, pleading with you, or just plainly greeting you or thanking
you, etc. This dimension to talk has commonly been labelled the
communicative function by the language teaching profession;
philosophers and linguists also call this dimension discourse function,
speech functions, speech acts and illocution. They all refer to the
intended effect that the speaker wishes to produce on those who are being
addressed. Whereas the informational function of intonation answers the
question What is being said?, the communicative or illocutionary function
addresses the question Why is it being said?
We have already alluded to this function also, in example (1.2), when we
discussed the different effects that a falling pitch and a rising pitch have on
the tag isn t he-.
(1.2) (Johns going out), isnt he

The fall suggests certainty, knowing; the rise suggests uncertainty,

querying. Generally speaking, a falling tone in an intonation unit that
contains major information denotes speaker-dominance: the speaker
knows and tells, orders, demands, etc. On the other hand, a rising tone in an
equivalent unit denotes speaker-deference: the speaker does not know and
so asks, does not have authority and so requests, coaxes, etc. Let us illustrate
further.
If you feel you are in the privileged position of authority and expect to be
able to tell people to do something, you might well say ( 1.16):
(1.16) Shut the window

with a falling tone. On the other hand, if you do not feel you are in such a
privileged position, you might resort to a request rather than an order; in
which case, you could still use (1.16), but with a rise, Try the two versions; for
the request, it is important to pitch the word Shut reasonably high, the word
the lower, and the first syllable of w indow lower again; the rise then begins
with the low pitch of win- and continues upwards in -dow. Notice, again, that
the difference of communicative function is realized solely through
intonation; the wording of ( 1.16) remains constant.
Admittedly, if you are making a request, you might well append a

18
 Intonation: what it is and what it does

politeness word like please. But notice that even with this addition, the
request is going to have a rising tone:
(1.16a) Shut the window, please

The rise that is begun on the tonic syllable continues upwards right through
the tail (-dow\ p le a s e ). The fact that the request can be either with or without
the word please, shows that it is the intonation that is the principal means of
realizing the request function, not just the word please.
Here is another example of contrasting communicative functions. (1.17)
would be interpreted as an exclamation if it was accompanied by a falling
tone:
(1.17) Isnt it hoi

It is typically used to comment on unexpectedly hot weather and is

frequently prefaced with an exclamatory gesture like phew ! But you will
have no difficulty in thinking of other situations where the degree of
hotness might have been unexpected; think, for instance, of the
astonishment somebody might show when the value of a microwave oven is
being demonstrated to show how quickly food can be made very hot.
Now, on the other hand, think of what the microwave oven demonstrator
might say when he or she offers heated food when the oven has been
inadvertently left on defrost; the other person expects hot food, but is
disappointed and protests. The demonstrator, temporarily mystified, is quite
likely to say (1.17) with a rise on hot as a genuine question.
Try the two versions of it; for the rising version, try to keep Isn t it quite
high, and start the rise on hot at a low pitch and raise the pitch as you say the
word. A slightly different rendering of the rising tone may well occur to you,
too: starting the rise on Is- and continuing upwards through -n't it hot, as
(1.17a):
(1.17a) Isnt it hot

It is still a question, but seems to add an element of surprise or puzzlement.

(Notice that hot is no longer the prominent word; hotness is now being
taken for granted; it is being treated as given. Isn t is prominent instead, in
contrast to the imagined It is of expectation.)
To return to the main theme, falls and rises indicate, in broad terms, the
communicative intent of the speaker. Thus intonation can distinguish
between statements and queries, orders and requests, exclamations and
questions, and the like, even when the actual wording remains constant. This
communicative function of intonation is a basic function of intonation too,
because whenever we say something, we have a purpose in doing so.

19
The intonation systems of English

4 .3 Th e expression of attitude
This third function is probably the most familiar, and it was certainly
considered as the primary function in the older, more traditional, studies of
intonation. It corresponds most clearly to the observation Not what they
said, but the way they said it. The way they said it usually refers to the mood
of the speaker or the attitude shown to the addressee or the message. A
message, a piece of information, can be given politely, grumpily, angrily,
warmly and so on.
But a message can also be given without any particular emotion, as a plain
piece of information, as is typical in newsreading. We can simply pass on a
piece of information, or ask for information without an overlay of emotion;
such a style is labelled either plain or neutral. This is a useful concept,
because we can then define the expression of attitude as departures from
neutral patterns of intonation. This function answers the question How is it
being said?
Intonation, however, is not the only means available to a speaker to
convey attitudes. First of all, there is the wider range of vocal effects and
paralinguistic features mentioned above: voice qualities, vocal effects (like
sighing, sniggering or humming), tempo and loudness. Secondly, there are
gestures of the face, hands and body; distance and proximity, eye contact or
the lack of it, also indicate kinds of relationship between speaker and
addressee and thus convey attitude. Thirdly, the choice of words can be an
indication of attitude; there are angry words like stupid, affectionate words,
swear words; words can be chosen to have a sarcastic effect or to flatter, and
so on.5
If you are angry, you will sound angry and employ appropriate gestures
and words and a range of paralinguistic features; but the sound of anger will
also be expressed in pitch - quite typically in a high pitch sustained
throughout the utterance, such as in (1.18). 1
(1.18) You stupid fool I Look what youve been and done
The effect of attitude is mainly to be found in the extent of a fall or rise and
in variations of pitch in the head and pre-head. A wide falling tone (falling
from a higher pitch than normal, to low), for instance, usually denotes
surprise, intensity, something unexpected; a narrow falling tone (falling
from lower than normal, to low), on the other hand, denotes mildness,
something expected. Try (1.19) with a normal, neutral fall on again, to
represent a plain statement; then with a wide fall, from high to low, to
represent surprise, enthusiasm, unexpectedness; and then with a narrow fall,
from mid-low to low, to represent expectedness, or even lack of interest:-
(1.19) The Conservatives won again
Your preferred intonation will indicate your attitude to the message. Notice,

20
 Intonation: what it is and what it does

again, that intonation can indicate a variety of attitudes without a change of

wording; in such cases, it becomes clear that intonation has been solely
responsible for the expression of attitude.

4.4 Syntactic structure

The fourth function relates intonation to the syntax of clauses. In English,
there are many cases where two syntactic patterns can only be distinguished
by intonation. One example has already been given: the distinction between
defining (or restrictive) relative clauses as in (1.8):
(1.8) My brother who lives in Nairobi

and non-defining (or non-restrictive) relative clauses in (1.9):

(1.9) My brother I who lives in Nairobi
The non-defining relative clauses like (1.9) have, incidentally, also been
called adding clauses, which perhaps more clearly indicates their function.6
The grammatical distinction between (1.8) and (1.9) plainly represents a
choice of meaning, which is overtly signalled by intonation in its spoken
form. (In its written form, a comma would be expected after brother in (1.9),
but since writers are notoriously inconsistent in their use of commas, there
is no guarantee that the distinction would be signalled in writing. In speech,
the signal is obligatory.) The parallel wording of the two structures is
disambiguated in speech.
Another frequently cited example is the parallel wording of two clause
types as in (1.20) and (1.21):
(1.20) She washed and brushed her hair
(1.21) She washed 1and brushed her hair

In (1.20), h air is deemed to be the direct object complementing both

washed and brushed; thus w ashed (and brushed) is transitive. In (1.21),
however, with an intonation unit boundary immediately after w ashed, the
word h air is deemed to be the direct object complementing brushed only,
thus leaving w ashed as intransitive - in the sense of simply washing oneself,
presumably only hands and face (but not hair!). So (1.21) means that the
person washed herself (but presumably not her hair) and then did
something about her hair. Intonation has thus made the distinction between
a transitive (1.20) and an intransitive (1.21) use of the verb wash, despite the
identical wording.
You may be able to think of other possible examples: look for verbs that
can be either transitive or intransitive and concoct a pair of sentences on the
model of (1.20) and (1.21). Take for example a verb like dress, hide, teach,
etc.; work out two meanings parallel to (1.20) and (1.21), e.g. She dressed an d
fe d the baby. (Again, it must be conceded that commas might well be used to

21
The intonation systems o f English

disambiguate these cases, but commas are part of the written mode of
English; in the spoken mode, we depend on intonation.)
It might also be argued that the equivalent in other languages (like
German, Dutch, French, Italian, etc.) would require the reflexive pronoun in
cases like (1.21); indeed, you could add herself to the English version. But
two points can be made in this connection: first, the English version without
h erself is perfectly normal, and possibly more frequent, and English speakers
are quite content to rely on the intonation to provide the sense; second, what
may be true of another language does not impinge on the linguistic
description of English. What is true of German will be incorporated into the
description of German, not into the description of English or any other
language. Similarly, what is true of English may, or may not, be parallel to
what is true of German, or any other language.
There are many more cases of two grammatical structures being
disambiguated by intonation in the spoken form of English, and these will
be dealt with in the course of Chapter 2. But the case of the reflexive
pronoun is also involved in a different kind of grammatical structure. Also,
the two illustrations of grammatical contrast so far involve tonality: one
intonation unit in (1.8) and (1.20), but two in the parallel wordings of (1.9)
and (1.21). In (1.22) and (1.23) the grammatical contrast is apparent in the
tonicity:
(1.22) He asked himself
(1.23) He asked himself
In (1.22), him self is a reflexive pronoun functioning as the direct object
complementing asked, thus asked is transitive, and him self is reflexive. In
(1.23) , however, it is an emphatic pronoun; (1.23) means that he himself did
the asking, he did not leave the asking to anybody else. Thus, in (1.23), asked
is intransitive, and him self is emphatic. Admittedly, it could be argued that
(1.23) can be expressed differently as He him self asked, but the wording of
(1.23) is perfectly normal, and possibly more frequent, and English speakers
are quite content to rely on the intonation to provide the sense. (Could the
written forms be distinguished to provide the sense in this case?)
Again, you may be able to think of other possible examples; look for verbs
that can be either transitive or intransitive and in their intransitive use could
be reasonably followed by an emphatic pronoun. I recall an occasion when
the hostess at a party was asked while she was in the kitchen if the guests
could begin on the (as yet uncut) cake; since she was not in a position to cut
the cake herself, she called out:
(1.24) Cut yourself

i.e. you do the cutting. The alternative, with tonic on Cut, was most definitely
not intended!

22
r

Intonation: what it is and what it does

Intonation has a grammatical role in disambiguating parallel wordings of

different syntactic structures. Whereas the first three functions of intonation
that we have discussed would be common to all languages (in general terms,
though not in details), this grammatical function may not be so. It is quite
likely that another language would employ overtly grammatical means to
differentiate between all its grammatical systems and not rely on intonation.
English, however, does use intonation for this purpose.

4.5 Textual structure

The fifth major function of intonation concerns longer structures than single
units of intonation and individual pieces of information. Obviously, an
intonation unit and a piece of information do not usually appear in isolation.
We have dealt with them in isolation for the most part in order to illustrate the
informational, communicative, attitudinal and grammatical functions without
any distracting complication. But we have already examined a bit of larger
text in (1.3) with its eight complete units and two abandoned units. The units
did not appear in isolation but in combination with each other as part of the
staging in the management of the message. How do these separate, discrete,
units of intonation hang together to form the text of discourse?
Naturally, the topic of the message is one factor that binds the information
together; another is the grammatical systems of reference and conjunction
that show that clauses and sentences belong together; and a third factor is
intonation.
To illustrate how intonation performs this function, think first of all of
how you know when one item of news has finished and a new one begins,
in newsreading. No one tells you, but you know. A new item usually starts on
a fairly high pitch: the first onset syllable is high and the general pitch level
of the whole of that intonation unit is relatively high. When that item comes
to an end, the general pitch level of its final intonation unit is relatively low,
the tone will fall to its lowest pitch and there is often a slackening of pace in
the final few words. A noticeable pause signals the end of that item. Then
comes the next item, and the first onset syllable of the initial intonation unit
of the new item is high and then comes a gradual descent in the general
pitch level of the following units until the lowest point is reached with the
final unit. This pattern of high start, gradual descent and low finish is typical
of newsreading and is detectable in other spoken genres, especially where
the discourse has been rehearsed or the semantic content prepared, as in the
telling of stories and jokes. The combination of pitch descent and pause
serves to combine units together but serves also to detach the final unit of
one item from the initial unit of the next.
This phenomenon has been called phonological paragraphing. It should
not come as too much of a surprise that there is an equivalent paragraph

23
 The intonation systems o f English

structure in spoken discourse to that found in the written mode. Paragraphs

in writing usually indicate the division of the larger text into separate smaller
topics. The same kind of division is, not unnaturally, found in spoken
discourse too. The intonation pattern described above is thus the equivalent
of starting a new paragraph on a new line. (Brown (1977) coined the term
paratone' in contrast to paragraph'?')
The key to phonological paragraphs is:

(1) The high pitch on the onset syllable of the initial intonation unit.

(2) The relatively high baseline of that initial unit; this means that the low
pitches are relatively high, compared to the low pitches in the final unit
of the paragraph.

(3) There is a gradual lowering of that baseline until the final unit is
reached.

(4) The depth of fall in the final unit is the lowest in the whole paragraph.

(5) There is usually a slowing down process in the final'unit.8

(6) There is a longer pause than is normally allowed between intonation

units.
7
"X .
/ ]
'l
I regret + putting the people out of the out of the South Side and central Edinburgh you ;
155 200 150 115 190 160 120 115 120 105 105 120-105 120-105 110 120-115 120-100 105 i

--------------------------------- ----------------------------------------------------- ---------------. J-

:: T b v ~ z - : = -------------------------------------
-"'"r
know + + 0.86 I dont think+ + 1.8 especially after the war you know after the + + 0.64
140.100 145 175 100 165-115 140-110 150 110 120 150 100

war when they started the + + 0.32 redevelopment and the + + 1.00 well the authority more or less
120 120 100 130-100 100 105-120-100-120 100 110 120 135 165-100 120 100

---------------------
----------------- ------------ - = =
made . it that everybody was to go outside you know + + 0.68 the gardens and houses but
140 100 125-100 100 100 100 130-95 130-100 90 130-110 100 120-90 120

Figure 1.1 A visual display of the information of extract (1.3). Original

information retained from Brown, Currie and Kenworthy (1980: 66) on
prominence (i.e. underlined syllables) and frequencies

24
 Intonation: what it is and what it does

Listen to the news and see if you can identify these six factors. You should
also be able to detect them in story-telling, and even in jokes. It is also
possible to find them in the more unfluent, spontaneous discourse of
informal conversation, although not so clearly. Figure 1.1 is a visual display
of the snippet of conversation (1.3) that we have examined before on a
couple of occasions; you will notice how high the initial onset syllable in
regret is pronounced in the initial unit of this paragraph; that d o n t is slightly
lower; especially lower again and w ar even lower; then authority is high
again, as the beginning of the new topic.
Brazil, Coulthard and Johns (1980) also provided specimens of recorded
text. Here is part of their Transcript A, arranged in phonological paragraphs
and with a simplified transcription: H = high baseline (or key); M = mid
baseline; L = low baseline. A teacher is addressing his class in the middle of
a lesson and wishes to introduce the subject of energy.
H 'Put vour oens
M 'down I'Pencils
L 'down II
H 'Now Ibe'fore 1came to
M '$ C h Q O l 1
L 'this 'morning 1
H 1had some 'cereal 1
M
L 1'had my 'breakfast II
H
M and 1had some 'toast 1and 1had an 'egg 1and
L
H
M 1had a cup of 'tea 1and 1had a 'biscuit 1
L
H
M and then 1came to
L school II
The first phonological paragraph simply consists of two intonation units,
and you see the progression from high to low. (The teacher actually followed
this up with a few more similar instructions to obtain the pupils attention.)
Then comes another paragraph which gives an outline of the immediate
topic, with a clear progression from high to low. The final unit of that
paragraph is immediately followed by a unit on a high baseline, and so you
can recognize that a new paragraph has been initiated. This third paragraph
carries on with a number of mid-baseline, units before a low one is reached;
this is one way of extending the length of a phonological paragraph.
This system of gradually lowering the baseline in successive intonation
units is an effective Way of indicating that the units belong together. But this
system can be interrupted to produce other effects. If a following unit is

25
The intonation systems o f English

pronounced on a higher baseline, against the expectation of a lower one, the

effect produced is one of contrast or surprise:
(1.26) H 'Wales played 'beat them 'one 'nil
M Bra'zii I and
L
If a following unit is pitched on the same baseline as the previous one, it
usually means that the second unit is deemed to be an addition to the first,
and that could be the interpretation of the sequence of mid-baseline units in
the example above, referring to items taken for breakfast.
If a following unit suffers a sudden drop in pitch, that has the effect of
indicating that the information was deemed to .be fully expected:
(1.27) H The 'Monster 'Raving 'Loony 'Partys 'candidate I
M
L 'lost his deposit
You will no doubt have noticed how the pitch variations in these
sequences of intonation units match the attitudinal function in individual
qnits: high for unexpected, mid for neutral, low for fully expected.?
However, there is a familiar, additional, meaning for low-pitched units
embedded in a higher-pitched context, namely asides, glosses, and the like:
(1.28) H And 'then he ac'cused me 1
M
L 1dont know 'where
H of 'stealing II
M
L he got the i'dea from 1

(1.29) H But then the 'next 'student 1

M
L a 'really 'clever 'bloke 1
H got 'everything 'right II
M
L

(1.30) H Hes 'coming 1

M to'nieht II
L he 'said 1

4.6 The identification of speech styles

The final major function of intonation that we need to consider is its role in
the way in which we can identify different speech styles, or genres. People
engage in a very wide range of differing language events; informal

26
 Intonation: what it is and w hat it does

conversation is one type of language event and is very different from other
types of dialogue like interviews, debates, interrogations, air-to-ground
communication by pilots, and so on. Monologues vary considerably too;
compare, for example, newsreading and prayer, or story-telling and poetry
reading, or a comedians monologue and a lecture!
If you switch the radio on (or the television - before the screen clarifies),
you can usually tell within seconds what kind of language event is taking
place. This is because newsreading, for example, somehow sounds different
from all other styles. We probably differentiate dozens of different styles
simply on the basis of the general sound of even just a few seconds of
spoken discourse. And we can usually manage this even if the actual words
are muffled, as they might be, say, in an adjoining room. There is something
about the general sound of particular language events that identifies them.
This general sound of a particular language event is known as its prosodic
composition. Differences in prosodic composition depend on a number of
features: degree of formality, number of participants, degree of privacy,
degree of semantic preparation, and whether the spoken discourse was
scripted or not. These features register in intonation, loudness, tempo,
rhythmicality, paralinguistic features and hesitation pauses. Rhythmicality
refers to degrees of rhythmic regularity, ranging from very regular, as in
poetry reading and prayer in unison (e.g. the Lords Prayer, said in public),
to irregular, as in informal, unfluent, conversation.
As far as intonation in particular is concerned, styles vary in the
proportions of falls and rises, relative length of intonation units and degree
of textual structure (phonological paragraphing). Prayer, noticeably,
manages without tone variation as a rule; tonic syllables are held level and it
is usually only the Am en that has a falling tone.
The fullest study to date on the prosodic composition of speech styles^
compared the following genres:

(i) informal conversations (private, unscripted)

(ii) informal narration of an anecdote (private, unscripted)

(iii) a news bulletin (public, scripted)

(iv) Bible reading in church (public, scripted)

(v) individual prayer (public, unscripted)

(vi) prayer in unison (public, scripted, but uttered by a body of people

simultaneously)

The prosodic composition of these six genres is displayed in Table 1.3:

27
The intonation systems o f English

Informal Anecdote News Bible Individual Prayer

conversation reading reading prayer In unison
p a ra g ra p h in g Yes Yes Yes
p u n ctu a tio n group n/a n/a Yes Yes n/a Yes
s h o rt u nits Yes Yes Yes Yes
to n e va ria tio n Yes Yes Yes Yes
high p ro p o rtio n
o f fa lls Yes
high p ro p o rtio n of
pre-tonic variation Yes Yes
forte Yes Yes
lento Yes Yes Yes
rhythm ic Yes Yes Yes
p a ra lin gu istic
fe a tu re s Yes Yes Yes Yes
hesitation Yes Yes Yes

table 1.3 The prosodic composition of different genres

A few points emerge. The recognition of phonological paragraphs is most

apparent in semantically-prepared discourse. It is not that it is impossible to
find paragraphs in other styles, but in semantically-prepared discourse it is at
its clearest. Secondly, intonation unit boundaries correspond to punctuation
in scripted discourse. Thirdly, the intonation units themselves tend to be
longer in semantically-prepared discourse, except where the delivery style is
fairly slow (lento) and loud (forte); units are generally shorter in informal
conversation when the message is being composed and relayed almost
simultaneously. Fourthly, there is a high proportion of falls in informal
conversation; this means that there are more individual, major, pieces of
information in that genre than, for instance, in narrating an anecdote or in
newsreading; in the latter styles, the speaker has a much stronger sense of
linking incomplete pieces of information with complete (see Chapter 4).
Fifthly, and possibly surprisingly, public, scripted, spoken discourse is
characterized by greater pitch variation in the pre-tonic segment; this
variation, no doubt, is intended to relieve any impression of dull, routine
repetition that could bore the hearers. Sixthly, newsreading eschews
paralinguistic features altogether: the newsreaders feelings are not allowed,
in the main, to intrude into the news. Prayer in unison shares this lack of
paralanguage, for the simple reason that a body of people simultaneously
engaged in speech do not have the opportunity to introduce their own
feelings. Finally, and expectedly, hesitation features in those genres that are
unscripted, even if a degree of semantic preparation has taken place.
Thus, it can be seen that intonation is a major identifying factor in the
prosodic composition of different genres of spoken discourse. This
particular function operates on a much larger body of discourse than the
other functions, but it is nevertheless recognizable.

28
 Intonation: what it is and what it does

5 . Conclusion
We have just considered six major functions of intonation in English. All of
them, except possibly the grammatical function, will manifest themselves in
all languages. A seventh function might occur to you, and indeed was
alluded to on the first page: a sociolinguistic function, in describing,
comparing and contrasting one accent (or dialect) with another. Such a
function relates to the user rather than the use of language, and on that score
will not be considered further in this book, interesting though such a subject
is. However, you cannot describe, compare and contrast the intonation
system of two accents until you have a basic framework for the description
of one. That is what this book will seek to provide.
Similarly, intonation can function in psycholinguistics, too. The intonation
of baby speech is quite different from childrens and adult speech. The
sociolinguistic and psycholinguistic aspects of intonation study have been
labelled the indexical function11 in the sense that they identify
characteristics of the speaker rather than of the message and the discourse.
Research into the indexical functions of intonation and indeed into the
genre and textual functioning is still at a relatively elementary stage and will
not feature significantly in this book.
This book will concentrate on presenting the body of knowledge that has
been established on the functions of intonation in conveying the meaning
of messages:

what is being said (information)

why it is being said (communicative functions)

how it is being said (attitudes)

which is being said (grammar; differentiation between syntactic choices)

It will deal with the structure and phonetic details of English intonation as
they form part of the three systems of tonality, tonicity and tone.

Notes
1. See, for example, Roach (1983:121).
2. See Halliclay (1967), Pike (1967) and Tench (1990).
(1.5) head tonic tail
'What are you 'going to do tonight
(1.6) pre-head head tonic tail
Im 'going into 'town this 'morning
(1.7) pre-head head tonic pre-head tonic i
Im 'going into 'town this 'morn i

29
 The intonation systems o f English

/'N
oq
pre-head head tonic tail

H
My 'brother who 'lives in Nai 'ro bi
(1.9) pre-head tonic tail pre-head head tonic tail
My 'bro ther who 'lives in Nai 'ro bi
4. See the analysis of this snippet of conversation in Tench (1990:181f.).
5. For a fuller discussion, see Tench (1990: 392-8; 440-6).
6. See Sinclair (1972) and Young (1980).
7. Brazil and Coulthard (1979) used the term pitch.sequences and demonstrated that
this kind of textual structuring exists not only in monologue but in dialogue, too.
8. Lehiste (1975, 1979) refers to this as pre-boundary lengthening.
9. See Brazil, Coulthard and Johns (1980: chapter 2) for a fuller discussion of this
system which they call key.
10. See Tench (1990: chapter 7; also 1988). See also Crystal and Davy (1975) for further
details of the characteristics of informal conversation, and Johns-Lewis (1986) for a
comparison of conversation with acting and reading aloud.
11. See Couper-Kuhlen (1986).

30

Ml
 Tonality
TH E U N ITS OF IN TO N A TIO N

1. Neutral tonality
Tonality is the system in intonation that divides spoken discourse into its
separate individual intonation units. Spoken discourse sometimes consists
of only one message or piece of information; indeed it might consist of only
one word, even only one syllable, e.g. Yes, No, Right, So?, etc. But it usually
consists of more than one piece of information, and often extends for as
much as an hour, or more, either as a monologue or as a series of turns in a
dialogue. The whole text is composed of many intonation units, each
bearing a single piece of information and representing the speakers
management of the information of the whole message. To give some idea of
quantities, a two-minute news bulletin has usually between 70 and 80
intonation units; admittedly, the pace at which they are delivered is fast in
newsreading, but even in relaxed informal speech, there would be
approximately 25 units per minute.
Each intonation unit contains one piece of information - as the speaker
perceives it. It is important to emphasize the perception and management of
the organization of information by the speaker. It is very easy to look at the
transcript of a record of speech and think you would have organized it
differently from the original speaker. You could well be right and if you had
had to give the same message using the same words you might well have
divided the material differently: but your division (or segmentation)
represents your perception, whereas the tonality of the original represented
that persons perception. (It is a recurring problem when introducing
intonation to get people to concentrate on one particular issue in intonation,
when they can think of many alternative renderings!)
Halliday also drew attention to the observation that very often intonation
units coincide with clauses, and this provides a very valuable starting-point
for the description of tonality. If a person wants to give a piece of
The intonation systems o f English

information, it is conveyed as a single unit of intonation, but it has to be

worded, and this is where grammar comes in. The clause is the most obvious
unit of grammar to handle a typical piece of information: the subject
represents the theme - what the message is about; the predicate represents
the rheme - what the message actually is. For instance, the syntax of
(1.10) A dog is a mans best friend

tells you what the theme is, i.e. the subject of the clause, a dog-, and then tells
you what the actual message is, i.e. the predicate, is a m a n s best fr ie n d (the
rheme).
It should be noted, however, that this coincidence of tonality and clause is
by no means a hard and fast rule. However, in Hallidays words (1967:18-19),
there is a
tendency for the tone groupi to correspond in extent with the clause; we may
take advantage of this tendency by regarding the selection of one complete tone
group for one complete clause as the neutral term ... That is to say, a clause that
consists of one and only one tone group will be regarded as neutral in tonality.

It is useful theoretically to lay a foundation for tonality on this basis: that a

single piece of information is handled in the grammar, typically by a single
clause, and in pronunciation by a single intonation unit.
It must also be emphasized that this is not a case of squeezing facts into
ready made theories. In one count,2 916 intonation units, out of a total of
1,880 in the whole data, corresponded to complete clauses - almost 5Q per
cent. And here is part of the transcript of a conversation between two
women on the subject of Bonfire Night, recorded by Crystal and Davy
(1975); note the frequency with which intonation units embrace the main
elements of the clause: there are nine clauses in the first eighteen completed
intonation units,
(2.1) A: were'looking'forward to \bonfire night
at \ least
the v children 'are
- - do vou inxdulge in 'this -
B: oh in . in v Sussex we 'did
- - Ive - in \fact
I 'went to 'one 'last x week
. b u t. it was .
A: xxx
'that was a bit /early
7wasnt it

32
 Tonality: the units of intonation

B: 'all the 'all the x iovs were

x]oy
was 'taken out of it
for / me
be'cause it was - a x huge
. \ bonfire
in a 'garden the 'size of this v room
- with big \houses
'all a x round
- and the 'bonfire was 'right x under
with its 'leaves a all 'dry
The concept of neutral tonality is a very useful starting-point. It embraces
a very important point: the functional equivalence of intonation unit, clause
and unit of information, linking up phonology with grammar and semantics.
Furthermore, this functional equivalence is realized to a large extent in
actual discourse.
Here is another text from Crystal and Davy (1975: 44) without the
intonation marks. See if you can guess what the intonation units would be;
mark the boundaries with a |; a dot represents a very brief pause; a hyphen,
or hyphens, relatively larger pauses; bracketed words represent the listeners
back channel.
(2.2) yes I remember there was a terrible story - horrifying story that
was told by a colleague of mine when I used to teach years ago -
who erm . this chap lived in erm a - a semi deta . detached house
and next door there was . a man whod just bought a new car -
and he was telling me that one morning he was looking through
the window - and this . man allowed his wife to drive the car very
unwisely and she was having a first go in it - (m ). and - he backed
it out of the garage - so that it was standing on the driveway . and
hed closed the garage doors - (yeah). and - she came out of the
house - to . take this car out and go shopping for the first time -
so she came out very gingerly - and opened the d o o r. and sat in
the car - and - e r . began to back . very very gently - taking great
care you see that she didnt do anything to this . to this new car -
and - - as she backed - there was an unpleasant. crunching
sound (la u g h s ) and she slapped on the brakes and looked around
frantically - and realized that she hadnt opened the . gates . th a t.
let on to the main road you see (o h ). and shed just backed into
these
In my experience, a group of people doing the same exercise would show
a degree of conformity of about 8G per cent. But why wouldnt they all agree
in the whole passage? There are two main reasons for individual diversity:
one is the imagined speed of delivery: the slower the pace, the shorter the

'33
The intonation systems of English

unit; and the second is exactly what we discussed above: peoples

perception of the organization of information will vary.

2. Marked tonality
We must now try to account for all those cases where intonation units do not
coincide with whole clauses. Marked tonality, i.e. any case that does not
correspond to neutral tonality, occurs either when two (or more) clauses fit
into a single intonation unit, or when - and this is much more frequently the
case - two (or more) intonation units are needed to cover a single clause.
The most obvious case is when a speaker presents two (or more) pieces
of information in a single clause as in (1.7):
(1.7) I'm going to town I this morning

Quite possibly, this m orning is added as an afterthought, as we mentioned

in Chapter 1. Examples can be found in the bonfire text: with big houses I all
around, two units added as afterthoughts to an already long clause.
On the other hand, two clauses may fit into a single intonation unit, as we
saw in the case of (1.4):
(1.4) He did. I saw it.

H e did, in this case, is said quite quickly, as given information; the new
information is contained in the second clause (or sentence, in this case). A
different case occurs where two verbs share the same complementing
object, as in (1.20); strictly speaking, that sequence constitutes two clauses.
Similarly, the verbs come, go, try and stop frequently precede another verb in
combinations like com e a n d see, go a n d look f o r it, try a n d stop me, stop a n d
think.
Two other, odd, cases could be mentioned here too, since they might look
like tonality contrasts. You could say the following words, He spoke to me
honestly, as a single clause and in a single intonation unit:
(2.3) He spoke to me honestly
Or, you could treat honestly as a separate unit:
(2.4) He spoke to me I honestly
in which case honestly is like an appeal to the addressee to believe the claim.
(It might be written, in fact, as two sentences: He spoke to me. Honestly!) This
distinction was once used in an advertisement:

HOW TO MAKE A FORTUNE

HONESTLY

By placing honestly on a separate line, the advertisers were able to exploit a

34
 Tonality: the units of intonation

feature of intonation successfully, since they knew that people would read
the wording in two ways and thus they got two messages across in one
advertisement.
Another similar example is the exploitation of two meanings of the word
so in the following: The read er will fin d them unusual a n d so puzzling. So
could be an intensifier like very.
(2.5) The reader will find them unusual I and so puzzling
but so could be a substitution of a clause like thus or consequently, meaning,
in this case, because they are unusual ( the read er will fin d them puzzling)-.
(2.6) The reader will find them unusual I and so I puzzling
(If you know phonetic transcription, you might find an accidental clash
between
(2.7) ai 'daunt 'nau
and
(2.8) ai 'daunt I 'nau
The first is a single clause; the second, a sequence of two.)
Other cases of marked tonality relate to the length and the structure of the
clause concerned.
Intonation units have an average of between two and three feet each, i.e.
two or three word stresses. Often, as we have seen, a unit might consist of
only one stressed syllable. The usual maximum number of stresses in a unit
is five, and this corresponds to the maximum number of elements in a single
simple clause: subject, verb, direct object, indirect object and adjunct, e.g.:
(2.9) The children wrote 'postcards to their 'parents from 'camp

(2.10) The 'office 'sends the 'students their 'grants in Oc'tober

This sentence could be said as one unit of intonation - and as one unit of
information - but it reaches the maximum of five feet. If a clause breaches
that maximum, it seems that it is automatically converted into two (or even
more) intonation units: for example (2.11):
(2.11) The 'train ar'riving at 'platform 'two I is the 'ten 'twenty from
'Swansea to 'London 'Paddington
The tonality division depends on the structure of the clause. In (2.11), the
subject is long enough to be separated off; in informational terms, the theme
has been separated intonationally fromlhe rheme, and this is very often the
case. This is illustrated in the bonfire text (2.1):
'all the 'all the 'joys were 'joy I was 'taken 'out of it for 'me
The first unit is the theme, the second the rheme: this is clear despite the
added complications of repetition and reformulation.
35
The intonation systems o f English

3. Lists
A list is a special kind of long clause. Each item is contained in a separate
unit,3 e.g.:
(2.12) He could speak English I French I and German
(2.13) She ate some soup I some bread and butter I and a piece of cake
Notice that b rea d a n d butter is treated in (2.13) as one list item; the speaker
perceives it as one piece of information. However, in another context, bread
and butter might be perceived as separate items, e.g.:
(2.14) We need milk I some bread I and butter
There is a potential for contrast here, as Halliday (1970: 36) has pointed
out: compare (2.15) and (2.16):
(2.15) We want red I white I and blue flags please
(2.16) We want red white and blue flags please
(2.15) refers to three kinds of flags whereas ( 2.16) refers to one kind of flag
with three colours. (2,16) is in fact an example of neutrgl tonality since the
intonation unit contains the whole clause.
However (2.12), (2.13), (2.14) and (2.15) all display marked tonality since
more than one unit was required for the list clause; each list item is treated
as one piece of information, but they are all contained, grammatically, in one
single clause.
Crystal (1969: 264) drew attention to a different kind of example involving
idioms with or. M ore or less is an idiom that means just about, roughly,
ndarly. Compare - talking of the potential punctuality, say, of a new train -
(2-17) Will it be more or less on time?
and
(2.18) Will it be more I or less on time?
The second case is rather like a list of two alternative possibilities. Similarly,
on e or two can be taken literally or as an idiom meaning some.
Compare - when offering a warm drink -
(2.19) Do you want one or two lumps of sugar?
and
(2.20) Do you want one I or two lumps of sugar?

Where you have the extra tonality boundary, you have what amounts to a list.

36
 Tonality: the units of intonation

4 . Marked theme
Marked tonality is also occasioned by deviations to the structure of a clause.
One deviation is to have a clause element preceding the subject with the
effect that that element becomes the theme, instead of the subject, as in
( 2.21):

(2 .2 1 ) This morning I Im going into town

The theme - the starting-point of the message, - is no longer I, but this
morning. This deviation is known as marked theme as opposed to the more
usual, neutral, theme, in which the subject, coming first, is the theme; if
anything (apart from conjunctions) precedes the subject, then that takes
over the role of theme, it is the starting-point of the message.4 Cases of such
marked theme always have a separate intonation unit, and thus there are
bound to be at least two units for such clauses. Consider these cases:
(2.22) Last night 1vou came in too late
(2.23) Whatever your excuse I you should be in by eleven
(2.24) If you go out in the evening I I want you in by eleven
(2.22) and (2.23) are cases of marked theme producing marked tonality.
(2.24) , in fact, has a clause as marked theme, and is thus an example of
neutral tonality, because each clause, the dependent one and the main one,
has its own intonation unit.
Other typical cases of marked theme in a single clause are:
(2.25) These ideas I you must reject
(2.26) Not until then I will I even think about it
(2.27) Scarcely even then I are you likely to find support
(2.28) Into my mind I came the thought of rejecting it

5. Adjuncts
The other kind of structural deviation that affects tonality is the addition of
certain kinds of adjunct either at the beginning or in the middle of a clause;
these items generally have their own intonation unit. If these adjuncts
appear at the end of the clause, they may either have an intonation unit of
their own or be incorporated into the unit of the preceding clause; this
choice of tonality, in the final position, affects tonicity also and so will be
considered more fully in Chapter 3. Here, we will consider them in initial
and medial positions, because in these positions they do affect tonality.
The adjuncts that affect tonality include linking adjuncts like however,
nevertheless, perhaps, o f course, unfortunately, etc. Consider (2.29) and
(2.30):

37
The intonation systems of English

(2.29) However I he ran the mile in four minutes

(2.30) He ran the mile I however I in four minutes
From the bonfire transcript (2.1) we have:
at least I the children are
in fact I I went to one last week

Also vocatives:
(2.31) Miss Smith I can you help me?
(2.32) Can you I Miss Smith I help me?
and viewpoint adjuncts like politically, technically, fro m a p erso n a l p oin t o f
view, etc.:
(2.33) Politically I the decision was a very clever one
(2.34) The decision I from a political point of view I was a very clever one.

6 . Tags
Clauses may be concluded with a tag as in its hot, isn t it. The tag has a
particular form: it consists only of an operator and a subject pronoun, it is
always in inverted order, and tense and number are determined by the main
clause. Tags may be either positive or negative. If the polarity of the tag is
reverse to that of the main clause - in other words, if negative follows
positive, as in the above example, or positive follows negative, the tag is
called a checking tag. If the polarity of the tag follows that of the main
clause - positive following positive (or, for some people, even negative
following negative), then the tag is called a copy tag.
The vast majority of tags have their own intonation unit. Example (1.2)
should therefore be transcribed with the following tonality,
(1.2) Johns going out 1isn't he
An example occurred in the bonfire transcript (2.1):
that was a bit early I wasnt it
Checking tags, i.e. those with reverse polarity, seem always to have
separate intonation units for themselves, and certainly from the intonational
point of view should, therefore, be regarded as separate clauses^ Thus, a
tagged clause, viewed as a sequence of two clauses, follows the general rule
of neutral tonality.
Copy tags, i.e. those with identical polarity, act somewhat differently. For
one thing, copy tags can only be accompanied by a rising tone, whereas
checking tags can be accompanied by either a falling or a rising tone, with a

38
 Tonality: the units o f intonation

consequent change of meaning. You can check this for yourself: you can say
(1.2) with either a rising or a falling tone on the tag, as you discovered in
Chapter 1; but you can only say the tag of (2.35) with a rising tone:
(2.35) Johns going out I is he
If you try the is he with a falling tone, it will sound distinctly odd.
The second difference from checking tags, is that copy tags do not
necessarily require their own, separate, intonation unit. You could say (2.35)
with a falling tone on out and then keep the pitch of the voice low without
allowing a rising tone:
(2.36), John's going out is he
Notice that there is quite a difference in meaning between checking tags
and copy tags. The use of checking tags implies an actual checking of the
proposition in the main clause: the speaker is pretty certain of the validity of
the proposition if the tone on the tag falls, whereas the speaker is less certain
about the proposition and needs confirmation if the tone rises - in both
cases, the proposition is being checked. The copy tag implies more of a
realization - even a sudden realization - of the significance of the
proposition; if there is no separate tonality for the copy tag, as in (2.36),
there is added a note of irritation.
Work out all the meanings with a different example:
(2.37a) They lost against \ Newport of all teams I /didnt they
(2.37b) They lost against \ Newport of all teams I \ did nt they
(2.37c) They lost against \ Newport of all teams I /did they
(2.37d) They lost against \ Newportof all teams did they
Try out these variations with examples of your own for (a) a degree of
uncertainty, (b) certainty, (c) a (sudden) realization of the significance of the
information, and (d) with a touch of irritation. Note that whereas (a), (b) and
(c) have two intonation units, the tonality is different for (d).

7. Tonality contrasts in grammar

We have constantly been making the point that changes in the tonality in an
utterance signal changes in the organization of information as perceived by
the speaker. We have thus been highlighting the role of tonality in the
informational function of intonation. Even the cases of marked tonality
relate to the way the speaker is managing the organization of information,
by having a theme different to the subject of the clause and by adding
adjuncts that comment on the information or relate to the addressee.
But we have also seen that intonation in English fulfils a grammatical

39
The intonation systems o f English

function too. First of all, intonation units often match clauses, even to the
extent that the boundaries of intonation units and clauses coincide. Look at
this example and think out the way you may say it:
(2.38) When you enter the waiting room is on the left

Did you need two tries? Most people do. Most people read it first as if
the waiting room is the direct object complementing the verb enter, and
then realize that that cannot be right because then there is no subject for the
verb is.
On a second try, you prepare yourself to treat the waiting room as the
missing subject of is, and that leaves you with enter as an intransitive verb
and as the end of the first clause. On this second try you would say when you
enter as one intonation unit with a boundary before the waiting room , and
thus the boundary between the two units coincides with the boundary
between the two clauses. (We concede, as we did in Chapter 1, that a comma
after enter would make the clause boundary clear; but we point out, as also
we did in Chapter 1, that commas belong to the written mode, but that it is
intonation that performs such a function in the spoken form, and secondly,
that writers are notoriously inconsistent in their use of commas in any case.)
The second way in which the role of intonation is seen in English
grammar is in disambiguating parallel wordings that contrast in syntactic
structure. We discussed two examples in Chapter 1 that involve contrasts in
tonality: (1.8) and (1.9), and (1.20) and (1.21).

7.1 Defining and non-defining items

Remind yourself of the difference between (1.8) and (1.9):
(1.8) My brother who lives in Nairobi |...
(1.9) My brother |who lives in Nairobi |...

In (1.8), the relative clause defines which brother is meant, i.e. not the one
who lives anywhere else; it restricts the reference. The relative clause in (1.9)
does not define which brother is meant; it adds extra information; it does not
restrict the reference, and you might justifiably assume that there is in fact
only one brother to refer to; as we have previously noted, such non-defining
clauses could well be labelled adding clauses. In spoken discourse, it is the
tonality - part of the intonation system - that differentiates between the
identical wording of the two cases. (Remember: any comma belongs to
writing.)
The distinction between defining and non-defining extends to phrases,
too. Compare
(2.39) The man dressed in black |(then stood up)
(2.39a) The man |dressed in black |(then stood up)

40
 Tonality: the units of intonation

and again
(2.40) The man in black |(then stood up)
(2.40a) The man | in black |(then stood up)

The tonality break after m an in (2.39a) and (2.40a) indicates that the
following clause or phrase does not define the man, but adds extra
information - which might be incidental or highly relevant.
Now compare
(2.41) The man with the dog sitting near the bus stop |(is in trouble)
(2.42) The man with the dog |sitting near the bus stop |(is in trouble)

The second, (2.42), seems to suggest that it is the man who is sitting near the
bus stop, as additional, non-defining information; but the first, (2.41),
strongly suggests that it is the dog that is doing the sitting, because the sitting
defines which dog is meant!
See how many meanings you can work out in the next example:
(2.43) The man. and the woman dressed in black (then stood up)

If you have no intonation unit boundaries until then, then you are
defining the man and the woman by distinguishing them from other men
and women who are not dressed in black. If you have a boundary after man,
then the defining clause is limited to the woman-, in other words there must
be more than one woman in the situation, but not necessarily more than one
man. And if you have a boundary only after wom an, then the following
clause dressed in black will be interpreted as an additional, but not a
defining, piece of information. Compare them for yourself:
(2.43a) The man and the woman dressed in black |(then stood up)
(2.43b) The man |and the woman dressed in black |(then stood up)
(2.43c) The man and the woman |dressed in black |(then stood up)

It may appear complicated at first, but it is in fact a straightforward matter of

intonation performing a grammatical function consistently and affecting the
meaning consistently too. \

7.2 Apposition
Apposition is the relationship between two or more items which are either
identical in reference or else the reference of one must be included in the
reference of the other. In
(2.44) Tom Jones, the singer, comes from South Wales

there is identity of reference between the two appositional items Tom Jon es
and the singer. In

41
The intonation systems o f English

(2.45) The Government, the prime minister and his cabinet, are pushing
for a quick decision

the p rim e m inister a n d his cabinet is included within the notion of The
Government.
There are two ways of saying the first example, and the difference is
parallel to that between defining and non-defining items. You can say it as
(2.44a) Tom Jones |the singer |comes from South Wales

in which case the singer is added as extra information. However, in

(2.44b) Tom Jones the singer |comes from South Wales

the singer does identify which Tom Jones is meant, since in South Wales
there are hundreds of Tom Joneses; the Tom Jones of Fieldings novel is not
meant either.
The distinction between defining and non-defining apposition is thus
indicated solely by intonation in the spoken mode, and consistently at that. (In
the written mode we note, again, an inconsistency in the use of commas.)
However, it might also be noted that non-defining apposition can be
introduced by a formula such as that is, that is to say, in other words, etc., as in
(2.44c) Tom Jones |that is the singer |comes from South Wales

Such formulas do not affect the intonation; the tonality remains the same for
non-defining items and the formula is incorporated into the intonation unit
accompanying the second item.
Quirk et al. (1972: 626-7) cite three cases of potential ambiguity between
instances of apposition and the more complicated types of
complementation. In speech, disambiguation is effected through intonation.
The first example is this:
(2.46a) They sent Joan a waitress from the hotel

As a single intonation unit, this would be interpreted as a waitress fro m the

hotel being sent to Joan, i.e. as a double transitive clause, Jo a n as indirect
object, a waitress fro m the hotel as direct object. But can you think of other
ways of saying that clause? What if you put an intonation unit boundary after
Jo a n !
(2.46b) They sent Joan |a waitress from the hotel

Now, in this rendering, Jo a n is the waitress from the hotel. But what if you
add a second boundary after waitress?
(2.46c) They sent Joan |a waitress |from the hotel

In this rendering, J o a n is still a waitress, but now she is sent away from the
hotel. Tonality is doing both grammatical and informational work in this
case, but it should make clear how intonation is sometimes absolutely crucial

42
 Tonality: the units o f intonation

in conveying a concise message.

Here is Quirk et al.s second example:
(2.47a) They considered Miss'Hartley a very good teacher

As a single intonation unit, this would be interpreted as a clause with a direct

object complementing considered, and a further object complement to Miss
Hartley, meaning: Miss Hartley was considered a very good teacher. But in
(2.47b) They considered Miss Hartley |a very good teacher

you have a case of non-defining apposition - an extra piece of information

is added.
Their third example is
(2.48a) They called Susan a waitress
As a single unit, this is parallel to (2.46a), so that Susan is an indirect object,
and a waitress is a direct object, both complementing called. But in
(2.48b) They called Susan |a waitress
a waitress is, once again, a non-defining appositional item.
Apposition may take the form of clauses as well as noun phrases, and in
many cases clauses in apposition become relative clauses. Here are some
examples from Quirk et al. (1972: 645-8):
(2.49) They put it where it was light |where everybody could see it
(2.50) He told them the news |that the troops would be leaving
(2.51) Their solution |to appoint a committee |is deplorable
(2.52) For them to pay him a commission | his suggestion |seemed an
excellent idea
These are all cases of non-defining apposition. An instance of defining
apposition corresponding to (2.50) might be
(2.53) The news that the troops would be leaving | has been announced
The clause that the troops would be leaving defined which news was to be
announced, but in (2.54) it does not:
(2.54) The news |that the troops would be leaving | has been announced
Apposition may also take the form of other clause elements; for example:
(i) predicates (see Quirk et al., 1972: 645-8):
(2.55) They summoned help |called the police and firg brigade
(ii) complements:
(2.56) She is better |very much better
(iii) adjuncts:
(2.57) Thirdly |and lastly |they would not accept his promise

43
T he intonation systems o f English

(2.58) They bought it cheaply |for three dollars

Finally, a few other cases where differences in tonality affect the
interpretation of syntactic structure:
(2.59a) John looked a curious man (subjective complement)
(2.59b) John looked |a furious man (apposition, implying because he
was ...)
Similarly
(2.60a) He died a happy man (subjective complement)
(2.60b) He died |a happy man (apposition, implying hed always been a
happy man)
Also
(2.61a) I rested content (subjective complement)
(2.61b) I rested |content (apposition, implying because I was content...)
(2.62a) Bill left very angry (subjective complement)
(2.62b) Bill Jgft |very angry (apposition, implying because Bill was very
angry...)
and finally
(2.63a) That's another thing I dont know yet (ranksh.ifted clause, post
modifying th in g )
(2.63b) Thats another thing | I dont know yet (apposition)

7.3 Verb phrases

A verb phrase can be simple: with a single lexical verb; or complex: with
more than one lexical verb, e.g. (You m ust) try to stop smoking. Some
instances of complex verbal phrases look very similar to a series of verbs, for
instance, compare
(2.64) He began smoking
(2.65) He began to smoke
(2.66) He stopped smoking
(2.67) He stopped [ to smoke

The pair (2.66) and (2.67) may look at first as if they are parallel to the pair
(2.64) and (2.65). (2.64) and (2.65) have different structures but share,
broadly, the same meaning. (2.66) is the real opposite to (2.64) and (2.65),
but shares the structure of only (2.64). (2.67) may look superficially the
parallel to (2.65), but it does in fact have a different structure, and quite a
different meaning. The verb phrases in (2.64), (2.65) and (2.66) are all
complex, but they are each a single phrase in a single clause. In (2.67), we

44
 Tonality: the units of intonation

have two simple verb phrases: the first belongs to a main clause, and the
second to the following purpose clause, which could be reworded as in
order to sm oke, which helps to make it clearer. Because (2.67) consists of two
clauses, the normal, neutral, tonality is expected: two clauses and two
intonation units.
However, there are a few cases where a complex verbal phrase has an
identical wording with a series of verbs in separate clauses. Here is an
example:
(2.68a) He came to hear about it
(2.68b) He came |to hear about it

(2.68a) involves a complex verbal phrase as an idiom meaning it happened

by chance th at..., but because it constitutes only one verbal phrase, it is the
predicate of only one clause; hence only one intonation unit. (2.68b), on the
other hand, has a main clause followed by a purpose clause (rather like
(2.67)), i.e. two clauses; hence, two intonation units.
Here is another example:
(2.69a) He left me to get on with the job
(2.69b) He left me |to get on with the job

(2.69a) has a single, though complex, verbal phrase: one clause, one
intonation unit; (2.69b) is a series of two clauses. You might notice an extra,
subtle difference in meaning in this pair; who, in each case, gets on with the
job?!
Other pairs of examples are:
(2.70a) Bill helped to avoid an accident
(2.70b) Bill helped j to avoid an accident
(2.71a) Bill looked to be on the safe side
(2.71b) Bill looked j to be on the safe side

7.4 Negative domain

Another well-known case of parallel wording of two different syntactic
structures offers entirely opposite meanings. If a main clause has a negative
and is followed by because and a reason, or by so (that) and a result, then
intonation - in the shape of tonality - performs the crucial role of
differentiating the syntactic structures and meanings. Look at the following
statement and try and work out two meanings for it which are entirely
opposite to each other:
(2.72) I didnt come because he told me

45
The intonation systems o f English

How many clauses are there? Two: I d id n t com e and because ... If you
consider the case of neutral tonality first, i.e. two intonational units, each
covering one clause, then the meaning is quite clear: the person did not
come, and a reason for not coming is added:
(2.72a) I didnt come | because he told me

If, however, you run the two clauses together into a single intonation unit,
you get quite a different meaning: the person did go, but not for the reason
that is given - he/she went for some other reason:
(2.72b) I didnt come because he told me

(This would normally be spoken with a fall-rise on told me, and implies: I
came for some other reason.)
It seems to be the case that the domain of the negative extends to the next
tonality boundary; thus in (2.72a), it is com e that is negativized, but in
(2.72b), it is becau se he told m e that is negativized. ((2.72b) could be
rendered as It is not because he told me that I cam e, implying: I came for
some other reason; but note that this is not the most obvious way to give
such a message; (2.72b) is, in fact, the most obvious way, with reliance upon
the effect of intonation.)
Now consider a so that clause:
(2.73) He didnt go so that he could get promotion

Like the preceding example, this sentence has a structure of two clauses, and
on that basis we would expect two intonation units:
(2.73a) He didnt go |so that he could get promotion

This suggests a meaning in which the person decided to stay in order to get
promotion. The domain of the negative extends to the next tonality
boundary and thus go is negativized. But in the marked tonality case - one
intonation unit covering the two clauses - it is the so that clause that is
negativized:
(2.73b) He didnt go so that he could get promotion

This rendering implies that the person did go, but not in order to get
promotion, but to gain some other result.
In both cases of marked tonality, (2.72b) and (2.73b), the negative is
transferred to the following clause; and marked tonality - a contrast in
intonation - is the typical way in which English speakers achieve this.

7.5 Report clauses

Report clauses are those that take the following form:
(2.74) He said he would come

46
 Tonality: the units o f intonation

(2.75) I reported that they had taken a decision

(2.76) They asked if he knew the decision
(2.77) He didnt know whether they had already decided
(2.78) We very much hope they will think about it

There is an initial verb of reporting (which has to be interpreted rather

broadly to include asking, knowing, hoping, etc.) and a clause containing the
content of what is reported. The two parts are usually linked by a
conjunction, e.g. that, whether, if, but the that conjunction may be omitted.
It is best to regard the two parts as a single clause consisting of the main verb
of reporting and a (rankshifted) clause as the direct object complementing
the reporting verb. Not all linguists describe report clauses in this way,6 but
it seems best to, since in all the above cases a single intonation unit covers
the whole statement. Read them for yourself. (Notice also that one single
intonation unit is still sufficient even if the reporting verb is placed at the
end, e.g. He would come, he said-, They h ad taken a decision, I reported, and
also if the statements were converted into interrogatives or imperatives, e.g.
Did he say he would com e? Ask i f he knew the decision.)
The difference between report clauses and direct speech is matched by
intonation. The direct speech version of (2.74) might have been He said, 7
will c o m e, which would be rendered in speech as two intonation units:
(2.79) He said | Ill come
with a little pause after said. Notice that the wording of reporting and direct
speech might occasionally be identical:
(2.80a) And then hell say whos coming (reporting)
(2.80b) And then hell say I whos coming (direct: A n d th e n h e w ill say,
'W h o is c o m in g ? ')

Compare reporting and direct speech in this pair:

(2.81) Tell me where your brother is (reporting)
(2.82) Tell me |where is your brother (direct: Tell m e , 'W h e re is y o u r
b r o t h e r ? )

The wording is almost the same, but again the intonation is quite different.
Now see if you can work out the difference in meaning between this pair
of examples with identical wording:
(2.83) Tell him to save time
(2.84) TeJ] him |to save time
(2.83) follows the normal pattern for report clauses and means the
equivalent of Tell him to hurry up. (2.84) seems to imply Dont let him work
it out for himself because that will waste time; by telling him, time will be

47
The intonation systems o f English

saved, i.e. main clause followed by a dependent purpose clause (in order
to).
Report clause structures with know need to be distinguished from linking
adjuncts, particularly you know.
(2.85a) You know its important

is a report clause (its im portant is the direct object complementing the verb
know), but in
(2.85b) You know |its important

th e you know is acting as a means of linking a previous utterance with the

new one in the following, main, clause.

7.6 Clause complements

Verbs vary in the kind of complementation they require in the rest of the
clause. Verbs may be intransitive and require no further complementation;
other verbs are transitive, some requiring a direct object as complement,
others requiring an additional, indirect, object complement as well.7 Some
verbs may sometimes operate as intransitive and sometimes as transitive, e.g.
en ter, as we saw above (p.40). Another example is wash, and we have already
seen how intonation helps in distinguishing wash (transitive) from wash
(intransitive). Remind yourself of the difference between (1.20) and (1.21):
(1.20) She washed and brushed her hair
(1.21) She washed 1and brushed her hair

In (1.20) she washed her hair (as well as brushed it); h er h a ir is the direct
object complementing the verb wash, which is thus transitive. In (1.21),
however, w ashed is intransitive, since the boundary separates that verb from
the following words. We also saw in the previous discussion that other verbs
which may sometimes be transitive and sometimes intransitive will also need
intonation to disambiguate between parallel wordings. This, as we saw then,
is intonation doing grammatical work.
There is another situation where intonation is necessary to disambiguate
parallel wordings that involve complements. In colloquial speech, the real
subject of a clause is sometimes left to the end, for effect, e.g. Theyre good,
a r e n t they, these arrow s (referring to the illuminated arrows at dangerous
bends in the road, that show up in the dark). This displacement of the
subject is much more common than most people realize, in informal speech,
and can lead, possibly, to confusion if the displaced subject occurs after
verbs which can either be transitive or intransitive. Halliday (1967: 42) gives
the following example:
(2.86a) Theyve left |the others

48
 Tonality: the units of intonation

in which the others is the real, but displaced, subject, with the meaning, The
others have left (actually). To leave is a verb that can either be intransitive as
it is in (2.86a), or transitive as in (2.86b):
(2.86b) Theyve left the others
In this case, the others is the direct object complementing left. (We concede
again the point that this difference might be shown clearly in the written
form by using a comma appropriately; but in the spoken form, it is
intonation that does the grammatical work.)
Here is another pair of examples:
(2.87a) Shes helping |Ann
(2.87b) She's helping Ann
You should be able easily to work out in which case Ann is helping or being
helped.

8. Intonation unit boundaries

One final, but obvious, question to ask is how we know where the
boundaries lie. In many circumstances there is no problem: if an utterance
consists of a single intonation unit, then it is bounded by silence before and
after. If an utterance consists of two or more units of intonation with clear
pauses which exactly coincide with sentence or clause boundaries, then
again there is no problem.
But problems do arise, and especially so in fast informal speech. Here is
an example from Brown, Currie and Kenworthy (1980: 42):
(2 . 88 )

I ve lived in this house ever since I was horn in fact thats

for well sixty years

The problem, they explain, arises with the phrase in fact. You can see that its
pitch is relatively low and comes after a fall; it also comes before a new
clause. (There is no point, of course, in appealing to commas, since this text
was spoken before it received any written form.) The pitch movements on

49
The intonation systems o f English

house, born and y ears suggest at least three intonation units; but what about
in fa c t ? It could either belong to the unit that contains ever since I was born,
or it could be part of the following unit, which would make it in fa c t thats
f o r well sixty years, or, thirdly, it could be a short intonation unit of its own.
Although many examples of this kind of problem arise in the course of
analysis of informal spoken text, there are principles with which we operate
so that most of the time we can identify the boundaries. Crystal (1969:
204-7) has the fullest account of the phonetic cues we use in the
identification of intonation unit boundaries. He claims that there are certain
regular patterns of features, but two principal ones. Firstly, there is a
perceivable pitch change at some point following a tonic syllable: either a
stepping up after a falling tone, or a stepping down after a rise; if the pitch
.of the tonic syllable was level, then either a stepping up or down would
signal the start of a new intonation unit. Secondly, there is either a (very)
slight pause or a change of pace in the flow of syllables; syllables at the end
of a unit tend to be relatively slower, but syllables at the beginning of a unit
have a tendency to speed up. Very often, all three features appear: a change
of pitch, a pause and a change of pace, but equally often only one feature
will be employed.
The identification of the boundaries is not necessarily a crucial issue. Pike
(1945) maintained that phonological units usually do have fuzzy edges;
how do you identify, for instance, the boundaries of consonants and vowels,
especially in the light of instrumental evidence of continual overlap? Other
linguists have maintained that it is not absolutely necessary to identify
boundaries precisely.8 Indeed, there is often a smooth transition from one
unit to another without a pause, and Halliday, for one, has incorporated this
feature into his description by talking of compound tones: a fall followed by
a rise is the most typical case.
Where there are cases of dispute, where Crystals phonetic criteria do not
produce a clear-cut boundary, then we have no option but to appeal to
grammatical or semantic criteria. This is because the vast majority of cases
do coincide with some kind of grammatical boundary (not just between
clauses, but between clause elements, instances of apposition, words in lists,
etc.) and do coincide with the organization of information into pieces. The
principle here is that you interpret difficult data by principles established in
easier data, but even then, some problems like Brown, Currie and
Kenworthys in fa c t remain unresolved.
Unresolved cases in the analysis of tonality should not undermine the
general theory that has been presented in this chapter. Such cases are
relatively infrequent, as a perusal of the transcription of intonation in
informal speech would testify. The general theory is that when people
communicate in speech, they must organize and manage their information

50
 Tonality: the units o f intonation

into discrete pieces of information which are then worded and formulated
grammatically and pronounced in intonation units. The theory also specifies
that a speaker's perception of their organization of information is presented
in a single unit of intonation, which is typically contained within a single
clause. Where there is this congruence of semantic, phonological and
grammatical units, tonality is said to be neutral; where there is not, tonality is
said to be marked, and this chapter has presented a comprehensive coverage
of marked tonality. There is always a congruence of units of information and
units of intonation.?
It must also be conceded that units of information do not always seem to
contain much in the way of information. Units often only contain exclama
tions like Coo!, My goodness me!, or displays of colloquial style like you know,
I m ean, y ou see or of back channel like mm, yes. Sometimes they only
contain running repairs when the speaker realizes that something they have
said does not correspond to standard grammar or that a better word could
have been chosen. The term unit of information has therefore to be defined
relatively broadly: it is the semantico-phonological unit for the development
of discourse, which handles not only information as propositional content
but also markers of style, expressions of attitudes and feelings, running
repairs, phatic communion like G ood m orning and How do y ou do? and
politeness formulas like please, thank you and d o n t m ention it.
See if you can work out a reasonable tonality analysis of the opening two
paragraphs of Treasure Island, imagining that you have to read it out to an
audience.
Squire Trelawney, Dr. Livesey, and the rest of these gentlemen having asked me to
write down the whole particulars about Treasure Island, from the beginning to the
end, keeping nothing back but the bearings of the island, and that only because
there is still treasure not yet lifted, I take up my pen in the year of grace 17-, and
go back to the time when my father kept the Admiral Benbow inn, and the brown
old seaman, with the sabre cut, first took up his lodging under our roof.
I remember him as if it were yesterday, as he came plodding to the inn door,
his sea-chest following behind him in a hand-barrow; a tall, strong, heavy, nut-
brown man; his tarry pigtail falling over the shoulders of his soiled blue coat; his
hands ragged and scarred, with black, broken nails; and the sabre cut across
one cheek, a dirty, livid white. I remember him looking round the cove and
whistling to himself as he did so, and then breaking out in that old sea-song that
he sang so often after-wards:-

Fifteen men on the Dead Mans Chest -

Yo-ho-ho, and a bottle of rum!

in the high, old tottering voice that seemed to have been tuned and broken at the
capstan bars.

51
The intonation systems o f English

Now try the next text, from Pygmalion, which is designed to be read
aloud, but from which the punctuation helps have been removed.
HIGGINS. There Thats all youll get out of Eliza Ah-ah-ow-oo! No use explaining As
a military man you ought to know that Give her her orders: thats enough for her
Eliza you are to live here for the next six months learning how to speak beautifully
like a lady in a florists shop If you are good and do whatever youre told you shall
sleep in a proper bedroom and have lots to eat and money to buy chocolates and
take rides in taxis If youre naughty and idle you will sleep in the back kitchen
among the black beetles and be walloped by Mrs Pearce with a broomstick At the
end of six months you shall go to Buckingham Palace in a carriage beautifully
dressed If the King finds out youre not a lady you will be taken by the police to
the Tower of London where your head will be cut off as a warning to other
presumptuous flower girls If you are not found out you shall have a present of
seven-and-sixpence to start life as a lady in a shop if you refuse this offer you will
be a most ungrateful wicked girl and the angels will weep for you [To Pickering]
Now are you satisfied Pickering [To M rs Pearce] Can I put it more plainly and fairly
Mrs Pearce
(Penguin edn: 45-6)

Notes
1. Tone group is Hallidays term for intonation unit; others refer to it as tone unit,
intonation group', breath group, phonological phrase, phonological clause, and
intonational phrase!
2. See Quirk et al. (1964); Tench (1990: 41-53)-
3. Halliday (1967: 17, 42-3) argued the case for a special listing pre-tonic, but this is
refuted in Tench (1990: 350-4). See also Crystal (1975: 19) on structural parallelism.
4. For a full exposition of theme and rheme, see Halliday (1985; chapters 3 and 8).
5. It has to be noted that this is not a view shared by all grammarians. A fuller
discussion will be found in Tench (1990: 110-16).
6. Sinclair (1972) and Young (1980) describe such structures as two consecutive
independent clauses, but the intonational evidence is one reason for rejecting such
a view.
7. For a full description of clause complementation, see e.g. Young (1980).
8. See Halliday (1967: 19 footnote), Gibbon (1976: 66), Brazil, Coulthard and Johns
(1980: 45-6). Brown, Currie and Kenworthy (1980) abandoned the attempt to
achieve principles of analysis in the identification of intonation unit boundaries, but
see Tench (1990: 170-83).
9. This is why Hallidays notion of compound tones with one and a half pieces of
information is not promoted here; it blurs the main principle. See Tench (1990:
166-70).

52
 Tonicity
THE FOCAL POI NT OF I NTONATI ON

1. Tonic syllables
In the previous chapter we looked at the function and the nature of the units
of intonation: in this chapter we look at the heart of these units, the tonic
syllable, which is the indispensable minimum. We have had reason to note
how some intonation units are abandoned as a speaker restarts what they
want to say; we describe units as being abandoned precisely on this point:
they lack a tonic syllable. If the speaker does not reach the tonic syllable in
any given intonation unit, then we do not have enough clues to decide on the
structure and the focus of information in that particular part of the message.
We will therefore begin this chapter with a consideration of the answers to
two obvious questions: what is a tonic syllable? and, what does it do?
It is usually affirmed that the tonic syllable is the most prominent, or
salient, of the stressed syllables in any given intonation unit. This traditional
statement has been challenged in the light of actual difficulties in attempts to
identify tonic syllables in recordings of natural, spontaneous, informal
speech. The challenge itself has been met with more vigorous
experimentation which has resulted in a much clearer understanding of the
nature of phonetic prominence in intonation units.1
Seven features appear to be significant for the identification of tonic
syllables:

1. pitch peak (i.e. maximum pitch height)

2. maximum pitch range
3. kinetic tone
4. loudness peak (i.e. maximum intensity)
5. decrescendo
6. tempo marking
7. pause

53
 Pitch peak clearly refers to the highest pitched syllable within a given
intonation unit; by maximum pitch range we recognize that a number of
syllables will exhibit pitch movement, but that one will exhibit greater
movement, than others. Kinetic tone also involves pitch movement; in this
case, we recognize that usually there is one syllable with a degree of
movement, even if all the others are fairly stable at a fixed pitch level.2 These
three features constitute the role played by pitch.
Loudness peak clearly refers to the loudest syllable within a given
intonation unit. By decrescendo, we refer to an observable decrease in
perceived loudness in the succession of syllables following a tonic syllable;
in other words, syllables following a tonic syllable are often less loud than
the syllables preceding the tonic, in the pre-tonic, segment. These two
features constitute the role played by loudness.
Tempo marking refers to the relative speed of delivery in the syllable
concerned; the tonic syllable is marked by an absence of quickening, i.e. it
remains as relatively long compared to surrounding syllables. Occasionally,
the syllable before a tonic is so long that it can be described as a drawl.
Furthermore, the identification of a tonic syllable may be aided by a slight
pause preceding it. These two features constitute the role played by timing,
although it must be conceded that the role of timing is much less
pronounced than either pitch or loudness. However, it can be seen that the
three traditionally labelled prosodic features - pitch, volume and duration
- together provide the basis of the way in which speakers produce tonic
syllables and learners identify them.
It is also maintained that different combinations of these features
produced different reactions. If six, or ail seven, features combine, maximal
prominence is produced which learners interpret as strongly contrastive. In
other words, if a speaker wishes to be seen as contrasting one part of their
message with something already said or implied, the tendency would be to
fire on all cylinders and use all features.
However, speakers are not always contrasting parts of their messages with
other items in the discourse, and are, consequently, content to use less than
all seven features, indeed to use only four features to project the tonic
syllable. The four features are kinetic tone, decrescendo, either pitch peak or
maximum pitch range, and, fourthly, either loudness peak, tempo marking
or pause. This is the phonetic basis of tonicity; in cases of uncertainty, the
analyst must consult semantic or syntactic criteria, as in the matter of
dubious intonation unit boundaries. However, in the vast majority of cases,
it is possible not only to identify a tonic syllable, but also to secure general
agreement among a group of listeners^
Figure 3-1 is an instrumental reading which shows the pitch, the volume
and the relative length of the tonic syllable. The top two graphs indicate
loudness, the bottom graph pitch, and time is indicated linearly.

54
;'GRAPH: 60 10 388 HZ LOG SCALE HIKDOH: 0 FRAMES
line: 1.135208 SECS FRAME: 46 EHERGlf: 2465 PITCH FREQ: 156 HZ
| Figure 3.1 // .1. Philip and |Jean are |getting married //

I The tonic syllable, then, is that syllable in a given intonation unit which is
7 made most prominent by a combination of pitch, volume and length. But
7 what is its function? Why is it there? Halliday (1970: 40) explains: The
j function of the tonic is to form the focus of information: to express what the
speaker decides to make the main point or burden of the message. This is a
vital contribution that Halliday has made to our understanding of the
function of intonation in any language. As we have already seen, the units of
intonation represent the speakers management of the organization of
information: one unit of intonation represents one unit of information. Now
we can add to the theory, and say that the tonic represents the focus of each
unit of information.
We can illustrate the function of the tonic quite easily from an example of
contrasting tonicity which was often used to help new learners of English.4
The following question can be asked in quite a number of different ways by
altering the tonic syllable; the most obvious way is to have the tonic syllable
as the last word:
(3.1a) Can you break an apple in two?

The intonation starts off quite high on the word Can, gradually descends
through y ou break an apple in, and then rises on two. Two is the tonic and is
made prominent by the combination of prosodic features referred to above.
But what difference is made if you shift the tonic to apple?
(3.1b) Can you break an apple in two?

55
The intonation systems o f English

The focus of attention switches to apple. (3.1b) could really only make sense
if people had already been talking about breaking things in two, and the
new focus of information is on a new item being offered for consideration.
Notice that by changing the tonic, you change the focus of information.
Try the same sentence with break as tonic:
(3.1c) Can you break an apple in two?
Wouldnt this suggest that the focus of information is now on breaking,
probably in contrast to some other action like cutting?
And wouldnt the next rendering suggest that you becomes the focus of
information, probably in contrast to somebody else?
(3.Id ) Can you break an apple in two?

And, finally, a fifth possibility would involve the tonic on Can:

(3.1e) Can you break an apple in two?
The focus of information is now on the ability of the person addressed.
Perhaps he or she has indicated the idea that breaking an apple in two
presents no problem; but then comes a query, or even a direct challenge, in
the form of (3.1 e).
As the tonic is changed, so is the focus of information. In each rendering
of (3.1), we are presented with only one unit of intonation, but with five
different focuses. In more technical terms, we can say that whereas the
tonality has remained constant, the tonicity has been changed each time.

2. Neutral tonicity
As with tonality, there is a neutral form in the tonicity system. The system of
tonicity is the range of choices in the position that the tonic syllable can have
in a given intonation unit. This was illustrated earlier, in Chapter 1, in the
discussion of the different possible renderings of
(1.5) What are you going to do tonight?
and has now been illustrated again with variations (3-la) to (3-le).
The neutral form of the tonicity system is to have the tonic syllable within
the last lexical item in the intonation unit. We described (3.1 a) as being the
most obvious way to intone that question, and you will notice that the tonic
syllable is in the last lexical item, two. If you would care to check through
Chapters 1 and 2, you will see that since we started marking the tonic syllable
from item (1.10) onwards, the vast majority of the examples have the tonic
syllable within the last lexical item. Look through and check for yourself.
Crystal (1973: 23) stated that, in his data, the tonic accompanied the last
lexical item 80 per cent of the time. In newsreading, the proportion reaches

56
 Tonicity: the focal point of intonation

as high as 88 per cent (Tench, 1990: 497), but it is noticeable that whatever
the genre of spoken discourse might be, this generally high proportion is
always retained.?
Halliday maintained that this propensity of the tonic to accompany the
last lexical item has significance for the structure of information within the
clause, and that this end position is the normal position unless there is good
reason for it to be otherwise. Notice that we have been careful to specify that
the item concerned must be a lexical item - not a grammatical item, not even
the last word. The lexical item has semantic significance and may often
contain more than a single word: compounds, e.g. station master, railway
track, signal failu re, etc., or phrasal verbs, e.g. to take off, to get in, to com e
across, etc., or idioms. Grammatical items are usually described as having
structural significance, like pronouns, articles, prepositions, auxiliary verbs,
etc., but Halliday includes all items that belong to closed systems, like
here/there, now /then, yesterday/today/tonight, together, again, etc. That is
why the most obvious way of saying (1.5) is to put the tonic not on the last
word tonight, but on the last truly lexical item, do.
The most obvious way of saying (3.1) was to place the tonic on two, as if
the context was as vague as What are you going to ask now? The context for
(3.1b), however, has to be much more specific: I can see that you can break
things in two - so ... Similarly, the contexts for all the other renderings would
be specific, although different in each case. The context provides the kind of
information that is given, and the tonic conveys what the speaker considers
to be new. Look again at (3.1b) to (3-le) and see what is given and what is
new information. The tonic syllable indicates what is new.
Halliday pointed out that the normal structure of given and new
information in a clause is to present the given first and the new last, and that
is why the tonic syllable usually comes at the end. But the context of (3-la) is
so vague that the whole of the clause could be regarded as new, and so we
now need to introduce the distinction between broad focus and narrow focus.

3. Broad focus
Imagine two or more people together and one of them suddenly makes a
decision and says:
(3.2) I think Ill go and have a cup of tea
The natural place for the tonic will be tea, as the last lexical item. None of the
information is given, in the sense that you can recover it from the context
(except, possibly, /). The whole sentence contains only new information, but
the tonic syllable cannot occur with each lexical item - otherwise you would
have several messages, instead of just one. The speaker says (3.2) as a single

57
The intonation systems of English

piece of information. The tonic indicates the end of the new information; it
is not as if the speaker is merely focusing on tea; he or she is focusing on the
whole plan of going and having a cup of tea. The focus is broad; it embraces
the whole clause.
Here is another example. Imagine two people together, and perhaps there
comes a lull in the conversation which is broken by a general question like:
(3.3) Where are you going on your holidays?
The natural place for the tonic is again the last lexical item, holidays. None of
the information is given (except, perhaps, y o u ) because this utterance is the
opening of a new topic. The question contains only new information; the
focus is, therefore, not confined just to holidays, but it embraces w here and
going, too. The focus is broad, embracing the whole clause.
A third example takes the form of a yes/no question:
(3.4) Did you have a nice time?
Again, this may be an opening, perhaps, of a new subtopic within the
general topic of holidays. The whole clause has broad focus.
Finally, an example with an imperative:
(3.5) Look at that man up there

This again has broad focus if it is said as an opening to a new topic of

conversation. Like the others, (3.2)-(3.4), the tonic falls on the last lexical
item. (T here belongs to a closed system and up is a preposition, and so they
do not count as lexical items.)
Broad focus has been illustrated above in four different clause types to
emphasize the point that information structure is not confined by
grammatical structure. It has also been illustrated as openings, but it is just as
relevant in the progress of a narrative. Here is an example of part of a
narrative from Crystal and Davy (1975: 44); each intonation unit is
reproduced on a separate line:
(3.6) (Yes I remember there was a terrible story - horrifying story that was told by a
colleague of mine when I used to teach years ago - who erm ...)

This chap lived in erm - a semi detached house

and next door
- there was a man
whod just bought a new car
- and he was telling me
that one morning
he was looking through the window
- and this . man
allowed his wife to drive the car
very unwisely
and she was having a first go in it
- and - he backed it

58
 Tonicity: the focal point o f intonation

out of the garage

- so that it was standing on the driveway
. and he'd closed the garage doors
- and she came out of the house
- to take this car out
very gingerly
- and opened the door
- and sat in the car
- and - e r. began to back
very very gently
taking . great care you see
that she didnt do anything to this . to this new car
- and - - as she backed
- there was an unpleasant. crunching sound
and she slapped on the brakes
and looked around frantically
- and realized
that she hadnt opened the . gates
. that. let on to the main road you see

Now read through the extract again and see how frequently the tonic
coincides with the last lexical item of each unit; notice, too, how often in
these cases of neutral tonicity the focus of information is broad, containing
exclusively new information - not always, but pretty often.
Broad focus, then, refers to instances of intonation units containing only
new information. However, there is the odd case of the reverse being true
too, where intonation units contain only given information. Heard , at a
conference was the following remark:
(3.7) Id love to attend the next one
meaning the next conference; one is functioning as a pronoun here and so
next is the final lexical item. The rejoinder to (3.7) came as follows:
(3.8) Youd love to attend the next one!?
said with some surprise, but with identical wording apart from the obvious
necessary substitution of pronouns. There is no new information as such in
(3.8) ; it is all given. In such cases, the focus in the rejoinder has to be treated
as parallel to that of the original. If (3.7) has broad focus, (3.8) must be
treated as having broad focus, too.

4. Narrow focus
Imagine, again, the situation in which one person decides suddenly to go
and have a cup of tea (3-2); if someone then says
(3.9) (Well) why dont you come and have a spot of lunch?
The only thing new is a spot o f lunch, you com ing (= the 1 going of 3.2) and
having are given since they are items that are recovered from the context.

59
The intonation systems o f English

The focus, now, is on only part of the information in the intonation unit; that
which is new is narrowed down to a spot o f lunch, hence narrow focus.
Notice again that the tonic indicates the end of the new information. Thus,
(3.9) illustrates - in the context of (3.2) - a narrow focus.
Then imagine the two people in conversation, in which the question of
(3-3) is raised, about the destination of one persons holiday. After handling
the response to that question, he or she may well return with the same
question:
(3.10) And where are you going on vour holidays?

The only new item is you/your, w here, going and holidays belong to the
context so they are given. The focus is narrow; the tonic falls again at the end
of the new information, in this case, a grammatical item. Notice that this
phenomenon of placing the tonic on words like y ou r and you is typical in
situations where the second person asks the same question of the first
person, e.g.:
(3.11) (Good morning | how arg you?)
Very well |thank you |and how are you?

The answer to D id you have a nice time? (3.4) might possibly be:
(3.12) We had an awful time

We having (= y ou having in (3-4)) and time are recoverable from the context,
and only aw fu l is new. Thus, the focus is narrow; indeed, very narrow,
consisting solely of one word.
And a possible response to Look a t that m an up there (3-5) could be:
(3.13) And look at that man
Again, the focus is narrow, confined to that, since all the other information
is recoverable from the context supplied in (3.5).
In the above examples, you will notice that narrow focus necessarily
assumes an element of given information. The focus itself involves new
information which might be found either at the end of the intonation unit,
or anywhere else. When narrow focus is located at the end of the intonation
unit, as in (3-9), it happens to conform to neutral tonicity - the tonic
accompanying the last lexical item. When narrow focus occurs elsewhere, it
conforms to marked tonicity, since the tonic accompanies either a non-
lexical item, as in (3-10), (3-11) and (3.13), or a non-final lexical item, as in
(3.12). Thus, whereas broad focus will, of necessity, require neutral tonicity,
narrow focus may happen to use neutral tonicity but is just as likely to
involve marked tonicity.
So far, given information has been identified with recoverability from the
context. However, it could refer to anything that is visible, or known, in any
situation. If we might return to the example of breaking apples in two, you

60
 Tonicity: the focal point o f intonation

could imagine a person actually doing this as their party piece; no other,
linguistic, context would be necessary to contextualize the challenge:
(3.Id ) Can you break an apple in two?

The words break, apple and two do not need to have been uttered, but they
are recoverable - not from the linguistic context, but from the physical
situation.
In the narrative, (3.6), the speaker does not mention the word drive until
he says:
- and this man |allowed his wife to drive the car
The last lexical item is car, but it has previously been mentioned and is
therefore given; the next-to-last lexical item is drive, which has not been
mentioned, but nevertheless the speaker still treats it as given. Why? The
speaker obviously wants to narrow the focus of the information down to
wife-, he excludes drive from new information on the assumption that
everybody knows that that is what you do with cars. Drive is implied; it is
part of general knowledge, in the context of buying a new car. It is, thus,
treated by the speaker as part of the given information in that intonation
unit.
Given information is recoverable either from the linguistic context, the
physical situation or general knowledge.

5. Marked tonicity
We have defined marked tonicity as the choice of tonicity that is not neutral.
Neutral tonicity refers to the tonic accompanying the last (truly) lexical item
in the intonation unit; marked tonicity typically takes two forms: the tonic
accompanying a non-final lexical item or accompanying a grammatical item.
The main function of marked tonicity is to carry those cases of narrow
focus that do not coincide with final lexical items, as in (3.10), (3.11), (3-12)
and (3.13), and in (3.1b)-(3.1e). Contrast is an obvious case of narrow focus,
as in the various renderings of (3-1). It emerges too in contrasts involving
grammatical items. For instance, the neutral tonicity of (3.14a) can be
contrasted in two, grammatical, ways:
(3.14a) I have been asking for ages
(3.14b) I have been asking for ages
which means something like Its not that I havent been asking - a contrast
in affirmative/negative polarity; and
(3.14c) 1have been asking for ages
which means something like Its not that Im now beginning to ask - a
contrast in tense.

61
The intonation systems o f English

It should not be too difficult to think of other contrasts involving

grammatical items. Take pronouns, for example:
(3.15) Its not what I think | but what you think

and prepositions:
(3.16) Think of what you can put Into it | not what you get o y i of it

and conjunctions:
(3.17) Remember | I said if

and even morphemes:

(3.18) Thats what youre exporting |and were importing

(3.19a) Shes an ex-girlfriend

as opposed to the normal

(3.19b) She's an ex-girlfriend

We must be careful in dealing with compounds. Compounds are words

composed, usually, of two otherwise independent words. To take a single
example, blackbird is a compound consisting of two words that have an
independent existence, black and bird. B lackbird is distinguished from a
black bird in a number of ways: semantically, blackbird refers to a single
concept, whereas black bird refers to two; grammatically, blackbird is unitary
and cannot be divided, whereas black bird can be divided (a black little
bird); orthographically, the compound is spelt as a single word, but the noun
phrase as two separate words; and phonologically, blackbird has a single
word stress, at the beginning, and black bird has two word stresses, one on
each word. Compounds are thus distinguished semantically, grammatically,
orthographically and phonologically from a parallel noun phrase.
Unfortunately, the orthographical distinction is not consistent. The
following compound can be spelt either as one word, two words, or
hyphenated: roadw orks, ro ad works, road-works. It is when a compound is
spelt as two words that a problem can arise in identifying neutral and
marked tonicity; the spelling ro a d works still represents a compound, as the
word stress, grammar and semantics testify.
In the narrative we considered above, (3.6), drivew ay occurs as tonic; it is
the final lexical item, which happens to be a compound, and is nowadays
usually written as a single word. Policem an and policew om an are also
examples of compounds normally written as single words, but the
compound p o lice car is usually still written as two words. So even if the
written form of (3-20) has the compound as two words, it must be treated as
a single lexical item:
(3.20) And there goes the police car

62
 Tonicity: the focal point o f intonation

And because the lexical item police car is final, (3.20) is a case of neutral
tonicity, despite its written appearance.
However, the unstressed part of the compound could, occasionally,
receive the tonic. Imagine a group of witnesses discussing an incident which
involved the police and their vehicles and a dispute arises over a detail. One
person might report that the ringleader of the group was taken away with
the rest of the group in a police van; but another disputes this:
(3.21) (A: He was taken with the rest of them in the police van
B: No he wasnt)
He was taken in the police car
This would be a case of marked tonicity because the normally unstressed ele
ment of the compound receives the tonic, rather like (3-18) and (3.19a) above.
Here are two more examples to illustrate tonicity and compounds. B ook
case can be written as book case and book-case, but the word stress is consist
ently on the first element. Thus (3.22) has neutral tonicity, despite the spelling:
(3.22) What we need is a new book case

and (3-23) has marked tonicity, again despite the spelling:

(3.23) He got a new book shelf | but what we need is a new book case
The compound chequ e book is usually written as two words. (3.24)
contains the compound in both neutral and marked tonicity:
(3.24) I've lost my cheque book |It's not a single cheque j but the whole
cheque book
Marked tonicity also emerges in other genres of spoken discourse, but not
necessarily conveying contrast. It might simply be a way of avoiding a tonic
on a repeated item. One genre familiar to many is the reading of football
scores. If there is an outright victory, neutral tonicity prevails; but in the case
of a draw, marked tonicity prevails. In (3.25)
(3.25) Wales three | Belgium one

the scores themselves are the final lexical items in each of the intonation
units: neutral tonicity. However (in the return match) in (3.26)
(3.26) Belgium one |Wales one
the tonic is placed on Wales, because the score one is being repeated and
treated as given. This is how all football followers can predict a draw before
the announcer reaches the actual score in the second intonation unit.6
A second example is the announcement of telephone numbers. Most
British numbers now have six or seven digits, and are usually uttered in two
groups of three or four. If the final digit is identical to the final digit of the
first group, it is treated as given; thus 751281 is usually rendered as

63
The intonation systems o f English

(3.27) seven five one |two eight one

In the case of football scores and telephone numbers, marked tonicity does
not represent contrast, but simply the avoidance of focusing on a repeated
item.
Marked tonicity also occurs regularly in another class of clauses. When a
clause ends in a common intransitive verb of motion or happening,7 that
verb - though it is the final lexical item - does not take the tonic. Consider
the following examples:
(3.28) The doctor's coming
(3.29) (Dont worry) Johns going
(3.30) An accident has happened
(3.31) A question was raised
(3.32) Discussions took place
(3.33) The postman called

The verbs com e and go do not add anything of significance in (3-28) and
(3.29), and they typically remain non-tonic even when in final position. The
verbs in the remaining examples are also relatively insignificant, because
what you expect of accidents is that they happen, and what you expect of
questions is that they are raised, and so on. The main information is the
doctor, Joh n , a n accident, a question, discussions, the postm an, and the
following verb simply fills the obligatory slot of predicator. Linguists have
often referred to the use of verbs in such cases as being semantically empty,8
because they do not add extra information.
The cases stand in contrast to the usual neutral tonicity of examples with
semantically richer words:
(3.34) The doctors running
(3.35) (Dont worry) Johns eating
(3.36) An accident is being investigated
.(3.37) A question was answered
(3.38) Discussions were announced
(3.39) The postman feM

These verbs inform; they are not merely fillers.

6. The tonicity of final adjuncts

When adjuncts occur in the initial position of the clause, they gain semantic
prominence, and like other classes of words become the (marked) theme of

64
 Tonicity: the focal point of intonation

the clause. But when they occur at the end, they are often least prominent -
but not always, and not all kinds either. The situation is a little more
complicated, and a number of points need to be made.
The first point to be made is that certain verbs require adjuncts of place as
their complements; verbs like put, sen d require a locative adjunct to
complete their sense, e.g. you cannot just say Put the books, you have to add
a locative element like here or down, etc. Thus a distinction needs to be kept
between locatives as essential complements to verbs - they have got to be
there - and locatives that add circumstantial information. Compare
(3.40) The books are here
(3.41) Put the books here
(3.42) Weve got some books here

In (3.40) and (3.41), here is complement to a re and put, but in (3-42) it

merely provides circumstantial information.
Circumstantial information is what adjuncts of place and time provide
most of the time. If the adjuncts belong to closed systems like here/there-,
now/then-, yesterday/today/tom orrow , they do not normally take the tonic,
unless a contrast is intended. Compare a usual rendering with a contrastive
rendering of the following:
(3.43a) Lets gq now
(3.43b) Lets go now (as opposed to a time in the future)

(3.44a) Thats all for today

(3.44b) Thats all for today (= you get more tomorrow)

(3.45a) I saw John yesterday

(3.45b) I saw John yesterday (as opposed to some other time)

(3.46a) Lets go there

(3.46b) Lets go there (as opposed to some other place)
There is another class of adjunct that acts in a similar non-tonic way; they
function as comment adjuncts, as in the following examples:
(3.47) Ive never read it though
(3.48) You must read it of course
(3.49) Hes not going to read it you know
(3.50) Theyve already read it however
(3.51) Ive begun to read it at least

There are a number of such comment adjuncts, but they operate in a

fairly closed system and that is why they do not usually take the tonic. But

65
 The intonation systems o f English

like the time and place adjuncts, they could do - producing an emphatic, if
not contrastive, effect, e.g.:
(3.48b) You must read it of course

They seem to be more effective, though, in initial position:

(3.48c) Of course you must read it

We have made two points so far: firstly, an adjunct, in some cases, is

required as an essential complement to a verb,; and in other cases, it provides
circumstantial information; secondly, some adjuncts belong to quasi-
grammatical closed systems, and if these happen to be in final position, they
usually do not take the tonic. Other adjuncts, like adjuncts of manner, are
lexical, i.e. they belong to open sets, and if they occur in final position are
likely to take the tonic:
(3.52) Lets go quickly
(3.53) He read it very well

A third point relates to yet another possible rendering of final adjuncts;

sometimes they can have an intonation unit all to themselves. For instance,
we could say (3.43) as
(3.43c) Lets go | now

with a falling tone on go and a slight rise on now. Compare also the
following:
(3.44c) Thats gM |for today
(3.45c) I saw John |yesterday
(3.42a) Weve got some books | h e r e ?
Strictly speaking, this is a matter of tonality, because an extra intonation unit
has been created; but it affects tonicity, too, because the adverb necessarily
becomes tonic in its own unit. We shall see, in the next chapter, that tone
plays a significant role in these examples, too. But at this juncture all we need
to note is that this pattern exists and that it gives more prominence to the
adjunct than it would otherwise have. Compare, then:
(3.54a) I dont want one now
with no prominence on now,
(3.54b) I don't want one j now

with more prominence on now, and

(3.54c) I dont want one now

with a great deal of prominence on now, suggesting a contrast with a point

of time in the future.

66

life
 Tonicity: the focal point of intonation

This threefold system of prominence operates with adjuncts of place and

comment, too. It also operates with another class of adjuncts, those that
indicate viewpoints. Compare the viewpoint adjunct, economically, in these
examples:
(3.55a) Were in poor shape economically
(3.55b) Were in poor shape |economically
(3.55c) Were in poor shape economically

Incidentally, a similar, but twofold, system of prominence operates with

final vocatives:
(3.56a) They're here John
(3.56b) Theyre here |John
with final glosses:
(3.57a) He shouldnt have done it the fool
(3.57b) He shouldnt have done it |the fool
and with final direct speech markers:
(3.58a) Dont run he said
(3.58b) Dont run [ he said
The effect of the extra intonation unit containing the final vocative, gloss or
direct speech marker is to give a little more prominence to the item
concerned.
A fourth point can now be made: some adjuncts can have two different
functions, and the different functions follow different tonicity patterns. An
adjunct like norm ally or fran kly can be used as an adjunct of manner, and if
it comes at the end of its clause, as in the following examples, it would tend
to be accompanied by the tonic:
(3.59a) He drives normally (i.e. in a normal manner)
(3.60a) I couldnt speak to him frankly (i.e. in a frank manner)
This is the pattern we would expect as in the other cases of manner we have
seen ((3.52), (3.53)).
But norm ally and fran kly can also act as comment adjuncts, and as is
typical of comment adjuncts, they do not take the tonic:
(3.59b) He drives normally (i.e. it is normal for him to drive)
(3.60b) I couldnt speak to him frankly (I have to be frank, I ...)
Other comment adjuncts include fortunately, unfortunately, thankfully,
happily, etc., and follow the same pattern for that function. An alternative
rendering to (3.59b) and (3.60b) is to grant the comment adjuncts their own
intonation unit to give them greater prominence, following the pattern of

67
The intonation systems o f English

final adjuncts described above:

(3.59c) He drives [ normally
(3.60c) I couldnt speak to him |frankly
with a falling tone on the first tonic, and a rising tone on the second.

7. Tonicity by default
So far, we have maintained a distinction between neutral and marked
tonicity depending on the position of the tonic within a given unit of
intonation. Neutral tonicity indicates broad focus where all the information
in the unit is new, but it may also indicate narrow focus if that which is new
information is found at the end of the unit. Marked tonicity indicates new
information that is either grammatical or not final.
We have also pointed out the oddity of cases like (3-8)
(3.8) Youd love to attend the next one!?
which was an echo of (3.7)
(3.7) Id love to attend the next one.

An echo is an identical wording of a part of, or the whole of, a previous

utterance, usually for the sake of expressing disbelief or surprise, or
checking against mishearing. Here are two more examples:
(3.61) (A: Jonathan won his race!)
B: He won

(3.62) (A: It costs eighty-three pounds)

B: Eighty-three pounds did you say
A similar case is when a person responds to a request for a repetition, e.g.:
(3.63) (A: Is this your handbag)
(B: Pardon)
A: Is this your handbag
Strictly speaking, in all these cases of echoes and repetitions, the
intonation unit only consists of given information, but the units are treated
as if they contain the same information structure of given and new as the
original. But there are two other cases where an intonation unit only consists
of given information: one is where the given information is the point of
contrast. Consider this example:
(3.64) Tom, Dick and Harry all got firsts | but Tom |was given the prize
The middle unit only contains given information, but it is now the point of
contrast. In (3.63), the second unit again contains only given information,

68
 Tonicity: the focal point of intonation

but the information is different, and is again contrastive:

(3.65) I see the moon |and the moon sees me
The second case of all-given information is insists, as Cruttenden (1986:
93-4) calls them. Here are some examples:
(3.66) (A: Why have you invited the Robinsons?)
B: But 1havent invited them!

(3.67) (A: How many tea bags did you use?)

B: I dont use tea bags

(3.68) (A: Has John read Slaughterhouse Five?)

B: No |John doesnt read books
All of Bs replies are insists; B insists that As presuppositions are not valid,
but note that Bs counter-presuppositional insists only contain information
that is given. Furthermore, B focuses on what he or she assumes to be the
basis of As presupposition and then negates it. The focus in the insists
usually correlates, therefore, with the given information in the original,
because it is the given information in the original that is the basis of As
presupposition.
We can thus explain the focus of these cases of all-given intonation units,
but there is one, final, case, which is much more difficult to explain.
In a radio programme about Edwardian days and ways in London, a
person was asked what they could remember about those days. The reply
came as:
(3.69) I dont know that there is anything to remember
Everything was treated as given, but why would the infinitive marker to
receive the tonic when it has absolutely no meaning of its own? We cannot
dismiss it as a rogue rendering, nor as a dialectal variation, since many odd
cases like this crop up.w
The explanation seems to be that the speaker is countering the whole
presupposition of the original question. In order to do this, he (it was a
man!) could not afford to put the tonic on a lexical item in case he gave
undue prominence to it and made it sound contrastive; the anything is the
given equivalent of the w hat of the original question, and he seemed to want
to avoid giving that prominence too. What he did, therefore, was to seek the
least semantically significant item in the unit and place the tonic there, in
order to avoid narrow focus and the possible interpretation of contrastivity.
In order to indicate a broad focus in all-given information, the tonic gets
placed on the least semantically significant item towards the end of the unit.
This is a genuine case of tonicity placement by default - avoiding all the
other potentially misleading options.

69
The intonation systems o f English

To illustrate tonicity by default further, we will take an example from

Cruttenden (1986: 93):
(3.70) Theyre not very expert | but there are lots of them
In this case, the speaker could have chosen to place the tonic in the second
unit on lots as the most lexical item, and this would have sounded quite
satisfactory. However, lots is treated as given in (3-70), and the speaker
wishes to avoid an interpretation of contrastivity in a re and them, and so the
tonic gbts placed on the least semantically significant item towards the end
of the unit, thus, by default, on of.
Some linguists11 treat all the cases presented in this section as cases of
default, but it seems to me that genuine cases of tonicity by default are those
cases where a speaker - as in (3-69) and (3.70) - places the tonic on the last
item of zero significance in the unit, in the deliberate attempt to avoid
potentially misleading interpretations of narrow focus.

8. Tonicity and grammar

Finally in this chapter, we consider a few cases where tonicity realizes a
syntactic contrast in parallel wording, in the same way as tonality
occasionally does.
One example was given in Chapter 1: the difference between
(1.22) He asked himself
(1.23) He asked himself

The explanation was given as a syntactic contrast between the reflexive and
the emphatic pronouns, which also affected the transitivity of the verb:
a sked in (1.22) is transitive, in (1.23) it is intransitive. Here is another
example with the same explanation:
(3.71) I have never taught myself
(3.72) I have never taught myself

A similar example can be found with the verb fe e l.12 In (3.73), we have an
example of a reflexive pronoun complementing a transitive verb, as in (3.71)
and (1.22):
(3.73): He felt himself (e.g. to see if he was bruised)

The tonicity is neutral in these cases because the tonic accompanies the last
lexical item. The marked version, (3-74)) in this case does not indicate an
emphatic pronoun (as in (3-72) and (1.23)), but an intensive complement to
an intransitive verb:
(3.74) He felt himself (i.e. he felt much more like his usual self)

70
 Tonicity: the focal point o f intonation

This version parallels other intensive complements as in He felt good, He felt

em b a rra ssed , etc., where the tonic would usually accompany the
complement itself.
A second example of tonicity contrast in syntax is provided b y . the
following pair:
(3.75) Shoot John
(3.76) Shoot John

In (3-75) , f ohn is a final vocative, which typically is left non-tonic (although,

as we have seen above, (3.57b), it could have its own intonation unit if the
speaker wished to give the vocative more prominence). In the written mode,
it would be preceded by a comma, but as we have often pointed out,
commas belong to the written mode only, whereas in the spoken mode, we
have to rely on intonation. (3.76) manifests neutral tonicity, with Jo h n as
direct object complementing the transitive use of the verb shoot. You will
now have noticed a number of cases where intonation plays a specific role
in distinguishing transitive and intransitive uses of the same verb.
The third example has already been dealt with above, in the section on
tonicity and final adjuncts; it will simply be mentioned again here to
complete the picture of tonicity contrasts in syntax. Adjuncts like normally,
fran kly, stupidly, happily can function as comment adjuncts as well as
adjuncts of manner. When they occur in final position in a clause, they are
potentially ambiguous; in the written mode, the ambiguity is usually
resolved by the presence or absence of a comma preceding them, but in the
spoken mode, we have to rely on intonation. Two further pairs can illustrate
the distinction; in each pair, the first item has neutral tonicity with the tonic
accompanying the adjunct of manner; the second pair has marked tonicity
with the tonic accompanying another item.
(3.77a) They didnt come happily (i.e. not in a happy mood)
(3.77b) They didnt come happily (i.e. 1am happy to say th a t...)
(3.78a) He presented me with the same information stupidly (i.e. in a
stupid manner)
(3.78b) He presented me with the same information stupidly (i.e. It was
stupid of him to ...)
We have now comprehensively reviewed the system of tonicity in English
intonation. Now check through the narrative of (3.6) and see if you can
explain the tonicity in each intonation unit; this chapter has covered all the
possibilities.
Here is an example of a tonicity exercise from a course in practising
English as a foreign language^ see if you can supply all the necessary
explanations. Each of the sentences overleaf is capable of being completed
in a number of different ways. The written form does not help, but the

71
The intonation systems o f English

spoken form would, by having the tonic syllable on the most appropriate
word (or words). For each of the continuations, which word(s) in the
original would have to have been tonic?
I cant see very well with my glasses now ...
(a) but John can
(b) but I can see well with yours
(c) but I can see quite well
(d) but I could when I bought them
(e) but John can see well with his
(f) but I can see well if I dont wear them
John's brother tried to buy both of the books ...
(a) not to borrow them
(b) not the magazines
(c) but he wasnt able to
(d) but one book had been sold
The early lenses werent made of glass ...
(a) but the later ones were
(b) although some of the early windows were
(c) although many people think they were
(d) they were made of crystal
John finds.it very difficult to wear contact lenses ...
(a) and so does Bill
(b) but Bill doesn't
(c) but Mary finds it easy
(d) but he has no difficulty in wearing spectacles

Notes
1. For a full discussion, see Wells (1986) and Tench (1990:201-14).
2. Wells (1986: 57) maintains that kinetic tone is invariably present; however, we must
allow for the existence of (so-called) level tones.
3. The semantic and grammatical clues to identifying the tonic, and therefore the
focus of intonation, are described in detail in Taglicht (1984).
4. Taken from Allen (1954).
5. See Tench (1990: 513) for further discussion and the one possible exception.
6. This has traditionally been the case on radio; television has begun to affect the
intonation of football result announcements, because the scores are presented on
screen before the announcement.
7. Cruttenden (1986: 83-4) calls such clauses event sentences.
8. Bolinger (1986: 120-3). Halliclay (1967: 38) also noted certain high frequency
collocations in intransitive clauses, e.g. My h ea d aches-, The d o o rs locked-, but this is
essentially the same phenomenon.
9. NB this pattern does not work with (3.41): Put the books \here, where the locative
element is complement to the verb.
10. See Cruttenden (1986: 93) and Ladd (1980: 81) for similar cases.
11. E.g. Ladd (1980).
12. This example is taken directly from Halliday (1967: 39).
13. Taken from Dickinson and Mackin (1969: 28-9).

72
 TONES
THE TUNES OF INTONATION

1. Primary and secondary tones

It is the tones of English intonation that most people are aware of. It should
by now be clear that English intonation comprises three separate, though
related; systems: tonality (intonation units), tonicity (the tonic syllables) and
tone. But people are much more aware of the third system than they are of
tonality or even tonicity. Tone refers to pitch movements, and that
observation which has frequently been commented on - that it is not what
they said, but the way they said it - relies primarily upon the pitch
movements. We are aware of rises and falls of pitch and high levels, low
levels, bouncing movements, calling tones and so on, which all contribute
to a wide range of different meanings.
Tone has been defined hitherto as the contrastive pitch movement on the
tonic syllable; so, for instance, in the standard example we took in Chapter 1,
the tone of
(1.10) A dog is a mans best friend

is the falling tone on the tonic, frien d. We now need to make a further
distinction, because it is obvious that there are other pitch movements
besides that which occurs on the tonic syllable. We make a distinction
between primary and secondary tones. Primary tones are the basic
contrastive pitch movements on the tonic, i.e. whether the pitch of the voice
moves up (rises), or moves down (falls), or combines a movement of down
and then up (fall-rises). Secondary tones are the finer distinctions of the
primary tones, i.e. the degree to which the pitch of the voice rises, falls or
combines a fall and a rise - whether there is, for example, a rise to a high pitch
or a mid pitch, or a fall from a mid pitch or a high pitch, etc. Secondary tones
also cover the pitch movements in the pre-tonic segment (the head and the
pre-head). (Pitch movements in the tail are not considered separately from
 The intonation systems of English

the pitch on the tonic because they are an extension to the tone itself.)
The primary tones of English, i.e. the fall, the rise and the fall-rise,
function, like tonality and tonicity, in the organization of information, but
they also feature, like the other systems, in a second function. Whereas
tonality and tonicity produce contrasts in grammar, the tone system
produces contrasts in the communicative, or illocutionary, function; that is,
they help to indicate differences between telling and asking, between
commanding and requesting, between congratulating and wishing, and a
whole host of similar functions that language is used to fulfil.
The secondary tones do not function, as such, in the organization of
information - or, for that matter, in either grammar or the communicative
functions. Their role is in the expression of attitudes, and the next chapter
will be devoted to a full description of all the possible pitch variations that
are discernible in English.
In this chapter we will concentrate on the role of the primary tones; we
will consider, first, their role in the organization of information and then
their role in the communicative functions. But even before we consider
these roles, it would first of all be as well to look at some of their
characteristic patterns of pitch.

2. Primary tones: falls, rises and fall-rises

We come now to the details of the pitch movements. One point must be
made clear: pitch is a relative matter. We do not talk about absolute pitch as
one might in music, but rather about levels and movements of pitch within
an individuals range of voice. An obvious example is that the actual range of
pitch of most adults is lower than that of children, and again, the actual range
of pitch of men is, on average, lower than that of women; In other words, in
absolute, musical, terms a low level of pitch for a child might in actual fact be
higher than a high level of a mans. The childs low level of pitch is measured
within the scale of the childs full range of pitch, and the mans high pitch
is measured within the scale of the mans range.
- y

childs range

------ womans range

------------mans range

It is quite useful to think of your own range of pitch as having what you
might call a high level, a low level and a mid level; it should also be possible

74

1
 Tones: the tunes of intonation

to imagine an intermediate range of pitch between high and mid, and

another between mid and low. These intermediate ranges could also be
thought of as having a notional midway point between high and mid and
between mid and low; this would provide a kind of five-point scale:
high
mid-high
mid
mid-low
low

Experiment with a word like No or Well and try and establish, at least, a
high, a mid and a low variety, keeping the pitch fairly steady. If you now say
No as a response to a question, the most natural way of intoning it would be
for it to be accompanied by a pitch movement falling from about mid-high
(or mid) to mid-low (or low). There are, of course, very many different ways
of responding with No, but the pitch movement just described could be
considered as the most normal or ordinary. This pitch movement is known
as the neutral fall, because no other meaning or connotation is added to that
of plain statement. It is the pitch you would expect if no other meaning or
connotation was intended. It is indicated by a downward-pointing line
before the syllable concerned.

2.1 Falls
The neutral fall can be contrasted with a high fall, in which the beginning
point is much higher. This sounds like a stronger, more determined
response; the symbol is placed higher. Compare (4.1) and (4.2):
high

(4.1) \No mid-high

(4.2) x No mid

mid-low

low

neutral fall (4.1) high fall (4.2)

Furthermore, the neutral fall can be contrasted with a low fall, in which the
beginning point is about mid-low, and the pitch drops to low. The symbol is
placed lower, and its meaning is something like mild or expected:

75
The intonation systems o f English

high
(4.3) x No mid-high
mid
mid-low
low \
low fall (4.3)

Compare all three. Remember that it is the neutral fall that is the primary
tone, and the high and low varieties are secondary tones. There are primary
functions associated with the fall, but secondary - attitudinal - functions
associated with the alternative degrees of fall.
If you take an example with a tail, you will find that there are two possible
ways of distributing the fall over the syllables: either the main drop of the fall
is carried through the tonic syllable, and then the unstressed syllable(s) will
carry it on to its final pitch level; or the tonic syllable keeps a fairly high pitch
and the fall is executed by a sudden drop to the low pitch in the following
syllable(s). Compare (4.4a) and (4.4b):
(4.4a) (4.4b)

Never Never

There is no difference in meaning, and so both can be symbolized as

(4.4) \ Never

There are, obviously, high and low varieties of a fall on a word like never.
(4.5) x Never
(4.6) .N ever
The (b) variety of the fall is quite likely to accompany the high fall. It is also
quite likely to accompany any tonic syllable that has a short vowel and a
voiceless consonant, as in nothing .
Let us now take an example with a pre-tonic segment, e.g.:
(4.7) I dont bexlieve it

The tonic syllable is the second syllable of believe; in a prepared intonational

transcription, a gap is left before the tonic syllable for the intonation symbol
to be inserted, as above. Otherwise, the symbol is placed through the letters:
believe. (4.7) can be rendered with a high fall (4.8) and low fall (4.9):
(4.8) I dont beNieve it
(4.9) I dont bex Neve it

with the same additional connotations as with (4.2) and (4.3).

76
 Tones: the tunes o f intonation

One of the most difficult tasks in intonation analysis is to discern the pitch
movement on the tonic syllable itself. In each of the renderings of I d o n t
believe it above, the pitch level of the unstressed syllable be- before the tonic
-lieve is likely to be fairly low. The pitch of the voice has then to jump up to
a higher level in order to effect a fall - you cant fall to low if you are at low
already! This inevitable jump up is often considered by the novice to be a
rise, and so a complete misconception can take place in the novices mind as
to what tone is actually being performed. This gives rise to the fairly familiar
complaint that I cant tell the difference between a fall and a rise. What you
must try and do is concentrate on the pitch movement of the tonic itself and
separate it in your mind from any preceding words or syllables. This,
however, is quite awkward when one or more syllables of a single word
belong to the pre-tonic and another syllable belongs to the tonic. The way to
do it is to first of all determine what the tonic syllable is, and any following
syllables, and then to imitate the pitch movement of that syllable; and
practice makes perfect. If there are any following syllables in a tail, they give
you a very valuable clue: if the tail remains low pitched, then the tone must
have fallen; if the tail is relatively high, then the tone must have risen.

2.2 Rises
We will now turn to the case of rising tones. Obviously, the beginning point
is low (or mid-low); then the pitch rises to mid or mid-high. (4.10) sounds a
little like a query, or possibly the beginning of a response which the speaker
is going to extend, like No, I d o n t think so
high
mid-high
(4.10) /No mid
mid-low
low
neutral rise (4.10)
The symbol is an upward-pointing line before the syllable concerned; it
might be tempting to add a question mark (?) to No, but as in the above
explanation, a rise can also indicate something like This is not all I want to
say; Im going to continue, in which case it might be tempting to add a
comma (,). It is best to resist such temptations since punctuation is a feature
of the written mode, and whereas it does correspond to a certain extent to
intonation, it is by no means as sophisticated.1
The rising tone as described above is its neutral form. Like the fall, it has
high and low varieties, depending on the extent of the rise.2 The high rise is
commonly associated with a stronger sense of querying, suggesting surprise
or even disbelief:

77
The intonation systems o f English

(4.11) 'N o

The rise finishes at a high point. The low rise finishes at about mid-low and
suggests a non-committal, or even grumbling, attitude:
high
(4.12) /No mid-high
mid
mid-low
low
high rise (4.11) low rise (4.12)

The neutral, high and low rises may accompany tonics and tails, with the
same range of meanings: compare
(4.13) /Never
(4.14) 7Never
(4.15) .Never
The rise requires a low starting-point and the main difference between the
three varieties is the end point of the rise in the tail. There is an (a) and (b)
form as in the fall, which is heard best in the high rise:
(4.14a) (4.14b)
Never Never

The (b) form would also be quite likely in cases of tonic syllables containing
a short vowel, with a voiceless consonant following.
The rise can also be demonstrated in an example with, a pre-tonic
segment:
(4.16) Do you really be/lieve it

The symbol comes in the middle of the word again, immediately before the
tonic syllable. It has a high variety:
(4.17) Do you really be/ lieve it
with the same meaning, of course, as in (4.11): and a low variety:
(4.18) Do you really be .lieve it

with a meaning parallel to that of (4.12).

Once again, there is a difficulty in the task of discerning the rise in the
tonic syllable, because in order to effect the rise, the pitch of the voice must
drop after be- to low so that the rise can begin. As with falls, the novice is
likely to think that this drop is part of the tone, but the same kind of advice
must be taken in the case of rises, too. If there is a tail, that is a valuable clue,
since if there has been a rise, the tail will finish relatively high itself.

78
 Tones: the tunes of intonation

2.3 Fall-rises
Finally, we turn to fall-rise tones, which comprise a sequence of a fall and a
rise even within a single syllable-. No said in this way means something like
I disagree, but theres more to think about; colloquially, it is often preceded
by Well to help reduce the sense of total disagreement. The symbol for the
fall-rise is v, a combination of the fall and the rise symbols, and the pitch
pattern looks something like:
high
mid-high
(4.19) vNo mid
mid-low
low
fall-rise (4.19)
With a tail, as in Never, there is again an (a) and a (b) form:
(4.20a) (4.20b)

Never Never
but a single means of symbolizing them: vNever
The first, stressed, syllable remains the most prominent and is thus
identified as the tonic, but the fall-rise movement is spread over the whole
tonic segment (i.e. tonic + tail). The distance between the fall and the rise
elements is increased the longer the tail is. If (4.21) is said with a fall-rise tone
accompanying the tonic syllable / (meaning As far as I am concerned, in
contrast to what others may think), the fall accompanies the tonic syllable I
and the rise is delayed until the final stressed syllable of the tail, producing a
kind of split fall-rise:
(4.21) v j dont believe it v /

I dont believe it
The fall-rise movement is spread over the whole of the tonic segment; the
rise is split off from the fall and accompanies either the final stressed
syllable, or if there is not one, the final unstressed syllable. Here is another
example of a split fall-rise, with stressed syllables in the tail:
(4.22) I dont v want to have to think about it

(I dont) want to have to think about it

and another example of split fall-rise, with no stressed syllables in the tail:
(4.23) Unvfortunateiv

(Un)fortunately
79
The intonation systems of English

3. Tones and the status of information

We have seen that when a speaker says something, they have to manage the
organization of information. This management entails the division of the
whole information into separate, consecutive pieces, which are represented
in speech by units of intonation (tonality): secondly, it entails the realization
of focus, either broad or narrow, in terms of new and given information,
which is all represented in speech by a choice, in the placement of the tonic
syllable (tonicity); and now, thirdly, we shall see that the management of the
organization of information entails the recognition of a status accorded to
each piece of information, which is represented in speech by choices in the
tone system.
Halliday is the linguist who has drawn most attention to this
phenomenon, although he never used the term status. He referred to
major and minor information (Halliday, 1967: 37) and various kinds of
dependence that one unit of information might have on another (ibid.:
34-5).3

3.1 Falls and rises

We noted above that one possible meaning of the rising tone is This is not
all I want to say; Im going to continue. There is widespread agreement to
this sense of information not yet being completed.4 One function of the rise,
then, is to indicate incomplete information.
A common sequence of tones in a pair of intonation units is a rise in the
first, to indicate incompleteness, followed by a fall in the second, to indicate
completeness, e.g.:
(4.24) He simply got /u |and went \ home
If you say the first unit of (4.24) with a rise, your listener will expect you to
continue; if you dont continue, the information will certainly sound
incomplete, left, as it were, hanging in the air.
Another common sequence of tones in a pair of intonation units is the
reverse: a fall in the first, and a rise in the second. In this case, the rise cannot
mean incomplete, because it does, in fact come at the end, e.g.:
(4.25) But he didnt go to \ bed |though he was /tired
This sequence is what Halliday termed major and minor information. The
main piece of information is contained in the first unit, and the second unit
contains an extra piece of information; in the speakers perception of the
message, not going to bed is treated as the major piece of information, with
reference to tiredness as minor, representing merely circumstantial
information.
We cannot label the fall, therefore, as necessarily indicating complete
information, since that is not the case in (4.25), even though it is in (4.24).

80
 Tones: the tunes o f intonation

Major/minor is one system in information; complete/incomplete is another: a

fall represents either major or complete and a rise either minor or
incomplete. One suggestion is to combine the status of major and complete
into a single category of primary information, and to combine minor and
incomplete into a single category of secondary information. There is a slight
problem with this otherwise neat solution: there is a variation to the rising
tone in incomplete information that is not available in minor information.
The variation to the rising tone indicating incomplete information is that
instead of a rising pitch on the tonic syllable there could be a slight jump up
to a mid-pitched level tone. If you count to four, you may either have a rising
tone on each of the first three numbers (and a fall on fo u r) or you may keep
a mid-level pitch on each of the first three, symbolized in (4.26b) by a
horizontal line at mid level:
(4.26a) /one | /two | /three | \four
(4.26b) - one | - two |- three | \four

This alternative type of rise, strictly speaking, runs counter to the definition
of tones given so far, in that there is no movement of pitch. However, its
existence is attested not only by Halliday, but by others too.5 It is recognized
here as a tone on secondary evidence: that it occurs on the tonic syllable,
and that it is an alternative to a variety that fully conforms with the definition.
This mid-level variety only occurs in a non-final unit of intonation; and it
is used as an alternative to the true rise to indicate incomplete information.
This is why intonation analysts are reluctant to consider incomplete and
minor information as a single category.
Notice that in both varieties of (4.26), the rise, or the mid-level tone,
indicates that the counting is incomplete, and that the fall indicates that it is
complete. (If you were counting to five, then fo u r would have an indication
of incomplete, of course, and/zt'e would take the fall.) This is how we intone
lists, e.g.:
(4.27) I lost my / passport | my /tickets | my /money |and the letter
for Mr \Tan

Notice how the mid-level tone is often used to indicate misfortune:

(4.28) I lost my - passport | my -tickets | my -m oney |that -letter |
the \lot
The mid-level tone may possibly establish itself as a separate contrasting
tone, say in another generations time, in which case we would have to revise
the description of English intonation. Just as the pronunciation of
consonants, vowels and word stress changes over a period of time,
intonation also gradually changes, and we may well be witnessing a change
in this respect, especially as the older intonation descriptions, like

81
The intonation systems of English

Armstrong and Ward (1926) and Palmer (1922), did not include reference to
a mid-level tone at all.
We can now recap. In a close sequence of intonation units, a rise indicates
either incomplete or minor information, depending on whether it precedes
or follows the fall. You will also recall examples from Chapter 3, in which a
final adjunct can either be contained in the intonation unit of the clause
preceding it, or it can have its own unit. In the latter case, the adjunct will
typically take a rise, as minor, circumstantial information; see, for example,
the following from Chapter 3:6
(3.42a) Weve got some \ books | / here
(3.43c) Lets \go | / now
(3.44c) Thats \aN |for to/dav
(3.45c) I saw \John | /yesterday (or: yester/day)
(3.55b) Were in poor \shape |eco/nomicallv
(3.56b) They're \ here | /John
(3.57b) He shouldn't have \done it |the /fool
(3.58b) Dont \run | he /said
(3.59c) He \drives | /normally
(3.60c) I couldnt \soeak to him I /frankly
. A problem of identification may well have occurred to you: how can you
tell the difference between a sequence of fall-plus-rise, as in the above, and a
split fall-rise that is spread over a number of syllables? Some cases of the fall-
rise tone are not problematical: if the pitch movement is contained in a single
syllable, it is a single tone, not a sequence of two: if the pitch movement has
a rise on a final unstressed syllable, it is the single fall-rise tone, because the
rise in the fall-plus-rise sequence would have to co-occur with a tonic, i.e.
stressed syllable. The problem arises when the rise occurs on a stressed
syllable: is it the second tonic in the fall-plus-rise sequence or is it the rise
component of a split fall-rise? The phonetic answer may not seem too helpful
at first to a novice: a single fall-rise tone has only one tonic syllable and the
rise component ends at a high level: the fall-plus-rise sequence has two tonics
and the rise ends at approximately mid level. The semantic answer may be
more helpful: the fall-rise tone relates to one piece of information in one
intonation unit, whereas the fall-plus-rise sequence relates to two pieces of
information in two intonation units. The question to ask, therefore, is: does
the rise element constitute the kind of information that can be described as
circumstantial? For example, in (4.21), the rise element occurs with believe it:
that cannot be considered, syntactically, as circumstantial information,
because it is the main verb: but what about (4.22)?

82
 Tones: the tunes o f intonation

The original (4.22) was given as an instance of a single fall-rise tone, which
in this case signified something like Whereas I dont want to have to think
about it, it looks as if I shall have to, against my wishes.
(4.22) I dont v want to have to think about it

The rise component ends high. In

(4.29) I dont \want |to have to /think about it

the rise ends in approximately the mid level: the second unit is minor
information compared to the major information in the first: it means
something like Thinking about it is not what I want to do.

3.2 Fall-rises

It is time now to look more precisely at the meanings of the fall-rise, and as
in the case of the rising tone, the fall-rise has different meanings depending
on whether it precedes a fall, or whether it itself is final.
When a fall-rise tone precedes a fall in a close sequence of two units, it
comes as a contrast with the ordinary rise. Compare (4.30), with a rise simply
indicating incompleteness, with (4.31), with a fall-rise:
(4.30) In the /kijchen |youll find a su rxprise
(4.31) In the vkitchen ] youll find a su rxprise

(4.30) and (4.31) represent a typical case of marked theme (see Chapter 2, p.
37), but whereas the rise in (4.30) merely leads on to the major information,
the fall-rise in (4.31). highlights the theme itself. The non-final fall-rise has this
effect on the theme of a clause - it highlights it. Here is another pair of
examples:
(4.32) Un/fQrtunately | he cant x come
(4.33) Unvfortunatelv | he cant x come

The rise in (4.32) merely acts as a means of connecting the comment adjunct
as theme to the remainder of the clause; in (4.33), the fall-rise draws
attention specifically to the comment adjunct itself - it is no longer merely a
connective, it is a comment in its own right and obviously gives expression
to some thought in the speakers mind (and if we knew, or invented, a.
context, we could speculate on what that thought might be).
This use of the fall-rise is not restricted to cases of marked theme,
however; its use highlights cases of neutral theme too:
(4.34) v i | x can come
(4.35) My v brother |wouldnt even x dream of it
In these examples, the clause subjects are the themes, and a speaker can
highlight them, if desired, by separating them off as a separate intonation

83
 The intonation systems o f English

unit from the unit containing the predicate. Thus a message in both cases is
managed as two, separate, pieces of information - theme and rheme, but
with the theme specially highlighted.
A whole clause can act as theme, when it precedes another; the first theme
is usually subordinate to the second, independent, one.7 Take (4.36) as an
example:
(4.36) If she /comes ] thatll make \five of us
A speaker could highlight the theme with a fall-rise:
(4.37) If she v comes |thatll make \five of us
or even
(4.38) vjf she comes |thatll make \fiye of us
In (4.38), the I f is particularly highlighted. (Notice that the fall occurs on the
tonic I f and the rise component is delayed until the (final) stressed syllable:
in this case, com es.)
And now we consider the use of the fall-rise in final position. It is generally
agreed8 that such uses of the fall-rise indicate some kind of implication.
Halliday once glossed the meaning as there is a but about it (Halliday, 1967:
27); he extended the meanings to include reservation, contrast, personal
opinion offered for consideration, and concession (Halliday, 1970: 26-8).
Here are some examples:
(4.39) Its v cheap (reservation: but thats not the only thing that is true
about it)
(4.40) It v looks expensive (contrast: but is it really?)
(4.41) Its worth convsidering (personal opinion: thats what I think)
(4.42) Let him v think about it (concession: at least, do that)
All of these examples can be summed up for convenience under the term
implication. Kingdon (1938: 59-60) expressed it thus: the fall-rise conveys
some insinuation in making the statement, expecting [the] hearer to
understand more than is said. The speaker does not have to verbalize the
insinuation, but assumes that the hearer can extrapolate the additional
message from the context, the setting or common knowledge. For instance,
if a person in a coffee shop says
(4.43) I dont v like coffee
the* implication might be I dont know why you brought me here, or
Although I dont like coffee, Ill show willing and drink some, or Do you
think they offer an alternative drink?, and so on; the speaker expects the
hearer to interpret the utterance correctly from whatever clues the speaker
thinks the hearer has. The point of using the fall-rise is that the additional
thought does not need to be expressed overtly.

84
 Tones: the tunes of intonation

We noted in Chapter 2, p. 46, a contrast in tonality between

(2.72a) 1didnt come | because he told me

and
(2.72b) I didnt come because he told me

and in a similar pair:

(2.73a) He didnt go |so that he could get promotion

and
(2.73b) He didnt go so that he could get promotion
We remarked then that the negativization of the reason in (2.72b) and of the
result in (2.73b) implied some other, unspecified, reason and result. They
would both be quite naturally rendered with a fall-rise on the tonic:
(2.72b) I didnt come because he v told me
(implication: I came for another reason)

(2.73b) He didnt go so that he could get provmotion

(implication: he went with another expectation)

Finally, a well-known example of a contrast between a fall and a fall-rise

associated with any-words, which was first raised by Lee (1956: 347).
Compare
(4.44) They dont admit \any students

with
(4.45) They dont admit vany students
The clear meaning of (4.44) is that all students are excluded, but what is the
implication of (4.45)? It is that some students are admitted; the implication
might even be that students who can be admitted are special in some way.
See if you can work out the implications of the following:
(4.46) He doesnt compose his songs just vanywhere
(4.47) He doesnt sing to vanyone
(4.48) He doesnt present them vany how (or: vany old how)
(4.49) He doesnt write vgnything
(4.50) He doesnt do it v&ny time
(4.51) He doesnt go vgnyway
The implications in each case refer to certain specific or even special places,
people, ways, things and times.
We are now ready to sum up the choices in the tone system for the
indication of information status. A rising tone before a fall indicates

85
The intonation systems o f English

incomplete information; after a fall, minor information. A falling-rising tone

before a fall indicates theme highlighting; after a fall, or independently, it
indicates an implication, an unspoken message that the hearer is expected to
interpret. A falling tone indicates major information, whether it is preceded
or followed by either a rise or fall-rise. In some respects it acts as a dominant
tone, which can be satellited by other tones. The system is therefore:
major: \
minor: / (in final position)
status of information incomplete: / (in non-final position)
implication: v (in final position)
theme highlighted: v (in non-final position)
and is illustrated as follows:
(4.52) Its suddenly become very \cold (major)
(4.53) Its become very \cold |all of a /sudden (major + minor)
(4.54) Its suddenly become very /cold ... (incomplete)
(4.55) All of a /sudden |its become very v cold (incomplete + implication)
(4.56) All of a v sudden | its become very \cold (theme highlighted +
major)

4. Tones and the communicative functions

The tone system is used in English not only for the indication of information
status but for another purpose, the indication of communicative functions.
By communicative function we mean the purpose for a given utterance
intended by the speaker. For instance, is the speaker telling me something or
asking me? Are they commanding me or requesting me to do something?
The tone system is a guide, in a general sense, to such communicative
functions.
The traditional view^ is that statements - to sound like real statements -
are accompanied by a falling tone, but genuine questions requiring an
answer of either yes or no are accompanied by a rising tone. Commands
have a fall, so do interjections, but question tags have either a fall or a rise
depending on the speakers sense of certainty or uncertainty. Questions
with so-called wh-items are usually accompanied by a fall, too. This
traditional view is largely borne out by extensive intonation analysis, but it
only scratches the surface.
Before we begin a detailed discussion, we will consider briefly a question
that must have come to mind. If the tone system of English realizes two quite
different functions in spoken discourse, how can you tell when a rise, for

86
 Tones: the tunes o f intonation

instance, is indicating information status and when it is indicating a

communicative function? The phonetic answer is easy in theory, but
sometimes difficult in practice: a rise is operating in its informational
capacity when it belongs to an intonation unit that is dependent upon
another. When indicating either incomplete or minor information, its
intonation unit is tied very closely - often with no pause - to another unit;
such a unit acts as a kind of satellite to the other. The same is true of the non
final fall-rise indicating the highlighting of a theme; it is closely attached to
another unit without pause. However, when a rise is operating in its
communicative capacity, it is operating in an independent intonation unit,
which is often separated from preceding and following units by a pause or
even silence, or, of course, by a change of speaker. Furthermore, the rise for
information status is usually confined to neutral rise, from low to mid;
whereas the rise for communicative functions not only uses a neutral rise -
and in that respect is wholly indistinguishable from its use in information
status - it does have the potential for high and low variations to add an
attitudinal dimension to the questions (or whatever).
In much informal conversation, with its false starts, hesitations,
abandonments, etc., it is often difficult to apply the criteria of dependent
and independent intonation units. For instance, when we considered a
snippet of informal speech in Chapter 1 there was an intonation unit (1.3j)
that was accompanied by a rise, but was preceded and followed by quite
lengthy pauses. However, no one would interpret the unit the gardens a n d
houses but as a question. But why not? For one thing, the whole discourse is
marked by these quite lengthy pauses, which are the consequence of
hesitation, looking for a word, thinking of what to say, etc. Secondly, the
context does not suggest that a question is expected. And, thirdly, the
grammatical structure of the words in the unit does not suggest that a
question is intended. Where the phonetic clues are missing, semantic and
syntactic clues help. It may well be that we rely on the semantic and syntactic
clues more heavily than we do on the phonetic clues in interpreting a unit as
a question; after all, we think we usually know when a question is being
asked. It may well be the case that the phonetic clues provide confirmatory
evidence rather than definitive evidence, especially in unrehearsed,
spontaneous, informal Speech.
From that discussion, you will notice that falling tones are associated not
with satellite, dependent, units of intonation, but with the more central,
nuclear, independent, units. This befits its designation of bearing major
information. This particular status of the fall is bolstered by the observation
that usually between half and two-thirds of all tones are falls. (We noted in
Chapter 1 that prepared or rehearsed discourse has a relatively high
proportion of rises, which demonstrates the planning of. incomplete

87
The intonation systems o f English

information with complete, but even so, half the tones are falls; the higher
proportion of falls is found in unprepared, unrehearsed speech, up to 65 per
cent.) Falls dominate, in both frequency and function, and this dominance is
manifest in the tone system in communicative functions, too.
The dominance of the falling tone is manifest in two ways: first of all,
even for the brief introduction to communicative functions given so far,
there are more functions that use the fall than use the rise.

Communicative functions: with fall;

statements /
yes/n o questions V
zc/z-questions V
question tags . V V
commands V
interjections V
Secondly, the fall is associated with the speaker knowing something, telling
something, and in the case of interjections, expressing their own feelings:
the speakers knowledge, authority and feelings dominate. On the other
hand, the rise is associated with the speaker not knowing and therefore
having to ask. The difference in the use of falls and rises in question tags
mirrors that distinction: a fall represents the speaker being pretty sure and
the rise the speaker being unsure. Generally speaking, a fall represents
dominance and a rise deference.
But, you may ask, what about those questions that begin with ioh-woTds
(who, whose, whom, what, which, where, when, why, how ) ? Surely, they are
questions that indicate the speakers lack of knowledge, but they are
accompanied with falls - and falls are associated with knowledge. The
answer is quite simple. In the case of the ^-questions it is only one part of
the whole proposition that is unknown. For instance, if I ask (example 1.5 in
Chapter 1)
(4.57) What are you going to \da tonight

it presupposes that I know that you are going to do something tonight; there
is only one part of the whole proposition that I do not know, but the rest I
do know. If, on the other hand, I ask
(4.58) Are you going to /do anything tonight

I am indicating that I do not know if the proposition (that you are going to
do something tonight) is valid or riot. In the wz/z-question, I know that you
have a plan; in the yes/n o question, I dont know.
Here is another example:
(4.59) When do you elect the Student Union \ President

88
 Tones: the tunes o f intonation

If I ask the question in the form of (4.59), it shows that I know there is to be
an opportunity to elect the Student Union President; the only thing I do not
know is when - but I know that the basic underlying proposition is valid. But
if I ask
(4.60) Do you elect the Student Union / President

then it indicates that I do not know about the system of Student Union
presidential elections; I do not know whether the basic, underlying, pro
position is valid or not.
The fall in (4.57) and (4.59) indicates my knowing; the rise in (4.58) and
(4.60) indicates my not knowing and furthermore my deference to the
knowledge that I presume my interlocutor possesses. So, although it may at
first seem that a falling tone in a ^-question counters the general principle,
it does in fact confirm it.
A fall indicates the speakers certainty or dominance in respect of
knowledge, authority and feelings; a rise indicates the speakers uncertainty
or deference to the knowledge - and, as we shall see, the authority and the
feelings - of the person addressed.
This explanation of .^-questions having a fall can even be illustrated in
very common questions like
(4.61) Whats the \time
(4.62) What's your \ name
(4.63) Where do you \live
(4.64) How \aie you
The fall in each of these questions represents the speakers presupposition
of the validity of an underlying proposition, even if it is so obvious that to
question it sounds odd: the time must be something, you must have a name,
you must live somewhere, you must be in some kind of condition, and in the
case of (4.65):
(4.65) Who \are you

you must be somebody!

Three points must be noted. The first is that we are, in fact, again handling
a speakers perception of the communicative function, whether that
perception is, in our estimation, accurate or not. A speaker may act as
knowing something, but in fact be quite mistaken, as in
(4.66) Twenty per cent means a \auarter
The speaker might in fact lie, but still acts as the one who knows and tells.
Similarly, a speaker can ask a question even when they know the answer,
what we call a rhetorical question; but the speaker has still got to use the
appropriate rising tone to make it sound like a question:

89
The intonation systems o f English

(4.67) Is that the right way to /do it

The second point is that there is always more than one way to intone an
utterance. What appears on the surface as a statement can be intoned as a
question; (4.66) could also be intoned as follows:
(4.68) Twenty per cent means a /quarter
meaning something like a genuine, but challenging, question: Is that what
you really think? And (4.67) could be intoned with a fall:
(4.69) Is that the right way to \d& it
meaning something like a genuine, but challenging, statement: I think that
is the wrong way to do it.
Probably every example that has been given so far is capable of being
intoned differently. But there is at least a tendency for declarative clauses,
^-interrogative clauses, imperative clauses and interjections to be intoned
with a fall to effect, respectively, statements, ^/z-questions, commands and
expressions of personal feelings; and for polar interrogative clauses to be
intoned with a rise to effect a yes/no question.
The third point is an extension to the second. Although there is an
observable tendency for declarative clauses, etc., to be intoned in a partic
ular way, it is equally observable that clause types and communicative
functions do not always match - as in (4.68) and (4.69) above. The tone
system, however, always indicates the communicative function, whatever the
clause type may be. It would be quite wrong to suggest that declarative
clauses are realized by falls; it is not the clause that is realized by a tone, it is
the communicative function. (4.68) and (4.69) above illustrate this. We can
also illustrate this with a consideration of the intonation to the following
imperative clauses:
(4.70) Stop \talking
(4.71) Take that silly \ look off your face
(4.72) Have a cup of \tea
(4.73) Stir \well
(4.74) Have a nice /time
(4.75) Say that a/gajn. (and Ill \hjt you)

(4.70) and (4.71) would no doubt be regarded as commands, and the

speakers authority is expressed in the choice of the fall. (4.72) and (4.73) are
also imperatives with a falling tone, even though they are not, strictly
speaking, commands; (4.72) might be interpreted as a recommendation,
(4.73) as an instruction, but in both cases, the speaker still considers him- or
herself to be the dominant partner in the exchange. (4.74) would seem
strange with a fall, because we know that we cannot command a person to

90
 Tones: the tunes of intonation

have a nice time: such a result is outside the speakers control; it is, in fact, a
wish, and because it is the listeners feelings that are affected, not the
speakers own feelings, a rise is the more appropriate choice of tone. (4.75)
is a threat; it is the opposite of a command, the speaker does not wish the
listener to repeat what they have already said. The rise, in this case, might
simply be an indication of incompleteness, because the substance of the
threat is yet to come.
It might also be noted that not all commands are necessarily issued in
imperative clauses, but they must be accompanied by a fall to indicate the
speakers authority, e.g.:
(4.76) Thou shalt have no other gods be\fore me
(4.77) All library books must be returned by \Fridav
(4.78) Silence in \court
(4.79) A\wav with you all

5. Dominance and deference in communicative functions

We have seen that the traditional description of the intonation of
communicative functions associates falls with statements, ^-questions,
commands and interjections - all displaying the speakers dominance in
respect to information, authority and the expression of personal feelings -
and rises with y e s /n o questions - displaying the speakers deference to the
hearers assumed knowledge.
There are, however, very many more communicative functions than those
just listed. The traditional list is based on the types of clauses recognized by
grammarians: declarative for statements, two types of interrogatives for two
types of questions, imperatives for commands, and verbless for interjections.
We have already made the point that there is, in fact, no one-to-one
correspondence between clause type and communicative function, that
declarative clauses can be used for questions, interrogatives for statements,
and imperatives for wishes, etc., and the second point that tones realize the
communicative functions rather than the clause types. We have also, on a
number of occasions, hinted at another point: that there are many more
communicative functions than those recognized in traditional grammars.
We have made reference to wishes, requests, instructions, recom
mendations, acknowledgement of the listeners personal feelings, and other
functions like greeting, bidding farewell, thanking, apologizing, con
gratulating, and so on - all things we use language for.
Now, there is no special tone for each separate communicative function;
the resources of intonation are simply not sufficient for so many different

91
The intonation systems o f English

functions. The tone system simply indicates the speakers status vis-a-vis the
hearer: either as dominant or deferent. All the communicative functions can
be grouped into three kinds: relating to knowledge in respect to
information, reality and belief; to authority in respect of influencing other
peoples action (suasion); and to social interaction. We will now consider
each of these three groups of communicative functions and show how the
tone system operates to indicate dominance and deference.10

5.1 Information, reality and belief

In the area of information, reality and belief, there is a little more to be said.
A fall indicates the speakers dominance (knowledge) and a rise their
deference to the presumed superior knowledge of the addressee. This is
best seen in pairs of contrasting intonations units, as in (4.66) and (4.68) and
in (4.67) and (4.69). The case of tags illustrates the general principle well:
(4.80) Hes finished with my \book | \ hasnt he

which sounds as if the speaker is pretty sure of the fact, and

\
(4.81) Hes finished with my \ book | / hasnt he

which sounds as if the speaker is not so sure.

In a study11 of the succession of communicative functions in an ordinary
piece of spontaneous informal dialogue the following functions were noted
with falling tones: statements, answers, explanations, reports, agreement,
acknowledgement, descriptions, suppositions, hypotheses, deductions.
They all presuppose the speakers dominance in knowing and telling.
The functions with an accompanying rise were y e s/n o questions, appeals,
and requests, which are all functions in which the speaker acknowledges
knowledge, or authority, in the addressee.
Other communicative functions that display the speakers knowledge,
and therefore are accompanied by a fall, are denials, affirmations and, of
course, disagreement. The expressions of doubt and hope are, like
suppositions and hypotheses, expressions of belief rather than knowledge;
in these cases, the speakers dominance is realized in the observation that it
is the speakers doubts, hopes, suppositions, hypotheses, etc. that are
expressed. Expressions of doubt, for example, are typically accompanied by
a falling tone:
(4.82) I doubt if he would \come
(4.83) I wouldnt have \thought so
(4.84) I cant imagine he \would

It is sometimes argued that the fall-rise properly expresses doubt. But this
is not the case. In (4.82), doubt is expressed lexically, and in (4.83) and
(4.84) , by other wordings. It is true that these three utterances could have

92
 Tones: the tunes of intonation

had a fall-rise, but the fall-rise itself does not mean doubt - otherwise (4.82)
would have to be considered as doubly doubting. The fall-rise simply means
that an extra message, an implication, is in the mind of the speaker, e.g.:
(4.85) I doubt if he would v come
(4.86) I v doubt if he would come

If we knew, or invented, a context, we could speculate on possible

implications: perhaps for (4.85), So wed better select someone else as goal
keeper, and for (4.86), but, of course, I cant be sure.
Similarly, it has been argued that the expression of possibility is indicated
by a fall-rise. Again, this is not so. Possibility can be indicated lexically as in
(4.87) or grammatically as in (4.88):
(4.87) Its \ possible
(4.88) I \ might be able to play

The expression of possibility is quite simply accompanied by a fall. As with

doubt, a fall-rise merely indicates implication: thus
(4.89) Its y possible

and
(4.90) I v might be able to play

might imply but I cant be certain.

But what difference would a rising tone make in these instances?
(4.91) Its / possible
(4.92) I / might be able to play

The communicative function is now quite different. Imagine that somebody

presents the proposition that the Monster Raving Loony Party could win a
local by-election, and your answer is (4.91). What you are doing is conceding
to your addressee that the proposition might be valid; you are deferring to
the possible validity of the addressees statement. Or, in the case of (4.92),
you are unexpectedly invited to participate in a football match as the
goalkeeper in a weeks time; you concede the possibility. A concession is a
kind of deference to somebody elses proposition; as such, it is signalled by
a rising tone.
Contradictions are also signalled by a rising tone. You may wish to contest
somebodys claim that you forgot all about some arrangement, by saying
(4.93) Oh no I /didnt
Contradictions are regularly accompanied by a rise; it seems to signal that
the other person may have grounds for making an assumption, e.g., in this
case, your non-appearance at the meeting, and you concede that.

93
The intonation systems o f English

An excellent example of contradiction is found in the traditional

pantomime where at one point a character at the front of the stage makes a
claim which the audience can see is mistaken. There is usually a baddy at
the back of the stage, and the audience, especially the children, are
encouraged to contradict the first characters claims, in a form like
(4.94) Oh yes there /is
or
(4.95) Oh no he /isnt

Challenges also regularly take a rising tone. In one sense, a challenge, like
a contradiction, is a concession to the other persons statement of fact, but
there is an element of enquiry about it, too: Are you sure of your facts? In
response to someones statement that her grandfather is travelling to
Timbuktu, you might respond:
(4.96) Your grandfathers travelling to Timbuk/tu

Because a challenge, by its very nature, is a strong expression, it is often

accompanied by a rise to a high level.
Concessions, contradictions and challenges all typically take a declarative
clause structure, but because of the sense of deference inherent in these
communicative functions, a rising tone is appropriate. Challenges can, in
fact, take the form of any clause type, because, in essence, a speaker who
challenges uses the wording of the original, and the rise indicates as much
as anything else: Is that really what you meant to say? If the original was
Lets go to the beach, the challenging response might be
(4.97) Go to the beach on a day like /this
The term echo question is often used when an interrogative clause is being
challenged. If you are asked if you are going to meet somebody whom you
happen to dislike, you might respond
(4.98) Am I going to meet / him
with the meaning of Can you really be serious in asking me such a
question? Wz-echo questions are common, with the tonic on the wh-iiem
itself:
(4.99) /What did you say youre going to do
(4.100) / When is he coming

which can either have the force of a challenge or be a genuine request for a
repetition of something you mistook or misheard.
Rises, of course, are most typical of enquiries, elicitations for information
on the validity of a proposition, which are usually structured as polar
interrogatives:

94
 Tones: the tunes of intonation

(4.101) Are you going a/wav for Christmas

But just as rises can accompany declarative clauses, falls can accompany
polar interrogatives. There is a kind of question that a speaker can employ
not so much to make an enquiry, rather to guide the listener to taking a
course of action or agreeing to an idea.
(4.102) Are you \satisfied now

really means I think you should be ..., and

(4.103) Can you pay me by \cheaue please

really means Thats what I want you to do. These conducive questions are
not genuine enquiries after information, but are opinions stated in disguise.
Guesses often take the form of a polar interrogative; but take a falling tone:
(4.104) Have you heard all this bexfore

If (4.104) had been accompanied with a rise, it would have been interpreted
as a genuine question, but the choice of a fall indicates the speakers guess
that the other person has indeed heard all this before.
Denials can follow the same pattern. We have seen above that a denial
takes a falling tone, because the speaker is sure of the facts. Denials may
appear in polar interrogative form as well as declarative, but the sense of
enquiry of the polar interrogative is countered by the fall:
(4.105) Have I ever let you \ down

meaning, of course, [I have never...12

Exclamations, as we noted in Chapter 1, often take the form of a negative
polar interrogative, e.g. Isn t it hot. A frequently used expression of
exclamation when looking at a young infant is
(4.106) Isnt he like his \mother

The falling tone makes this utterance unmistakably an exclamation.

Two other typical communicative functions that combine polar
interrogatives and falling tones are prompts and lead-ins. A prompt takes the
form of a question, but its real intention is to get someone to comment on a
matter. This is a technique often used by people interviewing on radio and
television; here is an example from Browns (1977) broadcast data:
(4.107) Do you bexlieve that prices can be curxtailed

What the interviewer was doing was prompting the politician to express an
opinion, i.e. I want you to tell me whether...
Lead-ins are similar. A preliminary utterance to telling a joke often takes
this form:
(4.108) Have you heard,the joke about the two poli\tjcians

95
The intonation systems o f English

What this really means is Im going to tell you a joke ab o u tS im ilarly , an

utterance like
(4.109) Do you know about Mr \Evans

said without giving you a chance to reply is in fact just another way of saying
Let me tell you about
There are three other types of questions that regularly take a falling tone.
The first is a repeat question, as in
(4.110) A: Do you want a cup of /tea
B: /Pardon
A: Do you want a cup of \tea

The fall in the response to Pardon seems to suggest something like What I
said was ..., which is, of course, a statement. A well-known elementary
course book for teaching English as a foreign language actually began with
such a dialogue (see Figure 4.1).
The second is an alternative question, as in
(4.111) A: Do you speak / French
B: \No
A: Do you speak \German then

The fall in the second interrogative seems to suggest Well, I will try... which
is, of course, a statement. This is similar to a list question, as in
(4.112) Do you speak / French [ /German |or Ktalian
where the rises seem to indicate incomplete items in a list, and the fall the
final item (see p. 36).
Finally, there are second-attempt questions, as in
(4.113) A: Are you /going
B: I dont \ know
A: Do you \want to go

The fall in the second interrogative suggests that the questioner is stepping
back in the pursuit of finding a basis to proceed, meaning something like
Well, what I need to know is ..., which is, of course, a statement.
We have seen, then, that a falling tone indicates the speakers knowledge
and certainty in respect of information, reality and belief: they know, and
tell. A rising tone, on the other hand, indicates a deference to the other
persons presumed knowledge: they dont know, and ask; or they
acknowledge some degree of validity about the other persons knowledge.
In general terms, a fall indicates the speakers dominance, a rise the
speakers deference.

96
 Lesson 1

Excuse me!

Yes?

Is thfs your handbag?

Pardon?
ft

Is thfs your handbag?

Yes, it is.

Thank you very much.

Figure 4.1

97
 The intonation systems of English

5.2 Suasion
In a similar way, a fall indicates dominance in the realm of suasion, i.e.
influencing peoples action; a fall indicates the speakers authority. Whether
there is justification for such authority, or not, is not the issue: the speaker
can act as one with authority. A rise indicates deference to the other persons
authority or decision. The difference between a command and a request
illustrates this, A person can only command if they have the authority to do
so - whether the authority is legal, moral, physical, etc. With a request, a
person asks another to do something, leaving the other person the ultimate
decision to act, or not.
Compare, for instance, the difference between
(4.114) Turn the \radio off

and
(4.115) Turn the /radio off
The second sounds much gentler, more polite. Indeed, a person who regards
themselves as having authority, may nevertheless decide to project
themselves as less authoritative by deliberately choosing to use a rising tone.
In general terms, then, in the realm of suasion a rise indicates deference to
the other persons authority and decision: they dont decide, and must ask.
We will now consider other communicative functions from this point of
view. A parent exercising authority over a child might say:
(4.116) Dont talk with your \ mouth full
On the other hand, that parent could try to achieve the same result by a
different tactic:
(4.117) Dont talk with vour / mouth full
A straight prohibition requires a fall: a plea, a rise.
What is the difference in tactics between (4.118) and (4.119)?
(4.118) Come \ on
(4.119) Come /on
The first could be interpreted as a demand: this is what you have to do; the
second is an example of coaxing: I do wish you would ... The first has a ring
of authority about it; the speaker has decided what you should do, and tells
you. The second has a ring of wishing or pleading: it may indeed be uttered
with a degree of forcefulness, but the speaker is coaxing rather than
demanding. Commands, prohibitions and demands suggest speaker-
dominance: requests, pleas and coaxing suggest deference - but, tactfully
deployed, may achieve the results desired by the speaker!
Advice and recommendations are usually accompanied by falls because
they contain a degree of authority on the part of the speaker. Suggestions

98

k i?
 Tones: the tunes o f intonation

and invitations take a rise for, by their very nature, they allow the other
person the final decision. Let us consider some examples:
(4.120) You should take a little \ break
(4.121) You could take a little / break
The combination of the fall with a clause containing should will certainly be
interpreted as advice: Thats what I think you should do. The rise and could
are more likely to be thought of as communicating\a suggestion: Thats one
possibility you could consider doing. The whole idea of advice contains the
element of respect for the authority a person has. Suggestions are alternative
courses of action, from which the other person can select; a suggestion may
also be a single course of action for the other person to consider: either way,
it is the other person who is left with the decision.
Now, if I follow the advice of (4.120) or the suggestion of (4.121), I might
possibly ask Well, what shall I do? The reply might be (4.122), or possibly
(4.123):
(4.122) How about a cup of \coffee
(4.123) How about a cup of /coffee
In this pair, the wording is parallel, but again the fall of (4.122) expresses a
recommendation and the rise of (4.123) an invitation.
Recommendations and invitations follow the same pattern, with the
speakers dominance in the former reflected in an accompanying fall, and
the speakers deference to the other persons decision in the latter in a rise.
(4.124) You ought to trv this new \coffee
(4.125) Would vou like to trv this new /coffee
(4.126) Have a \go
(4.127) Have a /go (= 'Would you like to ...?')
However, if the lexical verbs suggest, invite, request, p le a d are used, they
are not accompanied with a rise:
(4.128) I suggest a cup of \coffee
(4.129) I would like to invite you for \dinner
(4.130) We request the pleasure of your \companv
(4.131) I'm \ oleading with you
Why do these take a fall, when, after all, they actually contain the verbs that
refer to the communicative function? One answer might be their greater deal
of formality, but whatever the degree of formality, that should not alter the
basic type of communication that they are. A better answer is that these
utterances are really announcements of a suggestion, an invitation, a request
and a plea; announcements derive from a speakers decision: This is what Im

99
The intonation systems of English

going to do - Im going to And speakers decisions are indicated by falls.

Promises take a fall; the speaker has decided on a course of action and
seeks to guarantee it:
(4.132) Ill let you know to\morrow
(4.133) We promise you a quick rexolv
Threats take a fall, too; a threat has all the features of a promise, but with
a negative effect on the person addressed:
(4.134) (Say that a/gainl and Ill xhrt; you
(4.135) Dont vou \daretell lies
Offers take a rise: the speaker offers a course of action, but allows the
other person to decide:
(4.136) Can I / help you in any way
(4.137) Another cup of /coffee
Warnings also take a rise; although the speaker warns, it is the person
warned who has to take the decision; the speaker recognizes this with the
choice of the rising tone:
(4.138) Look where youre /going
(4.139) / Careful

Offers and warnings can be announced by using the lexical verbs offer
and w arn ; and as announcements, they are likely to be indicated by falls:
(4.140) We offer the best service in the \town
(4.141) Im \waming you
Appeals take a rise. Appeals are attempts by a speaker to get the other
person to reconsider a course of action, e.g.:
(4.142) (A: Ill have to throw this coffee axway)
B: You dont have to do /that

(4.143) (A: But the coffee tastes \awfuh

B: Its not /so bad
But the intention of an appeal is often to get the other person to consider
a course of action favourably; this kind of appeal often takes the form of you
see, y ou must understand, e.g.:
(4.144) You /see (Ive xgoi to pay him today)
(4.145) You must under/stand (we cant afxford a holiday)

The appeal may be directed at the other persons general knowledge, or

particular local knowledge, or knowledge of a particular word; this kind of
appeal usually takes the form of y ou know, e.g.:

100
 Tones: the tunes o f intonation

(4.146) We wanted to go to - you / know |that little place in \Jtaly |thats

independent

(4.147) (A: Im going to get the \tickets)

B: The /tickets
A: You / know |the tickets for the \circus

(4.148) Then they wanted to - you / know what I mean |what they did to
\Nixon
The use of the lexical verb ap p eal, however, as in the other cases above,
suggests an announcement rather than an appeal; hence a falling tone:
(4.149) Im appealing to vour common \sense
Thus, in general terms, communicative functions that display the
speakers dominance - authority, decision-taking, announcing - are
accompanied by a fall; those that display the speakers deference to the
others authority or their right to decision-taking take a rise.

5.3 Social exchanges

The third category of communicative functions is social exchanges. These
kinds of communication do not involve either the provision or elicitation of
information, nor are they primarily intended to influence other peoples
action; they are simply intended to establish and maintain relationships
between people; for instance, greeting and bidding farewell, introductions,
attracting attention, apologizing, sympathizing, wishing, thanking, and so
on. Once again, some of these functions seem naturally to require a fall,
others a rise, and yet others either. Let us take greetings and farewells as
examples.
Farewells are typically accompanied by a rising tone, e.g.:
(4.150) Good/bve
(4.151) Cheeri/p
(4.152) See you a/gain

However, if the parting is regarded as only temporary, a fall accompanies the

farewell, as if a piece of information is being given, e.g.:
(4.153) See you this \evening

Greetings take either a fall or a rise, e.g.:

(4.154) Good \morning
(4.155) Good / morning
What is the difference? By now, you should be expecting the fall to indicate
something akin to dominance, and the rise to deference; but in social

101
The intonation systems o f English

exchange, what can dominate, and what can you defer to? In general terms,
the answer is feelings. Falls tend to focus attention on the speakers feelings,
whereas the social functions in which the other persons feelings, or
attention, are in focus, are expressed by means of the rise. In farewells, we
attend to the other persons feelings: we are thinking of them, rather than
ourselves, when we bid them farewell. If you say G oodbye with a fall, you
invoke a very different situation:
(4.156) Good \ bye

It sounds very much like a dismissal in which the speakers feelings

dominate. The speakers feelings dominate even more provocatively in
(4.157) Good \riddance

which will only take a fall - a rise in (4.157) would sound like mocking.
But what is the difference between (4.154) and (4.155)? A rise in a greeting
seems to suggest an interest in the person(s) addressed; OConnor and
Arnold (1973: 66) describe it as bright and friendly. This is not meant to
suggest that the fall is unfriendly or unconcerned with the person(s)
addressed; all it means is that the greeting with a fall is a plain greeting,
possibly a little formal: I am greeting you. The rise seems to mean I am
acknowledging you as I greet.
The same difference can be detected in thanking. A fall means: I am
thanking you:
(4.158) \Thank you

A rise seems to mean I am acknowledging you as the one I thank:t3

(4.159) /Thank you

Greetings and thanking seem to take a fall as a neutral tone, but a rise to
indicate an interest in the feelings of the addressee. Greetings on the phone,
however, generally take a rise:14
(4.160) Hell/o

The speaker seems to be acknowledging the initiative taken by the caller,

and thus defers to the callers attention.
Welcomes take a fall; the speakers sense of welcome seems to dominate:
(4.161) \ Welcome
(4.162) How nice to \see you
(4.163) Come on \m

Good wishes take a rise; inevitably it is the feelings of the other person(s)
that are uppermost in the speakers mind:
(4.164) Happy / birthday

102
 Tones: the tunes of intonation

(4.165) Have a nice /time

(4.166) 1hoDe vou /Dass
(4.167) Do your /best
(4.168) Bon voy/age
(4.169) Remember me to your /father
(4.170) Give mv love to the /children ;

Toasts and congratulations take a fall; they are announcements of the

speakers feelings about somebody elses success or happiness. Typical
toasts are:
(4.171) Good \health
(4.172) To the bride and \groom
(4.173) Heres to a wonderful old \lady
(4.174) \Cheers

and congratulations:
(4.175) Well \done
(4.176) Congratulations on vour en\gagement
(4.177) 1was delighted to \hear about it
(4.178) May we congratulate vou on your recent apxppintment
Praise, appreciation, approval and disapproval are all accompanied by a
falling tone: it is the speakers feelings that dominate, e.g.:
(4.179) Thats \ great
(4.180) Its a \ lovely thought
(4.181) You shouldnt have gone to so much \trouble

Expressions of regret take a rising tone, as befits a situation in which the

feelings of the other person(s) are in focus:
(4.182) Im ever so /sorry

Apologies also take a rise, as they are regrets over what the speaker has
done:
(4.183) 1do beg your /pardon
(4.184) We wont let that happen a/gain

Sympathy also takes a rise, as a kind of regret over what has happened to
the other person(s):
(4.185) Thats a /pity
(4.186) Its a great /shame

103
The intonation systems o f English

(4.187) I was sorry to hear about your fathers /accident

When regret is communicated via exclamations, however, the exclamation

retains its falling tone:
(4.188) What a \njty
(4.189) How \awful
(4.190) What a terrible \shock

A request for forgiveness follows the normal pattern for requests, with a
rise. And granting forgiveness or reassurance is also accompanied by a rise
as the speaker acknowledges the other persons feelings:
(4.191) Thats ail / right
(4.192) It doesnt / matter

A request for attention, however, follows the normal pattern for

implication; in addition to soliciting somebodys attention, there is a
requirement of some service:
(4.193) Exvcuse me
(4.194) v Waiter
(4.195) v Nurse
A vocative can also function as a request for attention:
(4.196) Evlizabeth ,
(4.197) v John
An equivalent with a fall sounds more like a summons, in which a speakers
feelings or sense of authority dominate:
(4.198) ExNzabeth
(4.199) \John
(4.200) \Waiter
or even
(4.201) Exxcuse me
With a rise, they sound like enquiries:
(4.202) E/fizabeth (Are you there?)

Introductions usually take the form of a request (with a rising tone) and
an announcement (with a falling tone):
(4.203) /John | I would like you to meet my \father-in-law
(4.204) Miss /Jones |this is Mr \Evans

And finally, indicating different kinds of attention is also a function of the

104
 !j!

Tones: the tunes of intonation

tone system. This is often referred to as back channel; it is the means by

which one person indicates a degree of attention to what another is saying.
Back channel takes the form of signals like yes, no, oh, mm, spoken quietly
and usually with a fairly narrow range of pitch. A falling tone indicates
agreement and a rising tone a wish for the current speaker to continue:
(4.205) \m
(4.206) /m
An alternative to the rise, a mid-level pitched tone may simply mean: I am
still listening:
(4.207) -m

5.4 Conclusion
In the second half of this chapter, we have seen, from many examples, that
in addition to the role of expressing the speakers choice in respect of status
of information the tone system has a role in expressing the speakers choice
of type of interaction with the addressee. It is a simple system that covers a
wide range of communicative functions; it covers, in fact, all types of
communicative functions: a fall indicates the speakers dominance in
knowing and telling something, in telling someone what to do, and in
expressing their own feelings; a rise indicates the speakers deference to the
addressees knowledge, their right to decide, and their feelings. Calls for
attention require fall-rises because a need for service is implied; otherwise
calls with falls indicate a summons for attention (speakers dominance), and
with rises they indicate enquiries (speakers deference).

Notes
1. The inadequacy of punctuation symbols to. represent intonation is amply illustrated
in the work of many conversational analysts.
2. Palmer (1922) and Halliday (1967) both suggest that the high rise is a primary tone
in contrast with the low rise (described here as neutral). Most others do not make
such a distinction: see Tench (1990: 448-54) for a full discussion.
3. Hallidays scheme, together with Ladds (1980) exposition, are extensively discussed
in Tench (1990: 219-38).
4. This widespread agreement is shared by Armstrong and Ward (1931: 22), Pike (1945:
51-9), Kingdon (1958: 73, 221: prelusory), Crystal (1975: 35: non-final tonic),
OConnor and Arnold (1973: 88-9: non-final), Brazil et al. (1980: 86-90),
Cruttenden (1986:102-3), as well as Halliday (1967,1970).
5. See Halliday (1970: 62-3) for further examples, and also Brazil et al. (1980: 86-90,
for the oblique tone), Crystal (1975: 34-5) and Cruttenden (1986: 102-3).
6. See Halliday (1967: 45) for a similar case.
7. See Halliday (1985: 56-9) for the explication of clauses as themes.

105
The intonation systems of English

8. See Palmer (1922), Armstrong and Ward (1931), Pike (1945: 50), Kingdon (1958:
29-30), OConnor and Arnold (1973: 68-9: contrast, concession), Crystal (1975:
36). Ladd has a long and valuable discussion (1980: 145-62).
9. See OConnor and Arnold (1973), Gimson (1989), Thompson (1981), Baker (1982),
Hooke and Rowell (1982).
10. I am indebted to Leech and Svartvik (1994) for this threefold categorization of
communicative functions.
11. Tench (1990: 318-33).
12. See Hudson (1975) for a further discussion of these points.
13. See Knowles (1989: 195) for a similar treatment.
14. This was noted as early as 1945 by Pike (1945: 68); see also Leech and Svartvik (1994:
para. 358).

106
 Tone variations

1. Attitudinal meanings
We now come to that part of the intonation system that is most familiar to
everyone, but it has been necessary to show how this part actually belongs
to a much broader set of systems. We have already referred on a number of
occasions to that familiar observation about it not being what someone said
that concerned us but the way they said it. The way they said it actually
usually refers to the kind of attitude that is detected in the speakers voice,
whether the speaker was angry, or civil, or grumbling, or enthusiastic, and so
on. This attitudinal function, as has been frequently pointed out, is only one
of the kinds of meaning that intonation is used to express, but it is the kind
of meaning that most people are most aware of.
We have seen that intonation is used for managing the organization of
information - units, focus, status - and, incidentally, for distinguishing,
between various syntactic possibilities; people are less aware of these
functions because they are part of the ideational content of the message. (In
the same way, people are not so aware of grammatical constructions and
lexical choices - we do not, as a rule, think about them; we just use them.)
People are, possibly, more aware of intonations use in communicative
functions, because they have to respond to them in an appropriate way, but
people are most aware of the attitudinal function of intonation, because it
tells them something about the person who is speaking.
What are the kinds of variations to the tones that are used for attitudinal
purposes? There are variations to the degree of falling, rising and falling
rising, either greater or lesser than the neutral forms, and there are variations
in the pitch movements in the pre-tonic segment, i.e. the head and the pre
head. The two kinds of variations are called, as we noted in the previous
chapter, secondary tones, and the role they play can be described as follows:
once a person has decided to communicate in speech (rather than writing)

107
 The intonation systems o f English

and has formulated a message (in terms of lexis and grammar), that person
then has to encode that message phonologically in terms of
1 the consonants, vowels and stress patterns of the words
2 the rhythm and intonation of the syntax of the clauses
3 the units, focus and status of the successive pieces of information
4 communicative functions
5 and, if desired, by means of the secondary tones, an indication of a
state of mind.
There are, in fact, other considerations also, and a fuller picture will be
presented in the next chapter, but at this juncture it should be noted that the
expression of attitude is an optional element, whereas all the other elements
are essential to the communication of any spoken message. Attitudinal
expression is described here as optional simply because there are modes of
presentation in which it is not generally included; in newsreading, for
instance, it is kept to a minimum; in unison prayer, it is impossible. In
informal, spontaneous conversation, attitudes are expressed, but not
necessarily all the time; there will be periods of time in a conversation when
speakers are merely telling, or reporting, but there will be other periods
when they add their feelings to the messages. Notice that feelings are added

to items that are already, and necessarily, there: the' ideational, interactional
and textual components are obligatory, the attitudinal; component is
optional.
Linguists have not always thought of intonations role in expressing
attitude as being additional. Indeed, many of the best-known descriptions of
English intonation of the past have regarded the attitudinal function as
primary and central: Pike (1945), Crystal (1969), OConnor and Arnold
(1973). This is but a reflection of peoples greater general awareness of that
function than of the other functions. Other descriptions have placed the
organization of information and discourse functions as central, but have
nevertheless acknowledged the additional expression of attitude: Halliday
(1967), Brown (1977), Brazil et al. (1980), Ladd (1980); and Crystal (1975) also
now follows this line. All are nevertheless agreed that intonation is a means
by which attitude is expressed in speech.-

2. Intonational lexicons
What kinds of attitudes get expressed? This is not an easy question to answer,
and the inconsistencies between one description and another (and indeed
even within descriptions) are testimony to this difficulty. A novel suggestion
was made by Liberman (1979) that there is a kind of lexicon of intonational

108
 Tone variations

meanings. What he meant was that just as you can make a list of words, you
can make a list of intonation patterns, with each pattern having a particular
meaning; furthermore, just as words are composed of morphemes,
intonation patterns consist of morpheme-like components (pitch
movements in the head and the tonic) which can be assembled in different
ways to represent different meanings. Ladd (1980) followed up the same
idea and suggested that these intonational lexicons covered not only
expressions of attitude but other kinds of meaning, too. However, the idea
has been mainly confined to the attitudinal function, and that is how we shall
consider the notion of an intonational lexicon in this chapter.
The descriptions of English intonation listed above can all be presented as
intonational lexicons. We shall begin with an outline of OConnor and
Arnolds (1973) description, because it is perhaps the most famous in the
English language classrooms of the world and because it attempts to be fully'
comprehensive. Liberman himself commented that their description was the
nearest thing available to an adequate intonational lexicon (Liberman, 1979:
94).

2.1 OConnor and Arnold

Their description is set out in Table 5.1 on pp. 113-15. We will illustrate each
of the ten patterns with an example from their own drills. Pattern 1, which
they call the low drop, has, in its fullest form, a low pre-head, a high head
and a low fall:
(5.1) (A: Have you any news of / Malcolm)
B: Hes passed his e. xam

The pre-head H e s is low, there is a jump up at the onset syllable of the head,
p a ssed ; the pitch remains relatively high for his and e-; there is a small jump
down to about mid, or low-mid pitch for the tonic, which is then itself
accompanied by a short fall to a low pitch.1 The meaning of this pattern is
given as categoric, weighty, judicial, considered.
Pattern 2, the high drop:
(5.2) (A: Where on earth are my \sMppers)
B: I cant think Nwhats happened to them
The pattern of pre-head and head is the same as for pattern 1; the only
difference is that the fall in the tonic begins from the high pitch level of the
head and finishes at a low pitch. The meaning of this higher pitched fall is
given as conveying a sense of involvement, light, airy.
Pattern 3, the take off:
(5.3) (A: Let me get you some more \tea)
B: Youre very .kind
The pre-head (you re) and head (very) are kept low and the rise in the tonic

109
The intonation systems of English

is a low rise. Its meaning is given as encouraging further conversation,

guarded, reserving judgement, appealing to the listener to change his (sic)
mind, deprecatory, (in contradictions) resentful.
Pattern 4: the low bounce:
(5.4) (A: I \ hate |climbing /ladders)
B: Its all /light
The pre-head (It's) is low, the head (all) is high; the voice then jumps down
to low in order to effect a low rise in the tonic! Its meaning is given as
soothing, reassuring, hint of great self-confidence and self-reliance; (in
echoes) questioning with a tone of surprise and disbelief; (in non-final word
groups) creating expectancy about what is to follow.
Pattern 5, the switchback:
(5.5) (A: Do you 7smoke)
B; I do v sometimes

The pre-head (I) is low; the voice then jumps up to a high level on the onset
syllable of the head (do) and then falls during the production of the head -
in this case, it is only the single syllable, do; finally, the voice jumps back up
to high in order to effect the fall-rise in the tonic: the fall accompanies the
tonic syllable itself, some- and the rise is left to the final syllable, -times. Its
meaning is given as grudgingly admitting, reluctantly or defensively
dissenting, concerned, reproachful, hurt, reserved, tentatively suggesting: (in
echoes) greatly astonished.
Pattern 6, the long jump:
(5.6) (A: vM arv likes it)
B: Yes but I x dont
Again the pre-head is low (Yes but), the head (I) rises, and the tonic falls. Its
meaning is given as protesting, as if suffering under a sense of injustice.
Pattern 7, the high bounce:
(5.7) (A; Alans not \ here Im afraid)
B: Hes gone / home
The pre-head (H es) is low; the voice jumps to high on the onset syllable of
the head and remains high through the head (gon e); the voice then jumps
down to low or mid-low in order to effect a high rise in the tonic. Its meaning
is given as questioning, trying to elicit a repetition, but lacking any
suggestion of disapproval or puzzlement; (in non-final word groups) casual,
tentative.
Pattern 8, the jackknife:
(5.8) (A: Did you / like it)
B: I simply A h a t e d it
The pre-head (I) is low; the voice jumps to high on the onset syllable of the

no
 Tone variations

head and remains high for the remainder of the head (simply); then it jumps
down to mid or mid-low in order to effect a rise-fall, the rise accompanying
the tonic syllable, ha-, and the fall the tail, -ted it. Its meaning is given as
impressed, awed, complacent, self-satisfied, challenging, censorious,
disclaiming responsibility.
Pattern 9, the high dive:
(5.9) (A: Which are \our places)
B: x Theres | ,vours

This ninth pattern is actually a sequence of patterns 2 and 3, which we would

interpret as a sequence of two intonation units. OConnor and Arnold treat
it as a single pattern and give its meaning as appealing to the listener to
continue with the topic of conversation; expressing gladness, regret,
surprise.
Pattern 10, the terrace:
(5.10) (A: What re\ation did you get)
B: John and - George |seemed rather x keen

Pattern 10 has a low pre-head (not realized in the above example), a high
head (Jo h n a n d ) and then a mid-level pitch on the tonic (George); then it
must be followed by another intonation unit. The meaning of pattern 10 is
given as marking non-finality without conveying any expression of
expectancy.
The reasons for taking the trouble to exemplify all ten patterns are
twofold: first, to demonstrate some of the pitch variations involved in the
pre-tonic and the tonic; and second, to illustrate some of the attitudes that
get expressed by intonation.
The pitch variations, demonstrated above, include low, high, falling, and
rising heads, and high, low and rising-falling variations to the tones. The pre
head was uniformly low in all ten patterns: however, OConnor and Arnolds
system of description includes a supplementary set of ten patterns, all with
high pre-head, with the general meaning of emphatic. The emphatic set
also involves three other types of head: stepping, sliding and climbing heads.
The stepping head has a high onset, but each stressed syllable in the
remainder of the head steps down slightly in pitch:
(5.11) I 'simply 'dont know 'what to \do

The onset syllable is sim- and the remaining stressed syllables in the head are
d o n t and what, each on a slightly lower pitch than that of the preceding
stressed syllable. Its meaning is simply given as emphatic.
The sliding head has a high onset but it slides to a lower pitch; each
succeeding stressed syllable in the head also begins relatively high and slides
down, producing a series of falls before the main fall in the tonic. The

111
The intonation systems o f English

climbing head is the reverse, with a series of rises in the head, before the fall
in the tonic. These two variations to the pitch pattern of the head are said to
reinforce the sense of emphasis.
(5.12) I \simply \dont know \what to \ do
(5.13) I /simply /dont know /what to \do

Thus OConnor and Arnold have an intonational lexicon of twenty

patterns, with a range of meanings each. Two points can be made about the
range they present. The first is that in fact it is more complicated than it
appears above, because for all the patterns, different ranges of meanings can
be ascribed to statements, ^.-questions, yes/no questions, commands and
interjections.2 The meanings illustrated in the above thirteen examples all
relate to statements; different ranges of meanings are ascribed to the other
clause types.
The second point is that the range of meanings seems so wide, indeed too
wide to possess any general sense. You may indeed have wondered whether
some of the meanings were really appropriate. For instance, is the reply in
(5.3) above intended to encourage further conversation? Not necessarily,
since it might simply be a polite acknowledgement of the other persons
offer, and no more. Is the reply in (5.3) above guarded? This meaning seems
wholly inappropriate. Nor does it appear as intended to reserve judgement
to appeal to the listener to change his (sic) mind! It certainly isnt
deprecatory, nor does it convey a resentful contradiction. It is, as mentioned
above, nothing other than a simple, but polite, acknowledgement of the
other persons offer (see Chapter 4).
The problem with OConnor and Arnolds meanings is that they are as
diverse as the types of situations they had in mind. Very often it is not the
intonation pattern that has suggested a particular meaning, but the choice of
lexis and the situation itself in which the utterances are made. That is why
the range of meanings for (5.3) above is so wide and diverse. Check through
the ranges of meanings for the other patterns; look, for instance, at the range
credited to the eighth pattern: can you think of different situations in which
any of the suggested meanings would be appropriate, and different words,
too? A learner of the language may well be bemused by the indication of the
meaning of pattern 8 as either impressed or complacent or challenging or
censorious, and so on; how does the learner know which of these varied
ascriptions is being meant by any given instance of pattern 8? The truth is
that those ascriptions rely on the choice of lexis and a particular situation,
whereas the real meaning of the intonation pattern is something else.
 Tone variations

table 5.1 OConnor and Arnolds (1973) system of Intonation

1. Low drop: (a) no head; low fall

(b) low pre-head - high head - low fall
Attitude
In STATEMENTS: with no head, detached, cool, dispassionate, reserved, dull, possibly
grim or surly: with a high head, categoric, weighty, judicial, considered.
In WH-QUESTIONS: with no head, detached, flat, unsympathetic, even hostile; with a
high head, searching, serious, intense, urgent.
In YES/NO QUESTIONS: with no head (in tags used as independent comments),
uninterested, hostile; with a high head, serious, urgent.
In COMMANDS: with no head, unemotional, calm, controlled, cold; with a high head,
very serious, very strong.
In INTERJECTIONS: with no head, calm, unsurprised, reserved, self-possessed; with a
high head, very strong.

2. High drop: low pre-head - high head - high fall

Attitude
In STATEMENTS: conveying a sense of involvement, light, airy.
In WH-QUESTIONS: brisk, businesslike, considerate, not unfriendly, lively, interested.
In YES/NO QUESTIONS: willing to discuss but not urgently, sometimes sceptical; (in
question tags used as independent comments) mildly surprised acceptance of the
listeners premises.
In COMMANDS: suggesting a course of action and not worrying about being obeyed.
In INTERJECTIONS: mildly surprised, not so reserved or self-possessed as with the low
drop.

3. Take off: low pre-head - low head - low rise

Attitude
In STATEMENTS: encouraging further conversation, guarded, reserving judgement,
appealing to the listener to change his mind, deprecatory, (in contradictions) resentful;
in non-final word groups, deprecatory.
In WH-QUESTIONS: with the nuclear tone on the interrogative word, wondering, mildly
puzzled; otherwise, very calm, but very disapproving and resentful.
In YES/NO QUESTIONS: disapproving, sceptical.
In COMMANDS: (beginning with Dont) appealing to the listener to change his mind; (in
a few short commands) calmly warning, exhortative.
In INTERJECTIONS: sometimes reserving judgement, sometimes calm, casual
acknowledgement.

113
The intonation systems o f English

4. Low bounce: (a) low pre-head - high head - low rise

(b) high pre-head - low rise
Attitude
In STATEMENTS: soothing, reassuring, hint of great self-confidence and self-reliance; (in
echoes) questioning with a tone of surprise and disbelief; (in non-final word groups)
creating expectancy about what is to follow.
In WH-QUESTIONS: with the nuclear tone on the interrogative word, puzzled; (in echoes)
disapproving; otherwise, sympathetically interested.
In YES/NO QUESTIONS: genuinely interested.
In COMMANDS: soothing, encouraging, calmly patronizing.
In INTERJECTIONS: airy, casual yet encouraging, often friendly, brighter than when said
with the take off.

5. Switchback: low pre-head - falling head - fall-rise

Attitude
In STATEMENTS: grudgingly admitting, reluctantly or defensively dissenting, concerned,
reproachful, hurt, reserved, tentatively suggesting; (in echoes) greatly astonished.
In QUESTIONS: (in echoes) greatly astonished; otherwise, interested and concerned as
well as surprised.
In COMMANDS: urgently warning with a note of reproach or concern.
In INTERJECTIONS: scornful.

6. Long jump: low pre-head - rising head - high fall

Attitude
In STATEMENTS: protesting, as if suffering under a sense of injustice.
In ^//-/-QUESTIONS: protesting, somewhat unpleasantly surprised.
In YES/NO QUESTIONS: willing to discuss but protesting the need for settling a crucial
point.
In COMMANDS: recommending a course of action but with a note of critical surprise.
In INTERJECTIONS: protesting, surprised.

7. High bounce: low pre-head - high head - high rise

Attitude
In STATEMENTS: questioning, trying to elicit a repetition, but lacking any suggestion of
disapproval or puzzlement; (in non-final word groups) casual, tentative.
In WH-QUESTIONS: with the nuclear tone on the interrogative word, calling for a
repetition of the information already given; with the nuclear tone following the
interrogative word, either echoing the listeners question before going on to answer it
or (in straightforward, non-echo questions) tentative, casual.
In YES/NO QUESTIONS: either echoing the listeners question or (in straightforward,
non-echo questions) light and casual.
In COMMANDS and INTERJECTIONS: querying all or part of the listeners command or
interjection, but with no critical intention.

114
 Tone variations

8. Jackknife: low pre-head - high head - rise-fall

Attitude
In STATEMENTS: impressed, awed, complacent, self-satisfied, challenging, censorious,
disclaiming responsibility.
In WH-QUESTIONS: challenging, antagonistic, disclaiming responsibility.
In YES/NO QUESTIONS: impressed, challenging, antagonistic.
In COMMANDS: disclaiming responsibility, sometimes hostile.
In INTERJECTIONS: impressed, sometimes a hint of accusation.

9. High dive = High drop (2) + Take off (3)

Attitude
In STATEMENTS: appealing to the listener to continue with the topic of conversation;
expressing gladness, regret, surprise.
In QUESTIONS: very emotive, expressing plaintiveness, despair, gushing warmth.
In COMMANDS: pleading, persuading.
In INTERJECTIONS: intensely encouraging, protesting.

10. Terrace: low pre-head - high head - mid-level

Attitude
In ALL sentence types: (in non-final word groups) marking non-finality without conveying
any impression of expectancy.
In STATEMENTS AND INTERJECTIONS: (in final word groups) calling out to someone as
from a distance.

2 .2 Pike
We turn now to Pike (1945). Despite recent interest in American intonation,
Pikes description remains the only comprehensive attempt, and has been
admired as such by generations of American teachers of English. Crystal
described it as the first really thorough description of the intonation system
of any dialect of English (Crystal, 1969: 47) and Cruttenden concedes that it
has not been surpassed in America for comprehensiveness (Cruttenden,
1986: xi).
His orientation, like OConnor and Arnolds, was the attitudinal function.
He said as much in these two quotations:
In English, then, an INTONATION MEANING modifies the lexical meaning of a
sentence by adding to it the SPEAKERS ATTITUDE toward the contents of that
sentence (or an indication of the attitude with which the speaker expects the
hearer to react). (Pike, 1945: 21)

and:

115
 Contour (= tone) Descriptive label Excessive use
falls contrastive pointing'
falls to low (4) finality
2- 4 S: moderate,
contrastive pointing
Wh\ most frequent
Y/N'. insistent
1- 4 intense + unexpected .+ gushiness
contrastive pointing
3- 4 mild + detached + professional
contrastive pointing aloofness
3 -4 with creak mild unpleasantness grouchiness
falls to mid-low (3) non-finality, less
prominence
2- 3 mild, very frequent and
normal
endearment (with
female speakers)
1- 3 intense + unexpected
falls to mid-high (2) light (endearing baby talk
1-2 encouragement, etc.)
rises incomplete, needing
supplementation (from
speaker/hearer); polite,
cheerful
rises from mid-low (3)
3- 2 S: incomplete sequence
Y/N : question
3- 1 incomplete sequence
+ intense +
unexpected/polite
rises from low (4)
4- 3 incomplete, needing
supplementation +
deliberative
4-2 incomplete, needing deliberate hearty
supplementation + cheerfulness
deliberative + sequence
4 -1 incomplete, needing
supplementation,
deliberative + intense
rises from mid-high (2) incomplete sequence + insincere politeness
2 1
- mild, polite unctuous manner

t a b l e 5.2 Pikes (1 9 4 5 ) lexicon (S = statement; Wh = wh-question; Y/N = yes/no question)

fall-rises
rising from 3 to 2
2 -3 -2 implication, non-finality
1 -3-2 implication, non-finality +
intense/unexpected
rising from 4 to 3
2 -4 -3 implication, incomplete
+ deliberative
1 -4-3 implication, incomplete +
deliberative + intense/unexpected
3 -4 -3 implication, incomplete sequence +
deliberative + mild/detached
rising from 4 to 2
2 -4 -2 implication, incomplete sequence +
deliberative
1 -4 -2 implication, incomplete
sequence + deliberative +
intense/unexpected
rising from 4 to 1
2-4-1 implication, incomplete
sequence + deliberative +
unexpected/polite
1 -4 -1 implication, incomplete
sequence + deliberative +
intense

rise-falls
4-3-4 repudiation

complex
2- 4 + 3 -3 mildly poignant
3- 1 + 4 -3 encouragement
precontour 2 + 4 -4 suspended conclusion

level (in final position) strong implication (without

2-2 further differentiation)
3- 3
4- 4

117
The intonation systems o f English

For English, meanings of English intonation contours are largely of [this] general
type - ATTITUDES of the speaker (or, occasionally, imputed by the speaker to the
hearer). Most sentences or parts of sentences can be pronounced with several
different intonation contours, according to the speakers momentary feeling
about the subject matter. These attitudes can vary from surprise, to deliberation,
to sharp isolation of some part of a sentence for attention, to mild intellectual
detachment. (ibid.: 23)

Pikes description of American intonation rests on an analysis of four pitch

levels (not five, as used in Chapter 4), with falls, rises, fall-rises and rise-falls
beginning and ending on one particular pitch. Pitch level 1 corresponds to
high, and thus a 1-4 fall would be a fall from high to low; pitch level 2
corresponds to mid/mid-high, and thus a 2-4 fall would correspond to the
kind of fall that we have described as neutral; pitch level 3 corresponds to
mid-low/mid, and thus a 3-4 fall would correspond to a low fall. Falls can also
drop to only level 3, or even level 2.3 Rises have a corresponding range of
starting and ending points. The full description appears in summary form in
Table 5.2; it consists of a lexicon of 29 patterns (or contours as he preferred
to call them), including three complex patterns.
What is immediately apparent in a comparison with OConnor and Arnold
is that whereas Pikes description is as comprehensive - in fact, it includes
more patterns - its meaning labels are much less diffuse and in most cases
not specified to particular clause types. What Pike had done was find a much
more uniform, and simpler, system of attitude ascriptions.
However, the question arises about the nature of some of the ascriptions.
In what sense can incomplete sequence or question be called an attitude?
The same criticism could legitimately be levelled against OConnor and
Arnold at times, e.g. for pattern 10, the label marking non-finality without
conveying any expression of expectancy is hardly an attitude. What Pike
and, later, OConnor and Arnold realized is that intonation performs, in fact,
more than one function, but they never reached the systematic exposition of
the variety of functions as featured in this present description of intonation.
Not only is Pikes description more systematic than OConnor and
Arnolds, but it is also possible to display a kind of morphology of meanings.
It is possible to isolate individual meanings of most levels and movements
and thereby establish basic elements of intonation, somewhat analogous to
morphemes, which may combine to form words (Liberman, 1979:133). Pike
did not, in fact, display such a tonal morphology, but it can be abstracted
from the general description, and it must have existed in his mind as an
operating system, to produce the kind of consistency that appears in his
work. The abstracted tonal morphology appears in Table 5.3.

118
 Tone variations

A Contours
falls to 4 \4 finality falls from 2 2\ moderate
falls to 3 \3 non-finality one level falls X mild
falls to 2 \2 lightness falls from 1 1\ intense/
unexpected
rises from 3 3/ incomplete ' rises to 3 /3 incomplete
rises from 4 47 deliberative rises to 2 /2 sequences
rises from 2 27 polite rises to 1 /I intense (except
2 -1 = polite)
fall-rises V implication
(low) rise-fall A repudiation
levels (in
final position)
strong implication
2 -4 + 3-3 \ +- mild poignancy
3 -1 + 4-3 / +/ encouragement
precontour
2 + 4-4 suspended
conclusion
B Precontours
level 3 neutral
2 insistent
1 insistent + unexpected
4 highlighting of focus
slurred / protesting
insistent
descending
stress - - intense precision
deferred N (long precontour)

table 5.3 Pikes (1945) tonal morphology

2.3 Halliday and key

Halliday (1967, 1970) attempted to combine attitudinal meanings into one
general intonation system, which also included information structure,
syntactic contrasts and the communicative functions. The part of the general
system that controlled attitudinal meanings he called key, and key was
principally realized in the secondary tones. (It was Halliday who provided
the concept of primary and secondary tones.)
Unlike the description of English tones presented in Chapter 4, Halliday
proposed five primary tones: 1 (fall), 2 (high rise), 3 (low rise), 4 (fall-rise) and
5 (rise-fall). The rise-fall is typically used for strong assertions; and, in our view,
the high rise indicates a strength of feeling in challenging a statement or in
seeking information. The pre-tonic variations of pitch which Halliday included
in his description are rising, falling, level (high, mid and low), uneven

119
The intonation systems o f English

Pre-tonic Tone Tone number Descriptive label

mid, even fall 1 S,C,Wh,Ex: neutral
Y / N : strong (forceful
or impatient)
rising high fall 1+ strong or unexpected
falling low fall 1- mild or expected
glissando fall -1 forceful or querulous
rising
high/falling high rise 2 S, Ex: statement-question,
seeking confirmation;
contradiction, denial or
disappointing of expectation
W h : mild (tentative or
deferential)
Y / N : neutral

high/falling high fall - 2 Y/N :specifying the particular

high rise point of the query
low high rise -2 Y / N : intense, showing surprise
(incl. disapproval), concern
(mid, even) low rise 3 S: acceding to request,
unexpressed expectation;
reassurance
C: mild, request; polite
(with negative C)
low low rise 3 expressing unconcern,
uncertainty
falling to mid fall-rise 4 S: reservation,contrast,
personal opinion offered
for consideration
glissando low fall-rise 4 S: exclusive, contrastive,
falling strong reservation
rising rise-fall 5 S: asserting, expressing some
other form of commitment,
surprise, personal reaction
(favourable or unfavourable)
glissando low rise-fall 5 S: intense, showing awe
rising (which may be sarcastic)
or disappointment

table5.4 Hallidays 1967/1970 lexicon

(S = statement; C = command; Wh = wh-question; Y/N = y e s / n o question;
Ex = exclamation)

120
 Tone variations

(= OConnor and Arnolds climbing head) and listing (a series of low rises
before the tonic, each low rise indicating one listed term). Halliday proposed
a lexicon of fourteen patterns, as displayed in Table 5.4.
Hallidays labels are as pithy and systematic as Pikes, but he distinguishes
less than half the number of patterns. Halliday also has a system of neutral
tones for certain clause types: for example, whereas a fall is neutral for
statements, commands, rr/z-questions and exclamations, it is not neutral for
y es/n o questions: a fall with a y es/n o question is said to indicate
forcefulness or impatience. In this case, key is actually realized as a basic
tone, but in combination with a clause type in which it is not neutral.

2.4 Crystal
Crystal (1969) investigated the expression of attitude by asking people to
perform a set of sentences in a way that expressed a particular attitude. The
twenty attitudes that Crystal selected were: haughty, puzzled, am used,
p lea sed , questioning, w orried, dism ayed, disapproving, vexed, con
spiratorial, impatient, satisfied, grim, excited, precise, angry, matter-of-fact,
bored, irritated, apologetic. Ignoring, for our present purpose, matters of
tonality and tonicity, we note the following prosodic and paralinguistic
features required for Crystals system for the expression of attitudinal
meaning by intonation:
nuclear tone type
strong stressed syllables
high unstressed syllables
clipped syllables
drawled syllables
simple pitch range (syllabic): large step-up, slight/no step-up,
step-down
flattened syllables in tail
complex pitch range: narrow, wide
simple pitch range over a polysyllabic stretch: high, low
loudness: loud, soft
tempo: fast, slow
rhythm icality
tension: tense, lax
paralinguistic features

However, a very simple attitudinal lexicon appeared in Crystal (1975: 38),

based on seven tones occurring in both final and non-final tonics in
sentences. In this case other prosodic and paralinguistic features are
excluded (Table 5.5).

121
 Tone Position in sentence Descriptive label
level final tonic in sentence a b s e n c e o f e m o tio n a l
in v o lv e m e n t, which may be
interpreted as sarcasm, irony,
boredom, etc.
non-final tonic in sentence im p lic a tio n o f ro u tin e n e s s -
perhaps arising out of the level
tone in final position
low rise final tonic in sentence -
p e r s o n a l in c o n c lu s iv e n e s s
specific labels used here are
non-committal, unaggressive,
etc., which are a short remove
from polite, respectful, etc.
s o c ia l o p e n n e s s - specific
labels used here are casual,
friendly, persuasive, etc. and
(with appropriate kinesic
accompaniment) warning,
grim, etc.
non-final tonic in sentence a ttitu d in a lly n e u tra l
low fall final tonic in sentence a ttitu d in a lly n e u tra l
non-final tonic in sentence -
p e r s o n a l d e fin itiv e n e s s
specific labels used here are
abrupt, insistent, etc.
u n s o c ia b ility - specific labels
being cool, irritated, rude, etc.
high fall in any position d e fin ite e m o tio n a l
c o m m itm e n t - specific
labels
being emphasis, surprise,
warmth, selection depending
very much on kinesic
accompaniment
high rise in any position d e fin ite e m o tio n a l in q u ir y -
s p e c ific labels being query,
puzzlement, surprise, etc.
fall-rise in any position - doubt,
u n c e r ta in o u tc o m e
hesitation, etc., leading to
suspicion, threatening, etc.
rise-fall in any position d e fin itiv e o u tc o m e -
impressed, satisfied, smug,
etc., or the reverse, depending
on kinesic accompaniment

table 5.5 Crystals system (1975)

122
 Tone variations

This second version also recognizes one feature at least of Hallidays

description, namely the acknowledgement of neutral forms. But Crystal has
confined his attention to pitch movements in the tonic and neglected the
wider range of pitch movements in the pre-tonic.

2.5 Brown

Brown (1977) reviewed novelists descriptions of the way their characters

talk and, by comparing the prosodic and paralinguistic features required,
produced thirteen categories: replied/answered/said-, retorted/exclaimed-,
im p ortan t/p om p ou s/resp on sible-, depressed/m iserably/sadly-, ex c ited ;
anxious/w orried/nervous; shrill/shriek/scream ; warmly-, coldly-, thoughtfully-,
sexily-, crossly/angrily-, queried/echoed.
Browns selection of features includes lip setting, but does not specific
ally include step-up/-down, high unstressed syllables, strong syllables,
rhythmicality or flattened tail. She happens to have chosen nine attitudinal
categories not covered by Crystal, and only three which are parallel. It is
interesting to see that Browns and Crystals specifications are almost identical
for excited and angry, but differ markedly for w orried - indeed Crystals
worried subjects were marked by slow speech, Browns by rapid speech!
It is highly instructive to examine and compare the prosodic/paralinguistic
analysis of emotions as presented by Crystal and Brown. If we extract from
their lists those emotions that involve departures from normal pitch level and
from normal pitch range, and group them in such a way as to display those
departures, we will begin to see general characteristics associated with themi
(see Table 5.6). Some emotions involve only general pitch level, either high or
low: some emotions involve general pitch range, either wide or narrow; other
emotions involve various combinations; none in their lists involve the
combination of high and narrow.
The general characteristics of the departure from normal pitch level seem
to reflect the emotional, nervous condition of the speaker. With the high
level, the speaker is tense and emotions are aroused, and nervous tension is
heightened; with the low level, the speaker is either relaxed (satisfied;
important, pompous, responsible; cf. Pikes gloss deliberative) or
emotions are constrained or reined in and nervous tension is deflated.
The general characteristics of the departures from normal pitch range
seem to display the speakers relationship to the listener. With a wide range,
the speaker is warm and open towards the listener or, at least in the case of
shrill/shriek/scream , intends to impress, the listener of a need for a very
definite response. With a narrow range, the speaker is cold towards the
listener and may well not be interested in any response.
High and low, and wide and narrow have been discussed in relation
to tones up to this point, but these terms also seem to apply to heads as well.

123
T he intonation systems of English

C rysta l (1 9 6 9 ) B row n (1 9 7 7 )

HIG H haughty
am used
w orried an xiou s, w orried, n e rvo u s

HIG H puzzled excited

a n d W ID E pleased shrill, sh riek , scream
q u e stio n in g cross, angry

W ID E retort, exclaim
warm
excited q uerying, e ch o in g

LOW sexy
a nd W ID E

LOW dism ayed

d isa p p ro vin g im portant, pom pous, re spo n sib le
vexed
conspiratorial
im patient
satisfie d

LOW grim de pre sse d

a n d NARRO W

NARRO W bored cold

TABLE 5.6

Indeed, Crystals step-up and step-down involve heads, and his high
unstressed syllables can only refer to heads. Such reference to heads
immediately resolves some apparent anomalies between Crystals and
Browns analysis. For instance, Crystal does not credit excited with high, but
only with wide, whereas Brown describes excited as both high and wide;
however, Crystal does refer to a step-up which can occur in both tone and
head - thus, Crystals and Browns descriptions are almost identical.
Similarly, Crystal does not credit angry with high or wide, unlike Brown;
however he does refer to high unstressed syllables, and both agree on other
prosodic features like loudness, tempo and tension.

3. Intonational resources for attitudlnal meaning

In the review of the five intonation systems described in detail above, we have
seen that the speakers attitude is conveyed by a number of different factors.
These factors include the choice of lexis, the actual situation in which the
utterance takes place, the choice of other prosodic and paralinguistic features
besides those related to intonation, and fourthly, not mentioned hitherto,
body gestures, particularly of the face and hands. The fifth factor is intonation.

124
 Tone variations

We are now in a position to present the features of that fifth factor: the
intonational resources for attitudinal meaning are: pitch level, range of the
tone, different types of heads and the pitch of the pre-head. We already have
described in Chapter 4 the neutral forms of the falling, rising and falling
rising tones, which are basic features in the informational, syntactic and
communicative functions of intonation. Variations from these neutral forms
constitute those features of intonation that are used for the expression of the
attitudinal function.
Similarly, there are neutral forms of head and a neutral form of the pre
head, and variations from these forms are also used for the expression of the
attitudinal function of intonation.

3.1 Variations in tones

If the fall starts from a pitch level above mid-high, it is called a high fall; if the
fall starts from approximately mid-low, it is called a low fall. Thus, there is a
threefold classification of falls, which happens to be identical to Hallidays
distinction between medium (neutral), wide and narrow, which was
illustrated as follows:
1 medium [neutral] k
1 + wide K
1- narrow k (Halliday, 1970; 15)

This is parallel also to Pikes classification of falls: high (1-4) and low (3-4),
beside the moderate fall (2-4). Browns is also similar - although the
distinction is spread over two features, viz. pitch span, and placing in voice
range. However, this threefold classification of falls stands in
contradistinction to many other analyses which posit only two types of fall
(OConnor and Arnold; Crystal); the main point of contention against such
analyses is their unwillingness to acknowledge a full range of neutral forms.
Just as we assert three forms of a falling tone, we also assert three forms of
a rising tone. Besides the neutral rise that ends at mid/mid-high level, there
are also the low rise that ends at mid-low level and the high rise that ends
above mid-high. It is only Pike who also acknowledges such a three-fold
distinction in rises: 4-3 (deliberative), 4-2 (sequences) and 4-1 (intense).
(It must also be noted that Pike acknowledges other rises too, with a mid-low
start (3-2 questions, 3-1 questions plus intensity) and a mid-high start (2-1
politeness).) All other analyses offer just two forms of rise, whether they are
credited as two distinct neutral types (Halliday) or simply as high and low
varieties of a rise (OConnor and Arnold; Crystal, 1975).
The phonetic basis of the threefold classification of rises is the end point
of the rise, rather than the starting-point. It is observed that the actual
starting-point of the high rise can vary quite considerably, as Pikes analysis

125
The intonation systems o f English

highlights; but similarly, the end point of a fall can vary considerably too;
again, Pikes analysis highlights this (falls include 1-2, 1-3, 2-3 as well as
those falling to 4).
The high fall is variously glossed as intense, unexpected (Pike),
personal concern, involvement, liveliness ... more emotional, etc.
(OConnor and Arnold), vigorous agreement or contradiction ... strong
surprise, etc. (Gimson, 1989), strong, unexpected (Halliday), surprise/
redundancy contour - which includes a low but ascending head (Liberman;
redundancy in the sense that the speaker is protesting that the
informational content of the message should be regarded as self-evident).
There is general agreement about the meaning of this intonational form, and
we could adopt the label strong or intense - leaving the particular attitude
to be specified from lexical and situational factors, e.g. surprise, personal
concern, etc. It is tempting to use Browns description of a high fall as an
indicator [of] some positive attitude (Brown, 1977: 129). But since she
concedes that the surprise may be either excited or disagreeable, then
positive does not seem to be quite so appropriate.
The low fall too has a variety of glosses: mild, detached (Pike); cool,
calm, phlegmatic, detached, reserved, dispassionate, dull, possibly grim,
surly (OConnor and Arnold; and Liberman concurs); detached, unexcited,
dispassionate (Cruttenden). Again, there is very strong general agreement
about the meaning of this intonational form; we could safely adopt the label
mild and leave the specification of attitude to. the lexical choice and
circumstantial features of an utterance.
The high rise, i.e. rise to high, receives the same gloss from Pike as the high
fall: intense, unexpected. However, most other analyses draw attention to its
use as echoes, challenges and requests for repetition (OConnor and Arnold,
Halliday, Brown, Crystal); Palmer also characterizes its meaning as animated,
and Gimson as eagerness, brightness, enthusiasm, excitement, concern,
indignation. Again, we could adopt the label intense to indicate the meaning
of this intonational form and leave the closer specification of attitude to the
accompanying lexical choice and circumstantial features of an utterance.
The low rise, i.e. rise to mid-low, must be carefully differentiated from the
neutral rise, i.e. rise to mid/mid-high. The latter is very widely termed the
low rise, when no difference is suggested between it and a true low rise.
Yet the meaning of the rise to mid-low is often indicated in the course of
description of low rises in general: cf. OConnor and Arnold: reserving
judgement, guarded, reproving criticism, resentful contradictions,
deprecatory, wondering, calm, casual acknowledgement; also Cruttenden
(1986: 105-6): uncertainty, non-committal or even grumbling. Adapting an
example from Cruttenden (ibid.), we can illustrate the different meanings of
the three rises:

126
 Tone variations

(5.14) Hes \ passed | ""hasnt he (high rise: intense, surprise, challenging,

I would be surprised to hear otherwise)
(5.15) Hes \ passed | / hasnt he (neutral rise: question,
seeking confirmation, I would like to be sure)
(5.16) Hes \ passed | /hasnt he (low rise: non-committal, grumbling,
Well, let him be thankful for that at least - he could have failed!)
It seems that most of the attitudes associated with I
this low rise are
negative in some way, but the term negative would be misleading as a label
for this intonational form (cf. Browns observations on the high fall as
disagreeably surprised, above). Pike characterized this form as deliber
ative. It means that the speaker is carefully considering the matter in hand.
It does not state whether the judgement is favourable or unfavourable - that
must be gathered from the context (Pike, 1945: 54). His is an approach very
much in line with the approach taken here, but deliberative is too positive
a label for its typical use in British English. Cruttendens label non
committal seems to be the closest we can get to a general label, if it is
understood that it includes negative signals regarding the information or
the addressee; specific interpretations such as menacing, disapproving,
sceptical, would depend, as above, on lexical and situational factors.
The rise-fall has two varieties, although it is only Kingdon and Halliday
who have distinguished between them. The high rise-fall involves a rise from
about mid level to high followed by a fall. The low rise-fall involves a rise
from low/mid-low level to about mid pitch followed by a fall. In both cases,
the rise element carries the main force (see Halliday, 1970:11); and the total
pitch movement can be confined to a single syllable or split over several
syllables, in which case the highest point in pitch is usually reached in the
syllable following the tonic syllable.
high
mid-high
mid
mid-low
low
high rise-fall
The high rise-fall is acknowledged in most British studies and is glossed
variously as intensified (Palmer), enthusiasm, doubt, horror, sarcasm,
indignation, etc. (Gimson), impressed, awed, echoing, complacent, self-
satisfied, even smug, etc. (OConnor and Arnold), special emphasis
(Kingdon), committed, insistent, asserting (Halliday), definite outcome -
impressed, satisfied, smug, etc., or the reverse, depending on kinesic
accompaniment (Crystal, 1975). Along with most American studies, Pike

127
The intonation systems o f English

does not acknowledge the high rise-fall; but Bolinger (1986) does (for
Bolinger, it is Profile CA), as a sort of intensification of the fall. Intensified
is probably the best label for the high rise-fall, being stronger in expression
than the high fall.
The low rise-fall is omitted in most studies, but is acknowledged by
Kingdon and Halliday;4 the latter glosses it as intense, showing awe ... or
disappointment (Halliday, 1970: 32). He also observes that it is frequently
accompanied by breathy voice quality. The best label would seem to be
intensified, plus emotion, whether that emotion is awe or disappointment,
or anything else as indicated by lexical choice and other situational factors.
That Halliday readily acknowledges typical accompaniment of other voice
qualities (breathiness) is evidence that there is a greater involvement of
emotional expression.
There is also a low variety of the fall-rise. The neutral fall-rise, which
conveys thematic marking and implication in the informational function of
intonation, has a fall from about mid-high pitch to mid or mid-low, followed
by a rise. The low fall-rise begins its fall element at about mid pitch and it
falls more steeply and over a wider range [than the neutral fall-rise],
descending to a low pitch (Halliday, 1970: 18).
The low fall-rise is glossed either as exclusive, contrastive, expressing
strong reservation (Halliday) or as combining incomplete deliberation
with mild attention (or contrast) and detached attitude (Pike). Pikes gloss
may well reflect American usage, but Hallidays seems right for British usage.
The label strongly contrastive/implicational is thus proposed.

3.2 Heads
It is now time to look at heads in more detail and examine the part they play
in the expression of attitudinal meanings. As with tones, it seems necessary to
identify neutral forms of heads in order to make a distinction between forms
that do and do not carry expressions of attitude. Just as there are neutral forms
for both falling and rising tones, there are also neutral forms of heads that
accompany the neutral tones. In this respect, we follow the lead that Halliday
has taken, and we can adopt much of his description of neutral heads.
Heads may be level, ascending, descending (or mixed or glissando). Level
heads may be high, mid-high, mid, mid-low, or low; obviously, with
instrumental phonetic study it is possible to identify a gradience of pitch
levels for heads as it is with pitch levels for the beginning point of tones, but
since high, mid and low offer themselves as obvious candidates as pitch
heights, and since it is also possible to conceive of intermediate levels
between high, mid and low, these five pitch levels form a base from which to
describe functionally distinct pitch levels for heads. Ascending and
descending are chosen as labels for heads to maintain a distinction between

128
 Tone variations

the movement of pitch in heads and the movement of pitch in tones (i.e.
falling and rising). Glissando is a term adopted from Crystals work to refer
to smooth and usually fairly slow glides (1969: 164) in pitch which are
contained within a head; these are identical to what OConnor and Arnold
called sliding and climbing heads.
There is a fair amount of variation in the actual form of neutral heads.
Generally speaking, a neutral head is either level at mid, mid-high or mid-low
pitch, or gradually ascends or descends towards the level of the beginning
point of the tone. What must be borne in mind at this point, is that the pitch
level of the onset syllable is significant in terms of textual structure. Once the
onset pitch level has been given, the neutral form of the head then adjusts to
the beginning point of the tone, reaching about mid pitch - hence the
ascending or descending movement of pitch in the head. When heads
involve a number of accented syllables there may well be an arbitrary
mixture of ascending and descending syllables. The neutral heads before a
neutral fall may be illustrated, therefore, as follows:
mid-level
ascending _____--------------- ---
descending " ~k
mixed ' k etc.
The neutral heads before a neutral rise may be illustrated likewise, as
follows:
mid-level Y
ascending
descending
mixed
The end point of heads is adjusted slightly before high and low falls, so
that the final syllables of the head are fairly close in pitch to the beginning
point of the fall.

3.2.1 Low a n d h ig h h e a d s
A low level head before a fall and a high level head before a rise have the
effect of concentrating attention exclusively on the focus of information.
This particular sequencing is not acknowledged by Halliday, but Pike makes
reference to the effect, at least, of the low level head before a fall:
precontour four [= low level head] heightens the contrastive pointing of any(
succeeding falling primary contour by making a relatively large interval between
the precontour and the beginning of the primary contour; the greater the interval,
the sharper the contrast or pointing and attention. (Pike, 1945: 66)

129
The intonation systems o f English

Pike does not offer any similar observation about the inverted sequence
of high level head and rising tone, but the effect is, in fact, identical. In each
case, the focus is highlighted, at the expense of the semantic value of the
given section of the intonation unit. They can be illustrated as follows:
low head ________ K before a fall
high head r" before a rise
A high level head may also precede a fall, and a low level head may
precede a rise. The latter is specifically incorporated by Halliday into his
description of English under the label involvement:
Involvement may mean a desire to affect the decision, thus implying suggestion
or encouragement, or may imply some judgment as you ought (not) to, you
should have told me. (Halliday, 1967: 44)

The low level head before a low rise (as opposed to before a neutral rise)
features in OConnor and Arnolds description of English (pattern 3). The
low narrow rise mixes the relaxed, deflated, deliberative meaning with a
cold relationship towards the listener, thus producing the non-committal
attitude; however, the low level head adds involvement, and the whole
combination produces the kind of effect that OConnor and Arnold have
described in the following terms: reserving judgement, reproving criticism,
resentful contradiction, disapproving, menacing, scepticism - a rather
unpleasant catalogue of attitudes, augmented by Hallidays unconcern and
uncertainty.
The low level head before a high rise seems to express incredulity and
disbelief, mixing involvement with aroused emotions and a warm
relationship with the listener - at least in the sense that the utterance is likely
to produce a definite reaction from the listener.
The low level head before rises can be illustrated as follows:
(neutral rise)
u (low rise)
1/ (high rise)
The high level head before falls has a similar effect; Pike refers to
insistence and unexpectedness. It is neater to refer to all this as
involvement again, but to note the variations as a result of the combination
with neutral, high and low falls.
Cruttenden observes a sense of weightiness when the high level head is
combined with a low fall, for instance. The high level head before falls can
be illustrated as follows:
(neutral fall)
T\ (high fall)
(low fall)
130
 Tone variations

3 .2 .2 Wide d es c en d in g a n d a sc en d in g h e a d s
We now move on to consider another type of head, in which there is a
considerable movement of pitch, either descending from high/mid-high to
low/mid-low, or ascending from low/mid-low to high/mid-high. The
difference between these descending and ascending heads and those that
were considered to be variants of the neutral form of head is the degree of
pitch movement. This non-neutral form has a wide pitch range, and may be
conveniently referred to as the wide descending/ascending head. Both
possibilities can occur before both falling and rising tones.
The general effect of the wide descending head is a combination of the
meaning of wide (warm relationship with the listener or, at least, a strong
expectation of some kind of response) with one meaning of fall, viz.
authority. This agrees well with Pikes gloss insistent. Other glosses include
- with a following fall - categoric, considered, weighty, judicial,
dispassionate; often withdrawal, impatience (with statements), searching,
serious, intense, responsible, often impatience, irritability (with wh-
questions), more ponderous; often impatient (with jyes/no questions), firm,
serious, considered, weighty, pressing, dispassionate (with commands), and
great weight, emphasis (with exclamations) (all from OConnor and Arnold).
Again, this plethora of attitudes can be sorted out by reference to lexical
and situational factors, but the general meaning is as given above: warm
relationship with listener and/or strong expectation of response, with a
degree of authority, together with whatever is signified by the kind of tone
that follows - either falling or rising.6
The wide ascending head combines the general meaning of wide with
one of the meanings of rise, viz. appeal to the listener. Palmer has a general
gloss for this head (which he calls scandent):
As compared with a Superior Head (= high level head), the Scandent Head
generally expresses more animation. Compare the impressive:

SAdnli 5ei so: Gri: ino:mos wulvz!

with the vivacious: i

/ SAdnli 5ei so: Gri: ino:mos wulvz!

The first sentence implies: How horror-struck they must have been! What a
terrible situation for them!
The second sentence implies rather: How surprised they must have been!
What an interesting experience for them!
(Palmer, 1922: 76-7)

Note that Palmer is careful not to label this head as vivacious, or

petulant, for the precise attitude or emotion depends on the lexical and

131
The intonation systems o f English

Figure 5.1 Graph of stepping head in (He's the) stupidest m an I know

(Liberman, 1976: 177)

situational factors of an utterance. However, both Pike and OConnor and

Arnold describe the combination of wide ascending head and a falling tone
as expressing protests. This fits in well with the general meaning adduced
above, as protests indeed involve a strong expectation of some kind of
response, an appeal and a statement of fact, belief or reality.7

3 .2 .3 S t e p p in g h e a d s

OConnor and Arnold use the term stepping head for a wide descending
head in which the stressed syllable of each accented word is a step lower in
pitch than the previous one (1973: 73); the stepping adds a sense of
emphasis to what the speaker is saying. Liberman (1979: 174-7) does not
seem to acknowledge this, but accepts it as simply a variation of the more
normal falling movement. However, Pike recognizes the distinction: his
descending stress series is equivalent to the stepping head, and its meaning
is given as intense precision, or certainty.
A very clear picture of the stepping head is provided by Liberman, despite
the above comment: Figure 5.2 shows the movement of pitch in the
utterance (H es the) stupidest m an I know.
The graph shows the stepping very clearly, as opposed to a smoother
descending pitch movement. This type of head is best depicted as:

i------------K
but it can also precede a rise:
(5.17) 'Are you 'going to 'do it to /day

with the same degree of emphasis.

Pike and OConnor and Arnold only conceive of stepping heads as
stepping downwards; but stepping heads can step upwards too (see Crystal,
1969: 230). The stepping effect simply adds an extra degree of emphasis to

132
 Tone variations

the meanings of the wide ascending heads and is strongly related to another
quite independent prosodic feature, viz. degree of loudness of the accented
syllable. The ascending stepping head is depicted as follows:

_________ i-------------- 1 K
_________ ,-------------- 1---------------k "

3 .2 .4 G lis s a n d o h e a d s

Glissando takes the additional degree of emphasis a stage further. Glissando

involves a pitch movement, or a series of pitch movements, either
descending or ascending, before a tonic syllable, i.e. within a head; if a series
of pitch movements are involved, the pitch movement is constant
throughout the head and is dependent on accented syllables with noticeable
rhythmicality. The difference between a glissando movement and a tonic
movement is a matter of prominence.8
Glissando may involve a descending movement or an ascending
movement, or a series of each, but never mixed. The pitch level of the
beginning point of each descending movement may remain constant
throughout the head, but it may also gradually fall (Crystal, 1969: 221; Ladd,
1983). No such observation is made by Crystal or Ladd (or Halliday, for that
matter) for the reverse possibility: a gradual rise in the pitch level of the
beginning point of each ascending movement; but it is not inconceivable.
Halliday describes glissando heads as bouncing, uneven and swinging.
Before a falling tone, he acknowledges glissando ascending (uneven) pre
tonics (heads), but not glissando descending. Before a low pitched rise-fall,
he also acknowledges only glissando ascending (a swinging movement
with ... each foot rising ...,1970: 19); in fact, he takes this kind of head to be
the only possible head before a low-pitched fall-rise.
It seems to me that there is a gradience of meaning distinctions between
descending heads, stepping descending heads and glissando descending
heads. As we have noted above, the descending head indicates a warm
relationship with the listener or, at least, a strong expectation of some kind of
response, with a tone of authority. The stepping effect adds emphasis to this;
the glissando effect adds forcefulness - conceding no opportunity for contra
diction. Compare the following utterances with each of three types of head:
(5.18) ' The sun rises in the east

(5.19) It is time for you to go to bed .

(a) with a wide descending head: authoritatively
(b) with a stepping descending head: authoritatively, and emphatically
(c) with a glissando descending head: authoritatively, and forcefully

133
The intonation systems o f English

The glissando ascending lacks the sense of authority, but includes a sense
of appeal, a sense of demanding that something should be taken into
account. Hallidays example was
(5.20) He 'simply 'doesnt u n d e rstand
(1967: 42)
glossed as insistent; however, the effect of the glissando ascending head is
more like I appeal to you to believe my estimation of him. The two
utterances given above ((5.18), (5.19)) can also be rendered with glissando
ascending heads, with a meaning like I appeal to you to remember what Ive
told you ten times already (or I appeal to you to think it out logically, etc.),
expressed forcefully.
Halliday did not consider the possibility of glissando heads before a rising
tone, but Gimson did. The note of authority with forcefulness, or appeal
with forcefulness, is obviously combinable with the elicitation of
information, open suasion, etc. Examples from Gimson include:
(5.21) Are you \sure that \George and \Mary / know
(5.22) \Mind you \put your \hat and \coat / qjx
(5.23) /Will you be /coming to /see us on / Monday?
(5.24) /Dont make /such a / noise
(Gimson, 1989: 285)
(NB: Other examples of Gimsons remind us that all those different kinds of
heads may precede a variety of tones, including the fall-rise.)
The glissando heads can be depicted as follows:

(glissando descending)
\ \ \ K \ \ \ ^
(glissando ascending)
/ / / f\ / / / V

3.3 Pre-heads
Finally, we turn to pre-heads, the sequence of relatively unaccented syllables
before the onset syllable, or before the tonic syllable if there is no separate
onset syllable. Analyses vary quite considerably in the amount of attention
given to pre-heads. On the one hand, Halliday does not recognize a distinc
tion between head and pre-head at all, assigning both to an undifferentiated
pre-tonic segment. Pike did not spare much space for the consideration of
what he called proprecontour (1945: 68) and which he described as intro
ductory to the more important precountour (= head) (ibid.); it is usually
pitched low, at level 4 or 3, many examples of it being found scattered
through his whole description. On the other hand, Crystal (1969) distin
guishes between five types of pre-head, Liberman distinguishes between
three, and OConnor and Arnold merely between high and low pre-heads.

134
 Tone variations

OConnor and Arnold seem to consider the low pre-head as normal - it is

not indicated by any transcription symbol; but the high pre-head is said to
indicate emphasis (OConnor and Arnold, 1973: 36). It is illustrated by their
pair:
(5.25) The \fool
(5.26) T h e \fool

where the high pre-head in (5.26) makes the whole utterance more
exclamatory, more emphatic ... (1973: 36). There is certainly an extra
dimension added to the meaning by the adoption of a high pre-head instead
of a low one, and it does seem surprising that Halliday has neglected this
little bit of the intonation system completely.
By contrast, Crystal offers a clear and simple classification of five types of
pre-head, directly from detailed observation of his corpus (1969: 233-5). The
normal pre-head is approximately mid-low pitch, a little below that of the
onset syllable, which is given, typically, as mid. The high pre-head is pitched
above the level of the onset, and the extra-high pre-head above that again. The
mid pre-head is heard as being on the same pitch level as the onset, and the
extra-low pre-head is below normal low level. But Crystal does not venture
any meaning attached to any of these varieties, although he did note that high
unstressed syllables in the pre-head feature in his category amused.
There is general agreement that a pre-head at mid/mid-low level is a
neutral form in that it does not carry any expression of attitude. The neutral
form is, in fact, strongly influenced by the pitch level of the onset; it may be
higher than mid before high onset, and lower than mid before low onset.
Variations from this neutral form do carry meaning, however, and the
meaning is generally very simple. Most cases of the marked pre-heads
involve a pitch contrast with the immediately following onset syllable, or, in
the case of head-less intonation units, with the immediately following tonic
syllable: low pre-head before high onset or high beginning of a falling tone;
high pre-head before low onset or low beginning of a rising tone. And the
meaning is identical to those cases of high heads before rises and low heads
before falls: it directs particular attention to the informational content of the
head, or, in the case of head-less units, to the focus. Additional specification
of the attitude must be provided by the lexical content and the situational
factors relating to the utterances; what the marked low/high pre-head does
is simply to indicate that the utterance is marked attitudinally.
The intonation systems of English

4. Summary
The attitudinal potential in English intonation can be summarized as follows:

4.1 Tones
There are neutral forms of falling, rising and falling-rising tones, whose
meanings are determined by the informational, syntactic and com
municative functions of intonation. The falling and rising tones both have
high and low varieties which are, thereby, attitudinally marked; high is
strong/intense, low is mild with falls, and non-committal with rises. The
falling tone has further varieties: rise-falls, which can be either high, meaning
stronger/more intense (than the high fall), or low, meaning intensified,
plus emotional. The fall-rise has a neutral form which is high, and an
attitudinally marked form which is low, meaning strongly contrastive/
implicational.

Neutral Marked Meaning

A Tones
fall \ high v strong, intense
low N mild
high rise-faIIa stronger, more intense
low rise-fallA intensified, plus
emotional
rise / high 7 strong, intense
low y non-committal
fall-rise v lowv strongly contrastive/
implicational
B Heads/pre-heads
mid level
pitch contrast highlighting
high - (before rise) information
low _ (before fall)
high - (before fall) insistent
low _ (before rise) involved
descending (wide) \ warm/expectation
of response; authority
ascending (wide) / warm/expectation
of response; appeal
stepping \' /' emphatic
glissando \\ // forceful

table 5.7 Neutral and attitudinal marked tones and heads/pre-heads

136
 Tone variations

4.2 Heads and pre-heads

As with tones, there are neutral forms of heads and pre-heads, which have no
independent meaning from that indicated by the tone involved. Pitch
contrast is brought about by the combinations of low head/pre-head with
fall, high head/pre-head with rise, low pre-head with high onset, and high
pre-head with low onset meaning highlighting the immediately following
informational content, either in the head, or in the tonic. Descending
heads/pre-heads, if they are wide, mean warm relationship with listener
and/or strong expectation of response, with an authoritative tone;
ascending heads/pre-heads, if they are wide, mean warm relationship with
listener and/or strong expectation of response, with a note of appeal to the
listener. Stepping heads, both descending and ascending, add emphasis,
and glissandi, again both descending and ascending, add forcefulness.

Notes
1. The tonic actually begins in the middle of x-. /eg\zaemz/. The orthographical ver
sion could read e x \ x a m s , if it is felt necessary that this point should be made clear.
2. Note that OConnor and Arnold use labels from communicative functions for the
five different clause types.
3- Brazils notion of termination also exploits this variation in degree of fall (Brazil,
1975, 1978).
4. Pike does acknowledge the low rise-fall (4 -3 -4 ), but notes that it may be
accompanied by harshness and suggests repudiation as its meaning. But that
meaning seems to be tied to the solitary example he gives:

No, oh no, that cant be true

4-3-41 1 - 4-3-4| 2- -4 -3 (Pike, 1945:57)
The same intonational sequence with Yes, yes, that m u st b e so could easily yield the
meaning acceptance.
5. And Pike (see n. 4).
6. The glosses that Schubiger (1958) provides for the combination of wide descending
head with a rising tone are quite illuminating and bring out the effect of the rise:
friendly, reassuring, encouraging (with statements); rather complaining; tone of
entreaty (i.e. strong expectation of response, plus request) (with commands);
resentful, reproachful, perplexed (with u;/z-questions); entreaty, despair, concern
(with y e s /n o questions); and tenderness (with exclamations, e.g. G o o d old
London)-.
(5.27) Good old /London

(see Schubiger, 1958: 57)

7. It is only Palmer and Schubiger who isolate the combination of wide ascending
head with a rising tone for any comment: for Palmer it represented a more
animated form of y e s /n o question, besides echo questions and statements; for
Schubiger, it meant patronizing, condescending if the rising tone was low (e.g.
Lady (to charwoman, on Boxing Day):
(5.28) Come into the /drawing room
8. For a full discussion of the problems of identification, see Crystal (1969: 221-2).

137
 Intonation in a model of
communication

1. Models of communication
It has been the interest of many linguists to construct a model of the
processes by which we communicate with each other. In the simplest terms,
we can construct a flow-chart diagram that shows the direction of activities
frop a sender to a receiver, the co-ordination of brain and voice (or hands,
in the case of written communication) of the sender, and of ear (or eyes, in
the case of written communication) and brain of the receiver, and the
physical substance by which the communication is transmitted - sound
through the air, or marks (writing) on material (like paper).
However, there are other factors involved. One factor, of course, is the
language or dialect (or languages or dialects) in which the communication
is to be conveyed; secondly, the sender takes into account a number of other
factors in the context of the proposed communication. One such factor is
the receiver, or receivers: we adjust the form of a communication on account
of how old the receiver is, whether we know them and how well we know
them, whether we like them, how many there are of them, and so on. We also
take into account where we are and where they are, i.e. the physical settings,
the atmosphere of the encounter, i.e. the scene, and also the channel by
which we will communicate: face to face, by phone, by letter, etc. Also, our
psychological state affects the way we communicate; if we feel angry, we will
sound angry, and that could well affect our choice of vocabulary and the
pitch of our voice. A fourth factor is the purpose of the communication;
telling somebody off sounds quite different from advice; advice is different
from suggesting a course of action, which itself is different from a
recommendation, etc.; this aspect of communication was dealt with in
Chapter 4; it clearly involves intonation as well as grammatical and lexical
)
138
 Intonation in a model o f com m unication

choices. A fifth factor is the context of the communication both at a macro

and a micro level; a communication is sent in the context of what the sender
knows and what the sender thinks the receiver knows; the context also
involves what has just been said (or written) by either of the participants.
The sense of context might also be broadened to include the
communication event itself; we adjust the wording and intonation of a
communication to be appropriate for different kinds of events: for example,
we communicate in one way in the context of a cosy chat with a close friend,
and in quite a different way when addressing a stranger, or a crowd, or a
panel of examiners, or a very important person.
Thus a model of communication has to not only account for the linguistic
processes but also these sociolinguistic factors. In seeking to place
intonation in a model of communication, we are thereby looking at
intonations role, or roles, in more general terms than the details specified in
the preceding chapters. In order to do that, we must consider the general
features of linguistic communication.
Halliday (1973, 1978; see also Halliday and Hasan, 1985) distinguishes
between three metafunctions of language: the ideational, the interpersonal
and the textual. These are different dimensions of meaning that always
feature in any instance of the use of language. The ideational encompasses
both experiential and logical meanings. Experiential meaning refers to the
kind of meaning that language possesses for representing our perception of
the real world, whereas logical meaning refers to the types of relationship
between items (e.g. between nouns, or clauses, etc.). For instance, in the
following utterance,
There's a cup of tea in the pot if you want one
we know what the words and clauses refer to (experiential meaning) and we
recognize the relationship of the second clause to the first - a pragmatic
condition (logical meaning).
The interpersonal meaning refers to the interaction between the sender
and the receiver. In this case, the sender provides some information that is
clearly designed to be of benefit to the receiver, and it is quite likely that the
receiver will interpret the utterance as an offer. Thus a likely follow-up is the
action of pouring tea into a cup. Interpersonal meaning is thus a quite
different dimension to the ideational meanings.
The textual meaning refers to the way an utterance is worded, organized
and intoned to fit into its linguistic environment. The pronoun one is used
as a substitute for a cup o f tea to avoid unnecessary repetition. The context
provides the basis of the interpretation.
You will no doubt recognize some of the roles that intonation plays in
these different functions. Tonality provides the basis for the distribution of
information and is often crucial for disambiguating potential clashes of

139
The intonation systems o f English

meaning (Chapter 2). Tone is crucial in the interpersonal function (Chapter

4). Tonicity is crucial in the textual function, with its systems of identifying
new and given information (Chapter 3).
Thus the illustrative utterance above is likely to be intoned as
(6.1) Theres a cup of tea in the \oi | if you /want one

i.e. two pieces of information, one major, the other minor; a statement (the
falling tone); all new information in the first unit (broad focus), given
information (on e) in the second as well as new ( if you want).
But there are other ways of formulating that statement. Joos, in a famous
and entertaining work (Joos, 1961), attempted to disabuse people of the
impression that a language is a monolithic, unchanging, construct. His Five
Clocks presented five distinct styles of communication. The above utterance
was communicated by a male to a group of three women whom he did not
know, in a setting that was unfamiliar to both himself and them. This is a
typical situation for what Joos called the consultative style: the participants
in the event do not know each other well, and the group has an upper limit
of about six or seven. The speaker has to provide a good deal of background
information. In a more casual style, the participants know each other and do
not need to furnish as much background information; the speaker can rely
on insider knowledge, and might render the above as
(6.2) Dyou want a cup of /tea

Intimate style is even more informal and is characterized by what Joos called
extraction: the speaker extracts a minimum pattern from some conceivable
casual sentence (1961: 30). In this case, the participants know each other so
well there is hardly any need to provide background information. This can
be illustrated with
(6.3) cup o'/tea
(or even Cuppa?)
Besides these informal styles, more formal ones are also recognized. If the
speaker addresses a group that is too large for the consultative style, then a
formal style is adopted:
(6)4) You can get yourselves a cup of /lea I in the room next \door

In a formal style, the speaker does not expect the hearers to engage in a
conversation or discussion, and has to provide full background information.
Finally, Joos described a frozen style for print and for declamation ... the
reader or the hearer is not permitted to cross-question the author (1961: 39).
The announcement, spoken by a person in authority at a gathering of
people,
(6.5) Tea will be served in the large \ lounge |at three forty \five

140
 Intonation in a model o f com m unication

could as well be printed in the programme!

Throughout his study, Joos hints at the role of intonation in defining these
styles but offers no detailed descriptions. It is also noticeable, from a
phonological point of view, that precision of articulation contributes to the
definition of these styles, as the orthographic versions of (6.1), (6.2) and
(6.3) testify.
Each of these styles accompanies a different kind of situation. A term used
frequently in linguistic study to refer to varieties of situations is context of
situation. Halliday has been largely responsible for the theoretical
development of this term (see Halliday, McIntosh and Strevens, 1964;
Halliday, 1973, 1978; and Halliday and Hasan, 1985). He has identified three
general headings to describe the features of a situation that go beyond Jooss
early attempt. They are the field, the tenor and the mode of discourse.
1 The field of discourse refers to what is happening, to the nature of the social
action that is taking place: what is it that the participants are engaged in, in
which language figures as some essential component?
2 The tenor of discourse refers to who is taking part, to the nature of the
participants, their statuses and roles: what kinds of role relationship obtain
among the participants, including permanent and temporary relationships of one
kind or another, both the types of speech role that they are taking on in the
dialogue and the whole cluster of socially significant relationships in which they
are involved?
3 The mode of discourse refers to what part the language is playing, what it is
that the participants are expecting the language to do for them in that situation:
the symbolic organisation of the text, the status that it has, and its function in
the context, including the channel (is it spoken or written or some combination
of the two?) and also the rhetorical mode, what is being achieved by the text in
terms of such categories as persuasive, expository, didactic, and the like.

(Halliday and Hasan, 1985:12)

Every instance of the use of language, however long or short, can be

analysed in terms of its field, tenor and mode.
A more explicit instrument for the description of instances of the use of
language (or speech acts) is provided in the work of the anthropologist
Hymes (see, for instance, Hymes, 1977). He identifies the following
components:
1 Message form, the actual choice of lexis, syntax, phonology/
orthography with accompanying phonetic or graphic detail; compare,
for instance, the different wordings of (6.1) to (6.5).
2 Message content, the topic, which happened to be the same for (6.1) to
(6.5).
3 Setting, the time and place of a speech act, its physical circumstances.

141
The intonation systems o f English

4 Scene, the psychological setting, or cultural definition of an occasion

as a certain type of scene, e.g. a wedding, or a wedding reception, or a
best mans speech, or a toast to the bridesmaids, or a tea-break, etc.
5 Participants, the speaker/sender/addressor and the hearer/receiver/
audience/addressee, their respective ages, genders, statuses,
relationships.
6 Ends, both the conventionally recognized and expected outcomes from
a community perspective and the goals of the individuals involved, e.g.
the public recognition of a marriage and the promises taken
individually by the bride and groom.
7 Key, the tone,- manner or spirit in which an act is done, e.g. willingly,
grudgingly, as practice, etc. Hymes adds: The significance of key is
underlined by the fact that, when it is in conflict with the overt content
of an act, it often overrides the latter (as in sarcasm). The signalling of
key may be non-verbal, as with a wink ... (Hymes, 1977: 58). (It may also
be intonational, either linguistic or paralinguistic.)
8 Channels, the choice of oral, written, manual (as in sign language) and
other means of transmission.
9 Forms of speech, dialects, accents, registers, speech styles as, for
instance, when a speaker switches to a different accent (including
intonation) to initiate a humorous comment or anecdote.
10 Norms of interaction, the rules that determine specific behaviour and
proprieties in a speech act or event, e.g. when it is not acceptable to
interrupt or speak aloud, when it is acceptable to heckle or cheer (e.g.
in a best mans speech).
11 Genres, a conventionally recognized type of language event that is
identified by a unique combination of linguistic forms, thus enabling
members of a community to distinguish between, for example,
newsreading and other formal, solo, performances to an audience such
as recitations, sermons, announcements, etc., and between
conversations and debates, etc. The notion of prosodic composition
(see Chapter 1) is important in this respect.
The description of the situation can also be approached from a socio-
psychological perspective, seeking to understand a given situation and its
development from the actors-eye-view (Brown and Fraser, 1979: 56). Giles
and Coupland (1991: 13-20) report on studies of peoples perceptions
of social situations in which five dimensions emerge: co-operative-
competitive; intense-superficial; formal-informal; dominant-equal; task-
orientated-non-task-orientated. This is clearly a more elaborate system than
Jooss, but just as Joos considered intonation as one of the pieces of linguistic
evidence for different styles, Giles and Hewstone (1982) took phonological
and phonetic factors into account when speculating on the potential speech
patterns deriving from peoples assessment of a situation. Their chart of

142
 A B C D
Objective characteristics Friends chatting during Prosecuting and defence Welsh and English rugby Trade union and
of certain typical social coffee break lawyers in law court supporters in pub after management negotiation
situations (e.g., A-D) international match with crisis
mutually satisfactory
result

A1 B1 Cl D1
Possible cognitive Interindividual encounter Interindividual encounter Intergroup encounter Intergroup encounter
structures of social Co-operative Competitive Co-operative Competitive
situations by participants Informal Formal Informal Formal
Relaxed Tense Relaxed Tense
Equal Not equal Equal Not equal
Not task-related Task-related Not task-related Task-related

A2 B2 C2 D2
Potential speech patterns Low linguistic diversity High linguistic diversity Low linguistic diversity High linguistic diversity
Restricted code Elaborated code Restricted code Elaborated code
Verbal style Nominal style Verbal style Nominal style
Non-standard Standard pronunciations Non-standard Standard pronunciations
pronunciations pronunciations
Imprecise enunciations Precise enunciations Imprecise enunciations Precise enunciations
First name and informal Title and/or last name First name and informal Title and/or last name
address forms address forms address forms address forms
Attenuation of ingroup Accentuation of ingroup
speech markers (speech speech markers (speech
convergence) divergence)

table 6.1 Towards a model of speech as a dependent variable of social situations (from Giles and Hewstone, 1982: 200)
The intonation systems o f English

sample situations is reproduced in Table 6.1 with the phonological and

phonetic factors highlighted; intonation could have featured too.
Finally in this review of the ingredients of communication, we look at the
lexico-grammatical component of Fawcetts (1980) model of an interacting
mind. Fawcett divides up Flallidays three metafunctions into eight
functional components that, incidentally, help to identify the roles of
intonation; but in addition, he adds two further components. The first is for
those bits of language that help in expressing discourse organization, to
indicate a specific addressee or nominate the next turn-taker or indicate the
sequencing of information (e.g. firstly, on the one hand). The second is
metalanguage, as when a person overtly monitors their form of expression
by glossing (or in other words), reformulation, appeals (you know w hat I
mean?), and so on.

Hallidays metafunctions Fawcetts functional components

experiential
ideational logical relationships
negativity

interactional
interpersonal affective
modality
thematic
textual informational
discourse organizational
metalingual

TABLE 6.2

A speaker negotiates all ten kinds of meaning simultaneously, although the

last two are redundant unless they are marked overtly.
The value of Fawcetts interacting mind model is that it draws attention to
other components essential to the process of communication. The
communication itself draws upon a persons knowledge of the real world,
including the addressees presumed knowledge, and the persons affective
state. The communication is often supplemented by other semiotic codes,
like gestures, eye contact, physical distance, and intonational paralanguage.
The speaker often has a plan of communication, e.g. how to counter an
expected argument; but the plan may also involve a staging strategy; what to
say first, how to follow it up, etc. in blocks of information equivalent to
written paragraphs. Fawcett calls this' kind of planning the discourse
construction program. And finally, Fawcett acknowledges that a
communication can be accompanied by non-communicational behavioural
programs: The point of including this is to remind ourselves that not all

144
 Intonation in a model o f com m unication

options in behaviour necessarily involve communicating with another

member of our species, or any type of communication at all (Fawcett, 1980:
62-3).
We are now in a position to take an actors-eye-view of the communication
process, and in order to make the process clear, the first person singular will
be used, so that you can identify with it.

A need arises for me to say something. (By using the word say it is not
intended that writing is excluded; indeed the expression write down just
w hat y ou w ant to say indicates that saying is not confined to the oral
medium.)
I have at my disposal my language and a certain competence in it. I have
a stock of words (my lexicon) that I am confident in employing, but also I
acknowledge an acquaintance with other words with which I am not so
confident, and other words that I do not really understand. My active lexicon
is considerably smaller than the total stock of words available in English. I
am aware not only of the denotations of my active lexicon but also their
connotations, associations and collocations. I am also aware of differences
between standard and non-standard (dialectal) usages of some words and
the appropriateness of some words to a particular register. I have a stock of
syntactical patterns for the deployment of my total active lexicon and a list
of morphological shapes of the words for particular syntactic environments.
I am also aware that a few of those morphological shapes have variants in my
idiolect, e.g. the plural of form u la as either form u las or form u lae, the past
tense of kn eel as either knelt or kneeled. I know how to link words and
syntactic patterns with each other to produce coherent discourse. I have a
pronunciation system which enables me to articulate every single word I
wish to use, and every phrase, and every utterance; the system includes
consonants, vowels/diphthongs, syllable structures, word stress, rhythm tind
intonation. I also have variations in the system for rhetorical effect (like
joking in a non-standard accent) and for social effect, accommodating to or
disaccommodating from those I happen to be addressing. I have a script
(alphabet), a spelling and a punctuation system to record what I want to say
in writing.
I may also have a second (or third, etc.) language in which I am equally
confident. However, my competence in a second language may well be
distinctly more restricted. This degree of competence will determine my
effectiveness in any bilingual or second language situation.
I also have a knowledge of the universe, although I readily concede that I
do not know or understand everything. My (imperfect) knowledge extends
to an assumption of what my addressee knows and understands. It also
includes the context of my culture which enables me to assess most contexts
of situations that I encounter. It also provides me with notions of norms of

145
 The intonation systems o f English

interaction and interpretation and the appropriateness of types of genres for

particular occasions. I also readily concede that my knowledge of the
universe will diverge in both general and detailed ways from another
persons knowledge, that the context of my culture will be different from
others, and that I may be mistaken about any assumptions about others.
In wanting to say something, I draw upon my linguistic competence(s)
and my knowledge of the universe. I also decide how much I want to say and
what medium I want to use. If I want to say a lot, I will have to plan ahead
(discourse construction program); if I have very little to say, then there is
very little planning and I will probably proceed in a very a d hoc manner,
depending on how the situation changes and, in particular, how the topic
develops.
I know what I want to say; I recognize the topic. I assess the situation I find
myself in, and accordingly decide how I want to word what I want to say. I
take into account the physical setting, including the bystanders, and the
scene, the social occasion; also the participants, their number (it may be only
one participant other than myself, or even just myself in a soliloquy), their
age, gender, race, personality, interests, attentiveness, physical appearance
and any temporary mood, emotion or attitude, their relationship to me in
either a private or public capacity, their relationship with each other if they
number more than one, and their attitude towards me, any bystanders, and
the topic itself; what I hope to achieve, and what I think that they expect; my
feelings, as well as theirs; and what has just been said. All this influences my
decision about how I will say what I want to say.
Having assessed the situation, I encode what I want to say to the extent
that my competence in the language(s) allows. Particular items of my lexicon
are selected; they take particular morphological shapes for their functions in
the particular syntactical patterns I selected, which are joined together by
particular cohesive ties to form coherent discourse; particular phonological
or orthographical items are required; the brain activates particular neural
patterns and thus the encoding of the message is performed, by being
i converted into physical (either acoustic or visual) substance.
That physical substance is transmitted and received by your ear, or eye.
I monitor what I say, and if necessary, adjust it.
You decode, to the extent that your competence in the language(s) allows.
Particular items received are matched to your lexicon, grammar, phonology,
orthography and discourse competences, and thus, hopefully, a degree of
comprehension is achieved. However, some details may not match, e.g. the
connotations of a lexical item, the perception of a vowel sound (was it m an
or men?), the referent of a pronoun like those, etc.
Your comprehension of my message is then affected by your assessment
of the situation, along the same lines as above. Your assessment operates as

146
 Competence in
the language(s)
Need to Discourse
communicate Message construction
Knowledge of program
the universe

Reception Decoding Assessment o f

- Sound waves Comprehension situation
or from setting
handwriting/ lexis - scene
print, etc. grammar participants
discourse intention
phonology key
script/orthography linguistic
context
147

Figure 6.1 A model of the communication process

The intonation systems o f English

a filter, producing not just a comprehension of my message, but an

interpretation, and, possibly, a response. The interpretation develops as the
discourse proceeds. You also evaluate the message you have interpreted;
you might agree or disagree, accept it as true or regard it as a lie, consider it
interesting or insignificant, etc., but you do so in terms of your knowledge of
the universe.
The whole process can be captured in a flow chart (see Figure 6.1).
Fawcetts ten functional components can be identified in the flow chart:
the experiential, logical relationships, negativity and modality relate directly
to the knowledge of the universe (modality expresses an attitude to the
likelihood that a given proposition is true: Fawcett, 1980: 30); the
interactional relates to the speakers intention; the affective to key; the
thematic and informational to linguistic context; the metalingual to
monitoring; and the discourse organizational partly to the discourse
construction program and partly to linguistic context.

2. Intonation in the model

We are now in a position to view the contribution that intonation makes in
communication, in a global perspective. In order to help the reader identify
that contribution in the communication process, the first person singular
will again be used.

A need arises for me to say something.

I have at my disposal a competence in my language which includes all the
functions and systems of intonation. I may have a second (or third, etc.)
language, in which my competence in intonation may be equal to that in my
main, preferred, language; on the other hand, it may be precisely in the area
of intonation that I feel one of the limitations in that competence; in fact, my
intonation competence may be so seriously defective in that language, that I
use wrong patterns altogether and actually distort the message I seek to
communicate.
I also have at my disposal a knowledge of the universe which includes an
ability to select the most appropriate genre for a particular occasion. On an
occasion where informal conversation is appropriate, the prosodic
composition of my speech will be adapted to that and I will not sound as if
I am reading the news on the radio, or making an announcement at a
gathering of several hundred people, or praying, or lecturing, and so on. My
use of intonation will help people to identify the speech style, or genre, that
I have selected; see Chapter 1, section 4.6, and Tench (1988,1990: chapter 7),
Crystal and Davy (1969, 1975) and Johns-Lewis (1986).
I decide how much I want to say. If I am participating in an informal

148
 Intonation in a model of com m unication

conversation, any given turn of mine might consist of a single speech act
containing a single piece of information, encoded in a simple intonation
unit. On the other hand, I may want to say more than that, two, three, or
more intonation units, which will constitute a phonological paragraph. I may
wish to extend my turn to two or more phonological paragraphs, in which
case the paragraphing structure will become evident. Certainly in an
extended monologue, such structure will be evident; it helps people to
follow the development of my message, i.e. its textual structure; see Chapter
1, section 4.5 and Tench (1990: chapter 4) and Brazil and Coulthard (1979)
and Brazil et al. (1980), Brown and Yule (1983) and Pike (1967).
I decide how I wish to communicate. If I decide to speak, all the functions
and systems of intonation are brought into play. Even if I decide to write,
intonation still figures in the process. When I am writing a letter, I am often
conscious of the way I would speak it: the potential intonation will affect the
way I present information, the way I punctuate, the way I might emphasize
words or whole propositions and the way I paragraph the material. Even in
more formal writing, intonation plays a role; in a scholarly article Writing in
the perspective of speaking, Chafe (1986) wrote:
When writing is read aloud, of course, it does have intonation, though the reader
may use pitch contours quite different from the spoken norm. The very fact that
people assign various kinds of peculiar prosody in reading aloud might suggest
that such prosody is invented solely for that purpose, and that written language
is, in itself, devoid of intonation, stress, and pauses. But introspection suggests
that as both writers and readers we do assign such features to whatever we are
writing or reading. For example, when I wrote the last sentence I had in mind a
high pitch and strong stress on the word do'. You, as reader, may or may not have
read it that way, but in either case you are likely to know what you did. I am going
to assume that writers and readers assign intonation, stress, and pauses to
written language, though the writing itself provides less than optimal
representations of them.
Intonation is, of course, indicated to some extent - often with punctuation,
less often with italics. Again I invite you to consider how you read the last
sentence. (Chafe, 1986:18)
This is why the task of identifying the boundaries of intonation units in the
opening of Treasure Island was not an unreasonable exercise. Robert Louis
Stevenson knew that the art of good story-telling in print will match the art
of good story-telling in speech. In fact, a high proportion of all written
material will reflect the intonational properties of spoken discourse:
In speech each information unit is realized by an intonation unit. The equivalent
in writing is the unit boundaries that are expressed by punctuation marks - i.e.
commas, dashes, colons and semi-colons within sentences, and full stops
(periods in US English), question marks and exclamation marks at the ends of
sentences.

149
The intonation systems o f English

From here on we shall focus on intonation rather than punctuation. This is

because, even when the text is a written text, the reader tends to read into it the
intonation with which it would be spoken if it were read aloud - so that even with
a written text it is helpful to understand the meanings built into a spoken text.
(Fawcett, in preparation; original emphases)

To return to the involvement of intonation in my act of communication, I

assess the context of situation. In one very important respect, the setting has
to be taken very deliberately into effect. If the addressee is, or addressees
are, at some distance from me, I have to call them. The intonation of calling
is different from the intonation of settings in which all the participants are
within easy, unaided, talking distance. The communication in calling is also
different, in that the messages are usually short and the expected responses
are equally short. If I call somebody at a distance, especially if I dont know
where they are, I raise my voice in both volume and pitch and use two long
level pitches, the second usually a semi-tone lower (or two semi-tones lower
in repeated calls):
(6.6) "iona-than |~i(2na_than
Other, familiar, callings include:
(6.7) "Dinners -ready
(6.8) ~foo -ood
The pre-tonic is usually kept low, as in:
(6.9) Come and " get -it
The intonation of calling has not been dealt with in this book until this point;
more can be found in Pike (1945: 71-2), Liberman (1979) and Ladd (1980).
OConnor and Arnold (1973) recognize this intonation pattern, pattern 10
(The terrace) (see p. 115), but they do not illustrate it in any of their drills.
This pattern is also known as stylized intonation.
From my assessment of the context of situation, I decide on my
communicative intentions, express my feelings and take note of what has
been said already. My communicative intentions are expressed through my
choice of tone (see Chapter 4, sections 4 and 5), whether I am telling, asking,
bidding farewell, and so on. My feelings - if I wish - are expressed through
(non-neutral) variations of tones and pre-tonics (see Chapter 5). I note the
linguistic (and, at times, the situational) context through choices in tonicity
(see Chapter 3), especially in terms of new and given information.
I encode my message as successive pieces of information in intonation
units (tonality, see Chapter 2), each unit bearing a focus of information
(tonicity, see Chapter 3) and a status in relation to each successive unit (tone,
see Chapter 4, section 3). Occasionally, the precise choice in syntactic

150
 Intonation in a model of com m unication

organization is realized by tonality (see Chapter 2, section 7) and tonicity

(see Chapter 3, section 8).
And, finally, when I speak, I give away who I am. I not only give away my
views, feelings and intentions, and my perception of the place I occupy in
the encounter, but I also give away aspects of my identity: my voice gives
away an indication of my gender, age, social class and provenance. There is
thus an indexical function of intonation (and the rest of my pronunciation)
which includes my accent. The indexical function has not been included in
the treatment of intonation in this book because it does not directly affect
either meaning or paralanguage. It is manifest at the transmission stage of
communication. (This is not to deny, however, the possibility that the
addressee will evaluate the message in relation to the messenger. This most
certainly happens. We might, for instance, doubt the validity of a message
when we realize who the messenger is! But that is an operation of the
addressees knowledge of the universe, which is beyond the scope of
linguistic description.)
You will see that intonation features at every point in the process of
communication; it is impossible to account for any kind of linguistic
communication without it. From the process of reception by the addressee
to interpretation and evaluation, intonation is recognized, processed and
taken into account. Even in writing, and then reading, intonation plays a
part.

3. The intonation systems

The foregoing discussion of the contribution of intonation in the whole
communication process, and the detailed descriptions of the forms and
functions of intonation in English discourse, can all be displayed in a series
of system networks. We will make no specific provision for choices in genre
and medium, but the choices open to a speaker of English are the following
(Figure 6.2a and b):

Discourse construction program

r single phonological paragraph
a speaker's turn
multiple phonological paragraphs

phonological r single intonation unit

paragraph - multiple intonation units
(intonation group)

Figure 6.2a

151
 Encoding, after an assessment of the context of situation

- linguistic context
intonation
unit status of information
(tonality)
- interpersonal functions

- broad
linguistic focus /neutral tonicity/
context r on final lexical item
(tonicity) narrow
focus non-final lexical item
- /marked tonicity/
L non-lexical item

- neutral* /3/
- implication ------
- strong/4/
final in minor/2/
group
- major
status
(tone) - major -1
non-final - incomplete/2/ r neutral*
in group - neutral /5/J
- strong
- highlighted L intensified -
theme /3/

r dom inant---------
- mild
calling/29/
interpersonal
(tone)
neutral*

L deferent - strong ----------------------

- non-committal/10/
* neutral = not otherwise marked
// = realization (see p. 154) Attitude I:
tone and variations

Figure 6.2b
 neutral* / l/ neutral /17/
in siste n t/ll/ with - emphatic /18/
given info /12/ authority
- forceful /19/
neutral /6/ L expecting response
- neutral /20/
with with
appeal - emphatic /21/
emotion /7/
neutral /8/ - forceful /22/
important/13/
neutral* /2/
unconcerned /14/ - neutral /23/
given info /15/ with - emphatic/24/
authority
expecting response - - forceful /25/
expecting response - - neutral /26/
involved /16/ with - emphatic/27/
appeal
neutral* /9/ - forceful /28/

Attitude lla: lib:

level pre-tonics non-level pre-tonics
The intonation systems of English

Realizations
Pre-tonic (if present) Tone
1 mid level neutral fall They are going to come on \Monday
2 mid level neutral rise They are going to come on /Monday
3 descending neutral fall-rise They are going to come on vMondav
4 descending glissando low fall-rise They are going to come on vMonday
5 mid level high fall They are going to come on ^Monday
6 ascending neutral rise-fall They ate going to come on AMondav
7 ascending glissando low rise-fall They are going to come on . Monday
8 mid level low fall They are going to come on . Monday
9 mid level high rise They are going to come on -''Monday
10 mid level low rise They are going to come on .Monday
11 high level neutral/high fall They are going to come on \Monday
12 low level neutral fall They are -going to come on \Monday
13 high level low fall They are going to come on . Monday
14 low level neutral/low rise They are -going to come on /Monday
15 high level neutral rise They are going to come on /Monday
16 low level high rise They are -going to come on /Mondav
17 descending wide neutral/high fall They are xgoing to come on \Monday
18 descending stepping neutral/high fall They are going to -com e on \Monclay
19 descending glissando neutral/high fall They are \going to \come on \Monday
20 ascending wide neutral/high fall They are /going to come on \Monday
21 ascending stepping neutral/high fall They are -going to -com e on \Mondav
22 ascending glissando neutral/high fall They are /going to /come on \Monday
23 descending wide neutral/high rise They are xgoing to come on /Monday
24 descending stepping neutral/high rise They are going to come on /Monday
25 descending glissando neutral/high rise They are xgoing to \come on /Monday
26 ascending wide neutral/high rise They are /going to come on /Monday
27 ascending stepping neutral/high rise They are -going to -com e on /Monday
28 ascending glissando neutral/high rise They are /going to /come on /Monday
29 low level high level +
mid-high level They are -going to come on Mon-day
Figure 6.2c

154
 References

Allen, W. Stannard (1954) L iv in g E nglish Chafe, W. (1986) Writing in the

S p eech . London: Longmans. perspective of speaking in Cooper, R.
and Greenbaum, S. (eds), S tu d y in g
Armstrong, L.C. and Ward, I.C. (1926) A
W riting. London: Sage.
H a n d b o o k o f E n glish In to n a tio n (2nd
Couper-Kuhlen, E. (1986) A n In tro d u ctio n
edn 1931). Cambridge: Heffer.
to E n glish P rosody. London: Edward
Baker, A. (1982) I n t ro d u c in g English Arnold.
P ro n u n c ia tio n : A T e a c h e r s G uide. Cruttenden, A. (1986) In ton a tio n .
Cambridge: CUP., Cambridge: CUP.
Bolinger, D.L. (1986) In to n a tio n a n d its Crystal, D. (1969) P ro so d ic System s a n d
Parts. London: Edward Arnold. In to n a tio n in E nglish. Cambridge: CUP.
Bolinger, D.L. (1988) In to n a tio n a n d its Crystal, D. (1975) T he E nglish T o n e o f
Uses. London: Edward Arnold. Voice. London: Edward Arnold.

Brazil, D. (1975) D isc o u rse In to n a tio n I. Crystal, D. and Davy, D. (1969)

In vestiga tin g E n glish Style. London:
University of Birmingham: English
Longman.
Language Research.
Crystal, D. and Davy, D. (1975) A d v a n c e d
Brazil, D. (1978) D isc o u rse In to n a tio n II. C o n v ersa tio n a l English. London:
University of Birmingham: English Longman.
Language Research. Dickinson, L. and Mackin, R. (1969)
Brazil, D, and Coulthard, M. (1979) Varieties o f S p o k en English. Oxford:
E x c h a n g e S tru ctu re. University of OUP.
Birmingham: English Language Fawcett, R.P. (1980) C ognitive L inguistics
Research. a n d Social In tera ctio n . Heidelberg:
Julius Groos; Exeter: Exeter University.
Brazil, D., Coulthard, M. and Johns, C.
(1980) D isc o u rse In to n a tio n a n d Fawcett, R.P. (in preparation) H a n d b o o k
f o r the A nalysis o f S en ten ces in English
L a n g u a g e T ea ch in g. London: Longman.
Text, Vol. l\ S y n ta x : A System ic
Brown, G. (1977) L isten in g to S p o k en F u n c tio n a l A p p ro a ch . Extracts availably
E n glish (2nd edn 1990). London: from Computational Linguistics Unit,
Longman. University of Wales, Cardiff CF1 3TL.
Brown, G., Currie, K. and Kenworthy, J. Gibbon, D. (1976) P erspectives o n
(1980) Q uestion s o f In to n a tio n . In to n a tio n A nalysis. Frankfurt ant
London: Croom Helm. Main: Peter Lang.
Giles, H. and Coupland, N. (1991)
Brown, G. and Yule, G. (1983) D isco u rse
L a n g u a g e : C ontexts a n d C o n seq u en ces.
Analysis. Cambridge: CUP.
Milton Keynes: Open UP.
Brown, P. and Fraser, C. (1979) Speech as a Giles, H. and Hewstone, M. (1982)
marker of situation in Scherer, K.R. and Cognitive structures, speech and social
Giles, H. (eds), Social M a rk ers in situations: two integrative models,
Speech . Cambridge: CUP. L a n g u a g e S cie n c es 4: 187-219.

155
 References

Gimson, A.C. (1989) A n Introduction to the Lehiste, I. (1975) The phonetic structure of
P ro n u n cia tio n o f English, 4th edn. paragraphs in Cohen, A. and
London: Edward Arnold. Nooteboom, S. G. (eds), Structure a n d
Halliday, M.A.K. (1967) Intonation a n d Process in Speech Perception. Berlin:
G ra m m a r in British English. The Springer.
, Hague: Mouton. Lehiste, I. (1979) Perception of sentence
Halliday, M.A.K. (1970) A C ourse in Spoken and paragraph boundaries in
English: Intonation. London: Oxford Lindblom, B. and Ohman, S. (eds),
UP. Frontiers o f Speech Research. London:
Academic Press.
Halliday, M.A.K. (1973) Explorations in the
F un ction s o f L a ngu a ge. London: Liberman, M. (1979) The Intonational
Edward Arnold. System o f English. New York/London:
Garland.
Halliday, M.A.K. (1978) L a ngu a ge as Social
Semiotic. London: Edward Arnold. OConnor, J.D. and Arnold, G.F. (1961)
Intonation o f Colloquial English (2nd
Halliday, M.A.K. (1985) A n Introduction to
edn 1973). London: Longman.
F u n ctio n a l G ram m ar. London: Edward
Arnold. Palmer, H.E. (1922) English Intonation
(with Systematic Exercises). Cambridge:
Halliday, M.A.K. and Hasan, R. (1985)
Heffer.
L a ngu a ge, Context, a n d Text. Oxford:
OUP. Pike, K.L. (1945) The Intonation o f
A m erica n English. Ann Arbor:
Halliday, M.A.K., McIntosh, A. and Strevens,
University of Michigan Press.
P. (1964) The Linguistic Sciences a n d
L a n g u a g e Teaching. London: Longman. Pike, K.L. (1967) L a n gu a ge in Relation to a
Unified Theory o f the Structure o f
Hooke, R. and Rowell, J. (1982) A
H u m a n Behavior, 2nd edn. The Hague:
H a n d b o o k o f English Pronunciation.
Mouton.
London: Edward Arnold.
Quirk, R., Duckworth, A.P, Svartvik, J.,
Hudson, R.A. (1975) The meaning of
Rusiecki, J.P.L. and Colin, A.J.T. (1964)
questions, L a n g u a g e 51: 131-
Studies in the correspondence of
Hymes, D. (1977) F oundations in prosodic to grammatical features of
Sociolinguistics. London: Tavistock. English, Proc. Ling. IX: 679-91.
Johns-Lewis, C. (1986) Prosodic Quirk, R., Greenbaum, S., Leech, G. and
differentiation of discourse modes in Svartvik, J. (1972) A G ra m m a r o f
Johns-Lewis, C. (ed.), Intonation in Contem porary English. London:
D iscourse. London: Croom Helm. Longman.
Joos, M. (1961) The Five Clocks. New York: Roach, P. (1983) English Phonetics a n d
Harcourt, Brace and World. Phonology (2nd edn 1991). Cambridge:
Kingdon, R. (1958) The G roundw ork o f CUP.
English Intonation. London: Longman. Schubiger, M. (1958) English Intonation, its
Knowles, G. (1989) Patterns o f Spoken F o rm a n d Function. Tubingen:
English. London: Longman. Niemeyer.
Ladd, D.R. (1980) The Structure o f Sinclair, J. McH. (1972) A C ourse in Spoken
Intonational M ean in g. Bloomington: English: G ram m ar. London: Oxford UP.
Indiana UP. Taglicht, J. (1984) M essage a n d Emphasis:
Ladd, D.R. (1983) Phonological features of On Focus a n d Scope in English.
intonational peaks, L a ngu a ge 59: London; Longman.
721-59. Tench, P. (1988) The stylistic potential of
Lee, W.R. (1956) English intonation: a new intonation in Coupland, N. (ed.), Styles
approach, L in gu a 5: 345-71. o f Discourse. London: Croom Helm.
Leech, G. and Svartvik, J. (1994) A Tench, P. (1990) The Roles o f Intonation in
C om m unicative G ra m m a r o f English, English Discourse. Frankfurt am Main:
2nd edn. London: Longman. Peter Lang.

156

II
 References

Thompson, I. (1981) Intonation Practice. (ed.), Intonation in Discourse. London:

Oxford: OUP. Croom Helm.
Wells, W.H.G. (1986) An experimental Young, D. (1980) The Structure o f English
approach to the interpretation of focus Clauses. London: Hutchinson.
in spoken English in Johns-Lewis, C.

157

<
 Index

acknowledgement 92, 112 concession 93

adjunct 35, 37, 64-8, 82, 83 congratulations 103
advice 9 8 -9 consultative style 140
affirmation 92 context o f situation 141
agreement 92 contradictions 9 3 -4
announcements 99-101, 103, counting 81
104 Cruttenden 69, 70, 72, 115, 126-7
answers 92 Crystal vii, 7, 36, 50, 56, 108, 115,
apologies 103 121-3, 125-35
appeals 92, 100-1 Crystal and Davy 32, 33, 58-9,
apposition 41-4 148
appreciation 103
approval 105 decrescendo 54
Armstrong and Ward 7, 82 deduction 92
attention 104-5 default tonicity 68-70
attitude 7, 20-1, 74, 107-37 deference 18, 88, 89, 91, 105
demands 98
back channel 105 denials 92, 95
baseline 24-6 description 92
Bolinger 72 disagreement 92
Brazil et at 25, 105, 108, 137, 149 disapproval 103
Brown 24, 95, 108, 125, 126, 127 discourse functions 18, 108
Brown, Currie and Kenworthy dismissal 102
5-6, 49-50, 52 dominance 1 8 ,8 6 ,8 8 ,8 9 ,9 1 -1 0 5
Brown and Yule 149 doubt 9 2 -3

calls 115, 150 echoes 68, 94, 126

casual style 140 enquiries 94, 104
Chafe 149 exclamations 19, 95, 104
challenges 94, 126 explanations 92
clause 31-49
s e e also syntax falls 4-5, 18-19, 20, 28, 74, 75-7,
coaxing 98 81-3, 86-91, 92-105, 125-6,
commands 86, 88, 90, 98 136-7, 154
comm ent adjunct 67-8 fall-rises 4-5, 79, 83-6, 92-3, 128,
communicative functions 18-19, 136, 137, 154
74, 86, 87, 90, 91, 119 farewells 101

158
 Index

Fawcett 144-5, 148, 150 interpersonal meaning 139-40

focus intimate style 141
broad 57-9 introductions 104
narrow 59-61 invitations 99
se e also tonicity
football scores 63, 72 Jassem 7
forgiveness 104 Johns-Lewis 148
formal style 140 Jones 7
frozen style 140 Joos 140-1

Garrick 7 key 25, 119-21, 142

genre 17, 26-8, 56, 63-4, 142, 148 Kingdon 7 ,8 4 ,1 2 7 ,1 2 8
Giles and Coupland 142 knowledge 92
Giles and Hewstone 142-4 Knowles 106
Gimson 106, 126, 127, 134
given information 17, 19, 34, 57, Ladd 72, 106, 108, 109, 133, 150
59-61 lead-ins 95-6
glissando 129, 133-4 Lee 85
grammar 9 ,2 1 -3 ,3 2 ,3 9 -4 9 ,5 0 -1 , level tone 27, 72, 81, 105, 111, 122
57, 58, 60, 61-2, 70-2, 87, 90, lexical item 58-9
91, 119, 150-1 lexical tone 3-4
greetings 101-2 Liberman 108, 109, 118, 126, 132,
guesses 95 134, 150
lists 36, 81, 96, 121
Halliday vii, 7, 14, 31, 36, 48, 50, loudness 53-4
55, 57, 80, 84, 105, 108, 119-21,
125-35, 139, 141, 144 major information 28, 80, 86, 87,
Halliday and Hasan 139, 141 152-3
Hausa 10 minor information 80, 86, 152-3
head 12, 20, 109-15, 128-34, 137
hesitation 28 negative domain 45-6, 85
highlighted theme 83-4, 86 new information 17, 57
hope 92 news reading 20, 23, 27-8, 56,
Hymes 141-2 108
hypotheses 92 nucleus 12

ideational meaning 139 oblique tone 105

illocution 18 OConnor and Arnold 7, 108,
implication 84, 86, 93, 104 109-15, 118, 125-35, 150
incomplete information 80, 86, 87 offers 100
indexical function 29, 151 onset syllable 12, 14, 23, 109-12
information 7, 16-18, 39, 50-1,
108 Palmer 7, 82, 126, 127, 1 3 1 -2
information focus see tonicity paragraph see phonological
information status 80-6, 87 paragraph
information unit 51 paralanguage 2, 28, 121, 123, 144
s e e also tonality paratone 24
insists 69 pause 17, 23, 54
instruction 90 phonological paragraph 23-6,
interjections 86 27-8, 149

159
The intonation systems o f English

Pike vii, 7, 50, 105, 106, 108, sentence 5 -6

115-19, 123, 125-34, 149, 150 sentence intonation/'prosody 6
pitch peak 54 speech acts/functions 18, 141
pitch range 54, 121, 123 statements 86-90, 92
pleas 98 Steele 7
possibility 93 stylized intonation 150
praise 103 suggestions 9 8 -9
pre-head 12, 20, 109-15, 134-5, summons 104
137 suppositions 92
pre-tonic segment 12, 28, 73, 75, syllables 14-15
78 sympathy 103
prohibition 98 syntax 21-3, 87, 119, 150-1
promises 100
prompts 95 tag 4, 38-9, 86, 88
prosodic composition 27-8, 142, tail 13, 73
148 telephone numbers 6 3 -4
punctuation 9, 21, 22, 28, 34, 40, tempo 50, 54
42, 49, 71, 77, 105, 149 textual meaning 139-40
textual structure 23-6, 27, 149
Quirk et al. 4 2 -4 Thai 3
questions thanking 102
alternative 96 theme 3 2 , 3 7 , 5 2 , 8 3
conducive 95 threat 93, 100
echo 94 toasts 103
list 96 tonality 8, 9, 17, 18, 22, 31-52, 66,
repeat 96 139, 150, 151, 152-3
rhetorical 89 tone 8, 17, 18, 53, 73-106, 140,
wh- 86-90 150, 152-3
y e s /n o 86-90, 92 tone languages 3, 6
tonic 12, 17, 53-6
reassurance 104 tonic segment 13
recommendations 90, 98-100 tonicity 8, 17, 18, 22, 53-72, 140,
regret 103-4 150, 151, 152-3
report clauses 4 6 -8
reports 92 utterance 3 -6
requests 19, 92, 94, 98, 104
rheme 32, 52, 84 viewpoint adjunct 38, 67
rhythm 1, 7 vocatives 38, 67, 71, 104
rhythmicality 27
rises 4, 18-19, 20, 77-9, 80-3, warnings 100
86-91, 92-105, 125-8, 135, welcomes 102
136-7, 154 wishes 91, 102-3
rise-falls 127-8, 136, 154 word phonology 7, 11

160