I

First published in Great Britain in 1995 by

Arnold, a member of the Hodder Headline group
338 Huston Road, London NW1 3BH

© 1995 Norman Fairclough

All rights reserved. No part of this publication may be reproduced or transmitted in any
form or by any means, electronically or mechanically, including photocopying,
recording or any information storage or retrieval system, without either prior
permission in writing from the publisher or a licence permitting restricted copying. In
the United Kingdom such licences are issued by the Copyright Licensing Agency:
90 Tottenham Court Road, London W1P 9HE.

British Library Cataloguing in Publication Data

A catalogue entry for this book is available from the British Library

Library of Congress Cataloging-in-Publication Data

Fairclough, Norman, 1941-
Media discourse/Norman Fairclough.
p. cm.
Includes bibliographical references (p. 206) and index.
ISBN 0-340-63222-4. — ISBN 0-340-58889-6
1. Mass media and language. I. Title.
P96.L34F35 1995
302.23— dc20 95-17679
C1P

ISBN 0 340 58889 6 (Pb)

ISBN 0 340 63222 4 (Hb)

7 8 9 10 98 99 00 01 02

Composition by Phoenix Photosetting, Chatham, Kent

Printed and bound in Great Britain by Redwood Books, Trowbridge, Wiltshire
 CONTENTS

Acknowledgements vii
Note on transcriptions viji

1 Media and language: setting an agenda

:l
• .vj- 2 Approaches to media discourse 20
■ Linguistic and sociolinguistic analysis 21
A Conversation analysis 21
Semiotic analysis 24
Critical linguistics and social semiotics 25
Van Dijk: the 'social-cognitive' model 28
Ad -■ Cultural-generic analysis 31
:■
,A:l Desiderata for a critical analysis of media discourse 32

Communication in the mass media 35

The properties of mass communication 36
The economics of media 42
 The politics of media 44
Practices of media text production and consumption 48
4 Critical analysis of media discourse 53
Theory of discourse ■53
Analysis of communicative events 57
Analysis of the order of discourse 62
A sample critical discourse analysis 68
5 Intertextuality and the News 75
Discourse types 76
Discourse representation in media texts 79
Generic analysis of discourse types 85
Analysis of discourses in texts 94
6 Representations in documentary and news 103
Presences and absences in text: presupposition 106
Representations in clauses 109
Combination and sequencing of clauses 117
7 Identity and social relations in media texts 125
Medicine Now 128
High Resolution 135
The Oprah Winfrei/ Show 139
Today 142
8 Crimewatch UK 150
The Crimewatch format 151
Generic analysis 154
Voices 161
Discourses: official and lifeworld 164
Between state and people 167
999 169
9 Political discourse in the media 176
Creating a new political discourse 177
Bourdieu: the field of politics 181
The order of mediatized political discourse 184
Mediatized political discourse: further considerations 197
10 Critical media literacy 201
Bibliography 206
index 213
 ACKNOWLEDGEMENTS

My thanks to those who have looked at and commented on a draft

version of this book-A lan Bell, Mike Birch, Gunther Kress, Theo van
Leeuwen, Mary Talbot - or commented on presentations based upon
C lU ctp itri'S
' O 1 e. t j .-t-i j 3 i -j , j-—s. s i-
O clilCi -j ~ r \ i lU-ltiW i O iS u i’t cxitU C O iitc i^ u -c S aL ^ U c c i ’i iV itii'g cit'ct
College, Edinburgh, and those who attended a seminar on critical
discourse analysis at the London Institute of Education in April 1994.
My thanks also to students, visitors and colleagues in the Depart­
ment of Linguistics at Lancaster University, and to participants in
many fascinating interdisciplinary meetings, for a stimulating intel­
lectual environment. To Simon and Matthew for their precocious
critical media awareness. And above all to Vonny.
Thanks are due to the following for permission to reproduce copy­
right material: The British Broadcasting Corporation for extracts from
Medicine Noiv, High Resolution, and Today (all Radio 4). Mirror Syndi­
cation International for extracts from the Daily Mirror, 7 October 1992
and 14 January 1993. The Guardian for an extract, 7 October 1992. Rex
Features Ltd for extracts from the Sun, 24 May 1985 and 14 January
1993.
via
Note on transcriptions
The transcriptions of broadcast talk in this book are in some cases the
author's, in other cases taken from a variety of published sources.
Transcriptions generally do not divide speech into sentences, and
omit the punctuation found in written text. Other transcription con­
ventions vary - in, for instance, the extent to which they reproduce
hesitations.
The following conventions will be found here:
Pauses: represented as dots between words, like . this. Longer
pauses are shown by two . . or three . . . dots, or by giving the pause
as parts of a second, e.g. (0.5).
Hesitations: e, em, e:, e:m, or e:r, with the colon marking an elon­
gated syllable.
Simultaneous talk by two speakers is indicated by a square bracket.
 MEDIA AND LANGUAGE: SETTING AN AGENDA

Four events took place in roughly the first half of 1994, while I was
working on this book: Silvio Berlusconi's Fnrza Italia won the Italian
general election, in the UK Tony Blair was elected leader of the
Labour Party, between one and two million Hutu refugees fled from
Rwanda into Zaire in the space of a few days, and Rupert Murdoch
made a week-long trip to Delhi. It was generally recognized that
Forza Italia was a media creation (Berlusconi founded the party in
January, it won the election in March) and that Berlusconi's victory
was largely the result of his control of the Italian media - he owns
three television channels with a 40 per cent share of the audience, a
national newspaper, and Italy's biggest publishing company. Long
before the Labour leadership contest even officially opened, most of
the British media had already chosen Tony Blair as the successor to
the late John Smith. Blair's campaign was orchestrated by Labour's
own 'spin doctor' Peter Mandelson, and his attractiveness as a media
personality was seen (whatever his other virtues) as a major qualifica­
tion for the job. In mid-July the civil war in Rwanda, which had
received patchy coverage before, suddenly became the lead item on
2 Media and language: setting an agenda
television news (and in other media) for days on end, with extensive,
shocking coverage of suffering and death amongst the massive
numbers of Hutu refugees. And Murdoch's visit to Delhi was linked
to his acquisition of access to five satellite television channels beamed
at 2.5 billion people in fifty countries - more than two-thirds of the
world's population. The common theme of these events is the power
of the mass media. The power of the media to shape governments
and parties, to transform the suffering of the South (rooted in ex­
ploitation by the North) into the entertainment of the North, to beam
the popular culture of North America and western Europe into
Indian agricultural communities which still depend upon bullock-
power. The power to influence knowledge, beliefs, values, social
relations, social identities. A signifying power (the power to repre­
sent things in particular ways) which is largely a matter of how lan­
guage is used, but not only that: what made Rwanda 'good television'
for a short period in July 1994 was above all the availability of high-
quality film of the appalling human suffering.
This book has several objectives. The first is to set out a framework
for analysing media language which readers can use for themselves
to pursue their own interests in mass media. 1 hope to persuade
readers with a background in language studies of the particular fas­
cinations associated with analysing media language. And I hope to
persuade readers with a background in media studies of the value of
analysing mass media linguistically and in terms of discourse (I use
'discourse' for language use seen in a particular way, as a form of
bt: wÍLíi LIÍCi UCiUVV) • ., .' '-1” 1:t. . iStic anct
- ‘J “ " O '
discoursal nature of the power of the media is one good argument for
doing so. But I must stress that the approach to language adopted
here is a novel one, which links in well with issues which have been
widely taken up in recent media studies, such as intertextuality,
genre mixing, and identity.
A second objective is to argue a particular case: I hope to convince
readers that analysis of media language should be recognized as an
important element within research on contemporary processes of
social and cultural change, a theme which is attracting growing inter­
est in the social sciences. I have in mind, for instance, research on a
claimed transition from 'modern' to 'postmodern' society or from
'high modern' to 'late modern' society, research on shifts in cultural
values (e.g. in the direction of individualism or a 'promotional cul­
ture') and the constitution of social identities, on 'detraditionali-
zation' and changes in power relations and authority relations, and
so forth (see, for example, Beck 1992, Featherstone 1991, Giddens
 Media and language: setting an agenda 3
1991, Lash 1990, Wernick 1991). Given the focal position of the mass
media in contemporary social systems, there can be little argument
about their relevance to the study of sociocultural change. What will
be less obvious to most social scientists, and more contentious, is that
analysis of the language of the mass media can make a substantive
contribution to such research. I hope to establish this in this b o o k -
and my argument is again dependent upon a novel approach to lan­
guage analysis which rejects the arid formalism of past approaches. A
third objective, as I have already implied, is to highlight the linguistic
and discoursal nature of media power.
I shall be focusing upon the language of what we might call public
affairs media - news, documentary, magazine programmes, dealing
with politics, social affairs, science, and so forth. Many of my
examples are taken from the British media - the press, radio and tele­
vision - in the period 1992-3, and there are also some from the USA
and Australia. In the course of the book I shall refer to quite a number
of samples of media language, and I want to begin here with four
short examples which will give readers a sense of some of the main
concerns of the book, and provide a basis for the more theoretical dis­
cussion later in this chapter.
The first is from the beginning of an edition of the BBC current
affairs programme Panorama, concerned with the reprocessing in
Britain of nuclear fuel from overseas (BBC1, 10 August 1992). The
reporter, John Taylor, is pictured facing the camera, leaning against
the rail of a launch, with the ship referred to in the text at anchor in
lU'-
In the coming week this ship, the Shikishimi, will put to sea to guard a
deadly cargo on a dangerous voyage around the world. Its cargo will be
plutonium, one of the world's most toxic substances, and the raw mater­
ial of nuclear weapons. It will herald the start of an international trade in
plutonium centred around British Nuclear Fuel's reprocessing plant at
Sellatield. Critics say each shipment could be a floating Chernobyl.
Tonight Panorama asks: is the plutonium business worth the risk?

This extract is followed by the usual Panorama opening sequence

including the programme logo (a revolving globe) and signature
tune, and a sequence of images representing nuclear risk (including
the explosion of a nuclear bomb, and someone testing for radio­
activity with a Geiger counter).
Apart from the last sentence, which contains a question (is the plu­
tonium business ivorth the risk?), the extract consists of declarative sen­
tences - statements. (I shall use as little linguistic terminology as
4 Media and language: setting an agenda
possible, and the terms 1 do use are explained as we go along.) The
first three sentences are statements about what will happen in the
future. Despite the fact that future events are contingent on many
things and therefore uncertain, these are firm, categórica] statements
- that is the effect of using the auxiliary verb will - and there is no
qualification or 'hedging' (no 'probably' or 'maybe'). These categori­
cal statements are part of how a relationship between the reporter
and the audience, and social identities for reporter and audience, are
established at the outset of the programme. The reporter is projected
as a figure of authority, someone who knows (has 'the facts'), and
someone who has the right to tell. The authoritativeness of the lan­
guage works together with the authoritativeness of the image - a
well-known reporter directly addressing the audience on-camera -
and of the delivery, which is measured, emphatic (the reporter using
movements of head and hands to support vocal emphasis) and seri­
ous. The audience is projected as receptive, waiting to be told,
wanting to know.
But this is only part of the story, for reporter/audience identities and
relations are more complex. In addition to the knowledgeable reporter
informing the interested citizen, there is an element - more muted in
this example than the next one - oi the media artist entertaining the
viewer as consumer. This is evident in certain rhetorical, attention-
grabbing features: the direct question at the end, the metaphor of a
ffnpHng Chmiotnil which links reprocessing to the nuclear cause célebre of
Chernobyl in a witty and memorable phrase. It is also evident in the
choice of genre: the decision to represent the issue of the 'international
trade in plutonium' as a narrative, a story, about a projected voyage of
an individual ship - not something that has happened (as in most
stories) but something that is expected to happen. This story, with
pictures of the actual ship, makes for a more dramatic and entertaining
account than a description of the planned trade in general terms within
an expository genre, which might have been selected.
In any representation, you have to decide what to include and
what to exclude, and what to 'foreground' and what to 'background'.
In this case, certain details which you might have expected to be back­
grounded or excluded altogether - on the grounds that they are
common knowledge which a Panorama audience might be expected to
share - have been foregrounded: describing the cargo as deadly and
the voyage as dangerous, mentioning that plutonium is one of the
world's most toxic substances and the raw material of nuclear weapons. This
detail generates a sense of alarm, underlined by the reporter's
delivery which stresses the words deadly, dangerous and toxic, it is
 Media and language: setting an agenda 5
sensationalist. It also helps to build up a negative, critical view of the
trade early in the programme, as indeed does representing it as a trade
and a business rather than, say, as a transfer between countries. Notice
also that the trade is centred around the reprocessing plant at Sellafield
(already another cause celebre), rather than, say, involving Sellafield, or
/«dzem/Sellafield and Japan. The sentence beginning Cr/hcss/?!/ . . . is
interesting from this point of view. Given that the programme has
apparently already joined the cri tics, perhaps the role of critics say is a
'modal' one, to mitigate and disclaim responsibility for a damning
judgement by attributing it to unspecified others. The indirectness
and implicitness of the critical stance towards the 'trade' perhaps
shows a tension and trade-off between the evenhandedness required
of the reporter in his more traditional information-giving, authori­
tative role, and the more sensationalist demands upon the reporter as
entertainer. Tension between the objectives of giving information
and entertaining is widespread in the contemporary media.
This brief example shows how analysis of the language of media texts
- by which I mean what is said in broadcasts as well as what is written in
the press - can illuminate three sets of questions about media output:
1. How is the world (events, relationships, etc.) represented?
2. What identities are set up for those involved in the programme or
story (reporters, audiences, 'third parties' referred to or
interviewed)?
'*3>. W K p> f r p !? i H o n H ir w co b tj n h p | i . i/ o n n }-]-> in i/ r t ] i'n r j (n y

reporter-audience, expert—audience or p>olitician-audience rela­

tionships)?
I shall refer from now on to representations, identities and relations. A
useful working assumption is that any part of any text (from the
media or from elsewhere) will be simultaneously representing, set­
ting up identities, and setting up relations.
My second example comes from an edition of the ITV current
affairs programme This Week entitled 'Vigilante!' (10 September 1992),
which dealt with vigilante groups in Britain enacting their own justice
where they perceive the law to be ineffective. The programme opens
with a 'trailer' which gives brief versions of the vigilante stories to be
covered, followed by the usual This Week opening visual sequence
and signature tune, then the programme title 'Vigilante!' imposed on
a still picture of a silhouetted man carrying what appears to be an axe
handle. My extract comes after this. On the left 1 have given a rough
representation of visual images in the extract, and on the right the
language (reporter voice-over).
6 Media and language: setting an agenda
IMAGES LANGUAGE
Pictures of hills and valleys, As the coalmines of South
sound of choir Wales fall silent, the
blackened hills and valleys
grow green again. It's a
picture of peace. But in the
Groups of people converge on village of Penwyn, in July,
house, shouting an ugly scene was played
out following the violent
death of an elderly spinster.
When two teenage girls
from the neighbourhood
were charged with murder,
a mob of several hundred
local people converged on
Crowd in front of houses, the houses where the
gestures and shouts parents of the accused lived.
(Long pause filled with
shouting.) The dead
woman's complaints of
harassment had apparently
gone unheeded. The crowd
were enraged by reports
she’d been so brutally killed
Missiles picked up and thrown that she could only be
at windows, sounds of breaking identified by her
glass, crowd shouting and fingerprints. (Long pause
cheering filled with shouting.) A
shower of missiles drove the
families from their homes.
The police could do nothing
but help them to safety.

This extract takes one step further the tendency in the earlier one for
reporter and audience identities and relations to be on the
entertainer-consumer model. The genre is past-event narrative, and
the story is told through a combination of words and what the pro­
gramme identifies as a filmed reconstruction of the incident. The
extract, and indeed the programme as a whole, is on the borderline
between information and entertainment, and between fact and
fiction. The visual narrative of the film, in which the crowd is played
bv actors, is dramatic fiction.
 Media and language: setting an agenda 7
The images have primacy over the words in the sense that the
events related happen first visually (e.g. we see a missile thrown
before we hear a shower of missiles). (See Barthes 1977 and van Lee-
uwen 1991 on variable relationships between words and images.)
The linguistic account provides an interpretation of the images,
identifying the people in the crowd, the house and its inhabitants,
but also shifting between narrating events and providing setting and
background for them, often in the same sentence. An important part
of this is providing explanations of the crowd's behaviour.
There are also apparent inconsistencies between words and
images. The images show, first, groups of angry-looking people
walking purposefully along shouting, then a crowd of angry people
shouting and gesticulating in front of the lighted window of a house,
then some of them hurling missiles at the window, and glass
breaking. Responsibility for the violence is clear and unmitigated in
the film. In the linguistic account, responsibility is less clearly
attributed, and is mitigated. There are just three clauses (simple sen­
tences) which recount the incident itself. What is interesting is both
the way these are formulated, and the way they are positioned in the
account. The first (an ugly scene was played out) is vague about who did
what to whom, the third (A shower of missiles drove the families from their
homes) transforms the action of throwing missiles into an entity, a
shower of missiles, and does not indicate who actually did it. Only in
the second (a met of several hundred local people converged on Ike houses) is
the crowd represented as actually taking action, and then it is 'con­
verging on' (which implies a controlled action that does not entirely
square with the behaviour of a 'mob') rather than 'attacking' the
house.
What I'm suggesting is that the linguistic account is rather
restrained in blaming the crowd. True, it is referred to damningly
as 'a mob', but two sentences later it is referred to more neutrally
as 'the crowd'. What is significant about the positioning of these
event clauses is that they are separated by background explanatory
clauses. This both slows down the story and reduces the impact of
the violence; it also mitigates the actions of the crowd by framing
them with a great deal of interpretative, explanatory material.
There is, in short, an ambivalence in the representation here which
accords, I think, with an ambivalence in the programme as a
whole: it does not wish to defend unlawful violence, but it
presents the vigilantes as normally decent people frustrated by the
ineffectiveness of the law. The notion of 'good television' perhaps
favours the image of frightening violence in the film, which is
8 Media and language: setting an agenda
unambivalent, but which can be partly 'balanced' by mitigating lan­
guage. Once again, there is a tension between information and enter­
tainment.
My next example is from a programme in a BBC education series on
engineering, called The Works. Produced in collaboration with the
Engineering Training Authority, the series was designed to 'show
engineers in a creative light', mainly to secondary-school pupils. The
programme, entitled 'Slippery When Wet' (BBC2, i September 1992),
is concerned with liquids. The extract comes immediately after the
series opening sequence, which is done in a 'pop video' style, with a
fast-changing sequence of technical and scientific images
accompanied by loud synthesized music.

A liquid: a substance that can change its shape, but cannot be

expanded or compressed. These properties give liquids a special part
to play in the triumph of technology. Half the weight of this massive
aircraft is liquid, mostly kerosene, but also water, hydraulic oil, engine
oil, toilets, detergents, booze, and of course passengers, who are also
two-thirds liquid. It's a miracle it can fly at all. But without fluids, it
wouldn't work.

The language here is produced by an unidentified reporter in voice­

over. It is accompanied by highly complex interlocking images, music
and sound effects which give the programme a style which is quite
different from traditional forms oi television science. During the
course of this short extract, there are images of: a drop of water falling
in slow motion into a puddle, what appear to be blow-ups of water
molecules, liquid pouring into a vat, oil pouring into a glass, part of
an engine rotating at high speed, toothpaste being spread on to a
toothbrush, a hand 'painting out' the rotating engine to reveal an
aircraft taking off. At transitions between these images, they are
superimposed upon one another in several cases. Most of them are
accompanied by appropriate sound effects, and through most of the
sequence there is music. The overall result is noisy, fast-moving,
bewildering, and certainly attention-grabbing, an unusually enter­
taining form of broadcast science, providing a different resolution of
the information-entertainment tension from that of the last example.
The extract also illustrates another, related, tension, between
public and private: science and technology are part of public, institu­
tional life, as indeed is the whole business of producing television
programmes —but those programmes are received and consumed
overwhelmingly in private contexts, in the home, within the family.
p,,KHr’ life, ami nrivatp life involve different wavs of using language,
 Media and language: selling an agenda 9
and we find this tension realized in a combination, within the extract,,
of private and public language. The private element is actually most
striking in features of the extract which are not apparent in the tran­
scription: accent and delivery. The reporter has a Tyneside accent.
This is an accent which is more common among characters in broadcas t
drama than amongst political, science or education broadcasters; for
most people, it is associated with private life rather than public life. The
effect here is to weaken the boundary between the public and the pri­
vate, mixing the public world of science and technology with a voice
from ordinary' life. The delivery is also strikingly.' conversational in
rhythm, intonationandstress. The mixture of public and private isalso
evident in the transcription. The language in part has a semi-technical
character: terms like substance, properties and fluids are part of scientific
vocabulary, and the provision of formal definitions (like the definition
of a liquid at the beginning) is a scientific but not an ordinary language
practice. Notice, however, thatthereisnospecialistvocabulary which
a reasonably well-educated person might not understand. But there is
also some conversational language: booze, wouldn't work, describing
the aircraft as massive (the word is also foregrounded by being em­
phatically stressed), and the idiomatic formula it's a miracle it can . . .
(fly, in this case). We can describe this as a case of cnnversntioiw.lization
of the public language of science and technology (on cunversational-
ization in public language, see Fairclough 1994).
My final exam pie is taken from the luday programme winch is
broadcast every weekday morning on BBC Radio 4. The particular
programme I am using was broadcast during the 1992 UK general
election campaign (8 April 1992). The presenter, Brian Redhead (BR),
is asking representatives of each of the three main political parties
(Conservative, Labour, Liberal Democrat) why' an imaginary
'floating voter' should support them.

br: now our floating voter turns to you Brian Gould and he says look
(bg: yeah) I don't really' fane1/anoth er Conservative government I think
we've had enough of tha t bul l can't really bring myself to vote for you
because you've been out of office for so long you haven't got the experi­
ence if you get in the City might say do this lot know enough to run the
country I'm nervous that a vote for you would mean a vote for some
kind of flight from the pound
(answer from Brian Gould, question from br to Des Wilson, and
answer from Des Wilson omitted)
hr: Des Wilson thank you now . imagine this floating voter actually is
a mate of all three of you . knows you personally . and has sat up
he's a different bloke altogether this one's b e e n h e r e t h r o u p h t h e w h o l e
10 Media and language: setting an agenda
election he's listened to every blooming broadcast (one of panel: lucky
chap) he's fed up to the back teeth (one of panel: haven't we all) . and
he rings you up and he says the same question to each of you and I just
want a quick answer from each if you would . he says . hey
Chris . e:m . your campaign has been dreadful . I mean you've just
underestimated the intelligence of the electorate and particularly of
me . what would you why did you get it wrong
Conversationalization is much more marked in this case. The presenter
is constructed as an ordinary bloke talking to ordinary people, sharing
with them a common Tifeworld' (Habermas 1984), a commonsense
world of ordinary experience. One conversational feature is. the direct
representation of the talk of others, including an attempt to imitate the
voice of the (real or imaginary) original. Indeed, this whole item is built
around the presenter's simulation of the voice of the floating voter.
Conversationalization is also realized in a variety of linguistic features.
Most obvious are items of colloquial vocabulary (fancy, mate, bloke,
blooming) and the colloquial idiom fed up to the back teeth. Notice also that
altogether is used in a distinctively conversational way, in close associa­
tion with different, placed after a noun, meaning 'completely'. The
extract includes the colloquial use of the demonstrative pronoun this to
refer to someone previously m entioned (e.g. imagine this floating voter
actually is a mate of all three of you). There is also a feature of conversational
narrative in the use of narrative present tense (e.g., from earlier in this
interview, he comes back to you Chris Patten and he says).
These examples have identified two tensions affecting contem­
porary media language:
© the tension between information and entertainment
® the tension between public and private.
They are indicative of two tendencies:
• the tendency of public affairs media to become increasingly con-
versationalized
» its tendency to move increasingly in the direction of entertainment
- to become more 'marketized'.
Of course, a large part of media output is clearly designed as enter­
tainment (drama, soap operas, comedy shows, quiz shows, and so
forth), so what is involved here can be thought of as a shift in the
internal structure of the media, a relaxation of the boundary
between public affairs and entertainment within the media. This
shift can be seen in more general terms as part of an intensified
'marketization' of the media: because of increasing commercial
 Media and language: setting an agenda 11
pressures and competition, media are being more fully drawn into
operating on a market basis within the 'leisure' industry, and one part of
that is greater pressure to entertain even within public affairs output.
Marketization is a process affecting not only mass media. One fea­
ture of Thatcherism in Britain and parallel political regimes in other
countries is that more and more domains of social life have been
forced to operate on a more explicitly market basis - educa tional insti­
tutions including schools and universities, the health service, and
sections of the arts amongst them. Economic change has been
accompanied by cultural change, which has led some to refer to
contemporary societies as 'consumer' or 'promotional' cultures
(Featherstone 1991, Wernick 1991). Like many others, 1 regard these
developments as matters for concern. In the case of the media, for
instance, is the commercial imperative (especially in television) to
constantly entertain (Postman 1987), almost without regard to the
nature of the programme, compatible with the tradition of public
service broadcasting? If audiences are constructed, and competed
for, as consumers, even in news and current affairs programmes,
does this not negate the claims of broadcasting to constitute a public
sphere (Habermas 1989) in which people, as citizens, are drawn into
serious debate on the issues of the day? And if the media is not
sustaining a political public sphere, where else can it be constructed
in our mediatized society? (Perhaps the 'networking' associated
with, for instance, anti-road-building campaigns indicates that there
are other possibilities.) 1 return to these questions in Chapter 3.
Conversationalization, also, is affecting many other domains in
addition to the mass media - it is evident in interactions between pro­
fessionals (in a wide sense) and their publics or 'clients' in medicine,
education, politics and many other domains. A large-scale merging
of private and public practices is indeed a hallmark of contemporary
social life. I referred above to the particular structural properties of
mass-media communication which favour conversationalization -
the contradiction between the public nature of media production and
the private nature of media consumption (Cardiff 1980, Scanneli
1992). But there are also, as in the case of the shift towards enter­
tainment, more general social and cultural changes at issue. We
might see these in terms of tradition as an organizing principle within
societies becoming problematic (Giddens 1991), which entails prob­
lems with relationships based upon authority, an opening up and
democratization of social relations, a new public prestige for 'ordi­
nary' values and practices, popular culture, including 'ordinary' con­
versational practices.
12 Media and language: setting an agenda
We might also see a link between conventionalization and
marketization (the shift towards the consumer model and enter­
tainment). According to one view (Abercrombie 1991), the emphasis
has shifted in contemporary economies from production to con­
sumption, and this has entailed a change in authority relations which
favours consumers over producers, and a more general shift in social
relations in favour of ordinary people and their practices, culture and
values, including conversational language. While I certainly do see
a connection between the two tensions and between the two ten­
dencies, 1 shall treat them here as distinct, if overlapping.
1 have highlighted two aspects of the relationship between the
mass media and other parts of the network of social institutions they
operate within: their relationship to ordinary life (the 'lifeworld') and
the family on the one hand, their relationship to business and
commerce on the other. (The latter is a partly internal relationship, in
the sense that the mass media increasingly are seen as business. ) I see
the mass media as operating within a social system (Blackwell and
Seabrook 1993), which makes it important not to isolate particular
aspects such as these two tendencies from the way the media are
shaped by, and in turn contribute to shaping, the system overall. I
have already signalled a concern with the question of power: the
question of how the mass media affect and are affected by power rela­
tions within the social system, including relations of class, gender,
and ethnicity, and relations between particular groups like politicians
or scientists and the mass of the population, these issues have been
extensively discussed in media studies in terms of ideology (Hall 1977,
Hall ef al. 1978), and a major issue is how media language might work
ideologically (Fowler et al. 1979, Hodge and Kress 1979). Represen­
tations, identities and relations are of relevance to answering this
question: the ideological work of media language includes particular
ways of representing the world (e.g. particular representations of
Arabs, or of the economy), particular constructions of social identities
(e.g. the construction in particular ways of the scientific experts who
feature on radio or television programmes), and particular construc­
tions of social relations (e.g. the construction of relations between
politicians and public as simulated relations between people in a
shared lifeworld).
Two connected questions about the tensions and tendencies I have
highlighted are how they affect power relations within the social
system, and how they work ideologically. In respect of marketi­
zation, the increasing construction of audiences as consumers and
the increasing pressure on producers to entertain can be seen as part
 Media and language: setting an agenda 13
of a normalization and naturalization of consumer behaviour and
consumer culture which also involves advertising and the represen­
tation of people across the whole range of programmes (quiz shows,
soap operas, sport, drama, news, and so forth). There is considerable
diversity of voices,, but these diverse voices are so ordered that over­
whelmingly the system, with respect to consumption and con­
sumerism, is constantly endorsed and re-endorsed. Also, because
marketization undermines the media as a public sphere as 1 sug­
gested above, there is a diversion of attention and energy from pol­
itical and social issues which helps to insulate existing relations of
power and domination from serious challenge - people are con­
structed as spectators of events rather than participating citizens.
On the other hand, there is a major ambivalence in the case of
conversationalization. To put the issue rather baldly, do conversation-
alized discourse practices manifest a real shift in power relations in
favour of ordinary people, or are they to be seen as merely a strategy on
the part of those with power to more effectively recruit people as
audiences and manipulate them socially and politically? Fowler (1991:
57) takes the latter view: 'the ideological function of conversation is to
naturalize the terms in which reality is represented'. What he presum­
ably has in mind is the sort of example we have in the extract from the
Today programme: you haven' tgot the experience if you get in the City might
say do this lot know enough to run the country I'm nervous that a vote for you
would mean a vote for some kind of flight from the pound. This presupposes
that sudden international movements ot capital are judgements on
issues like government competence, rather than judgements on pros­
pects for profit. Notice the embedding of talk here: the whole of this
sequence is in the voice of the floating voter, which has embedded
within it the voice of 'the City', and these are in turn embedded in the
presenter's voice in the extract. Interestingly, all these voices are
conversationalized in similar ways. This not only helps naturalize the
ideological presupposition noted above, it also ideologically presup­
poses, in itself, that the presenter, the floating voter, and the City all
belong to the same lifeworld, the same world of ordinary experience,
along with the audience.
But conversationalization cannot, I think, be simply dismissed as
ideological: it might be ideologically invested or appropriated and
indeed often is, but it does nevertheless represent some degree of
cultural democratization. For example, in the extract from The Works
above, conversationalization helps to democratize technology,
making it more accessible to people, raising the status of the language
and experience of ordinary life by recasting science in their terms to a
14 Media and language: setting an agenda
degree, and rejecting the élitism and mystification which go along
with science as authorized specialists talking technical language.
Similar remarks might be made, for example, about conversational-
ization in politics. There is a real ambivalence about conversational-
ization, not simply a matter of its being sometimes ideological and
sometimes democratic. The fact that conversationalization is so
widely appropriated ideologically gives an aura of insincerity to even
the most innocent and exemplary instances of it. Conversely, even
where it is most clearly ideologically appropriated, the implicit claims
it makes about common experience and equality put these issues on
the public agenda - in certain circumstances, even hollow claims may
be challenged and redeemed in a way that would not happen if they
were not made at all. For instance, politicians can find themselves (in
Shakespearean terms) 'hoist with their own petard': if they claim to
be ordinary, they may find themselves evaluated as ordinary people
and found wanting, and unable to resort to traditional resources of
political mystique and charisma to protect themselves. In a very
limited sense, politicians are now more in the hands of ordinary
people, no matter how shallow their populist political rhetoric, even
if this 'people power' is systematically manipulated by the media.
1 understand ideology as 'meaning in the service of power'
(Thompson 1984, 1990) - ideologies are propositions that generally
figure as implicit assumptions in texts, which contribute to producing
or reproducing unequal relations of power, relations of domination.
I hey may be uvipiJcit, fur instance, m me presuppusiuon^ (umtrii-iui-
granted assumptions) of texts. Following work in French discourse
analysis (Pêcheux 1982, Williams forthcoming), I see presuppositions
as 'preconstructed' elements within a text, elements that are con­
structed beforehand and elsewhere. This links ideology to the pres­
ence of other, prior texts within a text (see Chapter 5). ideologies are
also implicit in the naturalized ways of organizing particular types of
interaction (e.g. the ways talking turns are organized in interviews).
To show that meanings are working ideologically it is necessary to
show that they do indeed serve relations of domination in particular
cases. A useful methodological principle is that the analyst should
always ask of any text whether and how it is working ideologically,
but expect answers to vary: ideology is more of an issue for some texts
than for others.
Exploring whether a particular implicit proposition or a set of proposi­
tions are working ideologically is one issue within a general set of ques­
tions that can be asked whenever one representation is selected over
other available ones, or whenever identities or relations are constructed
 Media and language: setting an agenda lb
in one way rather than another. The questions are (a) what are the
social origins of this option? where and who does it come from?
(whose representation is it, for instance?) (b) what motivations are
there for making this choice? (c) what is the effect of this choice,
including its effects (positive or negative) upon the various interests
of those involved?
It is possible to assess the importance of particular representations,
relations or identities for relations of domination without getting
involved in questions about truth. The question of whether a taken-
for-granted proposition helps produce or reproduce I'elations of
domination is independent of judgements about its truth or falsity.
Nevertheless, critical analysis cannot be indifferent to questions of
truth (Dews 1987, Norris 1992), whether it is a matter of how reports
falsify by omitting part of what vv'as done or said (Herman and
Chomsky 1988), or a matter of false ideological presuppositions. For
example, if a text presupposes that women are less intelligent than
men or black people than white people, it is an important part of the
analysis to point out that the ideological assumption is false.
Some readers may be persuaded of the case for investigating ques­
tions of power and ideology and the tensions between public and pri­
vate and information and entertainment in the mass media, but not
see the point of doing so with a focus upon language, and particularly
with a focus on what may seem irrelevant fine detail of the language
of a rather small number of texts. It is true that analysis of language
u iu i i u i p o m e s to in t
need to see language analysis as one of a range of types of analysis
which need to be applied together to the mass media, including com­
plementary forms of analysis which can generalize across large quan­
tities of media output (e.g. forms of content analysis as well as forms
of cultural and sociological analysis). But analysis of language has
certain advantages over other forms of analysis. The tensions and
contradictions I have referred to are manifest in the heterogeneity of
textual meanings and forms. Texts provide usually temporary and
short-lived ways of resolving the dilemmas into which people are put
by the tensions and contradictions which frame those texts. Textual
analysis can give access to the detailed mechanisms through which
social contradictions evolve and are lived out, and the sometimes
subtle shifts they undergo.
One objection that some media analysts may have to language
analysis is that it puts undue emphasis on the analysis of texts. The
trend in media studies has been away from analysis of texts and
towards analysis of reception of texts by audiences (Allen 1992,
16 Media and language: setting an agenda
Corner et al. 1990), though there are signs of a partial return towards
texts (Brunsdon 1990). This was a reaction against analyses of media
texts which postulated meanings and effects, including ideological
effects, without taking any account of how texts are actually
received by audiences. Media reception research has suggested that
texts do nor have unitary meanings, but are quite variously inter­
preted by different audiences and audience members, and may be
quite various in their effects. I fully accept the importance of recep­
tion studies for understanding meanings and effects. But reception
studies sometimes lead to a disregard for the text itself, which I do
not accept. It strikes me as self-evident that although readings may
vary, any reading is a product of an interface between the properties
of the text and the interpretative resources and practices which the
interpreter brings to bear upon the text. The range of potential inter­
pretations will be constrained and delimited according to the nature
of the text (Brunsdon 1990). If this is so, text analysis remains a
central element of media analysis, though it needs to be comple­
mented by analysis of text reception as well as by analysis of text
production.
Language analysis, then, can help anchor social and cultural
research and analysis in a detailed understanding of the nature of
media output. But only language analysis of a particular sort is cap­
able of making such a contribution. A rather arid, formalist analysis
of language, in abstraction from social context, still tends to dominate
' T H ' yl'xpk1rkmm-nkc of lm rr\\!? Th? t SOrf of d!?proneh CclP.HO^ be fho
basis for effective interdisciplinary work on the media. My view is
that we need to analyse media language as discourse, and the linguis­
tic analysis of media should be part of the discourse analysis of
media. Linguistic analysis focuses on texts, in a broad sense: a news­
paper article is a text, but so too is a transcription of a radio or tele­
vision programme. But discourse analysis is concerned with practices
as well as texts, and with both discourse practices and sociocultural prac­
tices. By discourse practices I mean, for instance, the ways in which
texts are produced by media workers in media institutions, and the
ways in which texts are received by audiences (readers, listeners,
viewers), as well as how media texts are socially distributed. There
are various levels of sociocultural practice that may constitute parts ot
the context of discourse practice. 1 find it helpful to distinguish the
'situational', 'institutional' and 'societal' levels - the specific social
goings-on that the discourse is part of, the institutional framework(s)
that the discourse occurs within, and the wider societal matrix of the
discourse. Discourse analysis can be understood as an attempt to
 Media and language: setting an agenda 17
show systematic links between texts, discourse practices, and
sociocultural practices. A detailed explanation of this view of
discourse analysis will be found in Chapter 3.
Let me say a little more about what is meant by text in this
framework. A first point is that I am using the word as it is often used
by linguists, for both spoken and written language-a transcription of
a broadcast is a text as well as a newspaper article. Second, in the case
of television it makes sense to include visual images and sound
effects as parts of texts, and to see linguistic analysis as part of what
has recently been called 'social semiotic' analysis (Hodge and Kress
1988, Kress and van Leeuwen 1990). Also, written texts in contem­
porary society are increasingly becoming visual as well as linguistic
texts, not only in the sense that newspapers, for instance, combine
words with photographs and with maps and diagrams, but also
because considerations of layout and visual impact are increasingly
salient in the design of a written page.
Third, the framework takes a 'multifunctional' view of texts, drawn
from the 'systemic' theory of language. Halliday (1978) argues that
what he calls the 'ideational', 'interpersonal' and 'textual' functions
of language are always simultaneously at work in any text, and even
in any particular sentence or clause. This ties in with my suggestion
earlier that representations, relations and identities are always simul­
taneously at issue in a text: the ideational function of language is its
function in generating representations of the world; the interper­
sonal function includes the functioning of language in the conrii-
tution of relations, and of identities. (The textual function relates to
the constitution of texts out of individual sentences - this will be
discussed later in Chapter 6.) The value of such a view of texts is that
it makes it easier to connect the analysis of language with funda­
mental concerns of social analysis: questions of knowledge, belief
and ideology (representations - the ideational function), questions of
social relationships and power, and questions of identity (relations
and identities - the interpersonal function). Representations are a
long-standing concern in debates about bias, manipulation, and
ideology in the media, but identities and relations have received less
attention. The wider social impact of media is not just to do with how
they selectively represent the world, though that is a vitally impor­
tant issue; it is also to do with what sorts of social identities, what
versions of 'self, they project and what cultural values (be it con­
sumerism, individualism or a cult of personality) these entail. And it
is to do with how social relationships are defined, especially social
relationships between the mass of the population who constitute
18 Media and language: setting an agenda
audiences for the most popular media output and people like politi­
cians, scientists, church leaders, and broadcasters themselves.
Another and related strength of a systemic view of text is that it
sees texts as sets of options. A text selects particular options from the
systems of options - the potential - available. On one level, these are
selections amongst available language forms, from the lexical and
grammatical potential: one word rather than another, or one gram­
matical construction rather than another (e.g. a passive rather than
an active sentence, or a declarative rather than an interrogative or an
imperative sentence - see Quirk et at. 1972). But these formal choices
constitute choices of meaning, the selection of options from within
the meaning potential - how to represent a particular event or state of
affairs, how to relate to whoever the text is directed at, what identities
to project. And these choices are in turn linked to choices at a
different level: what genres to draw upon in producing (or inter­
preting) a text, what discourses to use (see below). Such a view of text
encourages analysts to be sensitive to absences from the text - the
choices that were not made but might have been - as well as pres­
ences in it, as well as to weigh presences against possible alternatives
(e.g. how else might this have been put?). One should not, however,
be misled by the language of 'choices' and 'options'; this is a
framework for analysing the variability of language and its social
determinants and effects, and self-conscious linguistic choice is a
relatively marginal aspect of the social processes of text production
and interpretation.
1 should also mention here an important aspect or the analysis of
'discourse practice' in the framework for discourse analysis sketched
out above: intertextual analysis. This will be explained in Chapter 4.
The term discourse is widely and sometimes confusingly used in
various disciplines (Fairclough 1992a, Foucault 1978, van Dijk 1985).
It is helpful to distinguish two main senses. One is predominant in
language studies: discourse as social action and interaction, people
interacting together in real social situations. The other is pre­
dominant in post-structuralist social theory (e.g. in the work of
Foucault): a discourse as a social construction of reality, a form of
knowledge. My use of the term 'discourse' subsumes both of these,
and indeed sets out to bring them together. The first sense is most
closely associated with the interpersonal function of language, and
with the concept of genre (see Chapter 5, pages 85 ff.). The second
sense is most closely associated with the ideational function of lan­
guage, and with discourses - notice that in addition to being used as
an abstract noun for this general view of language in social use,
 Media and language: setting an agenda 19
discourse is used as a count noun (a discourse, several discourses) as a
category (alongside 'genre') within the intertextual analysis of texts
(see Chapter 5).
In the discourse perspective on media language which 1 have
sketched out above, the analysis of texts is not treated in isolation
from the analysis of discourse practices and sociocultural practices.
However, since this book is about media language, the focus will be
on texts rather than practices. Also, the focus will be on linguistic
aspects of texts, rather than other semiotic aspects such as visual
images in television. I shall, however, be alluding throughout,
though selectively, to the interconnection between the texts that are
in focus and other dimensions of the framework.
Chapters 2 to 7 will present a view of media discourse and a
framework for analysing it, and Chapters 8 and 9 will deal with case-
studies of particular types of media discourse. Chapter 2 will review
some of the most important previous work on media discourse, and
provide a set of desiderata for a satisfactory critical analysis of the
subject. Chapters 3 and 4 sketch out a social theory of media
discourse, with an account of communication in the mass media in
Chapter 3, and a description of the critical discourse analysis
framework which I use in the book in Chapter 4. Chapter 5 is con­
cerned with intertextual analysis of media texts, understood both in
terms of how media texts transform and embed within themselves
other texts, and in terms of how they draw upon and combine
together available discourses and genres. Chanters Aand 7 Heal with
the linguistic analysis of media texts, with Chapter 6 focusing upon
representational aspects of texts, and Chapter 7 focusing upon
aspects of texts that have to do with relations and identities. Chapter
8 is a case study of one television programme, Cnmewatch UK, and
Chapter 9 is a study of political discourse in the media.
 2

APPROACHES TO MEDIA DISCOURSE

My main objective in this chapter is to give a selective account of pre­

vious wuik uit nteuid discourse. Let me emphasize that this will cover
only one part of the literature 1 shall be drawing upon in this book: I
shall also be using material from media studies, social theory, and
elsewhere. Nor will the chapter attempt an exhaustive account of the
media discourse literature; 1 shall focus upon work which I have
found particularly fruitful in developing my own analytical
framework. I discuss the following approaches in turn: linguistic and
sociolinguistic analysis, conversation analysis, semiotic analysis,
critical linguistics and social semiotics, social-cognitive analysis, and
cultural-generic analysis. Separating the approaches in this way is
helpful for presentational purposes, but in fact there is a great deal of
cross-fertilization between them, and many analysts combine them.
(In describing these approaches, I shall sometimes use terms which
are not fully explained until Chapter 4. If these are unfamiliar,
readers may find it helpful to return to parts of this chapter after
reading Chapter 4.) The chapter concludes with a set of desiderata for
an adequate critical discourse analysis of media discourse, compiled
 Conversation analysis 21

on the basis of the review of approaches, which sets the scene for the
presentation in Chapter 4 of the framework I shall use in the rest of
the book.

Linguistic and sociolinguistic analysis

Ways in which language is used in the media may be or interest to

linguists for their own sake, as evidence, for instance, of particular
types of grammatical structure or particular intonation patterns. For
example, newspaper headlines have distinctive syntactic properties
which make them a grammatical oddity, and have long attracted the
attention of linguists (Mardh 1980, Straumann 1935). Media language
has also been analysed sociolinguistically, notably by Bell (1991).
Bell's work is unusual in that he is a practising journalist as well as a
sociolinguist. A number of the studies he has carried out use linguis­
tic and sociolinguistic analysis in ways which illuminate the sociocul­
tural analysis of news media, and I shall refer to them again in
Chapter 3. But much of his work is typical of 'variationist'
sociolinguistics in focusing upon correlations between variable
linguistic features and variable aspects of social context. In Bell
(1984), for example, he shows how the degree to which word-final-
consonant dusters are simplified in the language of radio reporters
(giving, for instance, Wes' Coas' coal as a pronunciation of West Coast
coal) varies between different New Zealand radio stations according
to the main occupational profiles of their audiences (whether they are
mainly manual, skilled, office or professional workers). The strength
of this work is its attention to linguistic detail, to the form and
'texture' of texts. (1 use the term 'texture' to refer broadly to the 'form'
as opposed to the 'content' of texts - see Fairclough 1992b.) But this
element in Bell's work operates with a rather narrow conception of
social aspects of media, and does not attempt to show systematic
linkages between language and sociocultural context.

Conversation analysis

Conversation analysis has been developed by a group of sociologists

-known as 'ethnomethodologists'. Ethnomethodology is an inter­
pretative approach to sociology which focuses upon everyday life as
a skilled accomplishment, and upon 'methods' which people use for
producing it (Garfinkel 1967). Some ethnomethodologists take a
22 Approaches to media discourse
particular interest in conversation and methods that people use for
producing and interpreting it (Atkinson and Heritage 1984). Conver­
sation analysts have concentrated mainly upon informal conversa­
tion between equals (e.g. telephone conversation), though more
recent work has given attention to institutional types of discourse
(Button and Lee 1987), including media discourse.
To illustrate the approach, I refer to studies of media interview
carried out by Heritage (1985), Greatbatch (1986) and Hutchby (1991).
Heritage focuses upon the 'formulations' used by interviewers in the
course of news interviews. This is one of his examples:
int: Would you be happy to see Prince Charles become King of Wales?
man : Well 1 couldn' I - you know I just couldn't care tuppence who
comes King and who don't like (0.5)
int: You don't think it makes any difference to you.

A formulation (such as the interviewer's second contribution here) is

a widely used device interviewers use to summarize what inter­
viewees have said. Formulations typically stress certain aspects of
what has been said rather than others, and often elaborate what has
been said by drawing out its implications. That happens here: the
interviewee didn't sax/ that it made no difference to him, but he did
arguably imply it. Heritage sees formulation as a technical device
which interviewers use to manage interviews within the constraints
under which they are forced to operate. One constraint is the pres­
ence of a lishTung audience: formulations are a way ui ensuring the
audience is constantly kept in the picture by clarifying what inter­
viewees say, drawing out implications, etc. Another is the require­
ment on interviewers to maintain a 'stance of formal neutrality':
alternative formulations provide a covert means of evaluating whatis
said, making things easier or more difficult for interviewees, pushing
the direction of the interview one way rather than another.
Heritage's view of properties of news interviews emphasizes
technical solutions to institutional problems. This is a valuable per­
spective because it shows how discursive practices are rooted in insti­
tutional structures and practices (one could fruitfully extend this
'back' to the political economies of institutions-see Chapter 3). But it
is not adequate on its own: to make sense of contemporary interview­
ing practices, one needs to recognize how they are shaped by, and
help shape, wider social and cultural shifts. Heritage emphasizes the
normative side of news interviews, what news interviews have in
common - their 'tacit ground rules'. But news interview is not a uni­
tary genre: there is considerable, culturally patterned, variation not
 Conversation analysis 23

only historically (1953 interviews were generally very different from

1993 interviews) but also in contemporary broadcasting, depending
upon the medium, type of programme, and particular style of the
interviewer. Greatbatch (1986) gives limited recognition to this vari­
ability. One of the ground rules of interviewing, normatively, is that
interviewees should confine themselves to answering questions, but
as he points out they don't always do so: sometimes they answer the
question and then introduce topics of their own, sometimes they
introduce topics of their own first and then answer the question,
sometimes they don't answer the question at all. One of Greatbatch's
examples is:

int: D'you quite like him?

eh: Well er I - think in politics you see: i - it's not a question of going
about liking people or no:t, it's a question of dealing with people.
And e:r I've always been able to deal perfectly well with Mister Wilson
and er - indeed he has with me.

The interviewee (eh ) does not answer the interviewer's question, but
begins by denying its relevance, says what he thinks the relevant ques­
tion is, and talks about that. However, Greatbatch considers such
examples as violations of the rule that interviewees should confine
themselves to answering questions: what he is interested in is when
interviewers sanction such violations, and when they tolerate them.
What this violational view misses is that some types of news interview
are now routinely seen as occasions where interviewees talk about their
own topics, subject to the weaker requirement that they at least coher­
ently link them to the interviewer’s questions. This is not violative
behaviour but part of a culturally significant shift in genre whose further
ramifications have been analysed by Tolson (1991 - see the discussion of
his work in the section on cultural-generic analysis below).
Conversation analysis actually shares with linguistic and socio-
linguistic analysis strengths in the detailed description of organ­
izational properties of media language. It has extended the resources
■■of descriptive linguistics through its analysis of the organization of
interaction (turn-taking, topic-control, formulation, etc.), though at
the same time it ignores many of the features which a linguistic
description would attend to. The focus is very much upon relational
aspects of conversation - the achievement of interaction - and ques­
tions of representation and associated linguistic features are given
relatively little attention. It is also resistant to linking properties of
talk with higher-level features of society and culture - relations of
power, ideologies, cultural values.
24 Approaches to media discourse
Semiotic analysis

By contrast, semiotic analysis does treat analysis of texts as a key

component of cultural analysis of media. I refer specifically here to
Hartley's important study of news discourse (Hartley 1982, see also
Fiske and Hartley 1978). Hartley's focus is upon the semiotic codes
and conventions which underlie both linguistic and visual aspects of
news stories. Analysis of visual codes attends to different modes of
presentation on television - the 'talking head' (newsreader or cor­
respondent looking directly at the camera), use of graphics and still
photographs, various types of 'actuality' or film report (film with
voice-over, the 'stake-out' with the reporter talking directly to
camera, the 'vox pop' with a member of the public talking to an
unseen reporter) - as well as the framing of pictures, camera
movements (pans and zooms) and the sequencing of shots. The
assumption is that choices among options available within visual
codes - including technical options relating to the camera-work -
carry social meanings (see also Tuchman 1978).
Hartley analyses a range of language-related codes and conven­
tions, including categorization of stories into a small number of major
topics, the effect of news values (as an 'ideological code') on the
treatment of topics, the assumption of consensus and the handling of
dissent, audience address - the operation of broadcasters as 'media­
tors' who translate news into the common-sense terms of audiences,
use oi a conversational communicative styie, the structuring of news
stories. A focus typical of semiotic analysis is upon ideologically
potent categories and classifications which are implicit in news texts,
and upon alternative or competing categories which are absent,
'suppressed'. For instance, it is a common observation that news
stories are personalized: the category' of individual personality is
widely evoked in news stories, whereas the category of social (and ,
especially class) subject is correspondingly suppressed. Or again,
many oppositions which appear on the surface of a text - for instance
between government and unions, management and strikers,
western allies and foreign dictators - can be assimilated to an under- *
lying opposition between 'us' and 'them'.
One very important achievement of this work is establishing that
analysis of texts is a significant part of sociocultural analysis of media, ;
by linking properties of texts to ideologies, power relations and cul- :
tural values. This general objective is taken up in critical approaches :
to linguistics and discourse analysis, which operate however with a
linguistically grounded conception of text. An obvious limitation of ;
 Critical linguistics and social semiotics 25
semiotic analysis in comparison with the linguistically oriented
approaches (linguistic, sociolinguistic, critical linguistic, social-
cognitive and cultural-generic analysis) is that it does not systemati­
cally attend to detailed properties of the texture of texts.

Critical linguistics and social semiotics

'Critical linguistics' is a type of discourse analysis which was

developed by a group based at the University of East Anglia in the
1970s (Fowler et al. 1979, Hodge and Kress 1979). Media discourse
is one of its main concerns (Fowler 1991, Trew 1979a, 1979b).
Critical linguistics is based upon 'systemic' linguistic theory (Halli-
day 1978, 1985). It brings to analysis of media discourse systemicist
views of the text already introduced in Chapter 1: the view of the
text as multifunctional, always simultaneously representing the
world (ideational function) and enacting social relations and identi­
ties (interpersonal function); seeing texts as built out of choices
from within available systems of options in vocabulary, grammar,
and so forth. Discourse is seen as 'a field of both ideological pro­
cesses and linguistic processes, and . . . there is a determinate rela­
tion between these two kinds of process' (Trew 1979b); specifically,
the linguistic choices that are made in texts can carry ideological
meaning.
Some of the most revealing analyses concern representation and
the ideational function, how events and the people and objects
involved in them are represented in the grammar of clauses (simple
sentences). The basic premiss is that coding events in language
entails choices among the models - the distinct process and parti­
cipant types - which the grammar makes available, and that such
choices are potentially ideologically significant. For example, on a
BBC Radio 4 Today programme (11 March 1993) the following
comment was made about 'cheap' Russian fish being 'dumped' on
the British market: 'the funny thing is it's not transferring itself to
the consumer at terribly low prices at all'. This might have been
worded as, for instance, 'the dealers involved in the distribution of
the fish are overcharging the consumer', coding the pricing of the
fish as an action process with a responsible agent (the dealers).
Instead, we have the distribution of the fish coded with an action
process verb (transfer) used reflexively, and the process of pricing is
transformed into a state (at terribly loiv prices). Responsibility and
26 Approaches to media discourse
agency are elided. If there were a systematic tendency in news
reports for such choices of process and participant types to leave
agency and responsibility unspecified in this way, one might
(depending upon the wider sociocultural context) see those choices
as having ideological meaning. See Chapter 6, pages 109-16 for more
detail.
1 suggested that in the above example pricing was transformed
from a process into a state. This sort of transformation is a 'nominal-
ization', changing a process into a nominal (i.e. noun-like) entity.
Another type of transformation is the shift of an active sentence into a
passive (e.g. from they are dumping fish on the market to fish is being
dumped on the market). The argument is that transformations such as
nominalization and shifting into the passive may be ideologically
motivated. For example, both allow the actor, the responsible agent,
to be omitted and, as I have just suggested, systematic elision or back­
grounding of agency may be an ideologically significant feature of
texts.
Trew (1979a, 1979b) has done some particularly fruitful work on
'discourse in progress' in newspapers - the transformation of mater­
ial from news agencies and other sources into news reports, and the
transformations a story undergoes from one report to another, or
from reports to in-depth analyses to editorials, over a period of time.
He refers to the coverage of police shootings in Zimbabwe in 1975 in
The Times. The headline of the first report (RIOTING BLACKS SHOT
U J D l _ T T O J~1 1 •! \_. .

agent but in an informationally de-emphasized position in the middle

of the headline, whereas the 'rioting' of those shot is foregrounded
(being placed at the beginning).

RIOTING BLACKS SHOT DEAD BY POLICE AS

ANC LEADERS MEET
Eleven Africans were shot dead and 15 wounded when Rhodesian
police opened fire on a crowd of about 2,000 in the African Highfield
township of Salisbury this afternoon.
(Trew 1979a: 94)

In the lead (first) paragraph, an agentless passive is used (were shot d

dead and . . . wounded), and the police are explicitly present only as
agents of opened fire on a rioting crowd, rather than as the ones who 1
shot dead the people. In an editorial, the event is transformed into .
The rioting and sad loss of life in Salisbury for which 'factionalism' is-..a
said to be responsible - the police as responsible agent is elided.
These are part of a more complex series of transformations over time
 Critical linguistics and social semiotics 27
which background police responsibility, and which are ideological
as well as linguistic processes: they assimilate problematic events to
preconstructed ideological frames for representing political relations
in southern Africa. The linguistic processes involve rewordings as
well as grammatical changes - notice loss of life replacing shot dead.
Such ideological-linguistic processes are also processes of struggle, in
which choosing to represent an event in one way may also be refus­
ing to represent it in other currently available ways. For further devel­
opment of this concept of transformation, see Hodge and Kress
(1979,1988, 1992). See also the discussion of transformations of texts
across 'intertextual chains' of discursive practices in Fairclough
(1992a).
Critical linguistics emphasizes the role of vocabulary choices in
processes of categorization. For example, a study of gender discrimi­
nation in media reporting might consider how differences in the
vocabulary used to refer to women and men assimilates people to
pre-existing categorization systems of an ideologically powerful sort.
Are women, for instance, systematically represented in terms of their
family roles (as 'wives' or 'mothers') or in terms of their sexual inter­
est to men? It is fruitful to combine such questions with analysis of
process and participant types: what sorts of participants in what sorts
of processes do women/men predominantly function as - for
instance, are both equally likely to function as actors in action pro­
cesses? and where they do function as actors, what particular cat­
egories ci process : i m v o i v e d it, for instance, smiling and
screaming, or debating and voting?). See Fowler 1991 chapter 6 for an
analysis along these lines.
A clause which codes an event (ideationally) in terms of a particular
type of process will also assess (interpersonally) the truth or probabil­
ity of the proposition so encoded, and the relationship between pro­
ducer and addressee(s). The concept of 'modality' is used in a very
general way to cover features of texts which 'express speakers' and
writers' attitudes towards themselves, towards their interlocutors,
and towards their subject-matter' (Fowler et al. 1979: 200). Choices of
pronouns, modal auxiliaries, speech acts, and many others, are
included within modality.
The limitations of critical linguistics have been quite widely
discussed, even by those involved and their sympathizers (Fowler
.1987, Kress 1989, Richardson 1987). In terms of the text-practice
distinction I introduced in Chapter 1, the focus is upon text and
(especially in the case of Trew) productive practices, but texts tend to
be interpreted by the analyst without reference to the interpretative
28 Approaches to media discourse
practices of audiences. Media studies has shifted its emphasis
away from text analysis to audience reception (recall the discussion
of this issue in Chapter 1), and this has not surprisingly led to
criticism. In terms of sociocultural practice, there tends to be a
rather monolithic view of the role of media in ideological reproduc­
tion which understates the extent of diversity and change in media
practices and media discourse. Although there is attention to inter­
personal (especially relational) aspects of texts, the emphasis is
perhaps rather too one-sidedly on representations, and I would
argue that issues of social identity ought to be foregrounded more
than they are. Although there are elements of intertextual analysis
of the constitution of texts in terms of discourses and genres, this
is underdeveloped compared with linguistic and above all gram­
matical analysis. And the linguistic analysis is very much focused
upon clauses, with little attention to higher-level organization
properties of whole texts. Mention of these limitations is not meant
to minimize the achievement of critical linguistics - they largely
reflect shifts of focus and developments of theory in the past
twenty years or so. See Hodge and Kress (1992) for a recent
attempt to 'update' the critical linguistics work of the 1970s.
A number of critical linguists have been involved in developing
the somewhat different approach of 'social semiotics' (Hodge and
Kress 1988, Kress and van Leeuwen 1990). in contrast with critical
linguistics, there is an interest in visual semiosis as well as lan-
fwhirh 11-1 r p or il ir n£ f-.h-
Also, productive and interpretative practices have become a major
concern, there is an orientation towards struggle and historical
change in discourse, and towards the development of a theory of
genre (van Leeuwen 1987, 1993) and the intertextual analysis of
texts. I shall draw upon some of this work and describe it more
fully in later chapters.

Van Dijk: the 'social-cognitive' model

In a series of studies, van Dijk (1988a, 1988b, 1991) has developed a

framework for analysing news (especially in newspapers) as
discourse which is similar in some ways to the view of discourse
taken in this book (and sketched out in Chapter 1). Discourse is con­
ceptualized in terms of three dimensions or perspectives (which I
have called text, discourse practice and sociocultural practice), and a
focus on discourse practice is seen as providing a way of linking
 Van Dijk: the 'social-cognitive' model 29
textual analysis to sociocultural analysis. Van Dijk's work, like
social semiotics, has made the important transition from text
analysis (which critical linguistics really still is) to discourse
analysis. Beyond that common ground, there are significant
differences, however. Van Dijk's analysis of practices of news pro­
duction and news comprehension has a social-psychological
emphasis on processes of social cognition - on how cognitive
'models' and 'schemata' shape production and comprehension -
whereas I focus (here and in other publications) upon how socially
available genres and discourses are drawn upon. Van Dijk's main
motivation for linking media texts to context is to show in detail
how social relationships and processes (e.g. the reproduction of
racism) are accomplished at a micro-level through routine prac­
tices, whereas my major concern is to show how shifting language
and discursive practices in the media constitute social and cultural
change. See Chapter 4 for a detailed account of my approach.
Van Dijk's framework analyses news texts in terms of what he calls
the 'structures of news', processes of news production, processes of
news comprehension. The analysis aims to show relationships
between texts, production processes and comprehension processes,
and between these and the wider social practices they are embedded
within. In analysing structures of news a distinction is made between
the 'macro' and 'micro' structures of news discourse. The former
relate to the overall content of a text - its 'thematic' structure - and the
overfill form of —its /5olifirnnfir/ strurtirre.
The concept of 'macrostructure' is central to the analysis of thematic
structure: the macrostructure of a text is its overall organization in
terms of themes or topics. It is a hierarchical organization, in the
sense that we can identify the theme of a whole text (and sum it up as
a single proposition), which can typically be spelt out in terms of a
few rather less general themes, which can each in turn be spelt out in
terms of even more specific themes, and so on. The schematic struc­
ture of a particular type of text is specified in terms of the ordered
parts it is built out of. Thus van Dijk suggests that a news report
typically has a headline, a lead, an 'events' element which covers the
main events of the story, and perhaps an element which gives verbal
reactions to the story, and a comment element (these last two
elements do not always occur as distinct sections). Each element of
schematic structure corresponds to a more general theme in the
. thematic structure. The headline of a news report formulates the
overall theme of a text. An important feature of the schematic struc­
ture of a text type are principles governing the way it orders thematic
30 Approaches to media discourse
content. In the case of news reports, there is a powerful 'relevance
principle' which requires more general information to come first, to
be followed by more detailed information. Thus the initial headline
and lead elements of news reports typically contain more general
information.
The 'microstructures' of news discourse are analysed in terms of
semantic relations between propositions - coherence relations of
causality, consequence and so forth. Microanalysis also identifies
syntactic and lexical characteristics of newspaper style, and rhetorical
features of news report, such as features which give reports an aura
of factuality.
The concepts of 'macrostructure' and 'schematic structure' are at
the centre of analysis of news production and comprehension, as
well as analysis of news structures. These wholistic structures are
seen to generate texts, and the interpretation of texts involves identi­
fying the wholistic structures which underlie them. Such structures
are intrinsic to the mental models of events and situations which
reporters bring to bear in interpreting events and source texts,
models which reporters try to convey to audiences in the way they
write reports, and models which audiences (readers etc.) draw' upon
in interpreting reports. This cognitive perspective helps to specify
how exactly the 'news values' that have been identified as shaping
news coverage influence the way particular reports are produced, it
also sheds light on how the texts which journalists get from news
agencies and other sources are transformed in producing a report, on
the forms in which news reports are memorized, and on the longer-
term effects they are likely to have on perception, cognition and
action.
This is a powerful integrated framework for nev/s discourse
analysis. Nevertheless, for my purposes it has a number of limita­
tions. First, the focus is on representations; social relations and
identities in news discourse - and the interpersonal function of lan­
guage - receive little attention. Second, texts are analysed linguistic­
ally but not intertextually, in terms of their constitution through
configurations of discourses and genres. A central feature of my
approach is the claim that linguistic analysis needs to be comple­
mented by intertextual analysis (see Fairclough 1992b, and Chapters
4 and 5). A third and related point is that van Dijk's work gives a one­
sided emphasis to news-making practices as stable structures which
contribute to the reproduction of relations of domination and racist
ideologies, which backgrounds the diversity and heterogeneity of
practices.
 Cultural-generic analysis 31

Cultural-generic analysis

Some British studies of media discourse have drawn upon work in

cultural studies associated with the Centre for Contemporary
Cultural Studies (at the University of Birmingham) to explore the
cultural and social import of ways in which media genres such as
interview or 'chat' are currently evolving (Montgomery 1990,
Tolson 1990). They take what Raymond Williams called a 'cultural
materialist' view of genre, seeing an innovation in genre as 'an
articulation, by technical discovery, of changes in consciousness
which are themselves forms of consciousness of change' (Williams
1981: 142), and regarding analysis of generic form as itself a mode
of cultural analysis. An important feature of this approach is that it
simultaneously attends to interaction (and relational features of
texts) and representation (see also van Leeuwen 1993). It draws
upon work by Coffman (1981) on how radio announcers address
audiences and the orientation in conversation analysis towards the
ongoing accomplishment of social relationships in talk, as well as a
Hallidayan multifunctional view of text (see Mancini 1988).
Montgomery's study of 'Our Tune', which 1 discuss in more detail
in Chapter 5, will serve as an example (Montgomery 1991). 'Our
Tune' was a very popular slot in a BBC Radio 1 show, in which the
D] (Simon Bates) summarized readers' ’letters in narrative form. Fol­
lowing the method of Labov (1972), Montgomery gives an account
of the generic structure of 'Our Tune' narratives in terms of
components, some obligatory and some optional, which occur in a
particular (though not totally rigid) order. In accordance with
common practice in narrative analysis, Montgomery distinguishes
the analysis of the story material from the analysis of its discursive
presentation. The latter involves those aspects of the narrative
which bring about the transformation of a private letter info a public
narrative, and those aspects which are oriented towards audience
reception of the story. Tensions which characterize media culture
are negotiated in the discursive presentation of this genre. For
instance, the tension between the public nature of media output and
the private circumstances of media reception (Scannell 1992 - see
also the discussion on this issue in Chapter 1) is concretely mani­
fested 'in a subtle blend of institutional and audience voices - pri­
vate discourses in a public space, public therapy on personal experi­
ence'. 'Our Tune' also tries to concretely negotiate the constant
tension in broadcasting between pressures to inform and pressures
to entertain by achieving balance between an entertaining narrative
32 Approaches to media discourse
style which draws upon fictional models (such as magazine stories),
and a commitment to truthfully recounting listeners' stories. These
examples point to the intertextual analysis of texts as often hybrid
configurations of genres and discourses which are realized in hetero­
geneous linguistic features.
The cultural-generic approach tried to relate changes in broadcast
genres to the evolution of the 'public sphere' of broadcasting (Haber­
mas 1989). Scannell (1992) has characterized the 'communicative
ethos' of broadcasting, emergent since the early days of radio, in
terms of the emergence of patterns of programming and a commu­
nicative style which accommodates to the private, domestic condi­
tions of media reception. Broadcasting genres have developed
simulated versions of informal conversational language (recall the
discussion in Chapter 1). Montgomery (1988) has investigated one
conventional feature, direct address of audiences, which is realized
textually in second-person pronouns, interrogative clauses,
imperative clauses, and so forth. He shows how audiences are con­
structed as complex and differentiated through the shifting direct
address of different sections of them. Such examples suggest that
features of genre are relevant to the construction of publics and of the
public sphere itself. Tolson, in a study of the evolution of interview
genre in documentary and talk shows, has argued that the generic
evolution of interview talk indicates a fragmentation of audiences,
and marks the demise of the 'general public' within the public sphere
of bro ad castin g Tolson^. wnrV show s hiov.r coporiiTicHtoiion in

the mixing of broadcasting formats (talk, variety, comedy) and

associated genres in intertextually complex and hybrid texts links to
wider tendencies in cultural change. These include the general cul­
tural validation of individualism (manifested in the 'personality
system' of the media), and the 'reflexivity' which has been taken as a
general feature of contemporary culture (Giddens 1991). Reflexivity
shows up as, for instance, 'self-reflexive metadiscourse' on the part of
talk show participants about television, about their own personalities
as constructs, about the apparent revelation of one's 'real self' in talk
shows as just a game.

Desiderata for a critical analysis of media discourse

In the final section of this chapter, I want to pull together from the
review of the literature a list of desiderata for an adequate critical
 Desiderata for a critical analysis of rneclia discourse 33
analysis of media discourse. This list will then provide a basis for the
elaboration of my own analytical framework in Chapter 4, and help
clarify how that framework relates to the literature. I should add that
no single book could reasonably hope to meet all these desiderata,
the list should rather be interpreted as pointing towards a pro­
gramme of research.

1. One focus of analysis should be on how wider changes in society

and culture are manifest in changing media discourse practices.
The selection of data should correspondingly reflect areas of vari­
ability and instability as well as areas of stability. (Cultural-generic
analysis. Compare conversation analysis, critical linguistics,
social-cognitive analysis.)
2. The analysis of media texts should include detailed attention to
their language and 'texture' (compare linguistically oriented
approaches with semiotics). It should also include detailed
analysis of visual images and sound effects (compare semiotics
and social semiotics with the other approaches.)
3. Text analysis should be complemented by analysis of practices of
text production and text consumption (compare social-cognitive
analysis with the other approaches), including attention to trans­
formations which texts regularly undergo across networks of
discourse practices (compare critical linguistics and social-
cognitive analysis with other approaches).
4. Analysis of texts and practices should be mapped or. to analysis of
the institutional and wider social and cultural context of media
practices, including relations of power and ideologies (compare
semiotic analysis, critical linguistics and cultural-generic analysis
with linguistic and sociolinguistic analysis and with conversation
analysis).
5. Text analysis should include both linguistic analysis and intertex-
tual analysis in terms of genres and discourses. it should be recog­
nized that texts are commonly hybrid intertextually with mixtures
of genres and discourses, and that such hybridity is manifest in
heterogeneous linguistic features. (Compare cultural-generic
analysis and social semiotics with other approaches.)
6. Linguistic analysis of texts should be conceived multifunctionally,
and be oriented towards representation and the constitution of
relations and identities as simultaneous processes in texts, and the
important relationships between them. (Compare cultural-
generic analysis and to a degree critical linguistics with other
approaches.)
34 Approaches to media discourse
7. Linguistic analysis of texts involves analysis at a number of levels,
including phonic, lexical, grammatical, and macrostructural/
schematic. (Compare social-cognitive analysis with conversation
analysis or critical linguistics.)
8. The relationship between texts and society/culture is to be seen
dialectically. Texts are socioculturally shaped but they also consti­
tute society and culture, in ways which may be transformative as
well as reproductive. (Compare more recent with earlier critical
approaches.)

\
\
 3

COMMUNICATION IN THE MASS MEDIA

Th 'iT'VTÏn Wit Q~
•°p
i\J U t V C l ’

theory of media discourse. Chapter 3 gives a general account of

communication in the mass media, while Chapter 4 gives a more
focused theoretical account of media discourse and a framework
for critically analysing it. The analytical framework, briefly alluded
to already in Chapter 1, is a version of 'critical discourse analysis'
(Fairclough 1989, 1992a, 1993). The theory set out in these two
chapters will be elaborated in greater detail, with examples, in
Chapters 5-7, and will form the basis for later chapters. Chapters 3
and 4 therefore have a key theoretical and methodological role in
the book as a whole.
It will perhaps be helpful to approach the question of what is
distinctive about mass communication in the first part of the chapter
through a comparison between communicative events in the media
and another type of communicative event. I shall refer for contrast to
medical consultation between doctors and patients (Fairclough
1992a, Mishler 1984). And given the diversity of media output, it will
help to have in mind one particular type of output. I have already
36 Communication in the macs media
indicated that the book is centred upon news, current affairs and
documentary, and in this chapter I shall be alluding mainly - though
not exclusively - to television documentary (see the excellent study in
Silverstone 1985).
Mass communication has certain special properties which distin­
guish it from other forms of communication, and which are partly
attributable to the nature of the technologies which it deploys
(Thompson 1990). These properties will be my first concern. But in
addition to such questions of medium and technology, an account of
communication in the mass media must consider the economics and
politics of the mass media: the nature of the market which the mass
media are operating within, and their relationship to the state, and so
forth. It is also important to attend to institutional aspects of media,
including practices of media text production within the institutions of
the press, radio and television, but also practices of media text con­
sumption and reception within the family and the home. A further
consideration is the wider sociocultural context of mass media
communication, the social and cultural structures, relations, prac­
tices and values which frame the mass media, shape mass media
communication, and are shaped by it.

The properties of mass communication

Communicative events airier m their Uiue-space parameters.

Whereas, for example, a medical consultation takes place with all par­
ticipants (centrally, doctor and patient) present together at a parti­
cular time and place, a communicative event in the media, such as a
television documentary, involves major temporal and spatial
disjunctions. The fundamental point is that the time and place of pro­
duction of a mass communication text is different from the time and
place of consumption, when an audience views or hears or reads it.
Indeed, a mass communication text is likely to be consumed in
various sorts of place and at various times, especially now with the
widespread use of video machines. And even the production of such
texts is often spatially and temporally disjoined - for instance, a docu­
mentary may take eighteen months to make and involve filming in
several countries. Satellite technology and the associated global­
ization of mass media, and the global domination of North American
and European media conglomerates, give a further twist to the
temporal and spatiai disjunctions of the media, in the sense that spa­
tial and temporal disjunctions are now often also major cultural
 The properties of mass communication 37
disjunctions - for instance, material produced in the USA or Europe
may be seen by audiences in India and South-East Asia.
These properties of temporal and spatial setting mean that a
communicative event in the mass media can actually be seen as a
chain of communicative events. In the case of a television docu­
mentary, for instance, there is not only the actual broadcast but also
the communicative events which constitute the production of the
documentary (a complex chain in its own right), and the viewing of
the documentary. The actual broadcast is in a sense a deficient
communicative event in that there is no direct communication
between broadcasters and audience. The chain can be extended to
include the source communicative events (such as political speeches
or interviews) which are transformed into the documentary on the
one hand, and subsequent communicative events (conversations,
reviews, etc.) in which the documentary itself is a transformed
source. (See below the discussion of sources in the section on prac­
tices of production and consumption.) Notice that such a chain con­
nects the public domain to the private domain: programmes are
produced in the public domain using predominantly public domain
source materials (e.g. political events), but they are consumed in the
private domain, mainly in the home and within the family. A crucial
property of the mass media is that they 'mediate' in this way between
the public and the private domains.
In fact the media have had a major impact on the boundaries
“ t 'L and priva1' n S u L l U ii H ' ; 5, i ’t’Uici vVU'i'C l i t t i l l li t
fundamental ways (Scannell 1992, Thompson 1990), as I indicated in
Chapter 1. Public events such as coronations or parliamentary
debates which were hitherto accessible only to those who attended
them have become accessible for universal private consumption by
being broadcast. Conversely, private events such as the private lives
of public figures (e.g. the British royal family) or the private grief of
bereaved parents have become public events meriting the status of
'news'. The media have helped restructure people's expectations
about the boundary between what Goffman (1969) called 'front' and
'back' region behaviour - behaviour for public consumption, versus
behaviour in private contexts. One example of this is the way in
which cameras have come to dwell upon the grief-stricken and tear-
stained faces of bereaved people in television news broadcasts.
The media have tried to bridge the gap between the public condi­
tions of media production and the private conditions of consump­
tion by evolving a 'communicative ethos' and a 'communicative
style' (Scannell 1992) which adjust towards the priorities, values and
38 Communication in the mass media
practices of private life. This includes the development of a 'public-
colloquial' language (Leech 1966), a public language for use in the
media which is modelled to varying degrees and in varying ways
upon the practices of informal, colloquial, conversational speech.
This is an important development which has already been referred to
and which will figure at various points in this book. (See the section
on the sociocultural context below for further discussion.)
Mass communication differs from other forms of communication in
the technologies it draws upon, which make possible its characteris­
tic temporal and spatial disjunctions. A medical consultation is a face-
to-face communication involving interaction through spoken lan­
guage and non-verbal communication (posture, gesture, expression,
touch). It is transient, whereas a television or radio programme
crucially is recorded in a permanent and reproducible form (Benja­
min 1970). The written notes which a doctor makes obviously do not
aim to represent the whole consultation; a consultation may be
recorded, and may even become a media event by being broadcast,
but these possibilities are not inherent properties of the genre. A tele­
vision documentary, by contrast, can be stored indefinitely, it can be
reproduced in any number of copies, and be used and reused for a
variety of purposes at different times and in different places. It can be
produced, distributed and consumed as a cultural commodity (see
the discussion of economics below).
There are obvious but important differences between different
types of media in their channels of communicarion and the techno­
logies they draw upon. The press uses a visual channel, its language
is written, and it draws upon technologies of photographic reproduc­
tion, graphic design, and printing. Radio, by contrast, uses an oral
channel and spoken language and relies on techn ologies o f sou n d
recording and broadcasting, whilst television combines technologies
of sound- and image-recording and broadcasting. The relationship
between the oral and visual channels in television is a major issue
which merits detailed attention case by case. In contrast with film,
television can be characterized in broad terms as verbally anchored,
with images mainly being used to support words (Ellis 1982; 129).
These differences in channel and technology have significant
wider implications in terms of the meaning potential of the different
media. For instance, print is in an important sense less personal than
radio or television. Radio begins to allow individuality and person­
ality to be foregrounded through transmitting individual qualities of
voice. Television takes the process much further by making people
visually available, and not in the frozen modality of newspaper
 The properties of mass communication 39
photographs, but in movement and action, it is a technology which
harmonizes with our contemporary culture's focus on individualism
and its orientation towards personality (see the section on the
sociocultural context below). Television as a technology also favours
action rather than contemplation, and foregrounds the present. Even
where programmes are prerecorded, the illusion of liveness and
immediacy is maintained. Rapid cutting between images generates
action and excitement, while close-up shots of people ('talking
heads') reduce social distance and convey an egalitarian ethos. The
condensed thirty-second combination of sounds and images in a
high-budget television commercial can stand as an archetype of the
capabilities of the medium, and indeed the dominant cultural form in
television - used as the basis of news programmes and soap operas
alike - is a sequence of disconnected short segments no longer than
five minutes in duration (on these and other properties of television,
see Ellis 1982).
Different types of communication involve different categories of
participant. In the case of medical consultations, the main parti­
cipants are obviously doctors and patients, though there may be
others (e.g. a nurse, ora relative of the patient). The categories of par­
ticipant in the media follow from the character of mass commun­
ication discussed above in mediating between public and private
domains. The main categories of participants in television docu­
mentaries, for instance, are reporters (a category of mediators), audi­
ence, and various categories of public domain Third party' who may
be involved - politicians, trade unionists, scientists and experts of
various other types, academics, and so forth. There is also, interest­
ingly, another important category of third party in contemporary
media, which emanates from the private domain - ordinary people
who may act as witnesses or represent typical behaviours or reactions
(commonly referred to as 'vox pop', an abbreviated form of the Latin
for 'voice of the people'). It is not simply the identification of parti­
cipants that is of analytical interest; a key question is how participant
identities and relations are constructed in various types of pro­
gramme. (See the section on sociocultural context below.)
An obvious and important feature of media events is the mass
nature of audiences. A television documentary is, in principle, avail­
able to the great majority of the population; there are powerful
economic imperatives towards audience maximization, particularly
in prime-time television (see the section on the economics of media
below), and audiences of around 12 million people in Britain, for
instance, are not unusual. Audience size underscores the potential
40 Communication in the mass media
influence and power of the media, and the interest that the state may
have in attempting to control it (see the section on politics below).
Moreover, media communicative events are sorts of monologues,
which is also of course germane to questions about the power of mass
media: audiences cannot directly contribute to the communication.
Whereas doctor and patient alternate in speaker and listener roles in a
medical consultation, media audiences only listen (or view or read).
Terms like 'communication' or 'interaction' are in a sense misnomers,
a point Thompson (1990: 228) makes in calling media discourse
'mediated quasi-interaction'. Media producers lack the simultaneous
feedback from audiences which is readily available in what people
say, fail to say, and in the ways in which they act and look in medical
consultations. As a consequence, producers postulate and construct
'ideal' audiences partly on the basis of guesses about audience
response drawn from experience and various types of indirect evi­
dence (such as programme ratings or market research). There is
much debate in this connection about questions of manipulation,
cultural domination and imperialism (especially where the cultural
gap between producer and audience is wide), and ideology.
All forms of mass communication give rise to questions about
access. In mediated quasi-interaction, the issue of which categories of
social agent get to write, speak and be seen - and which do not -
assumes considerable importance. There is no technical reason why
communities of various sorts (trade union branches, people living on
dii in n er-aiy nou^in^ ebidie, people Uj ci nuiiOLity culture)
could not produce their own videos and have them broadcast as
documentaries or news items. But this very rarely happens. Media
output is very much under professional and institutional control, and
in general it is those who already have other forms of economic, pol­
itical or cultural power that have the best access to the media (see the
discussion of sources on page 49). There do now appear to be various
moves to open up access - or perhaps, to put it more cynically, to
mitigate the unequal distribution of access. These include extensive
use of vox pop, radio phone-in programmes in which members of the
audience put questions to or even make comments on public figures,
audience discussion programmes, and access programmes in which
community groups or individuals are given space for their own
material. But some commentators see these innovations as quite
limited and marginal. Scannell (1992) notes, for instance, that
whereas public persons are called upon for their opinions, private
persons are generally called upon only for their experiences (though
compare the views of Livingstone and Lunt 1994). Following
 The properties of mass communication 41
Enzensberger (1970), one might say that the social relations of the
media inhibit the full exploitation of their potential as technologies.
Communicative events differ in the fields of social activity that
they represent, and in how they represent them. À communicative
event is itself a form of social practice, and what it represents are
other social practices, and more often than not other communica­
tive events. The question is, then, which (fields of) social practices
and which communicative events are represented in particular
types of communicative event. Medical consultations deal predom­
inantly with social practices in private life, for instance with
people's eating habits. Forms of mass communication such as tele­
vision documentary, on the other hand, deal with a wide variety of
social practices, mainly in the public domain, such as the social
practices of politics, education or law. The idea of mass commun­
ication as an extended chain of communicative events is again
helpful here, linking communicative events in the public domain to
communicative events in the private domain of media reception
and consumption.
The interesting question is, then, how public domain commun­
icative events are transformed as they move along the chain. Fol­
lowing van Leeuwen (1993), we can ask how one type of
communicative event 'recontextualizes' others - what particular
representations and transformations it produces, arid how these
differ from other recontextualizations of the same events. The gen­
eral point is that communicative events and social practices are
recontextualized differently depending upon the goals, values and
priorities of the communication in which they are recontextualized.
This raises questions of truth, bias and manipulation which have
been a major preoccupation in media analysis - see the section on
the politics of media below, in the analysis of texts, such differ­
ences of representation can be specified in terms of the use of
different 'discourses'. Notice that I am here using 'discourse' as a
count noun, with a singular and plural ('a discourse', 'several
discourses'): a discourse as a type of language associated with a
particular representation, from a specific point of view, of some
social practice. See also pages 18-19, and the analytical framework
in Chapter 4.
A medical consultation is an operational, instrumental type of
communication, concerned with getting things done. The patient
brings a problem to the doctor for specialist help, the doctor tries to
ascertain the precise nature of the problem, and to determine and
prescribe a course of treatment. Media events are generally rather
42 Communication in the mass media
less clear-cut in terms of the purpose and nature of what is going on.
In the case of a television documentary, for instance, on one level
what is going on may be an educational and informative process: the
programme is giving viewers a better understanding of some issue of
current concern. Documentaries, however, tend also to be persua­
sive: they try to get viewers to see things in a particular way. And
they also aim to be entertaining, to tell a good story as well as elabor­
ating a convincing argument, and to produce a pleasing film (on the
distinction between story and argument, see Silverstone 1985). Like
other sorts of programme, they are subject to a complex of economic
and political as well as cultural pressures.

The economics of media

The economics of an institution is an important determinant of its

practices and its texts. The funding system for the National Health
Service in Britain, for example, constrains the service doctors can pro­
vide for patients, and thereby shapes interactions between doctors
and patients, and the texts that are produced: the possibilities for
interaction are, for instance, severely reduced where doctors are
limiting the duration of appointments to five minutes! Similarly, the
intensely competitive commercial environment that the media
operate in at present shape media practices and texts (Inglis 1990,

The press and commercial broadcasting are pre-eminently profit­

making organizations, they make their profits by selling audiences to
advertisers, and they do this by achieving the highest possible
readerships or listener/viewer ratings for the lowest possible financial
outlay. Even non-commercial broadcasting organizations such as the
BBC are subjected to a parallel market logic: they are in competition
with commercial broadcasting, and they rely upon their ratings to
justify to the government and the public the licence fees which
people are required to pay.
Media texts and programmes are from this perspective symbolic,
cultural commodities, produced in what is effectively a culture
industry, which circulate for profit within a market, and they are
very much open to the effects of commercial pressures. The ratings
battle leads both to an increase in types of programme with high
audience appeal such as the 'soaps', and to attempts to increase
the audience appeal of other types of programme such as news,
current affairs and documentary. The process is often referred to as
 The economics of media 43
'going down-market'. This typically involves, in broad terms,
increasing emphasis on making programmes entertaining and cor­
respondingly less emphasis on their informative or educative quali­
ties (Postman 1987).
This affects both content and communicative style. For instance,
considerations of what will make 'good television' (though this
involves a complex of commercial and professional/aesthetic judge­
ments) are likely to loom larger in the choice of topics for documenta­
ries and in the ways in which topics are handled. The latter might
include more dramatic forms of presentation drawing upon fictional
m od els-as in the 'Vigilante!' extract discussed in Chapter 1 - a focus
upon media presenters as 'personalities' and the particular types of
personality that they cultivate, and the construction of an informal,
conversational relationship between presenter and audience. Pro­
ducers tend to see a shift towards the personal as increasing audience
appeal - a focus, for instance, in news programmes on the grief of
bereaved people - and the introduction of the topics of private life
tends to go along with the simulation of the communicative styles of
private life. But 1 shall argue shortly that market pressures are not the
only cause of such developments in communicative style.
Patterns of ownership are also an important, if indirect, shaping
influence upon media discourse. Ownership is increasingly in the
hands of large conglomerates whose business is the culture industry,
so that the media become more fully integrated with ownership inter­
ests in the national and international economy, intensifying their
association with capitalist class interests. This manifests itself in
various ways, including the manner in which media organizations
are structured to ensure that the dominant voices are those of the pol­
itical and social establishment (see the discussion of sources on page
49), and in the constraints on access to the media discussed earlier, it
is also more pervasively present in a pro-capitalist 'ethos', as
Williams (1975: 41) indicates in a statement about the global dom­
ination of television by US interests:

The commercial character of television has then to be seen at several

levels: as the making of programmes for profit in a known market; as a
channel for advertising; and as a cultural and political form directly
shaped by and dependent on the norms of a capitalist society, selling
both consumer goods and a 'way of life' based on them, in an ethos that
is at once locally generated, by domestic capitalist interests and authori­
ties, and internationally organised, by the dominant capitalist power.

This pervasive ethos is manifest, and analysabie, in media texts.

44 Communication in the mass media
The politics of media

Broadcasting organizations in the UK have, as conditions on their

licence to broadcast, public service obligations to provide impartial
and balanced coverage of social and political news, and educational
services. There is therefore a tension between the pressure to
increase ratings through opting broadly for more entertainment, and
the pressure to provide public service information and education.
The tension is more evident in Britain, where the public service tradi­
tion of the BBC is a strong one, than in the USA, where broadcasting
was commercially dominated from the start. But the public service
tradition in Britain is now under threat even in the BBC, because it is
obliged to enter a market where competitiveness has intensified,
especially with the arrival of satellite and cable television and
commercial radio.
Indeed, Habermas (1989) has pointed to a long-term demise of the
media as an effective political public sphere, a space for rational
debate and discussion of political issues, under the influence of a pro­
cess of commercialization which goes back to the nineteenth century.
He has referred to a 'refeudalization' of the mediatized public sphere,
in which audiences become spectators rather than participants, and
are addressed as consumers (of entertainment) rather than as citi­
zens. The intensified commercialization of the media in the past few
decades, especially since the advent of commercial television and
itidio. Svu? led to oim ibr onoiosos cind ?, dofoTic1? of public service
model (Garnham 1986). Scannell (1992) and Tolson (1991), however,
argue that the mediatized political public sphere is evolving, not dis­
appearing. Tolson contrasts an earlier 'paternalist' with a more recent
'populist' public sphere. Cardiff (1980) and Scannell (1992) have
traced the evolution in broadcasting of a communicative ethos which
is based upon an institutionalization of the conversational practices
of the private domain. Tolson, by contrast, sees a public sphere with
inner contradictions, vacillating between demands for information
and for entertainment.
I find Tolson's formulation a helpful one and will work with it
below. However, the concept of information needs to be treated with
caution. A great deal of media analysis has pointed to informationally
oriented aspects of media output (for instance in news programmes)
being ideologically shaped. In particular, representations in media
texts may be said to function ideologically in so far as they contribute
to reproducing social relations of domination and exploitation. Ideo­
logical representations are generally implicit rather than explicit in
 The politics of media 45
texts, and are embedded in ways of using language which are natur­
alized and commonsensical for reporters, audiences, and various cat­
egories of third parties - presuppositions and taken-for-granted
assumptions upon which the coherence of the discourse depends, or
the ordinary ways in which interviews are conducted.
I find it helpful to differentiate ideological aspects of discourse from
persuasive aspects, though both in different ways are political
aspects of discourse which problematize the idea of the media simply
'giving information'. A documentary, for instance, will typically
adopt a particular point of view on its topic and use rhetorical devices
to persuade audiences to see things that way too. Ideologies, by con­
trast, are not usually 'adopted' but taken for granted as common
ground between reporter and/or third parties and audience, without
recourse to rhetorical devices.
Where media analysis focuses upon ideological effects of media
discourse (critical linguistics, discussed in Chapter 2, is a case in
point), some form of complicity is suggested between the media and
dominant social classes and groups. But such complicity should not
be assumed. Rather, whether it exists and what forms it takes need to
be assessed case by case. The point is that while some sections of the
media can sometimes appear to be little more than tools of dominant
interests, the media overall are in a more complex and variable
relationship with such interests. There is sometimes direct conflict
between even mainstream media and government, or media and
capital. w here relationships of complicity do exist, they take a wide
variety of forms. There are cases of media moguls (people like Rupert
Murdoch, the late Robert Maxwell, or Conrad Black) directly manipu­
lating the media outlets they own in their own interests. There are
also instances in Britain of direct intervention by the state to control
media output - notoriously in the case of coverage of the crisis in
Northern Ireland - and in many other countries public broadcasting
is routinely controlled by the state. By contrast, the BBC in Britain has
rarely allowed itself to be directly politically manipulated - though it
did notoriously act as an instrument of the government during the
General Strike in 1926.
The state does have an interest in controlling media output. The
media, and especially television with its massive audiences, have
immense potential power and influence. This includes a mobilizing
power, as well as the ideological potential of the media (Enzensber-
ger 1970). Recent examples which are often referred to are the influ­
ence in the USA of television coverage of the Vietnam war in
swinging public opinion against the war and forcing the American
46 Communication in the mass media
withdrawal, the impact of television film of famine in Africa in forcing
governments to at least give the appearance of doing more about
'Third World' poverty, and the effect of television coverage of the
events of 1989 in the former socialist countries of eastern Europe on
the mobilization of popular protest movements. Attempts at state
control may be more or less direct. During the Gulf War, the military
exercised tight control over the media, determined that the experi­
ence of the Vietnam war would not be repeated (Kellner 1992).
Although in other circumstances the BBC may not suffer much direct
censorship, it has at times been subject to intense monitoring and
critique from government, notably in the Thatcher years of the 1980s,
which must at least have an inhibiting effect, as do the relationships
of mutual dependence, goodwill and trust which are built up
between journalists and government ministers and officials within
news-gathering networks (Tuchman 1978).
But in claiming that the media constitute a powerful ideological
apparatus, one is not necessarily suggesting that they are subject to
overt political manipula tion on a large scale. The history of the BBC is
an interesting case in point. Kumar (1977) points out that in the more
unstable and competitive climate which has obtained since the begin­
ning of the 1960s, the BBC has had to abandon its claim to be the voice
of a national cultural consensus. Its voice - personalized in its
announcers, newsreaders and presenters - has evolved in a populist
direction, claiming common ground (the 'middle ground' and a
shared 'common sense') with audiences, and often adootins a
cynical, challenging and even aggressive stance to a variety of official
institutions and personalities, including, for instance, government
ministers. But the common-sense assumptions and presuppositions
which the discourse of these key media personnel is built upon often
have a heavily ideological character - naturalizing, taking as obvious,
for instance, basic design features of contemporary capitalist society
and its consumerist values. The cynicism and aggressiveness
towards establishment figures is thus often at odds with the way in
which the discourse naturalizes establishment (dominant) ideolo­
gies. Putting it differently, the opening up of social relations (realized
in interpersonal aspects of language) is perhaps in contrast with the
continuing closure of social representations (realized in the ideational
aspects of language). (On the contrast between interpersonal and
ideational functions of language, see further Chapter 4.)
The concept of ideology often implies distortion, 'false conscious­
ness', manipulation of the truth in the pursuit of particular interests
(see Chapter 1). The only way of gaining access to the truth is through
 The politics of media 47
representations of it, and all representations involve particular points
of view, values, and goals. Accusations of 'bias' tend to overlook this.
But this does not entail a relativism which sees all representations as
equal. In media analysis one is always comparing and evaluating
representations, in terms of what they include and what they
exclude, what they foreground and what they background, where
they come from and what factors and interests influence their for­
mulation and projection, and so forth. 'The truth' in an absolute
sense is always problematic, and a source of much fruitless argu­
ment. But representations can be compared in terms of their par­
tiality, completeness, and interestedness, and conclusions can be
arrived at - and constantly are arrived at - about the relative
(un)truthfulness of representations. Needless to say, people always
make such evaluations from particular positions and points of view,
but these too can be compared in terms of how public-spirited or self-
interested they are. Truth is a slippery business, but abandoning it
altogether is surely perverse.
Ideological analysis of media has lost much of the prestige it had
during the 1970s, partly because of a changing political climate and
partly because of difficulties with this sort of analysis. It has been
criticized for assuming ideological effects of texts upon audiences
without actually investigating how audiences 'read' texts. Studies of
audience reception have now become very popular, partly at the
expense of ideological analysis (recall my comments on this in
Chapter 1). Ideolo gical analysis also tended to be reductionist in its
approach to texts, which are never simply ideology. But there is a
danger in the reaction against the ideological analysis of the 1970s
that its important insights will be lost.
My view is that media discourse should be regarded as the site of
complex and often contradictory processes, including ideological
processes. Ideology should not be seen as a constant and predictable
presence in all media discourse by definition. Rather, it should be a
working principle that the question of what ideological work is being
done is one of a number of questions which analysts should always
be ready to ask of any media discourse, though they should expect
the answers to be variable. Ideology may, for example, be a more
salient issue for some instances and types of media discourse than for
others. Media texts do indeed function ideologically in social control
and social reproduction; but they also operate as cultural commod­
ities in a competitive market (as 1 suggested earlier), are part of the
business of entertaining people, are designed to keep people pol­
itically and socially informed, are cultural artefacts in their own right,
48 Communication in the mass media
informed by particular aesthetics; and they are at the same time
caught up in - reflecting and contributing to - shifting cultural values
and identities. There is obviously overlap between these various
facets, but as well as differing in their relative salience between
different media texts, they may involve different aspects of the forms
and meanings of texts, and may result in texts which are contra­
dictory in their forms and meanings.

Practices of media text production and consumption

A further dimension of communication in the mass media is the insti­

tutional practices associated both with the production of media texts
and with the consumption of media texts. Processes of text pro­
duction are managed through sets of institutional routines. Media
organizations are characterized by routine ways of collecting and
selecting material, and editing and transforming source material into
finished texts (Bell 1991, Siiverstone 1985, Tuchman 1978, van Dijk
1988a). The production of a text is a collective process, involving jour­
nalists, producers, and various categories of editorial staff, as well as
technical staff. Bell estimates, for instance, that in a moderate-sized
press newsroom up to eight people may contribute to the production
of a story, and the story may correspondingly go through up to eight
versions. Siiverstone shows similar complexity in the production of
documentary. The journalist's first draft rnav be changed by the chief
reporter, the news editor, the editor, the chief sub-editor, a page sub­
editor, a copy sub-editor, or the check sub-editor (Bell 1991: 44-6).
Moreover, since a high proportion of source material is made up of
news items already produced by news agencies, a given story may
undergo a similiar process in each of several newsrooms before
appearing in a newspaper or on a news broadcast.
Consequently, news, documentary, and other types of media
discourse have a heavily embedded and layered character (Bell 1991:
50-5), in the sense that earlier versions are embedded within later
versions, and constitute so many layers within them. At each stage in
the construction of the story, earlier versions are transformed and
recontextualized in ways which correspond to the concerns, prior­
ities and goals of the current stage (recall my comments on represen­
tations in the previous section). But it is not simply earlier versions in
the production process that are transformed, recontextualized and
embedded in the final text: so too are the source communicative
events which stories are ultimately based upon - the interviews, the
 Practices of media text production and consumption 49
political speeches, the policy documents, and so forth. The pro­
duction of media texts can thus be seen as a series of transformations
across what I earlier called a chain of communicative events which
links source events in the public domain to the private domain con­
sumption of media texts.
With respect to sources, one striking feature of news production is
the overwhelming reliance of journalists on a tightly delimited set of
official and otherwise legitimized sources which are systematically
drawn upon, through a network of contacts and procedures, as
sources of 'facts' and to substantiate other 'facts' (Tuchman 1978).
These include government and local government sources, the police,
employers' organizations and trade unions, scientific and technical
experts from universities. Organizations which are not perceived as
legitimate (for instance, what are defined as 'extreme' political
groups or parties) are excluded or more rarely referred to. Ordinary
people, including rank-and-file members of organizations, feature as
ottering typifications of reactions to news, but not as news sources -
as Scannell (1992) puts it, they are entitled to their experiences but not
their opinions. The result is a predominantly establishment view of
the world, manifested textually in, for instance, ways in which the
reporting of speech is treated. Herman and Chomsky (1988) suggest
that where there is controversy, it is predominantly because there are
divisions within the establishment. The narrowness and inherent
conservatism of the network of legimitate sources can partly be
attributed to the wavs in which the media are ecortomicallv
embedded in and dependent upon the status quo in terms of
ownership and profitability (recall discussion of the economics of
media), and the dependence of journalists upon their sources consti­
tutes an inbuilt limitation on their campaigning zeal.
The consumption of media texts is characterized by its own institu­
tional practices and routines. Overwhelmingly, media texts are con­
sumed in private domain contexts, in the home and in the context of
family life. Research on media reception has shown the various ways
in which media text consumption may be embedded within domestic
life. Viewers may, for instance, in some cases give a television pro­
gramme their full attention, while in other cases watching television
may be an accompaniment to other domestic activities, such as
eating. Or again, watching television may be a solitary activity or an
activity engaged in collectively but in silence, or it may be embedded
within, and be the topic of, conversation among viewers. Such
variations are important in assessing the reception and effects of tele­
vision. Reception studies have also emphasized the variability of
50 Communication in the mass media
interpretations of, and responses to, television programmes: any dis­
cussion of 'the meaning' of a television programme needs to take
account of the variability of the meanings that may be attributed to it
by different categories of audience member.
It is fruitful to conceptualize media text consumption as well as its
production in terms of transformations across chained communica­
tive events. Evidence for audience interpretations of media texts is
predominantly the talk and writing of audiences, and media texts are
transformed in systematic ways into audience conversation (at
various distances in time and space from the original consumption of
the media text) and other types of audience discourse, written or
spoken. Such a perspective recognizes that the media constitute both
an important resource and topic for other types of discourse, and an
important formative influence upon them (Fairclough 1992b,
Thompson 1990).

Sociocultural context

One issue in discourse analysis is to what degree context is relevant to

investigation of discourse practices. Many analysts focus upon the
immediate situation of the communicative event (the 'context of
situation'), and maybe refer to some elements of institutional context,
but say little about the wider social and cultural context. My view is
t-hai this wider contextual matrix must be attended to because it
shapes discourse practices in important ways and is itself cumula­
tively shaped by them. This is particularly clear in the case of the
media.
Factors of institutional context alone can only give a partial under­
standing of media practices, in Chapter 2 I discussed Heritage's
(1985) analysis of media interview, which gives a powerful account of
how features of interview design (such as the way formulations are
used) serve to cope with institutional constraints. What such an insti­
tutionally oriented analysis cannot explain is certain recent changes
in media interview which seem to be part of wider sociocultural
changes affecting contemporary societies. I have in mind, for
instance, the way in which political interviews have changed
between the 1950s and the present in Britain from very formal inter­
actions between often anonymous reporters and public figures con­
structed in terms of their social status, to much more informal (often
conversational, sometimes combatitive) interactions between
presenters who are well-known media personalities in their own
 Sociocultural context 51
right, and public figures who are also painstakingly constructed, by
promotional apparatuses, as personalities (Tolson 1991). Whereas
the relationship between interviewer and interviewee once faithfully
reflected status-based authority differences, it is now much more
open and negotiable, with politician and presenter often talking as
equals. The personalities of presenters are in many cases fashioned
from models in private life - as I suggested earlier, presenters often
project themselves as inhabiting the same common-sense world as
their audiences, using a communicative style partly based upon
properties of conversation. In accordance with these changes, the
discourse of political interviews has changed substantially.
The point is that such developments are not just features of media
interview. Documentary has broadly shifted from a focus on general
social issues in which people figured as representative of social
types, to a concern to construct the people it represents as indi­
viduals with their own personalities (Corner 1991, Tolson 1990).
And the shift towards greater informality and more conversation­
like ('public-colloquial') discourse is a general one not only in the
media but in many domains of public discourse, including medical
consultations (Fairclough 1992a, Mishler 1984). They are part of gen­
eral changes in social relations and cultural values which have been
discussed in terms of individualism, 'detraditionalization' (Giddens
1991) and 'informalization' (Featherstone 1991), affecting relations
of authority, relations between public and private domains of social
|;fe -vnH j Uq copstructkm 0f self-idi'rdk"’ The medic arc shamed hy
the wider society, but they also play a vital role in the diffusion of
such social and cultural changes, and this should be one focus in
analysis of media discourse. Obvious issues for attention here
include changing constructions of gender relations, race relations,
and class relations.
Changes in media discourse also reflect, and help to diffuse, con­
temporary 'promotional' (Wernick 1991) or 'consumer' culture, the
way in which models of promotion (of goods, institutions, parties,
personalities, and so forth) and consumption have spread from the
domain of economic consumption to the public services, the arts, and
the media. I have referred to the increasing salience of entertainment
in various sorts of media output, and on how audiences are increas­
ingly being constructed as consumers - with leisure being con­
structed as consumption - rather than as, say, citizens. Similarly, in
government leaflets aimed at the public and dealing with such mat­
ters as welfare benefit rights, the influence of advertising and pro­
motional genres is increasingly evident, with the public again being
52 Communication in the mass media
constructed as consumers rather than - or as well as - citizens, even
though it is their civic rights that are at issue.
Media texts constitute a sensitive barometer of sociocultural
change, and they should be seen as valuable material for researching
change. Changes in society and culture manifest themselves in all
their tentativeness, incompleteness and contradictory nature in the
heterogeneous and shifting discursive practices of the media. The
framework for critical discourse analysis introduced in Chapter 4 is
designed to capture these properties of media discourse, and provide
a resource for linking discourse analysis to social-scientific analysis of
sociocultural change.
 CRITICAL ANALYSIS OF MEDIA DISCOURSE

Tins chapter sketches out the iramework which i shall use for
analysing media discourse in the rest of the book, drawing upon the
account given in Chapter 3 of communication in the mass media. This
is a version of the theory of 'critical discourse analysis' which I have
developed in previous publications (Fairclough 1989, 1992a, 1993).
The chapter first briefly describes in general terms the theory of
discourse I am operating with, and then focuses attention upon
media discourse. The framework is described and then applied to an
example.

Theory of discourse

Recent social theory has produced important insights into the social
nature of language and how it functions in contemporary societies.
Social theorists have tended to put such insights in abstract ways,
without analysis of specific texts. To develop a form of discourse
analysis which can contribute to social and cultural analysis, we need
54 Critical analysis of media discourse
to combine these insights with traditions of close textual analysis
which have developed in linguistics and language studies - to make
them 'operational', practically usable, in analysis of specific cases.
Some of the critical approaches discussed in Chapter 2 have begun to
do this. Critical discourse analysis is an attempt to learn from them
and improve on them, in line with the desiderata at the end of
Chapter 2 (see Fairclough 1992a for a more detailed account).
Calling the approach 'critical' is a recognition that our social prac­
tice in general and our use of language in particular are bound up
with causes and effects which we may not be at all aware of under
normal conditions (Bourdieu 1977). Specifically, connections
between the use of language and the exercise of power are often not
clear to people, yet appear on closer examination to be vitally impor­
tant to the workings of power. For instance, ways in which a conven­
tional consultation between a doctor and a patient is organized, or a
conventional interview between a reporter and a politician, take for
granted a whole range of ideologically potent assumptions about
rights, relationships, knowledge and identities. For example, the
assumption that the doctor is the sole source of medically legitimate
knowledge about illness, or that it is legitimate for the reporter - as
one who 'speaks for' the public - to challenge the politician. Such
practices are shaped, with their common-sense assumptions, accord­
ing to prevailing relations of power between groups of people. The
normal opacity of these practices to those involved in them - the invi­
sibility of their ideological assumptions, and of the power relations
which underlie the practices - helps to sustain these power relations.
'Discourse' (as I pointed out in Chapter 1) is a concept used by both
social theorists and analysts (e.g. Foucault 1972, Fraser 1989) and
linguists (e.g. Stubbs 1983, van Dijk 1985). Like many linguists, I shall
use 'discourse' to refer to spoken or written language use, though 1
also want to extend it to include other types of semiotic activity (i.e.
activity which produces meanings), such as visual images (photo­
graphy, film, video, diagrams) and non-verbal communication (e.g.
gestures). Recall the discussion of the 'social semiotics' approach in
Chapter 2. In referring to use of language as discourse, I am signalling
a wish to investigate it in a way that is informed by the social theory
insights mentioned above, as a form of social practice.
Viewing language use as social practice implies, first, that it is a
mode of action, as linguistic philosophy and the study of pragmatics
have recognized (Austin 1962, Levinson 1983). It also implies that
language is a socially and historically situated mode of action, in a
dialectical relationship with other facets of the social. What 1 mean by
 Theory of discourse 55

a dialectical relationship is that it is socially shaped, but is also socially

shaping - or socially constitutive. Critical discourse analysis explores
the tension between these two sides of language use, the socially
shaped and socially constitutive, rather than opting one-sidedly for
one or the other.
Language use - any text - is always simultaneously constitutive of
(1) social identities, (2) social relations and (3) systems of knowledge
and belief (corresponding respectively to identities, relationships
and representations in the terms introduced in Chapter 1). That is,
any text makes its own small contribution to shaping these aspects of
society and culture. In particular cases, one of the three might appear
to be more important than the others, but it is a sensible working
assumption tha t all three are always going on to som e degree. Lan­
guage use is, moreover, constitutive both in conventional ways
which help to reproduce and maintain existing social identities, rela­
tions and systems of knowledge and belief, and in creative ways
which help to transform them. Whether the conventional or the cre­
ative predominates in any given case will depend upon social circum­
stances and how the language is functioning within them.
The relationship between any particular instance of language use -
any particular text - and available discourse types may be a complex
and (in the terms of the last paragraph) creative one. It is always pos­
sible to find relatively straightforward instances of particular
discourse types - a conventional and typical political interview on the
radio, for instance. But many texts are not so simple. They may
involve complicated mixtures of different discourse types - a political
interview which is in part rather like a friendly conversation and in
part like a political speech, for example. Given my concern in this
book with changing discursive practices in the media, such complex
texts are of particular interest.
The critical discourse analysis approach thinks of the discursive
practices of a community - its normal ways of using language - in
terms of networks which I shall call 'orders of discourse'. The order of
discourse of a social institution or social domain is constituted by all
the discursive types which are used there. The point of the concept of
'order of discourse' is to highlight the relationships between different
types in such a set (e.g. in the case of a school, the discursive types of
the classroom and of the playground): whether, for instance, a rigid
boundary is maintained between them, or whether they can easily be
mixed together in particular texts. The same question applies to
relationships between different orders of discourse (e.g. those of the
school and the home): do they commonly overlap and get mixed
56 Critical analysis of media discourse
together in language use, or are they rigidly demarcated? Social and
cultural changes very often manifest themselves discursively
through a redrawing of boundaries within and between orders of
discourse, and I shall be showing that this is true of the media. These
boundaries are also sometimes a focus of social struggle and conflict.
Indeed, orders of discourse can be seen as one domain of potential
cultural hegemony, with dominant groups struggling to assert and
maintain particular structuring within and between them.
It is useful to distinguish two main categories of discourse type,
which are constituents of orders of discourse: genres, and discourses.
A discourse is the language used in representing a given social prac­
tice from a particular point of view. Discourses appertain broadly to
knowledge and knowledge construction. For instance, the social
practice of politics is differently signified in liberal, socialist and Mar­
xist political discourses; or again, illness and health are differently
represented in conventional ('allopathic') and homoeopathic medical
discourses. A genre, by contrast, is a use of language associated with
and constituting part of some particular social practice, such as inter­
viewing people (interview genre) or advertising commodities (adver­
tising genre). Genres can be described in terms of their organizational
properties - an interview, for instance, is structured in a quite
different way from an advertisement. See Kress and Threadgold
(1988) and van Leeuwen (1993).
The analysis of any particular type of discourse, including media
discourse, involves aii alieiTiawon betvvccn twin, complementary
focuses, both of which are essential:
• communicative events
® the order of discourse.
On the one hand, the analyst is concerned with the particular, with
specific communicative events, for instance a particular newspaper
editorial or television documentary. The concern here is always with
both continuity and change - in what ways is this communicative
event normative, drawing upon familiar types and formats, and in
what ways is it creative, using old resources in new ways? On the
other hand, the analyst is concerned with the general, the overall
structure of the order of discourse, and the way it is evolving in the
context of social and cultural changes. The focus here is upon the con­
figuration of genres and discourses which constitute the order of
discourse, the shifting relationships between them, and between this
order of discourse and other socially adjacent ones. These are not, let
me stress, alternatives, but complementary perspectives on the same
 Analysis of communicative events 57
data which we can shift between d uring analysis. My presentation of
a framework for critical analysis of media discourse will discuss the
two perspectives in turn.

Analysis of communicative events

Critical discourse analysis of a communicative event is the analysis of

relationships between three dimensions or facets of that event, which
I call text, discourse practice, and sociocultural practice. 'Texts' may be
written or oral, and oral texts may be just spoken (radio) or spoken
and visual (television). By 'discourse practice' I mean the processes of
text production and text consumption. And by 'sociocultural prac­
tice' 1 mean the social and cultural goings-on which the communica­
tive event is a part of. The analytical framework is summarized in a
diagram on page 59.
Let me briefly link this analytical framework to the discussion of
communication in the mass media in the last chapter. The section on
practices of media text production and consumption dealt with
aspects of discourse practice. Most of the chapter was concerned with
various aspects of sociocultural practice: mass communication as a par­
ticular type of situation, the economics of the media, the politics of
the media, and the wider cultural context of communication in the
mass media. These two features are addressed further below.

Texts
The analysis of texts, the properties of which were little mentioned in
Chapter 3, covers traditional forms of linguistic analysis - analysis of
vocabulary and semantics, the grammar of sentences and smaller
units, and the sound system ('phonology') and writing system. But it
also includes analysis of textual organization above the sentence,
including the ways in which sentences are connected together ('cohe­
sion'), and things like the organization of turn-taking in interviews or
the overall structure of a newspaper article. I shall refer to all this as
'linguistic analysis', though this is using the term in an extended
sense. For details on these types of analysis, see Chapters 6 and 7.
Analysis of texts is concerned with both their meanings and their
forms. Although it may be useful analytically to contrast these two
aspects of texts, it is in reality difficult to separate them. Meanings are
necessarily realized in forms, and differences in meaning entail
differences in form. Conversely, it is a sensible working assumption
58 Critical analysis of media discourse
that where forms are different, there will be some difference in
meaning.
As I have already indicated in Chapter 1 , 1 work with a multifunc­
tional view of text. This sees any text, and indeed even the individual
clauses and sentences of a text, as simultaneously having three main
categories of function, each of which has its own systems of choices:
ideational, interpersonal, and textual. This view of text harmonizes with
the constitutive view of discourse outlined above, providing a way of
investigating the simultaneous constitution of systems of knowledge
and belief (ideational function) and social relations and social identi­
ties (interpersonal function) in texts. Or, in the terminology of
Chapter 1, with representations, relations and identities. So, for
instance, in analysing a sentence in a written text, the analyst might
focus upon how three aspects are articulated:

* particular representations and recontextualizations of social prac­

tice (ideational function) - perhaps carrying particular ideologies
• particular constructions of writer and reader identities (for
example, in terms of what is highlighted - whether status and role
aspects of identity, or individual and personality aspects of
identity)
« a particular construction of the relationship between writer and
reader (as, for instance, formal or informal, close or distant).

representations, categories of participant, constructions of parti­

cipant identity or participant relations which are not found in a text.
Analysis of text needs to be multisemiotic analysis in the case of the
press and television, including analysis of photographic images, lay­
out and the overall visual organization of pages, and analysis of film
and of sound effects. A key issue is how these other semiotic modal­
ities interact with language in producing meanings, and how such
interactions define different aesthetics for different media.

Discourse practice
The discourse practice dimension of the communicative event
involves various aspects of the processes of text production and text
consumption. Some of these have a more institutional character,
whereas others are discourse processes in a narrower sense. This was
shown in the discussion of mass media communication in Chapter 3:
with respect to institutional processes, 1 referred to institutional
 Analysis of communicative events 59
routines such as editorial procedures involved in producing media
texts, and how, for instance, watching television fits into the routines
of the household; but I also referred to discourse processes in the nar­
rower sense in discussing the transformations which texts undergo in
production and consumption. I shall call these respectively 'institu­
tional processes' and 'discourse processes'. (One could also include
here more psychological and cognitivist concerns with how people
arrive at interpretations for particular utterances - 'interpretative pro­
cesses'. As I indicated in Chapter 2, van Dijk works with a more
cognitively oriented framework which is otherwise rather similar in
conception to mine.)
The analytical framework of critical discourse analysis is summar­
ized in Figure 1.
The visual representation of the relationships between the three
dimensions of communicative events in the diagram is significant: I
see discourse practice as mediating between the textual and the social

text production

TEXT______

te x t c o n s u m p t i o n

DISCOURSE PRACTICE

SOCIOCULTURAL PRACTICE

Fig. 1: A framework for critical discourse analysis of a commun­

icative event
60 Critical analysis of media discourse
and cultural, between text and sociocultural practice, in the sense
that the link between the sociocultural and the textual is an indirect
one, made by way of discourse practice: properties of sociocultural
practice shape texts, but by way of shaping the nature of the
discourse practice, i.e. the ways in which texts are produced and
consumed, which is realized in features of texts. Ixfotice also that, as
we have just seen with the distinction between 'institutional pro­
cesses' and 'discourse processes', discourse practice straddles the
division between society and culture on the one hand, and discourse,
language and text on the other.
In referring to 'the nature of the discourse practice', I have in mind
particularly the polarity alluded to earlier between broadly conven­
tional and broadly creative discourse processes, involving either a
normative use of discourse types (genres and discourses) or a creative
mixture of them. This is where the two perspectives within critical
discourse analysis - on the communicative event, and on the order of
discourse - intersect. The question here is how the communicative
event draws upon the order of discourse (normatively or creatively),
and what effect it has upon the order of discourse - whether it helps
reproduce its boundaries and relationships, or helps restructure
them. Creative discourse practice can be expected to be relatively
complex, in terms of the number of genres and discourses mixed
together and the way they are mixed together. But complex discourse
practice may also become conventionalized - for instance, there are
ifC Vv ii* UC'<~U fi t'- j 1 Cjli.l i c c G i t V G i \UOi. tai COli u J li lu CiO-To Ot Ai IjlG o Gi
informing, persuading, and entertaining.
In very general terms, a conventional discourse practice is realized
in a text which is relatively homogeneous in its forms and meanings,
whereas a creative discourse practice is realized in a text which is rela­
tively heterogeneous in its forms and meanings. Of course, it is the
particular nature of the creativity of the discourse practice and of the
heterogeneity of the text that is of interest in a specific analysis - and
their relationship to the sociocultural practice that frames them. Also
in general terms, one would expect a complex and creative discourse
practice where the sociocultural practice is fluid, unstable and shifting,
a nd a conventional discourse practice where the sociocultural practice
is relatively fixed and stable. As I suggested in Chapter s, media texts
are sensitive barometers of cultural change which manifest in their
heterogeneity and contradictoriness the often tentative, unfinished
and messy nature of change. Textual heterogeneity can be seen as a
materialization of social and cultural contradictions and as important
evidence for investigating these contradictions and their evolution.
 Analysis of communicative events 61
The focus on discursive creativity, hybridity, and heterogeneity in
my analysis of media discourse in this book corresponds to the nature
of the times. We are living through a period of rapid and continuous
change in society and culture, the media play a significant role in
reflecting and stimulating more general processes of change, and the
practices of the media are correspondingly in constant flux. This
includes the discursive practices of the media. The general point to
emphasize is that creativity in discursive practices is tied to particular
social conditions - conditions of change and instability. The term
'creativity' can be misleading in its individualistic connotations:
discursive creativity is an effect of social conditions, not an achieve­
ment of individuals who have particular (creative) qualities.
I want to contrast the linguistic analysis of texts (in the extended
sense I explained above) with the mtertextual analysis of texts (see
Bakhtin 1986, Fairclough 1992a, Kristeva 1986). Intertextual analysis
focuses on the borderline between text and discourse practice in the
analytical framework. Intertextual analysis is looking at text from
the perspective of discourse practice, looking at the traces of the
discourse practice in the text. Intertextual analysis aims to unravel the
various genres and discourses - often, in creative discourse practice,
a highly complex mixture - which are articulated together in the text.
The question one is asking is, what genres and discourses were
drawn upon in producing the text, and what traces of them are there
in the text? To use a familiar example, the traces in a documentary text
of a mixture or genres or information, persuasion and entertainment.
Intertextual complexity in the mixing of genres and discourses is
realized linguistically in the heterogeneity of meaning and form.
Linguistic analysis is descriptive in nature, whereas intertextual
analysis is more interpretative. Linguistic features of texts provide
evidence which can be used in intertextual analysis, and intertextual
analysis is a particular sort of interpretation of that evidence - an
interpretation which locates the text in relation to social repertoires of
discourse practices, i.e. orders of discourse. It is a cultural interpreta­
tion in that it locates the particular text within that facet of the culture
that is constituted by (networks of) orders of discourse. The linguistic
analysis is, in an obvious sense, closer to what is 'there' on paper or
on the audio- or video-tape, whereas the intertextual analysis is at
one remove in abstraction from it. Consequently, in intertextual
analysis the analyst is more dependent upon social and cultural
understanding. This can seem problematic to those who expect more
'objective' forms of analysis, though it is easy to overstate the object­
ivity of linguistic analysis. Nevertheless, linking the linguistic
| 62 Critical analysis of media discourse
’ analysis of texts to an intertextual analysis is crucial to bridging the
.1 gap between text and language on the one hand, and society and cul-
ture on the other. See Chapter 5 for further discussion.

Sociocultural practice
Analysis of the sociocultural practice dimension of a communicative
event may be at different levels of abstraction from the particular
event: it may involve its more immediate situational context, the
wider context of institutional practices the event is embedded within,
or the yet wider frame of the society and the culture. All of these
layers may be relevant to understanding the particular event - and
indeed particular events cumulatively constitute and reconstitute
social and cultural practice at all levels. Many aspects of sociocultural
practice may enter into critical discourse analysis - recall the various
aspects discussed in Chapter 3 - but it may be useful to broadly
differentiate three: economic, political (concerned with issues of
power and ideology), and cultural (concerned with questions of
value and identity).

The framework which I have sketched out here is compatible with

various different emphases. One might, for instance, choose to focus
on discourse practice, either on processes of text production, or on
processes of text consumption. One might alternatively choose to
focus on text, as I have done It is. I believe, imnortant to maintain the
comprehensive orientation to communicative events which is built
into the framework, even if one is concentrating upon only certain
aspects of them in analysis. My emphasis will be upon linguistic
analysis of texts, intertextual analysis of texts, and selective sociocul­
tural analysis. Through intertextual analysis I link up with issues of
discourse practice, but I am not concerned in this book with direct
analysis of production or consumption of texts. The discussion of
sociocultural practice is selective because I am not writing as a
sociologist or cultural analyst, but as a discourse analyst with an
interest in these other types of analysis.

Analysis of the order of discourse

I come now to the second of the twin perspectives within a critical

discourse analysis of the media, the order of discourse - how it is
structured in terms of configurations of genres and discourses, and
 Analysis of the order of discourse 63
shifts within the order of discourse and in its relationship to other
socially adjacent orders of discourse.
I referred earlier to the positioning of the media between public
orders of discourse and private orders of discourse, and to the way in
which the media transform their source public discourse for con­
sumption in domestic settings. This mediating position, and the
external relations between the order of discourse of the media and
socially adjacent public and private orders of discourse such as those
of books and magazines, is the key to understanding the media order
of discourse and the internal relations between its constituent genres
and discourses. The order of discourse of the media has been shaped
by the tension between its contradictory public sources and private
targets, which act as contrary poles of attraction for media discourse;
it is constantly being reshaped through redefining its relationship to -
redrawing its boundaries with - these public and private orders of
discourse. Moreover, the negotiation and renegotiation of the
relationship between public and private discursive practices which
takes place within the order of discourse of the media has a general
influence on the relationship between these practices, and between
the public and the private in an overall sense, in other domains of
social life. Research on media orders of discourse is thus of more than
parochial interest, because it impinges upon major changes in society
and culture. Similar remarks apply, for instance, to the (renegoti­
ation within broadcast media discourse of the relationship between
the more traditional order of cbocouise ui public service broadcasting
and the commercial order of discourse of the market and con­
sumerism.
The general point here is that the relationship between institu­
tions and discursive practices is not a neat or simple relationship.
Different institutions come to share common discursive practices,
and a particular discursive practice may have a complex distribution
across many institutions. For instance, advertising may be rooted in
the orders of discourse of commodity production, distribution and
consumption, but it has come to be an element in the orders of
discourse of diverse institutions - education, medicine, the arts, and
so forth. It follows that discourse analysis should always attend to
relationships, interactions and complicities between social
institutions/domains and their orders of discourse, and be sensitive
to similarities in social organization and discursive practices
between different institutions. Although the media may be a parti­
cularly clear case of such fluid relationships between institutions,
this property is widely shared.
64 Critical analysis of media discourse
It should also be emphasized that media discourse may shape
socially adjacent orders of discourse as well as being shaped by them.
For instance, television formats have considerable cultural salience,
and one finds them as models in a variety of public domains. An
example would be the way in which the celebrity-interview format is
now quite widely used in higher education for introductory books on
the thinking of prominent figures, as well as in magazines (such as
the left-wing political magazine Red Pepper). Postman (1987:91) refers
to the influence of other television genres on education, including the
television version of 'discussion' which he characterizes in these
negative terms:

. . . each of six men was given approximately five minutes to say

something about the subject. There was however no agreement on
exactly what the subject was, and no one felt obliged to respond to any­
thing anyone else said. In fact, it would have been difficult to do so,
since the participants were called upon seriatim, as if they were finalists
in a beauty contest, each being given his share of minutes in front of the
camera.

Media discourse also^ influences private domain discourse practices,

providing models of conversational interaction in private life which
are originally simulations of the latter but which can come to reshape
it. A complex dialectic seems to exist between the media and the con­
versational discourse of everyday life.
V>£tnrop-n r>.f 1 'l-'i. •
tions between discourses and genres within the media order of
discourse, maybe difficult to disentangle, but the distinction between
these two concerns in analysis of orders of discourse is a useful one.
Both external and internal relations include choice relations, and chain
rela tions. What I have said so far appertains to choice relations. (Let me
remind readers of the point made towards the end of Chapter 1 that
'choice' does not here imply free choice on the part of participants -
selection among alternatives is generally socially conditioned.) Ex­
ternally, the issue is how the order of discourse of the media chooses
within, and appropriates, the potential available in adjacent orders of
discourse. Internally, the issue is to describe the paradigms of alterna­
tive discursive practices available within the order of discourse of the
media, and the conditions governing selection amongst them. Discur­
sive practices are functionally differentiated, providing contrasting
formats for the main types of output in the media. Thus there are
different discursive practices for news, documentary, drama, quiz and
'soap' programmes on television, and there are different discursive
 Analysis of the order of discourse 65
practices for hard news, soft news, comment and feature articles in
newspapers. (As these two examples show, the classification of func­
tionally different discursive practices may be at various levels of gen­
erality.) But there are also alternatives for any given type of output
whose selection is governed by different conditions, which I come to
shortly.
I have already referred to chain relations in suggesting that a
communicative event in the mass media can be regarded as in fact a
chain of communicative events. Such chains are partly internal - the
process of text production within a media institution is a chain of
communicative events - and partly external - the source communica­
tive events at one end of the chain lie outside the media, as do the
communicative events (conversations, debates, reports) which
media texts may themselves, be sources for. A description of the
media order of discourse is concerned to specify what communicative
events, internal and external, are chained together in this way; and
the sorts of transformations that texts undergo in moving along such
chains, and how earlier texts in the chain are embedded in later ones.
Choice relations and chain relations intersect in an account of the
order of discourse: one needs to specify the choice relations that
apply at each link in the chain.
The distinction between choice and chain relations suggests a
refinement of the intertextual analysis of texts discussed in the
section on discourse practice above. Part of the intertextual analysis
ol a text is concerned with uni a veiling mixtures ui genres and
discourses which are in a choice relationship in the order of
discourse. But the intertextual analysis of a text is also concerned with
embedding - with how the transformations which texts undergo in
shifting along chains leave traces in embedding relations within
texts. See Chapter 5 for further discussion.
In trying to arrive at a characterization of the media order of
discourse, the analyst constantly has in mind two important ques­
tions, which may receive different answers for different parts of this
complex order of discourse: (a) how unitary, or how variable, are
media discursive practices? and (b) how stable, or how changeable,
are they?
The questions are linked: typical of a settled and conservative
society are unitary and stable discursive practices, typical of an unset­
tled society are variable and changeable discursive practices. There
are also more local institutional pressures towards unitary practice -
standardized formats reduce production costs, and conform to audi­
ence expectations. In describing the order of discourse, one is trying
66 Critical analysis of media discourse
to capture the particular balance that exists between what Bakhtin
called 'centripetal' (unitary and stable) and 'centrifugal' (variable and
changeable) pressures, and in which direction that balance is tending
(Bakhtin 1981). The variability question links back to my mention
above of alternatives for a particular type of output, such as television
documentary. Where there is variability, selection between alterna­
tives may, for instance, involve political and ideological differences
and struggles, attempts to cater for different 'niche' audiences, as
well as differences of professional or artistic judgement. Variability is
also an issue in text consumption: what orders of discourse do audi­
ences draw upon to appropriate media texts? Do they, for instance,
talk or write about them in the genres and discourses of private life, or
in those of public (e.g. academic) domains they are familiar with?
And what social factors are relevant to that choice?
Changing media discursive practices, and their relation to wider
social and cultural changes, are, as I have indicated, a particular con­
cern of this book. Change can be conceptualized in terms of shifting
external or internal, chain or choice relations. An account of the
media order of discourse should particularly highlight major points
of tension affecting internal or external boundaries. I have already
identified some as central themes of the book: the public/private
boundary and the privately oriented communicative ethos of broad­
casting, with extending use of a conversational, public-colloquial
discourse style; the boundary between public service and informa­
tion on one side and the market on the other , with the construction of
audiences as consumers and the colonization of even the news media
by entertainment; and, related to this, the boundary between fiction
and non-fiction, with non-fictional programmes such as docu­
mentary often drawing upon fictional, dramatic, formats.
I want to use the term discourse type for relatively stabilized con­
figurations of genres and discourses within the order of discourse.
One issue here is that genres occur in particular combinations with
discourses - particular genres are predictably articulated with parti­
cular discourses. For instance, party political broadcast is a genre
which predictably draws upon economic discourse, discourse of law
and order and educational discourse, but not, for instance, on the
discourses of science, cookery, or craft (e.g. knitting). But discourse
types also standardly involve configurations of genres rather than a
single genre. So, for instance, a party political broadcast may
combine political oratory, interview, and simulated fireside conver­
sation. Or again, 'chat' has emerged as an important studio-based
discourse type in television, involving an articulation of elements of
 Analysis of the order of discourse 67
conversation with elements of entertainment. A major concern here
is capturing the distinctive discourse types which have emerged in
the order of the discourse of the media, such as chat, or what passes
for discussion on television (recall the above quote from Postman),
and properties which cut across types, such as the 'communicative
ethos' identified by Scannell (1992). Another concern is a historical
focus on the stabilization and destabilization of the configurations
which constitute discourse types.
One might see the mass media as an interrelated set of orders of
discourse, in that the orders of discourse of television, radio, and
the press are distinct in important ways which relate to differences
of technology and medium while also having significant similari­
ties (recall the discussion of differences between media on pages
38-9). There are also sufficient differences between different
outlets to distinguish at a more detailed level separate orders of
discourse for different newspapers, radio stations, or television
channels.
The media order of discourse can, I think, usefully be examined as
a domain of cultural power and hegemony. The media have in the
past often been described as if they were dominated by stable unitary
practices imposed from above. This is certainly not an adequate char­
acterization of the contemporary media, though it may have some
truth for certain aspects of media practice, and was markedly less
inadequate thirty years ago than it is now. It implies a code model of
f-Vi-Djy»esy-Jj pr of rji epp.n yc-p.- jjy’ -rmprlp •Ip) Q? 2 JpLlr"?_lfc'OT' Ox kVCll'
defined, unitaiy and stable codes which dictate practice.
It does not, however, follow that because the code model is inade­
quate, questions of power and domination do not arise. One
common picture of contemporary media stresses cultural diversity - a
view of the media as highly pluralistic in practices, with no single web
of power running through the whole system. This would perhaps
compromise entirely the notion of a media order of discourse, or at
least lead to a very different model of it as a mosaic of practices.
Another possible approach, however, is to ask how the relative
diversity and pluralism of the media might itself operate within a
system of domination. Gramsci's concept of hegemony (Forgacs 1988,
Gramsci 1971) is helpful here as a theory of power and domination
which emphasizes power through achieving consent rather than
through coercion, and the importance of cultural aspects of domi­
nation which depend upon a particular articulation of a plurality of
practices. The issue with respect to a hegemony model becomes one
of whether and how diverse discursive practices are articulated
6S Critical analysis of media discourse
together within the order of discourse in ways which overall sustain
relations of domination. See Fairclough (1992a) for a discussion of
code, mosaic and hegemony models. Britain faces a war to stop pedlars, warn MPs

Various aspects of the critical discourse analysis framework are elab­

orated in the following three chapters. Chapter 5 focuses upon
discourse practice, specifically discourse processes, intertextual
CALL UP FORCES
analysis of texts, and the order of discourse. Chapters 6 and 7 are con­
cerned with the linguistic analysis of texts, the former dealing with
representation and the ideational function as well as the textual func­
IN DRUG BATTLE!
By DAVID KEMP
tion, the latter dealing with relations, identities and the interpersonal
THE armed forces should be called up to fight off a massive invasion by drug pushers,
function. 3 MPs demanded yesterday.
Cocaine pedlars are (he greatest threat ever faced by Britain in peacetime — and could
4 destroy the country's way of life, they said.
The MPs want Ministers to consider
A sample critical discourse analysis ordering the Navy and the RAF to track
5 suspected drug-running ships approach­
ing our coasts.
On shore there should be intensified
1 conclude this chapter with a sample analysis to make the critical 6 law enforcement by Customs, police and
discourse analysis framework a little more concrete. My example is a security services.

repiort which appeared in 1985 in the British newspaper, the Sun, Profits
about a government document on hard-drug abuse. It is reproduced
7 The all-party Home Affairs Committee
in Figure 2. visited America and were deeply shocked
by what they saw.
My objective is to give readers a quick overview of how the In one of the hardest-hilling Commons
framework applies in a particular case, so I shall be very selective in. 8 reports fur years, the committee—chaired
by Tory lawyer MP Sir Edward
my comments (for instance, not referring to consumption at all), and Gardner— warned gravely:
^ ■Is f.-irrvi hu .;■>warlike
cerramiy will nut attempt a mil analysis phe example is more fully ^ threat from the hard-drugs industry.
The traffickers amass princely incomes
analysed in Fairclough 1988). from the exploitation of human weakness,
I shall shift slightly from the order in which I presented the boredom and misery.
They must be made to lose everything —
framework, first analysing this as a communicative event in terms of their homes, their money and ail they
possess which can be attributed to
discourse practice and text, but deferring discussion of sociocultural prac­ their profits from selling drugs. Jr
Sir Edward said yesterday: "We believe
tice until after I comment on what the example indicates about the 9 that trafficking in drugs is tantamount tu
media order of discourse. murder and punishment ought to reflect
this."
The Government is expected to bring in
clampdown laws in the autumn.
The communicative event
The discourse practice here involves transformations of source texts -
most obviously the Committee report, but also presumably a press
conference or interview alluded to in the penultimate paragraph -
into an article. The text is likely to have gone through a number of
versions, as it was transformed across a chain of linked communica­
tive events. For a reconstruction of such a transformational history in
detail, see Bell (1991). The discourse practice is complex, in the sense
that it articulates together features of the source discourse (the report) Fig. 2: Extract from the Sun, 24 May 1985
70 Critical analysis of media discourse
and features of the target discourse, the discourse of consumption,
the informal, colloquial language of private life.
This is shown in an intertextual analysis of the text, an analysis
which looks at the text from the perspective of discourse practice,
aiming to unravel the genres and discourses which are articulated
together in it. I shall focus on discourses, in particular how official
discourses of drug trafficking and law enforcement are articulated
with colloquial discourses of drug trafficking and law enforcement,
within a genre of hard news (described below). Compare the article
with a short extract from the source report:

The Government should consider the use of the Royal Navy and the
Royal Air Force for radar, airborne or ship surveillance duties. We
recommend, therefore, that there should be intensified law enforce­
ment against drug traffickers by H.M. Customs, the police, the
security services and possibly the armed forces.

In part, the Sun article draws upon the official discourses which are
illustrated in this extract. This is most obvious where the report and
the Committee chairman are directly quoted, but it is also evident
elsewhere.
What is striking about the text is that these contrasting official and
colloquial discourses are both used within what is traditionally called
'reported speech' - or more precisely, the reporting of the source
written document. Although the direct quotation is marked as
COllUilld iitrCLiy iiUii't lift itCUi 1
report actually said and the Sun's transformation of it into colloquial
discourse is not always clear. For instance, the main headline is in the
form of a direct quotation, though it is not in quotation marks. The
newspaper itself seems to be taking on the prerogative of the Com­
mittee to call for action, though its call is translated into a colloquial
discourse, becomes a demand rather than a recommendation, and
loses the nuances and caution of the original (the Government should
consider the use of becomes call up!).
To show some of this in detail, I now move to linguistic analysis
of the text, though in this case I shall focus upon certain relatively
superficial linguistic features of vocabulary and metaphor. In
accordance with the complex discourse practice and intertextual
relations, this is a relatively heterogeneous text linguistically. For
instance, in the directly quoted sections the article uses the same
term (traffickers) as the report to refer to those dealing in drugs,
whereas elsewhere it uses colloquial terms not found at all in the
report - pushers and pedlars. But even in the parts of the article
 A sample critical discourse analysis 71
where the report is summarized rather than quoted, official discourse
is sometimes used - for instance armed forces, law enforcement, and
security services. Compare forces in the headline with armed forces in the
first (lead) paragraph; the former is an expression from colloquial
discourse, whereas the latter belongs to official discourse.
Why does the article use such pairs of terms? Perhaps because it is
translating official discourse into colloquial discourse and thereby
giving a populist force to official voices, but at the same time preserv­
ing the legitimacy of official discourse. The position and point of view
of the newspaper is contradictory, and that contradiction is regis­
tered here in the heterogeneity of the language. Hall el al. (1978: 61)
refer to a trend in media towards 'the translation of official
viewpoints into a public idiom' which not only 'makes the former
more "available" to the uninitiated' but also 'invests them with
popular force and resonance, naturalizing them within the horizon of
understanding of the various publics'. Notice that use of colloquial
vocabulary in the Sun article has both ideational and interpersonal
functions: it draws upon a particular representation of the social
reality in question, but at the same time the newspaper, by using it,
implicitly claims co-membership, with the audience, of the world of
ordinary life and experience from which it is drawn, and a relation­
ship of solidarity between newspaper and audience. (These implicit
claims are modulated, however, by the use of the vocabulary of
official discourse as well.) Thus this vocabulary is simultaneously
ruricuondi with respect to repu'tseiiicilions, iuciiUties, ancl relations. Ir
is also worth noting how a visual semiotic works together with lan­
guage: it is colloquial and not official discourse that dominates the
visually salient headlines.
Notice also the metaphor of dealing with drug traffickers as
fighting a war. Although the metaphor does occur at one point in the
report,- it is elaborated in the Sun article in ways which are wholly
absent from the report - the mobilization (again using a colloquial
term, call up) of armed forces in the headline, and the representation
of drug trafficking as an invasion in the lead paragraph. The
metaphor is also significant in terms of the newspaper's implicit claim
to a relationship of solidarity and common identity with the audi­
ence. It draws upon war as an evocative theme of popular memory
and popular culture, claiming to share that memory and culture. The
metaphor also links this text intertextually to popular media coverage
of the drugs issue over a long period, where the representation of the
issue as a waragainst traffickers is a standard feature of the discourse.
It is an ideologically potent metaphor, construing drugs in a way
72 Critical analysis of media discourse
which helps to marginalize other constructions from the perspective
of oppositional groups - drugs as a symptom of massive alienation
associated with the effects of capitalist reconstruction, unem­
ployment, inadequate housing, and so forth.

The order of discourse

Turning to the second of the twin perspectives within critical
discourse analysis, what does this example indicate about the order
of discourse? The discourse type is a 'hard-news' story from the
popular press. As a hard-news story, it is different in genre from
other types of article which are in a choice relation within the order of
discourse - soft-news stories, comments and features. It has the
typical generic structure of a hard-news story: a 'nucleus' consisting
of a headline (in fact both a major and a minor one) and a lead para­
graph which gives the gist of the story; a series of 'satellite' para­
graphs which elaborate the story in various directions; and a final
'wrap-up' paragraph which gives a sense of resolution to the story
(Media Literacy Report 1993). in this discourse type within the order
of discourse of the Sun (and other similar tabloid newspapers, though
not the broadsheet newspapers), this genre is standardly articulated
with the combination of official and colloquial discourses 1 have
discussed above. So the discourse type here is a relatively stabilized,
and recognizable, one.
A.,. _ 1 ____ . : ............ ______ _ 1 ............ . - t C .. 7, ' ................... l - i . : ... • . 1 . I.V ..
("M l K.’ i J ' / i l . ' v u ' ? LV M C ittai u a p c x i i_»i C/iLM C- iC iC illU iiO Id? a t e p u u iic.-*

colloquial' nature of the style - indicative of a redrawing of bound­

aries between (external) public and private orders of discourse within
the media order of discourse to produce this hybrid style. One feature
of chain relations which is striking in this case is the way in which the
. source text is transformed into, and embedded in, the article. 1 have
already commented in this regard on the ambivalence of voice, an
ambivalence at times about whether the article is giving the words of
the report or the newspaper's (radically transformed) reformulation
of them. I suspect this ambivalence is common in this discourse type.
It is linked in this case, and more generally, to a mixing of genre - the
combination of the informative hard-news genre with elements of
persuasive genre. Notice in particular that the main headline is an
imperative sentence which, as 1have already indicated, functions as a
demand. In addition to reporting, the Sun article is characteristically
also campaigning for particular policies and actions. Another feature
of chain relations is the way the article is intertextually linked into
another chain which consists of previous coverage of the drugs issue
 A sample critical discourse analysis 73
in the popular media. This sort of chaining is a quite general fea­
ture of media texts.
Let me finally comment, briefly and partially, on the socio­
cultural practice which has framed the stabilization of this sort of
discourse type, summarizing points which I made in the last
chapter. The newspaper is mediating source events in the public
domain to a readership in a private (domestic), domain under
intensely competitive economic conditions. The maximization of
circulation is a constant preoccupation, in a wider economic con­
text in which the accent is upon consumption and consumers and
leisure, and a wider cultural context of detraditionalization and
informalization which are problematizing traditional authority rela­
tions and profoundly changing traditional constructions and con­
ceptions of self-identity. These features of sociocultural context
have shaped, and are constituted in, the complex discourse prac­
tice that 1 have described, and the shift towards that discourse
practice which has taken place over a period of time. The discourse
practice mediates between this unstable sociocultural practice and
heterogeneous texts.
Turning to the politics of this type of article, one important likely
effect of the translation of official sources and official positions into
colloquial discourse is to help legitimize these official sources and
positions with the audience, which in this case means within
sections of the British working class. (Notice, though, that one

articles, to see what the effects actually are in detail.) in the terms I
used earlier, this would seem to be a powerful strategy for sustain­
ing the hegemony of dominant social forces, based upon a hybridi­
zation of practices which gives some legitimacy to both official and
colloquial discourses (though the preservation of the former
alongside the latter perhaps covertly signals their continuing
greater legitimacy, while using the latter as a channel for official
'messages'). At the same time, the newspaper, as I have pointed
out, not only takes on a persuasive role in campaigning for (its
version of) the report's recommendations, but also, through the
war metaphor, helps to sustain and reproduce dominant ideo­
logical representations of the drugs issue.
I have suggested that this example is representative of a relat­
ively stable discourse type, but the restructuring within media
discourse of boundaries between public and private orders of
discourse, and the emergence of various forms of public-colloquial
discourse, are striking features of the modern media which invite
74 Critical analysis of media discourse
historical analysis. What we have here is a creative articulation of
public and private orders of discourse which has become conven­
tionalized. But the picture is rather more complex, in the sense thatin
the context of constant renegotiation of the public/private boundary,
the heterogeneity of texts such as this might under certain circum­
stances be perceived as contradictions, and the relatively stable
discourse type might come to be destabilized.
 5

INTERTEXTUALITY AND THE NEWS

In this chapter I shall elaborate upon what I said in Chapter 4 about

intertextual analysis of media texts, and in particular apply such
analysis to some sample texts. Recall that intertextual analysis is a
bridge between the 'text' and 'discourse practice' dimensions in the
critical discourse analysis framework. It is an analysis of texts from
the perspective of discourse practice, and more specifically from the
perspective of 'discourse process' - in terms of the ways in which
genres and discourses available within the repertoires of orders of
discourse are drawn upon and combined in producing and consum­
ing texts, and the ways in which texts transform and embed other
texts which are in chain relationships with them.
This chapter will be centred around analyses of sample media texts
which focus upon three aspects of intertextual analysis: the analysis
of'discourse representation', of how the speech and writing of others
is embedded within media texts; generic analysis of discourse types,
alternative theories of genre, and analysis of narrative; analysis of
(configurations of) discourses in texts. First, however, I shall develop
what I said in Chapter 4 about 'discourse types'.
76 Intertextuality and the nexos
Discourse types

I introduced the concept of 'discourse type' for the configurations of

genres and discourses which actually occur, and which may become
more or less stable and conventionalized within orders of discourse.
In accordance with the twin perspectives discussed in Chapter 4, the
focus in analysing discourse types may be either on orders of
discourse and the stabilization - and destabilization - of discourse
types within them; or upon communicative events, and particular
configurations of genres and discourses articulated together in parti­
cular communicative events. Discourse types may involve complex
configurations of several genres and several discourses, or may be
closely modelled on single genres and discourses. In addition to
genre and discourse, it is helpful to have other categories available for
the intertextual analysis of discourse types. I find the following
useful: activity type, style, mode, and voice (see Fairclough 1992a for a
somewhat different account). Discourse types also differ in how they
handle intertextual chain relations.
These categories are not, however, all of equal status. A genre is a
way of using language which corresponds to the nature of the social
practice that is being engaged in; a job interview, for instance, is
associated with the special way of using language we call 'interview
genre'. Discourse types are generally most easily characterized in
generic terms. In fact, many analysts would use the term 'genre' in
the way in which i am using discourse type (e.g. Kress and
Threadgold 1988). I am reluctant to do so, because discourse types
often draw upon two or more genres - some types of job interview,
for instance, have developed a discourse type which mixes interview
genre with informal conversation. However, genre is the overarching
category for analysing discourse types, and some discourse types are
closely modelled on single genres.
I have defined a discourse as a particular way of constructing a
particular (domain of) social practice (compare Foucault 1972, Gee
1990). Discourses are relatively independent of genres, in the sense
that, for instance, a technocratic medical discourse might show up
in interviews, lectures, news items or textbooks. There are, never­
theless, compatibilities and incompatibilities between genres (or
genre mixes) and discourses, and one aspect of analysing a
discourse type is to uncover these. It is necessary to specify both
which fields (topics, subject-matters) are associated with a genre
(see Chapter 6), and which discourses are drawn upon to construct
these.
 Discourse types 77
Styles, modes and voices (Bakhtin 1986) are ways of using language
associated with particular relationships between producer and audi­
ence (writer and reader, speaker and listener). Modes are associated
with particular media (spoken or conversational versus written modes).
Voices are the identities of particular individual or collective agents.
One feature of genres or combinations of genres is the styles, modes
and voices associated with them. Finally, the compositional structure of
a discourse type, its organization as a structured sequence of parts
(activities), I shall refer to as its 'activity type' (Levinson 1979). Many
accounts of genre emphasize activity type, and the term genre is often
used in effect to mean activity type. This is, I think, a mistake, because it
easily leads to the other categories I have introduced being ignored.
So far I have been referring to the axis of choice in the intertextual
analysis of discourse types - to (combinations of) genres, discourses,
styles, modes and voices selected from those socially available within
orders of discourse. Discourse types also differ in terms of the types
of intertextual chain relation they enter into with other discourse
types, and ways in which texts are transformed along chains and
embedded within subsequent texts in the chain (Foucault 1972). As I
shall suggest in detail below, for instance, one striking feature of
news discourse is the way in which it weaves together represen­
tations of the speech and writing of complex ranges of voices into a
web which imposes order and interpretation upon them. The
treatment of discourse representation (in the terms of traditional
grammar, 'reported speech ) is quite different in news and, tor
instance, the law, or ordinary conversation.
There are no definitive lists of genres, discourses, or any of the
other categories I have distinguished for analysts to refer to, and no
automatic procedures for deciding what genres etc. are operative in a
given text. Intertextual analysis is an interpretative art which
depends upon the analyst's judgement and experience. Of course,
the analyst does have the evidence of language, and the most satis­
factory intertextual analyses are those where the identification of
genres, discourses and other categories in a text is supported by
features of and distinctions within the language of the text. The
labelling of intertextual categories may, moreover, be at varying
levels of specificity. Different analysts might, for instance, depend­
ing upon their purpose and focus, identify the same text generically
as interview, media interview, news interview, aggressive type of
news interview, and so forth.
I see texts as drawing upon these intertextual categories, rather
than, say, 'containing' or 'instantiating' them. They are models, ideal
78 Intertextuality and the news

types, which are part of people's productive and interpretative

resources. In identifying a category as operative in a text, the analyst
is suggesting that it will be oriented to by those who produce and con­
sume the text. One consequence is that a category may be marked or
evoked by even the most minimal textual cue - a single word, a detail
of oral delivery, a detail of visual appearance in the case of a written or
televisual text. Moreover, particular textual features may be ambi­
valent, or polysemous, with respect to intertextual categories - it may
be unclear, for instance, which genre a word evokes, or it may evoke
more than one. The categories can be thought of as models which are
summations of people's textual experience, and in drawing upon
them people are making intertextual and historical links with prior
texts or text types within their experience.
In the terms of the critical discourse analysis framework introduced
in Chapter 4, the analysis of discourse types cuts across textual
analysis, analysis of discourse practice (processes of text production
and processes of text consumption), and sociocultural analysis. In
addition to my concern here with intertextual analysis, discourse
types require linguistic analysis and analysis of productive and con-
sumptional situations and routines. And sociocultural analysis needs
to address such issues as the relations of power that underlie the
emergence and continuity of particular discourse types, ideological
effects that might be associated with them, ways in which they con­
struct social identities, cultural values that they project, and so forth.
prorn th*? perspective of dis'^onrse practice, p-v^sibilities for cic-
ative reconfiguration of genres and discourses seem unlimited, but
these creative processes are in fact substantially constrained by the
sociocultural practice the discourse is embedded within, and in parti­
cular by relations of power. It is therefore important to combine a
theory of discourse practice which highlights the productivity of
discourse types with a theory of power - as I indicated in Chapter 4 ,1
find the Gramscian theory of hegemony a particularly useful one.
Part of the (cultural) hegemony of a dominant class or group is hege­
mony within the order of discourse - control over the internal
and external economies of discourse types, i.e. over how genres and
discourses are articulated together to constitute discourse types, and
the boundaries and relationships between discourse types within
orders of discourse. In this regard, analysis is needed of the relation­
ships - of complementarity, opposition, resistance, etc. - between
discourse types within and across orders of discourse. My objectives
in this chapter, however, will be more modest. I shall use particular
samples of media discourse to develop what I have said so far about
 Discourse representation in media texts 79
particularities of intertextual chaining associated with discourse
types, analysis of generic aspects of discourse types, and analysis of
discourses and combinations of discourses in discourse types.

Discourse representation in media texts

A very high proportion of media output in news, current affairs and

documentary consists of the mediation of the speech or writing of,
mainly, prominent people in various domains of public life - politi­
cians, police and lawyers, many categories of experts, and so forth.
Sometimes such people speak for themselves - they may write
articles in newspapers, they may be interviewed on radio, or they
may be filmed and interviewed on television. Sometimes their
discourse is represented by newsreaders or reporters. Here is an
extract from a radio news slot that constitutes part of the BBC Radio 4
Today programme, broadcast every weekday morning between 6.30
a.m. and 8.40 a.m. It illustrates how a single news item commonly
weaves together representations of the discourse of a number of
people. Another way of putting it is in terms of 'voices': a complex
web of voices is woven. (Stressed words are italicized.)
newsreader: Libya has told the United Nations that it's willing to let
the two men accused of the Lockerbie bombing come to Scotland to
stand trial. The position was spelt out in New York last night by the
Foreign Minister, Omar Al-Muntasir, when he emerged from a meet­
ing with the Secretary-General, Dr Boutros-Ghali.
omar al-muntasir: The answers we have received from the UK and
the US through the Secretary-General are very acceptable to us and we
see them as a positive e: answer and enough guarantees to secure a
fair . trial for these two suspects once they submit themselves to e:
such jurisdiction.
newsreader: Libyan officials at the UN, faced by the threat of more
sanctions, said they wanted more time to sort out the details of the
handover. Relatives of the 270 people who died on Flight 103 in
December 1988 are treating the statement with caution. From the UN,
our correspondent John Nian.
correspondent: Western diplomats still believe Libya is playing for
time. However on the face of it Libya does appear to be inching closer
to handing over the two suspects. If this initiative is only a delaying
tactic, its aim would be to persuade the waverers on the Security
Council not to vote for the new sanctions, in what is likely to be a close
vote. However the UN Secretary-General is reported to have been
taking a tough line with Libya, demanding that it specify exactly when
80 Intertextuality and the news
the two suspects would be handed over. The Libyan Foreign Minister
has promised a reply on that point later today, but he's asked for more
time to arrange the handover. Meanwhile the West has maintained the
pressure on Libya. The Foreign Secretary Douglas Hurd, and the
American Secretary of State Warren Christopher, have both reiterated
the threat of sanctions. Western diplomats say that unless the two
suspects are handed over immediately, a new resolution will be tabled
tomorrow.

(Today, BBC Radio 4, 30 September 1993).

It is also worth noting how this story appears in the news 'headlines'
at the beginning of the programme - rather a misnomer, because they
are more like lead paragraphs in newspapers than headlines: Libya
has now told the United Nations that it is willing to see the two men accused of
the Lockerbie bombing stand trial in Scotland, but it cannot meet the deadline
to hand them over.
The voices here - those speaking or whose speech is represented -
are: the BBC (differentiated into the newsreader and the BBC UN
correspondent), the Libyans (differentiated into 'Libya', Libyan
officials at the UN, and the Libyan Foreign Minister), 'the West'
(differentiated into western diplomats, the British Foreign Secretary
and the US Secretary of State),, the UN Secretary-General, and an
unspecified reporter (the UN Secretary-General is reported to have been
taking a tough line). Those who actually speak are the newsreader, the
correspondent, and the Libyan Foreign M inister Fn addition, speech
(answers and guarantees) is attributed to 'the UK' and 'the US' by the
Libyan Foreign Minister, though there is no representation of what
was said beyond identification of the types of speech act. Threat in
faced by the threat of more sanctions is similar, though in this case the
voice is not identified - nearer the end of the report the (reiteration of
the) threat is attributed, to the British Foreign Secretary and the US
Secretary of State, though here again only the speech act is identified.
There are also two cases where representations can rationally only be
based upon what has been said, yet they are not formulated as repre­
sentations of discourse: Relatives of the 270 people who died on Flight 103
in December 1988 are treating the statement with caution, and Western dip­
lomats still believe Libya is playing for time. The first is formulated as a
representation of action (treating . . . with caution), the second as a
representation of thought (believe) - what they are doing and
thinking, rather than saying. This gives us another voice, that of the
relatives. (See discussion of the analysis of the 'population' of a text in
Talbot 1990.)
 Discourse representation in media texts 81
An important variable in the representation of discourse is the
degree to which boundaries are maintained between the repre­
senting discourse and the represented discourse - between the voices
of the reporter and the person reported. One way of maintaining
boundaries is to allow people to speak for themselves, as the Libyan
Foreign Minister does. Even if this does not happen, it is possible to
quote what was said directly. In writing, direct quotation is marked
off by quotation marks, and even in speech reporters can use into­
nation to put a quote in vocal 'quotation marks'. Direct quotations
also preserve the original wording, not, for instance, changing the
tense of verbs, the person of pronouns, or 'deictic' words such as this
and here. Compare the direct quotation She said: 7 want you here now'
with the summary She said she wanted them there then. In traditional
grammar, the former is called 'direct speech' and the latter 'indirect
speech'.
In this example, apart from where the BBC correspondent and the
Libyan Foreign Minister speak for themselves, boundaries between
reporting and reported voices are not strongly maintained. The
represented discourse is integrated into the representing discourse,
summarized rather than quoted, using indirect speech in many
cases. One feature of indirect speech is that although it is expected to
be accurate about the propositional content of what was said, it is
ambivalent about the actual words that were used - it may simply
reproduce them, or it may transform and translate them into dis-
i v V n c h fl f r n o r o p p c i l v !- h p r,p p n r [ p r / c y n i r p A .t i in f p r p s H 'C 'f r

example is: Libyan officials at the UN, faced by the threat of more sanctions,
said they wanted more time to sort out the details of the handover. Is the hand­
over the Libyan formulation, or a translation of what the Libyans actu­
ally said into another discourse? We can compare this formulation
with the one used by the Libyan Foreign Minister: once they submit
themselves to such jurisdiction. Other reports suggest that there is in fact
a clash of perspectives and discourses in this political confrontation,
that 'the West' talks in terms of Libya handing over the suspects,
whereas Libya talks in terms of giving them the option of standing
trial in Scotland.
Reports are rarely even-handed with all the various voices repre­
sented. Some are given prominence, and some marginalized. Some
are used to frame others. Some are legitimized by being taken up in
the newsreader's or reporter's voice, others are not. Equity and bal­
ance cannot be assessed by merely noting which voices are repre­
sented, and, for instance, how much space is given to each; the web
of voices is an often subtle ordering and hierarchization of voices. For
 82 Intertextuality and the nexus

instance, hand over is used six times, as a noun or a v e rb n ■

representations of the discourse o f the Libyan offidak m T - f m
Foreign Minister, the UN Secretary-General wo r , S' be Libyan
it is also given the legitimacy o f being used in the Z le T t fT h f T T *
anchor person (who reads the opening headlined m ™ T°day
and the BBC UN correspondent. It is located in + ' h e n ew sreader,
prominent positions in the story, in the headH^ IrdormationaUy
prominent and authoritative summary of the wblcb gives a
end of the story. By contrast, the other formulation^story, and at the
in a backgrounded position (in a subordinate c l a u s T ° 5 ° nCe'
statement by the Libyan Foreign Minister, which is h'n the
minent position in the middle of the report Th<= f eU m anunpro~
newsreader's first sen tence (let the two men & 0naiu]adon in the
valent between these two alternatives • • • s and. trial) is ambi-
Another more general point about positioning in thn
Libyan voices are more prominent in the earlier , 7 ^ ° rtls that
whereas in the second half of the report, from the B B C r r M ^ rep° rt'
dent, the voices of 'the West' and the UN both correspon-
of the Libyan position - are dominant. The last (, P° rtrayed as critical
Meanwhile, w rap up th e re p o rt w ith w e s te rn v o ice s fro m
te n c e s u m m a risin g w h a t is im p licitly a w e ste rn d is m L s a l “ f T ' '
Libyan overture, and containing a threat. i m ssaJ of the
'balance' which are important to creatine- an ; P1 6 aPPearances o f
ivity, it is often easy to divide voices into protagTnL^and ° f ° bjeCt"
(Martin 1980} - m this case, 'the Wprk and r - if S and antago»ists
Sentence connectors (feuroer, meanwhile) and"
are markers of the ordering of voices in the BBC f tm njunctl° n (b»0
report (on these and other cohesive devices see Hnim TeSp° ndent,s
1976). The first and second sentences are linked with 7
^ HaSan
sets up a contrast between what western dink™ r u h° Wever- This ;
doing and what Libya appears to be doing The second 6 le,Ve. Libya is
fences are interesting. The second sentence is the d and third sen- :
voice, not a representation of another voice Re ^orresP°ndent's
are generally authoritative, but this one is rln,,ui P °r J rs statem ents #
of it, appear to be), so there is little conviction expressed M (? t *
actually moving towards a 'handover'. ('Hedges' n thai Ljbya is d
ing dow n what you say to reduce its riskiness e ^ t f ° T ton' 3
less assertive or more polite. See Brown and' LevinsoTw y« I T * 3
fences 2 and 3 are also m a contrastive relationshin m , V/8'' Sen~ i
marker of it, in that there is an implicit shift in seni ° USbthere is no |
voice of the western diplomats in the formulation o U ih ^ ^ ^ %
persuade the waverers on the Security Council not to vote for n lw s ^ '
 Discourse representation in media texts 83

However in sentence 4 sets the 'tough' voice of the UN Secretary-

General against the hypothetical manipulative 'aim' of Libya. Sen­
tence 5 is the only one in the correspondent's report that represents a
Libyan voice, though the but in the sentence implicitly contrasts pos­
itive and negative sides of the Libyan Foreign Minister's response to
the UN Secretary-General - his 'promise' and his request for more
time. Finally, meanwhile draws a line between these diplomatic moves
and what 'the West' is doing, using the latter to frame and to
minimize the former.
Analysis of 'framing' draws attention to how surrounding features
of the reporting discourse can influence the way in which repre­
sented discourse is interpreted. Framing can be blatantly manipula­
tive. For instance, instead of said in Libyan officials . . . said they wanted
more time . . ., we might have had (in certain newspapers, for
instance) claimed or even made out - reporting verbs which question
the truthfulness of what the officials said. This report is not blatant,
however; it is more subtle in its framing. On the face of it, the Libyan
voices are treated equitably, being given the headline and the lead
(opening sentences) in the report as well as the recorded statement by
the Libyan Foreign Minister. In accordance with the ethos of public
service broadcasting, the report is designed as 'objective' and 'bal­
anced'. Nevertheless, framing and the subtle management of audi­
ence interpretation are there.
Consider, for instance, how the Libyan Foreign Minister's
_iicu<aCicik3-
tically, it is preceded by a sentence which both gives the time, place
and situational context of the statement, and formulates what it says
(the position was spelt out - where 'the position' is that Libya is willing
to let the two men go to Scotland to stand trial). But the gist of what
the Foreign Minister actually says is that the UK and US have given
Libya acceptable guarantees that the men would receive a fair trial,
and their going to Scotland is referred to only hypothetically and non-
factively in a backgrounded subordinate clause (with the conjunction
once). The newsreader's framing thus points the audience towards a
misleading interpretation of what was said.
There is further interpretative framing of Libyan voices after the
Libyan Foreign Minister's statement. The statement that Libyan
officials wanted more time is framed by faced by the threat of more
sanctions, which might imply a ploy to deflect more sanctions
(compare the correspondent's statement western diplomats still believe
Libya is playing for time). And the following non-attributed for­
mulation of the position of the relatives of those who died (relatives
84 Intertextuality and the news
. . . are treating the statement with caution) frames the newsreader's
whole representation of the Libyan initiative, and casts doubt upon
it. Yet no evidence is given to justify this authoritative statement
about the views of a large number of people. It is significant that this
unsubstantiated piece of interpretation is positioned after the
carefully substantiated account of the Libyan position: the aura of
objectivity has been established, and interpretation now perhaps
stands a good chance of passing as fact.
Let me mention just one more instance of framing and interpreta­
tion management, the correspondent's sentence about the UN
Secretary-General. The first part of the sentence (before the comma)
is a partly attributed interpretation of what the Secretary-General is
claimed to have said (partly attributed in the sense that source is not
identified), which frames the following 'demand' that Libya specify
exactly when the two suspects would be 'handed over'. The choice of
reporting verb, demanding, of course further frames and reinterprets
what the Secretary-General is reported to have said. Readers are
firmly guided towards an interpretation of Boutros-Ghali's words
which is negative for Libya.
The report from the Sun newspaper discussed in Chapter 4 makes
quite an interesting contrast with the Radio 4 news report. The for­
mer is much more populist than the latter, and there is a much clearer
sense of the source document being translated into a simulated con­
versational discourse - recall the use of (drug) pushers alongside the
document's traffickers, and the elaborated metaphor of a 'battle' to
tight off' the 'invasion' of the drug pushers. Another contrast is that
the Sun shifts overtly from report to campaign - notably in the
headline Call Up Forces in Drug Battle!, which the radio news does not
do. The style of the radio news is much closer to its official sources - in
contrast to the other more magazine-type elements of the Today pro­
gramme, which are populist though in a rather different way from
the Sun.
My main point above has been to suggest that news reports (and the
same is true for instance, in documentaries) include mechanisms for
ordering voices, subjecting them to social control. The mere fact that a
plethora of voices is included in media treatments of social and political
issues does not entail an absence of control, merely that the question of
how voices are woven together, how they are ordered with respect to
each other, becomes decisive. Television adds an important extra
dimension in that the voices are often given bodily forms, movement
and action through film, which means that relationships within a pro­
gramme as a visual experience, and questions of how a variety of
 Generic analysis of discourse types 85
voices may be given a visual closure, need to be addressed. But atten­
tion to the weave of the web is a common concern across press, radio
and television.

Generic analysis of discourse types

My objective in this section is to compare three different conceptions

of genre in terms of their value in analysing discourse types in the
media. Rather than opting entirely for one of these and rejecting the
others, I want to argue that the analysis of discourse types should
incorporate the insights of all three, but that it needs to do so in a way
which avoids the overly rigid view of genre which characterizes the
first of them. The final part of this section is a discussion of narrative.

Genre: schematic view

The first view of genre I shall call the schematic view, in which a genre
is seen as what I am calling an activity type: a schematic structure
made up of stages, either all obligatory or some obligatory and some
optional, which occur in a fixed or partially fixed order. This view of
genre is common amongst linguists, and perhaps the best-known
example is in a paper on analysis of conversational narratives by
Labov and Waletzky (1967). It has also been widely used by systemic
!? | i c f-c sDC o p m j u } 1 w j ,A^ j i c f p t i 1 i a ra m y l jn -n ei o nr t V*. c > c o p Cnp Q r oi m

intense debate about the teaching of genre (see Cope and Kalantzis
1993, Threadgold 1989).
I have already illustrated the schematic view of genre in discussing
the article on drugs from the Sun (Figure 2, page 69) at the end of
Chapter 4. On page 7 2 1referred to the 'generic structure' of the article
as consisting of: Headline + Lead + Satellites + Wrap-up. The order
of these four stages is fixed - the Lead cannot, for instance, precede
the Headline. As the example showed, the Headline can be complex -
the article actually had a major and a minor Headline. It also had a
two-paragraph Lead - the first two paragraphs summarize the gist of
the whole story. There are then eight Satellite paragraphs which
elaborate various aspects of the story. It is characteristic of satellite
paragraphs that there is little progression within them - they could be
quite extensively reordered without disrupting the story. They all
link back to the Lead, but they are relatively independent of one
another. There is, however, some embedding in this case: three of the
paragraphs are direct quotations and obviously group together.
86 Intertextuality and the news
Another detail is that the schema allows for the Satellite paragraphs
to be divided by sub-headlines - Profits in this example. The Wrap-up
(which is the one optional stage in this generic schema) consists of the
final paragraph of the story, which offers some sort of resolution in
the form of future government action. (See Media Literacy Report
1993 for extensive analysis of this sort, and van Dijk 1988a.)
This sort of schematic analysis is good at showing the routine and
formulaic nature of much media output, and alerting us, for instance,
to the way in which the immense diversity of events in the world is
reduced to the often rigid formats of news. But there is a suggestion
of its limitations in the main headline of this article. As I pointed out
in Chapter 4, it is an imperative sentence with the force of demand.
There is, 1 suggested, a multiplicity of purpose here, with the news­
paper campaigning and trying to persuade people as well as give
them information about the Committee's findings. Yet simply
labelling this as the headline stage of the news report genre fails to
capture this.

Generic heterogeneity: sequential and embedded

Van Leeuwen (1987) proposed the second of the three conceptions
of genre I want to discuss, precisely to deal with problems of this
sort. He suggests that the social purposes of journalism, part overt
and part covert, are complex and contradictory - the production of
descriptions which can be seen ns impartial and objective, but also
entertainment, social control, and legitimation. These are social
constraints on journalistic practice which are negotiated through
selecting options within a network of generic strategies, depending
upon the relative salience of these different social purposes on par­
ticular occasions. Accordingly, journalists' stories vary extensively
in their activity-type structures according to van Leeuwen, and
rigid generic schemata cannot account for the actual diversity of
output.
This is one of his examples, taken from the Australian (Sydney)
newspaper Daily Mirror. I have followed van Leeuwen in numbering
the separate clauses (simple sentences) of the article:
1 'When Mum first took me to school
2 I started to cry
3 because I thought I would never see her again.'
4 'But after a few days I really loved school' - Mark, aged six.
5 Mark, now 10, quickly discovered starting school wasn't as scary
as he thought.
 Generic analysis of discourse types 87
6 Mark was one of the many children teacher-turned-author Valerie
Martin spoke to when writing From Home to School, a book dealing
with the first day.
7 'The first day at school can be a happy and a memorable one',
Valerie said.
8 'But the secret is getting ready and preparing now.'
9 Valerie said the main problems for new pupils were separation
from families, meeting large numbers of children they didn't
know and conforming to a classroom situation.
10 Here are some of Valerie's suggestions to help take the hassle out
of the big day.
11 Over the next few days try to get your child used to:
- putting on and taking off clothes
- tying shoe laces
- eating and drinking without help
- using a handkerchief
12 Valerie says it is important your child knows how to:
- use and flush a toilet
- ask for things dearly
- say his or her name and address
- cross a road safely
13 On the first day it is important not to rush children.
14 Valerie says give them plenty of time to get ready, eat breakfast,
and wash and dean their teeth.
15 If possible, get everything ready the night before
16 because children become unsettled if they have to rush.
17 Vinci in tally dun t wony ii you or your child ci'icb , valeric suys.
18 'It won't last long.'

Van Leeuwen describes the staging of the article as follows. The

article begins (clauses 1—5) with Narration, a story about an indi­
vidual child, but then after a transitional sentence (6) shifts into Expo­
sition as Valerie - the expert - explains aspects of starting school
(7-9), but then after another transitional sentence (10) shifts again
into 'Adhortation', urging parents to do certain things (11-18). Actu­
ally the picture is rather more complex, because the Exposition and
Adhortation stages have, according to van Leeuwen, a 'double struc­
ture': as representations of Valerie's discourse they can be called
Report or Description - the Exposition and Adhortation comes from
Valerie not the reporter, though van Leeuwen suggests that these are
purposes of journalism even if they are usually made covert through
being mediated via 'experts'.
One important feature of the analysis is that the stages are differen­
tiated on the basis of linguistic features. Each stage can be characterized
88 Intertextuality and the news
in terms of a bundle of linguistic features, and the bundles change as
we move from stage to stage. For instance, in the Narration clauses
(1-5) the conjunctions are mainly temporal (tohen, after), the tense is
past, the processes are actional (took, started to cry) or mental (thought,
loved), the characters in the story are the themes (initial elements,
most often subjects) of the clauses (Mum, I), and reference is to speci­
fic individuals. By contrast,' in the Exposition clauses (7-9), the con­
junction but is non-temporal, the tense in the represented discourse
clauses in 7 and 8 is present, the main processes are relational (pro­
cesses of being), and the themes of the clauses are topics rather than
people (the first day at school, the secret), and reference is generic rather
than specific (e.g. the first day at school refers to any and all first days at
school). The analysis is quite complex - see van Leeuwen's article for
more detail, and Chapters 6 and 7 for some of the terminology.
A key point about this analysis is that the stages are predictable
parts of a generic schema, but are actually themselves what are
usually thought of as different genres (e.g. Narrative, Exposition).
Van Leeuwen is suggesting that the complex social constraints on
journalism and its multiple purposes commonly manifest themselves
in generically heterogeneous texts - and indeed that heterogeneity is
the norm rather than exceptional. This is, 1 think, an important
insight, and the sort of generic complexity van Leeuwen's method
shows up is something to look out for in analysing media texts. But I
don't think it tells the whole story about generic heterogeneity. Its
limitation is that it deals only with what I havp called elsewhere
(Fairclough i992a: 118) sequential and embedded forms of intertex-
tuality - where different generic types alternate within a text, or
where one is embedded within the other (as Exposition and Adhor-
tation are embedded within Report in the example above). What it
does not account for is mixed intertextuality, where genres are
merged in a more complex and less easily separable way, within
stages of an activity type.

Generic heterogeneity: polyphonic

One example of mixed intertextuality was the extract I discussed in’
Chapter 1 from the BBC Television education programme 'Slippery
When Wet', from a series on engineering called The Works (page 8). I
suggested that the extract was an instance of conversationalization of
public (scientific) discourse, and involved the mixing of scientific;
exposition with features of conversation - a sort of conversational;
modulation of the genre of scientific exposition. It is not possible to
 Generic analysis of discourse types 89
isolate distinct parts of the extract as either exposition or conversa­
tion. I pointed out, for instance, that even where the words as they
appear in the transcription seem to be straightforwardly scientific
exposition, the voice which speaks them, in terms both of accent and
delivery, conversationally modulates the exposition. There are also
instances of conversational vocabulary (e.g. booze) in expositional
clauses. In contrast with the sequential intertextuality described by
van Leeuwen where stages are realized in configurations of con­
gruent linguistic features, in mixed intertextuality we find con­
figurations of non-congruent, contradictory linguistic features. In
addition to this mixture, however, when we bring the visual images
and sound effects into the picture it seems that the genre of scientific
exposition is simultaneously modulated in another direction. Images
and sounds are articulated together in a complex and fast-moving
way which is reminiscent of pop video, and which in broad terms
adds the modulation of entertainment to the mix of scientific expo­
sition and conversation.
This may seem to be a rather exceptional example, and it may give
the impression that mixed genre texts are rather unusual in the
media. I don't think that is so at all. I referred in Chapters 1 and 3 to
conversationalization, the colonization of the discursive practices of
the media by private domain practices, as a pervasive feature of the
contemporary media. Conversationalization is precisely the sort of
modulation I have referred to in this example, productive of mixed
<* ....... r • '. ^ i • «.
iü îiifi' u i iiu irj.l£ A iu d iifÿ
Ti .
v V iacl-f t i l e
.. t -
ic.-u
t * ................. * • ........................................ .. ...t . . . .
i d , ci;:> i i w e i C j i . c m i
17j -
. . ...„I.. .
jy U t.y -

phonic (Bakhtin 1986). For example, various forms of media inter­

view, including political interviews and chat show interviews with,
for instance, show-business celebrities, are now standardly conver­
sationally modulated, to varying degrees. Indeed, the three-way mix
which we have in a rather unusual form in the extract from 'Slippery
When Wet' between more traditional discourse practices in the public
service tradition, elements of conversation, and elements of enter­
tainment, is generally evident in a great deal of media output, as I
suggested in Chapter 1 when identifying two major simultaneous
tendencies in discourse change affecting media: conversational­
ization, and marketization.
This third view of genre focuses less upon activity-type structures
associated with genres than the other views, bringing also questions
of genre-associated styles, modes and voices more into the picture.
For instance, conversational modulation of scientific exposition is
a matter of both style and of mode - a mixing of the language of
private-domain relationships with the language of public-domain
90 Intertextuality and the neivs
relationships, and a mixing of the language of face-to-face interaction
with the language of mass communication. Also, one aspect of con­
versational modulation, the reporter's accent, projects a particular
identity and voice.

It is not my intention to choose between these three views of genre. All

of them contain insights about media discourse. In part this is because
of the diversity of media output: it is possible to find cases which have
the formulaic properties predicted by the schematic view of genre,
cases where diversity of purpose is mapped on to the staging of the
activity-type structure of genre as in van Leeuwen's account, and cases
where there is a more radical deconstruction of genre boundaries, and
emergent new genres such as the radical-television-science genre of
'Slippery When Wet'. In part it is because the three views of genre to
some extent complement one another. For instance, van Leeuwen's
position does not entail that the formulaic structure which has been
attributed to hard-news stories, Headline + Lead + Satellites ( +
Wrap-up) is wrong. The same story can have this formulaic structure
and the sort of sequentially intertextual generic structure van
Leeuwen discusses, they are not mutually exclusive. Van Leeuwen's
position is better seen as a claim tha t such a terribly abstract formula or
schema does not tell us enough about stagingin activity-type structure
- though it does tell us something. Similarly, the second and third
views focus on different aspects of generic heterogeneity - van
Leeuwen's view on heterogen pity in activity-type stagmg, the- thud
view on heterogeneity in the styles, modes and voices associated with
genres. Again, the same text can have both sequential and mixed inter­
textuality. What I am suggesting, then, is that it is helpful to keep all
three views of genre in mind in doing analysis, and consider their
relevance and usefulness for each piece of analysis.

Narrative
A substantial proportion of media output consists of narratives.
According to Swales (1990), narrative is a 'pre-genre': it is broadly
genre-like in being a way of using language associated with a parti­
cular category of purposeful social activity, but it is so pervasive a
way of using language and there are so many distinct types of nar­
rative (each of which can be seen as a genre in its own right) that it
would be misleading to treat it as an ordinary genre.
Journalists themselves talk about 'stories', not only in connection
with news programmes but also, for instance, in documentaries,;
 Generic analysis of discourse types 91
(Silverstone 1985), even applying the term to items which are not
really narratives at ail. One obvious reason why narratives are so
prominent in the media is that the very notion of reporting centrally
involves recounting past events, i.e. telling the story of what hap­
pened, and much of media output consists of or includes reports. But
the social purposes of journalism are, as we saw above, complex;
journalists don't only recount events, they also interpret and explain
them, try to get people to see things and to act in certain ways, and
aim to entertain. The concept of a 'story' suggests this multiplicity of
purpose, in that we normally think of stories as forms of enter­
tainment and diversion, and often fictional rather than factual, in
fact, not all news stories have this character. There may be important
variations linked to class. The 'stories' of news are far more story-like
in this sense in outlets for predominantly working-class audiences:
'stories are for those who, because of their social status and
education, are denied the power of exposition, while exposition is for
those who have been given the right to participate in the debates that
may change society' (van Leeuwen 1987: 199).
Theories of narrative standardly distinguish two facets of a nar­
rative: (a) the actual story, a basic, chronologically ordered series of
events including the participants (or 'actants') involved in them; and
(b) the presentation, the way is which the story is realized and
organized as a particular text (Toolan 1988). Presentation is often
called discourse, but adding another meaning to that term is confus­
ing. The story element of a narrative raises issues of representation
and issues relating to the ideational function of the language,
whereas the presentation element raises issues of identity and of rela­
tions relating to the interpersonal function.
Montgomery (1991) makes use of this distinction in his analysis of
'Our Tune', referred to in Chapter 2 (page 91). The ideological focus
of the stories is the family: a story typically involves a problem arising
out of family life, which is resolved within and through the family.
This gives stories a Complication + Resolution structure, where the
Complication involves some destabilization of family harmony and
the Resolution is restoration of harmony, which may be cyclically
repeated within a story.
One aspect of presentation in narrative is the issue of generic struc­
ture or staging discussed in the previous section. Montgomery uses
and develops the influential work of Labov (1972) on conversational
narrative. This approach stresses that the staging of a narrative
involves not only 'event-line' (story) elements concerned with
recounting events, but also non-event-line elements which relate to
92 Intertextuality and the news
presentation, 'the management of the discursive event as a bounded
whole', or to its reception by the audience. The salience of non-event-
line elements in media narratives is a measure of the degree to which
stories are mediated by presenters. They include Framing elements
which manage the transition between the rest of the programme and
'Our Tune', Focusing elements which give a preliminary indication of
what a narrative is about, and Situating elements which define the
temporal and spatial parameters of the events narrated. Let me
illustrate another element, Orientation. Orientations function on the
one hand to orient the audience behind the experience of a character,
to generate empathy towards the character. For example:
so you can imagine
not only has she tried to top herself and got herself taken to hospital
but now as she's recovering from that she's had the biggest blow or one
of the biggest blows you can have

Or, on the other hand, they function to anticipate audience reaction:

and one night
you guessed it
she took half a bottle of pills
Other elements include Evaluations, which take the form of general
maxims on the model of 'these things do happen'; and Codas, which
make the transition from (the time frame of) the narrative to the
present. Some of tfmsa Nements are positional!}' rigid (e.g. Framing
comes at the beginning and the end of a narrative), others (e.g. Ori­
entation or Evaluation) are positionally quite flexible.
Other aspects of narrative presentation discussed by Montgomery
are to do with the relationship between Simon Bates and the writer of
the letter (the Broadcast Narrator and the Epistolary Narrator), and
Simon Bates and the audience. There is a contradiction and a tension
between the genres of confessional personal letter that is the material
for 'Our Tune' and public narrative within the programme, which
manifests itself, for instance, in the handling of judgements of the ^
Epistolary Narrator's behaviour. Self-assessments in a confessional
mode (e.g. I loent haywire) cannot be directly transformed into public
narrative without acquiring a strong condemnatory force. Hence the
frequency of hedges which tone down reported self-assessments (e.g.
1guess and a little in so I guess Marianne went a little haywire). The relation­
ship between Simon Bates and the audience makes sense of one inter-
esting linguistic feature of 'Our Tune', the frequency of interpolations- :
which reformulate an immediately preceding formulation, such as the.
 Generic analysis of discourse types 93
elder sister in er the sister the elder sister became the person who looked after
everybody. Such interpolations show an ongoing orientation by Bates to
possible interpretative problems or misunderstandings by the audi­
ence. Along with the extensive non-event-line elements in the generic
staging, they are indicative of the extent to which the Epistolary
Narrator's tale is mediated by Bates. The interpolations also have an
identity function, providing 'a repetitious signalling of the DJ's role as
"honest broker" of the story materials' (Montgomery 1991: 164).
A further feature of narrative presentation is the extent to which
the material dealt with is presented as scandalous or risky and
needing to be negotiated with delicacy. Montgomery suggests that
this tendency is indicative of the influence upon 'Our Tune' of
fictional narrative models, creating a tension between pressure to
fictionalize and sensationalize the narratives, and pressure to estab­
lish their authenticity and truth. A tension, in broader terms,
between entertainment and information. Recall in this connection
the extract from the This Week programme 'Vigilante!' discussed in
Chapter 1. 1 suggested that in that case the tension manifested itself
in a contradiction between the dramatic and sensational represen­
tation of events in the film, and the more cautious, explanatory and
expositional representation of events in the linguistic commentary.
If the analysis of narrative presentation draws attention to how nar­
ratives are fictionalized in response to pressures to entertain, it also
draws attention to factuality as a property of narratives which is discur­
sively achieved. ri here is a range or devices wiiiiin die rhetoric of
factuality which are standardly drawn upon in the production of, for
instance, news stories, involving visual and aural semiotics as well as
language, including the layout of the newsroom, the opening
sequence and theme music of the news programme, the appearance of
the newsreader. One objective here has to be the creating of a sense of
authority, though even in news that may come into conflict with the
pressure to entertain. (For example, at the time of writing, there is
some discussion in the British press about whether changes to the for­
mat of the prestigious ITN News at Ten, which includes a dramatic
opening sequence and a 'space-age' newsroom, have undermined the
authority of the well-known newsreader Trevor Macdonald.) Within
the language, the attribution of news statements to authoritative
sources is a key part of the rhetoric of factuality, profoundly affecting
the structuring of news texts with respect to the construction of com­
plex embedding relationships between voices (interviews, reports,
film sequences, and, of course, discourse representation). The section
on discourse representation above provided some illustration.
 94 Intertextuality and the news
iJ|
Another important area here is modality, and in particular the
If prevalence of various linguistic realizations of categorical modalities
hi which make strong truth claims. Recall the discussion of the Panorama
\ extract in Chapter 1. Also important is presupposition: as I show in
Chapter 6, presupposition helps to authenticate the new by locating
\' it within a matrix of purportedly given (presupposed) information.

Analysis of discourses in texts

Discourses are, as I have indicated, constructions or significations of

some domain of social practice from a particular perspective. It is
useful to identify discourses with names which specify both domain
and perspective - for instance, one might contrast a Marxist political
discourse with a liberal political discourse, or a progressive educa­
tional discourse with a conservative educational discourse. I shall
illustrate the analysis of texts in terms of discourses using press cover­
age of an air attack on Iraqby the USA, Britain and France on 13 January
1993 (two years after the Gulf War), referring to 14 January editions of
five British newspapers: the Daily Mirror, the Sun, the Daily Mail, the
Daily Telegraph and the Guardian, and the Guardian Weekly for the week
ending 24 January 1993. I focus upon two issues: the 'congruent' as
opposed to 'metaphorical' selection of discourses for formula ting who
did what to whom and why within the attack; and the role of confi­
gurations of discourses in the construction of these events.
The distinction between congruent and metaphorical discourses is
the extension of a terminology used by Halliday (1985). A congruent
application is the use of a discourse to signify those sorts of experi­
ence it most usually signifies; a metaphorical application is the exten­
sion of a discourse to signify a sort of experience other than that
which it most usually signifies. The distinction is a rough one, but a
useful one. Metaphorical applications of discourses are socially
motivated, different metaphors may correspond to different interests
and perspectives, and may have different ideological loadings. The
following examples (headlines and lead paragraphs) illustrate how
congruent and metaphorical discourses combine in the coverage of
the attack.
Spank You And Goodnight
Bombers Humble Saddam in 30 Minutes
More than 100 Allied jets yesterday gave tyrant Saddam Hussein a
spanking - blasting missile sites in a raid that took just 30 minutes.
(Sun)
 Analysis of discourses in texts 95
Saddam's UN Envoy Promises Good Behaviour After Raid by US, British and
French Aircraft
Gulf Allies Attack Iraqi Missiles
More than 100 aircraft blasted Iraqi missile sites last night after the
allies' patience with Saddam Hussein's defiance finally snapped.
(Daily Telegraph)
In the examples I looked at, it is a discourse of military attack
that is congruently applied (e.g. jets or aircraft blasting missile sites
and Gulf Allies Attack Iraqi Missiles in the examples above). Not sur­
prisingly, we find such formulations in all the reports, along with
expressions like 'retaliate' and 'hit back' (e.g. Iraq To Hit Back, in
the Sun) which represent these events as a contest between two
military powers. But there are distinctions to be drawn. Whereas
the Guardian, the Daily Mail and the Daily Telegraph use what one
might call an 'official' discourse of military attack - that is, they use
tire sort of language that might be used in official and military'
accounts - the Sun and the Daily Mirror (and exceptionally the
Daily Telegraph in the example above: more than 100 aircraft blasted
Iraqi missile sites) use a fictional discourse of military attack, the
discourse of stories about war (whether purely fictional, or fiction­
alized versions of fact), which highlights physical violence. The
Daily Mirror is particularly rich in expressions for processes of
attack which link to this discourse: blitz, blast, hammer, pound, blaze
into action, (warplanes) scream in. While the attacks are mainly for­
mulated as acuon by aircraft or 'the allies against Iraq' or specific
targets (e.g. 'missile sites' or 'control centres'), both the Daily Mir­
ror and the Sun also formulate them in a personalized way as
directed at Saddam Hussein (The Gulf allies struck hard at Saddam
Hussein, 'Spot Raids Give Saddam Pasting', allied warplanes have bombed
the hell out of Saddam Hussein).
The main headline and lead paragraph from the Sun above show
that formulations of the attack do not by any means draw only upon
military discourse: Spank You And Goodnight (notice the play on
, -'Thank you and goodnight' which makes a joke even of this serious
event) and More than 100 Allied jets . . . gave tyrant Saddam Hussein a
spanking. This is a metaphorical application of an authoritarian
discourse of family discipline which is a prominent element in repre­
sentations of the attack - Saddam as the naughty child punished by
his exasperated parents. The Guardian editorial sums it up as an act of
punishment against a very bad boy who thumbed his nose several times too
- often - also notice the allies' patience . . . finally snapped and good
behaviour in the Daily Telegraph example, both consistent with this
 96 Intertextuality and the neios Analysis of discourses in texts 97
disciplinary discourse. (One might also be tempted to read 'spank­ Humble Saddam and Retribution in the Gulf. I think these can be read in
ing' in terms of a discourse of sexual 'correction'.) The attack is for­ terms of a (Christian) religious discourse, though 'retribution' also
mulated several times in the reports as 'teaching Saddam a lesson' evokes a legal discourse. Another discourse which features only once
(for instance, The allies launched 114 war planes to teach defiant Saddam a here - attributed to a Whitehall official by the Daily Mirror - but was
lesson in the Daily Mirror, and Let's hope he's learnt his lesson, attributed quite common in coverage of hostilities in the Gulf at the time, is a
to a US official in the Sun). This is again consistent with the discourse discourse of communication exchange, of 'signals' sent through mili­
of family discipline, or disciplinary discourses more generally. So too tary actions. (In the words of the Whitehall official, If Saddam does not
with Toe The Line O r. . . We'll Be Back!, the main page-one headline in get this message . . . he knows there will be more to come.) Again, both the
the Daily Mirror. Such conditional threats ('do x or we'll do y', 'if you Sun and the Daily Mirror draw upon a discourse of dangerous-animal
do x - or don't do x - we'll do y') occur several times in the report. control in their editorials: the air strikes are intended to 'curb'
A related but rather more specific metaphorical discourse that is Saddam, and if he doesn't learn this time, he will have to be put down for
evoked is that of the disciplining of young offenders, juveniles found good like the mad dog he is (Daily Mirror), the tragedy is that zoe did not
guilty of crimes (with the focus on crimes of particular sorts, such as ’■ finish him off last time (the Sun).
'joy-riding'). A British government official is quoted in most reports An important distinction within a report, which takes us back to
as saying that the attack was a short, sharp and telling lesson for ( the discussion of discourse representation earlier in this chapter, is
Saddam. This evokes the expression used by the British Conservative ( between discourses which occur in represented discourse attributed
government in the 1980s, when it tried to develop the policy of ; to the Voices' of others in quotations or summaries, as opposed to
delivering a 'short, sharp shock' (in the form of incarceration in discourses which are unattributed and are drawn upon by the 'voice'
highly disciplined quasi-military institutions) to juvenile offenders. ) of the report itself. However, a key question (which requires histori­
The same group of discourses is indicated in reasons given for the % cal research and research on production processes) is where the
attack. The headline for a report on pages 2-3 of the Sun is He Had It s discourses of reporters come from. By comparing attributed and
Coming, and the lead paragraph refers to the pasting that Saddam % unattributed formulations within and across reports, one can often
Hussein has been asking for. According to the Daily Mirror, Saddam * see the same discourses being drawn upon by reporters and official
had pushed his luck too far. These formulations evoke a conversational § sources. For example, the discourse of correction (in Spank You And
or 'lifeworld' version of an authority-based discourse of discipline, f Goodnight, also in the Dnihi Mirror inside-nacre headline A Span kin $
referring to what is elsewhere frequently formulated in the reports as f Not A Beating and in the headline of the Guardian editorial More A
the 'provocations' of the subordinate party in this disciplinary : Smack Than A Strike) may have originated in a statement by a US
relationship, i.e. Saddam (note also formulations such as Saddam | official: It's just a spanking for Saddam, not a real beating. Similarly the
'goading' or 'taunting the West'). Disciplinary formulations such as Sun headline He Had It Coming and more generally 'teach-Saddam-a-
'provocations' alternate with legalistic formulations such as ' lesson' formulations apparently echo official sources - the Sun quotes
'breaches' and 'infringements' (of the UN ceasefire conditions). f a White House statement: Saddam had this coining. Let's hope he has
The metaphorical application of such discourses is a very prom- -1 learnt his lesson.
inent feature of these reports, and in assessing that application one Official influence upon media formulations is built up over the
might wonder whether such a disciplinary relationship applies or . longer term rather than just on a day-by-day basis; 'teach-Saddam-a-
ought to apply in relations between nations, or indeed whether the - lesson' formulations had been widely used officially and by the press
relations between nations ought to be personalized as they consisten- • for a period before the attack, and similarly official sources including
tly are here: the target of discipline is Saddam, not Iraq or the Iraqi t President Bush had spoken of 'patience running out' in the weeks
government - whereas its source is mainly 'the (Gulf) allies' or 'the before the attack. (Also, this relatively minor Iraq crisis was intertex-
West', and rarely George Bush (the American president at the time). lually linked to earlier ones including the Gulf War, and fed discours-
Other discourses are metaphorically applied, though they are less || ally from them. See the discussion of 'discourse-historical' method of
prominent in the reports. One is evident in the Sun headline Bombers f -. analysis in Wodak 1990.) The Guardian, the Daily Telegraph and the
Humble Saddam in 30 Minutes as well as the Daily Mail headlines Allie Daily Mail are more likely generally to use such formulations only as
98 Intertextuality and the news
attributions. There is a constant flow between official sources and the
media: the latter may take up the discourses of the former, but the
former also design their statements and press releases to harmonize
with discourses favoured by the media. Bruck (1989) points out that
the influence of official discourses on media discourse depends upon
the discourse type - it is, for instance, likely to be greater in news
reports than in editorials or features.
While the discourses and specific formulations of certain favoured
sources are massively present and foregrounded, those of o th er- and
especially oppositional - sources are either omitted altogether from
some reports, or backgrounded. For example, the Labour Party left­
winger Tony Benn described the attack as the last piece of gunboat diplo­
macy of a lame-duck US president according to the Guardian, but that
was the only report of Benn's comment, and it was backgrounded
(positioned in a single paragraph in the middle of a report in the
bottom left-hand corner of a centre page). Formulations of the attack
attributed to the Kuwaiti government, which draw upon a discourse
of disease and surgical intervention (bursting the abscess of the Baghdad
government according to the Sun, removing the Iraqi cancer according
to the Daily Telegraph), were quite widely reported though back­
grounded. By contrast, only the Guardian reports formulations of
Saddam's actions prior to the attack as 'acrobatics' and 'mere
fireworks' (from the newspaper al-Ahram, Cairo), and 'clownish'
behaviour (al-Thaurah, Damascus), constructing Saddam in the less
threatening role of a clown/nerfnrmer (downs don't generally merit
bombing). Significantly, the same reports highlight the 'double
standards' applied by the West, in not reacting as vigorously to the
plight of Muslims under attack by Serbians in Bosnia, or of the 400
Palestinians extradited by Israel and isolated in No Man's Land
between Israel and Lebanon at that time, in defiance of a United
Nations resolution. Why no air attacks on Israel?
If selection between alternative congruent and metaphorical
discourses is one issue, another is configurations of discourses, how
discourses are articulated together within discourse types. Bruck
(1989), for instance, suggests that five main discourses were drawn
upon by the Canadian media in the mid-1980s in their coverage of dis­
armament, peace and security issues, which he calls: the discourse of
state leaders, bureaucratic-technical discourse, scientific-technical
discourse, the discourse of victims, and the discourse of survival . The
first three are dominant discourses, the last two oppositional dis­
courses. The analysis of news output is concerned with both the
selections made between these discourses, and the ways in which
 Analysis of discourses in texts 99
they are articulated together, which between them allow the analyst
to describe the range of discursive practice in the coverage of these
issues.
The following report was inset in a double-page spread in the Daily
Mirror of 14 January, dealing with the attack in Iraq. (Major reports
are often made up in this way of combinations of smaller reports, and
the relationship between articles on a page in such cases is worth
attending to.)
The Mother Of All Rantings
Evil dictator Saddam Hussein promised Iraq last night they were
winning a new great victory - just like they had in the 'Mother of All
Battles' in Kuwait. His pledge came three hours after allied aircraft
pounded his missile sites.
In a ranting, confused speech, he told his nation on television that a
new jihad - holy war - had begun. He urged the Iraqis to fight 'in the
name of God' . . . and he promised them they would humiliate the
allies.
Saddam called the allies 'the infidels' and said they were 'under the
influence of Satan'. And he raged: 'Every aeroplane of the aggressors
in the Iraq sky shall be a target for us and we shall fight in the name of
God and down their aircraft. The aggressors will be defeated.'
Reading stiffly from hand-held notes, he said: 'The criminals have
come back. But tonight they came back without any cover, not even a
transparent one.
'They came back for the purpose they never spoke about the first
time in their evii aggression, namely to impose colonialism.'

This report includes a new and clearly oppositional configuration of

discourses for formulating the attack, attributed to Saddam himself:
an Islamic religious discourse (infidels, under the influence of Satan), and
political discourses of aggression and colonialism. The reference to
the absence of any 'cover' obliquely cues also a legal discourse - the
attacks were condemned as 'illegal' by those who opposed them.
Notice also that the allies are referred to here as criminals.
However, this oppositional configuration of discourses is framed
within a larger configuration by the dominant discourses I have
discussed above. Saddam's speech is firstly formulated and summed
up in the headline in an ironic play upon his own (in)famous descrip­
tion of the Gulf War as 'the mother of all battles', with rantings
evoking discourses of madness and political fanaticism. In the lead
paragraph, Saddam is referred to as an evil dictator, deploying the
religious and political discourses I referred to earlier as part of
the anti-Saddam armoury. The summary of Saddam's speech in the
WO Intertextuality and the news
second paragraph is framed by the initial thematized phrase In a
ranting, confused speech, and similar framing devices are used where
Saddam is directly quoted - notice the choice of raged as a reporting
verb, and reading stiffly. In the first two of these cases there is again an
evocation of the discourse of madness. The net effect of the framing
of Saddam's oppositional discourses with the dominant ones is to
undermine and ridicule the former.
Diverse discourses are articulated together in the naming and
identification of both the protaganists and the antagonists, though to
quite different effects. The identification of the protaganists caused
some difficulty in that the USA, Britain and France were claiming to
act to enforce a United Nations resolution, but neither the 'no-fly'
zone they had imposed on southern Iraq nor the attack had been
endorsed by the UN. The attackers are referred to in the reports as
'the Gulf allies', 'the West', and most frequently'the allies'. 'TheGulf
allies' is problematic in that the alliance which fought Iraq in the Gulf(
War was actually divided on this later attack, and none of the Arabj
members of the alliance was involved. 'The West' is problematic
because a number of members of 'the West' were also critical. 'The
allies', with its reassuring evocation of the Second World War, seems
to have been the least problematic label. The Guardian also refers
'correctly' to 'the United States, Britain and France'. A number of
other identifications were used elsewhere: in the Guardian Weekly 'the
coalition' and 'the US and its allies' were used, and President Clinton
was quoted ns supporting the international community's actions. The .
variety of these formulations, the range of discourses they draw
upon, and the instability of naming practices here, are indicative of
the difficulty in constructing an identity for the protaganists.
By contrast, the considerable range of expressions used to refer to
Saddam Hussein shows a number of discourses working together to
discredit him, as in the following editorial from the Sun on 14 January:
Wipe Out The Mad Menace
At long last, Allied warplanes have bombed the hell out of Saddam
Hussein.
The Iraqi madman has pushed the West too far.
He has played a dangerous game and now he must pay the price.
Four times Saddam has sent raiding parties over the border into
Kuwait.
M en ace
His boast that Iraq planned to 'recover' Kuwait was the last straw.
The tinpot tyrant could not be allowed to cling onto power a moment
longer.
 Analysis of discourses in texts 101
He is an international terrorist, a constant menace to peace.
The tragedy is that we did not finish him off last time.
Go get him, boys!

This is discoursal overkill: a remarkable range of discourses are ar­

ticulated together in the verbal annihilation of Saddam Hussein. The
density of the assembled discourses is no doubt attributable to the
fact that the genre is editorial rather than news report: this is an
apologia for the attack, based upon a thorough discrediting of
Saddam Hussein. He is referred to as a madman, a menace to peace, a
tyrant, a terrorist, a blusterer (cf. his boast), and a figure of ridicule (the
implication of tinpot), yet at the same time a calculating politician
(who has pushed the West too far, and played a dangerous game), and
actions against him are formulated in terms of discourses of legal
retribution (he must pay the price), war fiction (bombed the hell out of) and
even westerns (wipe out, finish off, go get him, boys). We find the range
of discourses extending further elsewhere - he is referred to, for
instance, in the terms of religious and ethical discourses as 'evil' and a
'coward'.
A configuration of discourses is put to different effect in the editor­
ial in the Guardian Weekly, where evaluation of competing discourses
is itself a topic. The editorial is a critique of the attack, under the
deadline What Signal Will He Read< The editorial is a dialogue with
opposing positions represented by different discourses. Thus it
r-p fp re bn p ’h H r| ic h in r f» s jf-e p ff f r n r n — hrnri}vo$h.'yhi tfjfk o f T o n c h o io
Saddam a Lesson, and attributes the discourse of delivering a signal to
Saddam to what it calls the tough-minded (this discourse generates the
expression 'coercive bombing' in another report in the same edition).
It does, however, in its own voice draw upon some of the dominant
discourses for formulating Saddam and his actions (he is evilly brutal
Saddam, with a record of provocation - though perhaps deliberate). The
editorial also formulates the attack, tentatively, in a different
discourse: Mr Bush's likely desire to settle accounts before leaving office.

Other terms which are roughly equivalent to 'discourses', but derive

from different theoretical frameworks and traditions, are quite
widely used, including schemata, frames, and scripts (from cognitive
psychology), metaphors, and vocabularies. I have discussed meta­
phorical applications of discourses, and for the most part the dis­
courses I have referred to are realized in the vocabulary of texts.
Aspects of grammar may also be involved in the realization of
discourses. For instance, 1 noted earlier that conditional threats (e.g.
102 Intertextuality and the news
Toe The Line . . . Or We'll Be Back) are a feature of disciplinary
discourse, and these are realized in particular syntactic constructions
(in this example, imperative clause + or + declarative clause).
Analysis of collocations in texts (patterns o f co-occurrence between
words) is a way of linking analysis of discourses to the linguistic
analysis of texts (Sinclair 1992). Configurations of discourses identi­
fied in the analysis of discourses may be realized in - condensed into
- collocational relations in phrases or clauses (Fairclough 1991).
Collocations are often a good place to look for contradictions in texts.
For example, in the editorial from the Sun above, the following collo­
cations occur: mad menace, tinpot tyrant, the Iraqi madman has pushed the
West too far. Mad evokes the discourse of madness whereas menace
evokes the discourse of political extremism, and the collocation
bonds the two discourses together in a detail of the text. Similarly, the
Iraqi madman has pushed the West too far compacts together the
discourse of madness and the discourse of political calculation.
Both selections amongst available discourses and selection of parti­
cular ways of articulating them together are likely to be ideologically
significant choices. There may, for instance, be various ways of
rationalizing the decision to construct relations between 'the West'
and a 'Third World' country like Iraq as relations between a teacher
and a recalcitrant child, but such a construction implicitly evokes an
imperialist and indeed racist ideology of relations between nations,
which contributes to the continuity of imperialist and neo-colonialist
relations in practice. Of course, in accordance with what Tsaid about
ideology in Chapter 3, one cannot assume ideological effects conse­
quent upon selections of discourses, merely that the question of
potential ideological effects is always worth raising.
 6

REPRESENTATIONS IN DOCUMENTARY
AND NEWS

Tn tÇ’THS o f 0f r - f nr o f hIDcHd dT5COUTÇ0 ID vTOd.UCf^d

in Chapter 4, the focus now shifts from discourse practice and inter-
textual analysis of texts - the concern of Chapter 5 - to linguistic
analysis of texts. As I indicated in the section on texts in Chapter 4, I
am working with a broad and enhanced understanding of linguistic
analysis which includes, for instance, analysis of relationships
between sentences, or of relationships between speaking turns
within a dialogue. This chapter will deal with one of the three pro­
cesses which I have argued are always simultaneously going on in a
text: representation. The other two, construction of relations, and
construction of identities, are discussed in Chapter 7. This chapter
Will accordingly be mainly concerned with the ideational and textual
functions of language in texts, and Chapter 7 with the interpersonal
function.
The focus, then, is upon how events, situations, relationships,
people, and so forth are represented in texts. A basic assumption is
that media texts do not merely 'mirror realities' as is sometimes
naively assumed; they constitute versions of reality in ways which
104 Representations in documentary and news
depend on the social positions and interests and objectives of those
who produce them. They do so through choices which are made at
various levels in the process of producing texts. The analysis of repre­
sentational processes in a text, therefore, comes down to an account
of what choices are made - what is included and what is excluded,
what is made explicit or left implicit, what is foregrounded and what
is backgrounded, what is thematized and what is unthematized,
what process types and categories are drawn upon to represent
events, and so on. Questions about the social motivations for parti­
cular choices, and about ideologies and relations of domination, are a
constant concern in the analyses of such choices in this chapter. A
longer-term goal in analysis of representation is description of net­
works of available options from which such choices are made. See
Halliday (1985) for an account of the systemic grammatical
framework which 1 draw upon here.
There are two major aspects of representation in texts, discussed in
turn in this chapter. In a logical terminology, the first has to do with
the structuring of propositions, the second with the combination and
sequencing of propositions. The first is concerned with how events
and relationships and situations are represented. Actually, since the
analysis here is linguistic rather than logical, we shall be looking at
this question with respect to the clause, a term linguists use for a
grammatically simple sentence. Clauses roughly correspond to pro­
positions- in many cases, a clause will consist of a single proposition.
So the question will be, how is this clause structured in terms of the
process (typically realized in its verb), participant (typically realized
in its nouns and nominal groups), and circumstantial (typically
realized in adverbials) elements it contains? What choices have been
made from among the possible types of process, participant and cir­
cumstance? For instance, what appears in one text as a relational pro­
cess without an agent (life gets harder) might appear in another text as
a causative process with an agent (the profiteers are making life harder).
The question is, what motivates one set of choices over another?
Turning to the second aspect of representation in texts, concerned
with the combination and sequencing of propositions, again, since the
analysis is linguistic rather than logical, we shall be looking at this in
terms of the combination and sequencing of clauses. We can broadly
contrast two levels at which choices are available here (van Dijk 1988a):1

1. Local coherence relations between clauses. The initial question is:

(a) how are clauses combined together in to the complexes of clauses
that are generally referred to as sentences? Actually, 'clause
 Representations in documentary and news 105
complex' is a better term tor spoken language, because it is often
difficult to find units which correspond to the sentences of writ­
ten language (Halliday 1989). Two further questions within local
coherence relations are:
(b) what relations of cohesion are set up between such complexes of
clauses (sentences - see Halliday and Hasan 1976)? and
(c) what forms of argumentation are used within different texts?
2. Global text structure. This involves the sort of analysis of genre
and activity type that 1 introduced in Chapter 5. The question here
is what choices are made between alternative available activity
types or generic schemata in a given text.

At both levels, questions of sequencing - what precedes or foiiows

what, and why - are a central concern, and for that reason 1 also
include in this part of the treatment of representation the thematic
(and more generally informational) structuring of clauses. A major
concern here is what element of a clause is thematized, placed in ini­
tial position, and why.
This chapter links with Chapter 5 not only in the analyses of
activity type and generic structure, but also, importantly, with the
analysis of discourses. The analysis of discourses shares with the
analyses of representation in this chapter a common concern with
choices that are made in texts in the representation, signification and
construction of reality, and social motivations for these choices. The

resource both for spelling out how properties of and differences

between discourses are realized in the language of the text (in
vocabulary, metaphor, grammar), and for specifying the representa­
tional options which are available within a particular discourse.
My concern is with the analysis of representation in particular
texts, but for certain purposes within media research it will make
sense to combine such 'microanaiysis' with other forms of text-
oriented analysis. If, for example, one wishes to analyse the media
coverage of a particular issue such as a war, microanaiysis alone will
not give the necessary overview. What may also be needed is some
form of content analysis which allows, for instance, a generalized
comparison of how 'enemy' forces and 'our' forces are represented
(see, e.g., Herman and Chomsky 1988, Kellner 1992 on the Vietnam
War and the Gulf War). I would, however, argue that close textual
analysis is a valuable complement to content analysis in such cases.
Before engaging in analysis of what is in the text, however, one
needs to attend to the question of what is excluded from it.
 106 Representations in documentary and news
II
Hi!
Presences and absences in texts: presupposition

Unsurprisingly, analysis of representation is mainly analysis of what

is 'there' in the text. But it is also important to be sensitive to absences
from the text, to things which might have been 'there', but aren't - or,
and this really comes down to the same thing, to things which are
present in some texts appertaining to a given area of social practice,
but not in others. For instance, one revealing comparison within the
collection of reports about air attacks on Iraq by the USA and its allies
discussed in Chapter 5 is between texts which include the topic of
civilian casualties and texts which exclude it. Another more specific
example of significant absence is from a speech by the British Prime
Minister, John Major (Carlton Club, London, 3 February 1993), in
which he said, 'I increasingly wonder whether paying unem­
ployment benefit, without offering or requiring any activity in return,
serves unemployed people or society well.' The absence here is the
word 'Workfare' - the name of the American scheme which requires
unemployed people to carry out usually menial work in return for
social security benefits. At the time, Workfare was arousing impas­
sioned responses and a great deal of hostility, and Major seems to
have been 'testing the water' without wanting to clearly position
himself as a target for that hostility. Much of the media coverage
treated the absence as a matter of note, but the government was still
left with room for manoeuvre and the capacity to hedge over whether
Workfare was really what Major had in mind. A more general absence
which coirunenldtors nave noted is an absence of historical context in
most news stories; news is standardly constructed in terms of events
which are treated as more or less isolated from prior or subsequent
events - isolated from history (Herman and Chomsky 1988, Pilger
1992).
Actually, it makes sense to differentiate degrees of presence, as it
were, rather than just contrasting what is present and what is absent.
We might think in terms of a scale of presence, running from 'absent'
to 'foregrounded': absent - presupposed - backgrounded - fore­
grounded. If something is presupposed, it is in a sense present in the
text, but as part of its implicit meaning. If something is explicitly
present in a text, it may be informationally backgrounded, or
informationally foregrounded. The distinction between background­
ing and foregrounding will be discussed in the section dealing with
combination and sequencing of clauses.
Any text is a combination of explicit meanings - what is actually
'said' - and implicit meanings - what is left 'unsaid' but taken as
 Presences and absences in texts: presupposition 107
given, as presupposed. Presuppositions anchor the new in the old,
the unknown in the known, the contentious in the commonsensical.
A text's presuppositions are important in the way in which it posi­
tions its readers or viewers or listeners: how a text positions you is
very much a matter of the common-sense assumptions it attributes to
you. The presuppositions of a text are part of its intertextuality (and
in that sense belong with the concerns of Chapter 5): presupposing
something is tantamount to assuming that there are other texts
(which may or may not actually exist) that are common ground for
oneself and one's readers, in which what is now presupposed is
explicitly present, part of the 'said' (Fairclough 1992a). Presuppo­
sitions are what French discourse analysts call 'preconstructed'
elements in a text, elements which have been constructed elsewhere
in other texts (Pecheux 1982, Williams forthcoming).
In various types of reports and narratives, presuppositions help
establish represented realities as convincing. For instance, the
opening parts of a documentary need to establish for the audience a
reality, a world, which carries conviction as authentic. The following
extract occurs near the beginning of a documentary in the Channel 4
Critical Eye series, broadcast on 15 October 1992. Called 'Wind of
Memory', the programme documents the genocide of the Indian
population of Guatemala.
reporter (voice-over): Santiago de Plan. The Sutowilas Indians call
this place (Indian language term) the heart of the world. Perhaps that's
because the ancient traditions are still so strong in this village. Maybe
it's also because Santiago de Plan is the sanctuary of Machimon.
Part of the authenticity in this case is achieved through film, music
and sound effect, and through the voice of the reporter (which
sounds authentic as Guatemalan). But achieving authenticity is also
partly a matter of positioning the viewer through presupposition as
someone who is already familiar with the culture and community
depicted. The effect of the definite article in the Sutowilas Indians is to
presuppose the existence of this group of Indians, that is, take it as
given knowledge for the audience. The propositions of the two
because-clauses are also presupposed: it's taken for granted that the
ancient traditions are still strong, and that Santiago de Plan is the
sanctuary of Machimon. It is also presupposed that there are ancient
traditions, and knowledge of the existence of Machimon, and of
Machimon being a god, are presupposed.
Another example is the opening of the documentary 'A New Green
Revolution?', a television documentary programme in the science
108 Representations in documentary and news
documentary series Horizon, which was broadcast in January 1984 on
BBC2. The programme is about the social and economic effects on the
'Third World' of breeding new high-yielding varieties of staple crops.
See Silverstone (1985) for a detailed analysis. Here are the opening
words, followed by an extract which occurs shortly afterwards.
Keith griffin (Sync): The difficulty is that if we persist with our current
line, looking for technological solutions to socio-economic problems,
then we will run out of time. These problems of impoverishment,
inequality, social tension, of conflict, will explode.
keith griffin (voice-over): Normally the crisis in the Third World,
poverty, inequality, hunger, is a silent crisis. Only occasionally does
the crisis of the peasantry erupt in the form of violence and civil
discord.
There are many presuppositions here which again draw the viewer
into the common-sense assumptions, the world-view, upon which the
programme is founded. For instance, it is presupposed that there is a
difficulty, that our current line is a bad one (implied by persist with), that
this line is attributable to all of 'us', that we are trying to achieve
something in a limited amount of time, that what we are trying to
achieve is the avoidance of an explosion (we need this presupposition
to make a coherent connection between the two sentences of the first
extract), that there is a crisis in the Third World, that there is a crisis of
the peasantry. In addition, there are presuppositions associated with
the major categories drawn upon here, such as 'the Third World'.
Actually, it is hoc simply presupposed that the Third World exists, it is
presupposed that the expression Third World is the appropriate
designator for — the name of - the countries concerned. Similar
presuppositions hold for 'the peasantry', 'impoverishment',
'inequality', 'social discord': both the category and its relevance to the
point at issue are presupposed, taken as common sense. (For some, a
presupposition is a particular category of implicit proposition, upon
which the truth or falsity of a presupposing sentence depends. Pre­
supposition is contrasted with entailment and implicature, other
categories of implicit meaning. See Levinson 1983.)
The unsaid, the already said, the presupposed, is of particular
importance in ideological analysis, in that ideologies are generally
embedded within the implicit meaning of a text rather than being
explicit (Fairclough 1989, chapter 4). Consider, for example, the pre­
supposition I have just suggested, that the 'current line' is attribu­
table to all of 'us'. The fudging of the boundary between the
generality of the population and its government or other powerful
 Representations in clauses 109
agents or élites, such that the actions or practices or values of the latter
are generalized to the former, is commonplace in the media. In so far as
it legitimizes and so helps to reproduce relations of domination through
assuming a consensus that doesn't exist, it can be seen as having an
ideological function. The discussion of local coherence relations at the
end of this chapter is also relevant to the issue of presupposition.

Representations in clauses

I come now to the first of the two major aspects of representation dis­
tinguished above, representation in clauses. As I have already sug­
gested, when people represent in language events, actions,
relationships and states, the people and objects involved in them, the
time and place and other circumstances of their occurrence, and so
forth, there are always choices available. Partly these choices are a
matter of vocabulary: the vocabulary one is familiar with provides
sets of preconstructed categories, and representation always
involves deciding how to 'place' what is being represented within
these sets of categories - shall I call the violent death of people at the
hands of others a 'killing', 'murder' or 'massacre'? It may also be a
matter of metaphor: shall I call it a 'holocaust' or an 'extermination'?
But these choices are also partly a matter of grammar. The grammar
of a language differentiates a small number of 'process types' and
par '■} t"
difference between an action (with a causal actor) and an event (with­
out a causal actor) is a difference in reality, in the nature of things, but
that is not so, at least in any simple sense. When people represent in
language something that happens, they have to choose whether to
represent it as an action or an event. Recall the example on page 25
from the Today programme, commenting about 'cheap' Russian fish
being 'dumped' on the British market: 'the funny thing is, it's not
transferring itself to the consumer at terribly low prices at all'. There
is no actor, no one responsible for the prices, in this form ulation-as if
the fish distributed themselves and set their own prices.
There are two points to make about such relatively low-level
choices in texts - low-level in the sense that they involve single
clauses, and even single words within them. First, there are often
systematic patterns and tendencies in particular types of text and par­
ticular discourse types. Work in critical linguistics, for instance (Fow­
ler et al. 1979, Hodge and Kress 1979), has suggested that some
newspapers systematically background the involvement of the police
HO Representations in documentary and news

in violence and other forms of undesirable social behaviour. Cumula­

tively such representational practices may have significant ideological
effects. Second, as I indicated above, such choices may realize contrasting
discourses. For instance, the example at the end of the last paragraph is
indicative of the sorts of linguistic realization associated with a major
divide in the representation of economic and social problems and dis­
asters (e.g. unemployment, violence, environmental disaster): between
discourses which foreground (often hidden) causality, responsibility and
even conspiracy, and discourses which represent such problems and
disasters as a matter of fate, happenings beyond human control.
The grammar of English differentiates the following process types:
Action, Event, State, Mental Process, Verbal Process (Fairclough
1992a, Halliday 1985). An Action involves both the participant-types
Actor and Patient (person or thing affected by action): the Actor does
something to the Patient. A typical Action clause has a transitive
structure (Subject + Verb + Object) (e.g. police kill 15, child breaks win­
dow). An Event involves just one participant, which may either be
affected by what happens and hence a Patient (e.g. 15 die, window
breaks), or be in an active, causal relationship to what happens, and
hence an Actor (e.g. victims screamed). Events have an intransitive
structure (Subject + Verb). A State is 'being' (e.g. 15 are dead) or
'having' (many have serious wounds), and has an 'equative' structure
(Subject + Verb + Complement). A Mental Process involves the
participant-types Senser - the person who experiences or undergoes
the mental process - and Phenomenon —what impinges on con­
sciousness from outside. There are mental processes of cognition
(e.g. Thatcher realizes it's time to go), perception (e.g. Thatcher sees the
writing on the wall) and affect (e.g. Thatcher wants to go). Finally, a
Verbal Process involves an Actor and a participant-type we might call
Verbiage - what is said (Thatcher says it's time to go).
I want to illustrate the analysis of process and participant types -
and try to show what insights it can yield - with three extracts from
the opening few minutes of 'A New Green Revolution?'. The first,
part of which I have already used above, occurs shortly after the pro­
gramme begins:
Keith griffin (voice-over): Normally the crisis in the Third World,
poverty, inequality, hunger, is a silent crisis. Only occasionally does the
crisis of the peasantry erupt in the form of violence and civil discord.
narration : Millions of poor people in the Third World may not be
silent much longer. They're caught up in an economic system which is
steadily driving them towards red revolution. Agricultural technology
is a crucial part of that economic system.
 Representations in clauses 111
The second extract begins shortly afterwards:
kezth griffin (Sync): Wherever one looks throughout the world one
sees rising political tensions and violence and civil disturbance. The
hope would be that if an appropriate technology could be found and
introduced which increased the demand for labour, while at the same
time increasing production of the things that the poor consume, that
this would help to diminish the social tensions.
narration: Several international agricultural research centres, fun­
ded by Western aid, are looking for new techniques to help solve the
problems of hunger.
In the Philippines at Los Banos, the International Rice Research
Institute was set up to help Asia increase rice production.
The scientists there are working against the odds and against the
clock.
As individuals, most of them do genuinely want to help the poor -
but is that what they achieve?
The third extract again follows shortly afterwards. (I have included a
representation of visual images: CU is 'close-up', MS 'medium shot',
WS 'wide shot', MCU 'medium close-up'.)
MIX to pipes in slum area of narration: Everywhere in the
Manila, pan to WS slums Third World life in rural areas
gets harder - and poor people
flock to the city. The urban
poor get poorer.
CU child standing in pipe
Slum area, mother and child When rice prices go up, hunger
and unrest grows.
In the city, the people can
usually be kept in their place.
MS Filipino soldiers marching
towards camera
WS remote mountain village (zoom But in remote rural areas, out
in) of the eye of the regime, the
New People's Army, dedicated
to supporting poorer people
and small farmers, plans
violent revolt.
guerrilla: 'Standing position!'
NPA guerrillas weapons training
MCU they raise their machine guns
WS Borlaug and Knapp by plots narration: Have the scientists'
new techniques helped to
increase or to decrease this
violence and tension?
112 Representations in documentary and news
The people who are in focus here are the poor in countries like ,
Bangladesh. The first thing to notice is that events and situations that
involve the poor are often worded in a way that doesn't directly refer
to them - the poor don't figure as a participant in the process. The
first extract illustrates this: 'crisis', 'poverty', 'inequality', 'hunger',
'violence' and 'civil discord' are all situations affecting or involving
the poor, but they are worded without direct reference to the poor.
These are all 'nominalizations': that is, processes that have been
turned into noun-like terms (nominals) which can themselves func­
tion as participants in other processes (e.g. 'the crisis of the
peasantry' is Patient in the Event process, 'does the crisis of the
peasantry erupt'). When a process is nominalized, some or all of its
participants are omitted - that is why the poor don't figure explicitly
in 'hunger', 'violence', and so forth, in this example. A lot of nomin­
alizations in a text, as there are in this case, make it very abstract and
distant from concrete events and situations (Hodge and Kress 1979).
The second point is that where the poor are explicitly referred to, it
is not as Actors - as people who are doing something - but more as
Patients - as people who,are affected by the actions of others - or as
participants in States. Again, the first extract illustrates this, twice in
the penultimate sentence. In they're caught up in an economic system,
they, the poor, is Patient in a passive clause which lacks an explicit
Actor (and so doesn't specify who or what is responsible for their
being so caught up); in which is steadily driving them towards red revolu­
tion, them is Patient, with an abstract rsnrnin a!i z a ti o n ('economic
system' - substituted by which) as Actor. Both of these are Action pro­
cesses. Notice here the nominalization red revolution within a direc­
tional Adjunct towards red revolution: a revolution implies the poor
actually doing something on their own behalf, but wording it as a
nominalization backgrounds the active role of the poor, and wording
them as Patient of the clause foregrounds their passivity. There are
other examples of the poor as Patients in the other extracts: most of
them do genuinely want to help the poor, the urban poor get poorer, the people
can usually be kept in their place.
If the poor are not the active agents, the Actors, who are? There are
three main types of Actor. First, nominalizations, such as ecotwmic
system in the example just referred to, or in the second extract: an
appropriate technology . . . which increased the demand for labour, and pre­
sumably: 'finding an appropriate technology' (this in the text) would
help diminish the social tensions. Second, the scientists, and the
scientists collectively as 'centres', in the second extract, for example:
several international agricultural research centres . . . are looking fur new
 Representations in clauses 113
techniques, As individuals, most of them (i.e. scientists) do genuinely want
to help the poor. And third, the New People's Army, which is (dedicated
to) supporting poorer people and plans violent revolution. Notice how the
activity of the poor is again backgrounded in the former by poorer
people figuring as Patient while the New People's Army (which presum­
ably in fact consists of poorer people) is Actor.
In fact there are only two Actors in the third extract, the New
People's Army and, exceptionally, the poor, in the poor people flock to
the city, interestingly, the Action here is one more usually associated
with sheep - notoriously passive - than people, so the exception
does not really contradict what I have said so far. What is striking in
this extract, though, is how processes which might have been
worded as Actions, in ways that foregrounded agency, causality
and responsibility, are worded in ways that background them: life
. . . gels harder, the urban poor get poorer, rice prices go up, hunger and
unrest grows. The first two are State processes and the second pair
Event processes with Patients. They all background and mystify
who or what caused the processes referred to - one might, for
instance, have had, instead of rice prices go up, an Action process,
when rice producers (or shops, or governments - it is precisely not
clear who) raise rice prices. What such choices have in common -
choices of process type, the choice of a nominalized rather than a
clausal process, and the choice of a passive clause with a deleted
Actor rather than an active clause (there are quite a few examples
here) - is the capacity to background and in some cases to mystify
agency and responsibility.
Let me turn now to choice of categories (Hodge and Kress 1979),
and questions of vocabulary. Notice that the poor are categorized in
various ways in these extracts: as poor (the poor, the urban poor, poor
people, poorer people), as peasants (the peasantry), and as the people. They
are not, for instance, categorized as 'the oppressed' (as in the title of
Paolo Freire's celebrated book on literacy, Pedagogy of the oppressed
(1972). The main categorization is in terms of poverty - in other
words, in terms of their condition, rather than in terms of the
relationships of exploitation implied by 'the oppressed'. One won­
ders whether this - along with the positioning of the poor I noted
above, as passive rather than active participants in events - is how
these people see themselves, or whose way of seeing them it is - their
own government's? that of overseas governments or agencies? The
general point is that one should also ask where the media get their
categorizations from, both those that are explicit in the vocabulary,
and those that are implicit in how people or things figure in process
114 Representations in documentary and news

types. Recall the discussion of the sources of formulations and

discourses used in the media in Chapters 4 and 5.
Generalizing from this example, we may say that there are always
alternative ways of wording any (aspect of a) social practice, that
alternative wordings may correspond to different categorizations,
and that such alternative wordings and categorizations often realize
different discourses. In this case, for instance, a discourse of poverty
is drawn upon and realized in the wording which constrasts with a
discourse of oppression, which might have been drawn upon, but
significantly was not. Recall also, however, my comments in Chapter
5 on collocation: by focusing upon patterns of co-occurrence in the
vocabulary of a text, the analyst can show how different discourses
can be condensed together in short phrases within a text which easily
pass unnoticed (e.g. referring to Saddam Hussein as a mad menace).
A further sphere of choice is metaphor (Lakoff and Johnson 1980).
Contrary to common assumptions, metaphor is not just a literary
device. Choice of metaphor may be a key factor in differentiating
representations in any domain, literary or non-literary, including
even scientific and technological. Notice, for example, the metaphors
in the first extract from 'A New Green Revolution?'. A contrast is set
up between the usual silence of the poor in the 'Third World' and their
occasional, and likely future, eruptions. Actually, as I have already
pointed out, the poor are partly implicit rather than explicit here; the
Narration does say millions of poor people in the Third World may not be
silent much longer, but Griffin refers to a silent crisis, and the crisis of the
peasantry as erupting, in the form of violence and vioii discord. Meta-
phorizing the usual inaudibility of the poor as silence begs the
question of whether they are inaudible because they have nothing to
say (implausible, I would suggest) or inaudible because the media of
communication do not represent what they have to say. Except when
there is an eruption: political activism on a scale which cannot be
ignored is metaphorized in terms of that other eminently
newsworthy category, the large-scale natural disaster. The poor are
attributed with two political options: inert passivity and silence, or
irrational and uncontrollable explosion (also implicit in the other
wording of this option, red revolution). See Montgomery et al. (1989)
for an interesting discussion of the metaphorization of elections as
war, and how the metaphor shapes an electoral campaign and its
media coverage.
Van Leeuwen (1993) suggests that representation can be seen as
the recontextualization of social practices. For any text, the analyst
can explore how represented social practices are recontextualized
 Representations in clauses 115
within the social practice which the text is part of. Recontextualiza-
tion entails transformation - represented social practices are trans­
formed in ways which are determined by the concerns and values
and purposes of the text in its own social practice. The same social
practice (e.g. coping with poverty in the Third World', or the induc­
tion of children into the school system) will be differently recontex­
tualized, and differently transformed, in different texts.
Texts can be compared in terms of their relationship to the sort of
account of a social practice which an ethnographer might produce.
Van Leeuwen suggests eight primary elements of a social practice: its
participants, their activities, the circumstances (time and place) of activi­
ties, the tools and dress prescribed, the eligibility criteria for participa­
tion, performance indicators for activities (e.g. should they be
performed quietly and slowly, or quickly and efficiently), and
reactions of participants to each other. I have referred above to parti­
cipants and activities (i.e. 'processes') and in passing to circum­
stances. Texts recontextualizing a given social practice can be
compared initially in terms of which elements they include or exclude
and what relative weight they give them. Then, for each element,
there are alternative ways in which it may figure in a text, and the
analyst can show what choices have been made. We have seen above
some of the sorts of choices available for the participant and activity
elements.
Let me give an example of other choices for the participant
element, and make some points about the circumstance element,
drawing upon news reports of an important speech by prominent
Conservative right-winger and ex-Cabinet Minister Lord Tebbit at
the Conservative Party Conference in 1992, in which he galvanized
Conservative opposition to the further economic and political inte­
gration of the European Union envisaged in the Maastricht Treaty. I
shall use these reports to make a link between this and the next
section. Here is the front-page report from the Daily Mirror, and the
headline and first two paragraphs of the Guardian front-page report,
both for 7 October 1992:

HE HASN'T GOT A PRAYER

Major in Crisis as Tebbit twists knife
By Alistair Campbell, Political Editor
John Major was facing a leadership crisis last night after being savaged
by ex-Tory Party chairman Norman Tebbit.
Lord Tebbit warned the Prime Minister that he hadn't a prayer of
staying in power unless he changed his policies. His vicious attack at
116 Representations in documentary and news
the Tory Conference in Brighton was cheered by delegates and he was
given a standing ovation.
Mr Major could only look on in dismay as his stance on the Maas­
tricht Treaty and the economy was crucified.
Lord Tebbit piled on the agony by later claiming he was simply
offering a 'hand of friendship'.
Butin the same breath he warned the Premier that if he didn't get the
economy right - he could kiss goodbye to his job.
Daily Mirror

MAJOR TOUGHS OUT EURO-MUTÍNY

The Conservative leadership last night began cracking the whip over
Euro-sceptical MPs after successfully gambling that it could crush Lord
Tebbit's head-on challenge to the Maastricht treaty on the Brighton
conference floor.
The Chancellor, Norman Lament, will face the conference over the
economic crisis and the Exchange Rate Mechanism controversy tomor­
row, but John Major's leadership is far from out of the woods.
Guardian

Participants are, of course, prominent in both reports, but whereas

participants in the Daily Mirror report are referred to only indi­
vidually and by name (Major etc.), participants in the Guardian are
also frequently referred to collectively, either in terms of their func­
tion (the leadership, ministers) or impersonally in terms of location at or
T.w- jf-T-j : r-i bln r-. rv rj-n f o r ,o r < i ',c* { th ( * q n in T P T tp rí’ trip

assembly, the platform - in the sense of the leaders on the platform).

These locational designations blur the distinction between parti­
cipants and circumstances in van Leeuwen's framework, and are part
of a greater concern with circumstances in the Guardian report. Notice
also the temporal circumstances in paragraphs 1 and 2 of the Guard­
ian report (last night, tomorrow), as well as the locational circumstance
in paragraph 1 (on the Brighton conference floor). The Daily Mirror report
has only one reference to location, in the second sentence of para­
graph 2, and two temporal circumstances (last night and later, in sen­
tences 1 and 5). These differences of linguistic detail are, I think,
significant in helping to project profoundly different constructions of
politics. The Daily Mirror projects a simple, relatively static and
decontextualized confrontation between individuals (indeed person­
alities), the Guardian projects a more complex game played out over
time and largely involving collective agents amongst whom distinc­
tions of function and status (partly represented in spatial terms, e.g.
'platform' versus 'floor') are important.
 Combination and sequencing of clauses 117
Combination and sequencing of clauses

I shift now to the second major aspect of representation, combination

and sequencing of clauses (propositions). As a way into some of the
issues here, 1 shall discuss the most prominent parts of the coverage
of the Tebbit speech in two media outlets: the beginning of the report
on the main evening news bulletin on BBC1 on 6 October 1992, pre­
ceded by the section of the opening headlines which dealt with the
issue; and the front-page report in the Daily Mirror on 7 October
(reproduced above).
At the Conservative Party Conference, Lord Tebbit urged the
government to abandon the Maastricht treaty and put Britain first,
second and third. The Foreign Secretary admitted the party could
break itself over Europe and urged it to give that madness a miss.
The former Cabinet Minister Lord Tebbit drew noisy cheers from a
sizeable minority of representatives at the Conservative Party confer­
ence today when he urged the government to abandon the Maastricht
treaty and negotiate a new one with no mention of economic, monet­
ary or political union. He mockingly talked of the terrible wounds
inflicted on industry and home owners as we established our creden­
tials as good Europeans. And he said that politics, like charity, should
begin at home. But the Foreign Secretary, Mr Douglas Hurd, said it
was not in Britain's interests to be on the sidelines when the security
and prosperity of Europe was being decided. And he said that the
European summit in Birmingham would focus on bow fhp community
could learn to do fewer things better. The conference passed a motion
calling for the government to continue to build an open and outward­
looking community.
BBC1 Nine o'clock News
Van Leeuwen's view of representation as recontextualization can be
applied, as I have already suggested, to the analysis of clauses, but it
is also useful in comparing combinations of clauses in texts. One
aspect of social practices which may or may not be prominent in
recontextualizations is Reactions - ways in which participants react to
other participants. For present purposes, I shall distinguish as topics
in the reports the Speech itself (and other speeches in the debate) and
Reactions to it. Also, as van Leeuwen points out, recontextualization
of a social practice is likely to involve to a greater or lesser degree
Evaluations of it, in this case especially evaluations of the wider pol­
itical effects and consequences of the Tebbit speech. With respect to
reporting of the Speech itself, I shall also differentiate: direct Quota­
tions from it, Summaries of it, and Formulations of it. Whereas
118 Representations in documentary and news
Summaries give the gist of what was said, Formulations give interpreta­
tions (recall the discussion of discourse representation in Chapter 5).
In terms of these distinctions, the BBC1 report can be represented
as follows (S stands for 'sentence'):
HEADLINES: SI: SPEECH (Summary)
S2: SPEECH (Summary 4- Quotation)
REPORT: SI: REACTION + SPEECH (Summary)
S2: SPEECH (Summary + Quotation)
S3: SPEECH (Summary)
S4: SPEECH (Summary)
S5: SPEECH (Summary)
S6: SPEECH (Formulation)
and the Daily Mirror report as follows:
HEADLINES: SI: EVALUATION or SPEECH (Formulation)
S2: EVALUATION + SPEECH (Formulation)
REPORT: SI: EVALUATION + SPEECH (Formulation)
S2: SPEECH (Formulation)
S3: REACTION [SPEECH (Formulation)] + REACTION
S4: REACTION + SPEECH (Formulation)
S5: SPEECH (Formulation + Quotation)
S6: SPEECH (Formulation)
As the coding indicates, the main headline of the Daily Mirror is
ambiguous. Since the same wording appears in a formulation of
Tebbit's speech in the "econH paragraph of the report (Lord Tebbit
warned the Prime Minister that he hadn't a prayer of staying in power . . .),
we could analyse the headline as a formulation of what Tebbit said.
But there is no indication in the headline that these words are
reported, so it could also be analysed as the reporter's Evaluation of
the effect o f Tebbit's attack on Major's political position. Another
point to notice is that although we have what appear to be two
Summaries of Tebbit's speech in sentences 2 and 6 introduced by the
reporting verb warned, they are so distant from what Tebbit actually
said in his speech that they need to be taken as Formulations. In
sentence 3 we have two Reactions, with a Formulation of Tebbit's
speech embedded in the first of them (His vicious attack at the Tory
conference in Brighton). The embedding is shown with square brackets
in the analysis. The Daily Mirror text is accompanied by a photograph
of Major in what appears to be an attitude of prayer, with the caption:
OH LORD: John Major can only pray his luck will change as he listens to
Norman Tebbit tearing his policies apart. This consists of a Reaction (OH
LORD) followed by an Evaluation with a Formulation of Tebbit's
 Combination and sequencing of clauses 119
speech embedded in it. There is an obvious play on words in the
headline, where he hasn t got a prayer is used in the sense 'he hasn't got
a chance'; notice too that OH LORD is not only the conventional way
of beginning a prayer, but also a way of expressing a reaction of
dismay, and a play on Tebbit's title. There is also a much smaller
photograph of Tebbit, confidently smiling, with the caption:
VICIOUS: Tebbit.
The two reports are strikingly different. The BBC1 report predom­
inantly consists of Summaries of the speeches in the debate, with no
Evaluations and only one instance of Reaction. The Daily Mirror
report, by contrast, focuses very much on Evaluations of the political
effects of the speech and Reactions to it, and reports of Tebbit's
speech are mainly interpretative Formulations rather than Summar­
ies. Its main emphasis - not surprisingly given its general critical
stance towards the government - is on the damaging political effects
of the speech rather than the speech itself.
But there is more to the comparison of the reports than the relative
frequency of the different Evaluations or Reactions or the other
topics. There are various ways, in addition to sheer frequency, in
which topics can be relatively foregrounded or backgrounded. It is
always worth first of all looking at the relative positioning of different
topics within the generic structure of the text. Topics which appear in
the informationally prominent headline position in television or
press reports, or in the lead (i.e. the first) paragraph in a press report,
are informationally foreerounded (recall the discussion in Chanter 5).
In this case, the overall emphasis in the Daily Mirror report on pol­
itical effects is sharply focused in the positioning of Evaluations in the
headline and lead paragraph.
Positioning in headlines and leads is a matter of what I called 'glo­
bal text structure' in the introduction to this chapter. Foregrounding
is also partly due to 'local coherence relations'. In sentences (clause
complexes), main clauses generally foreground information,
whereas subordinate clauses generally background it. This is especi­
ally so when the main clause precedes a subordinate clause. So in the
second Daily Mirror headline (Major in crisis as Tebbit twists knife), the
reported political effect is both in the main clause and at the begin­
ning of the sentence, whereas the formulation of the speech is in the
subordinate clause at the end. Exactly the same is true of the lead par­
agraph (John Major was facing a leadership crisis last night after being
savaged by ex-Tory Party chairman Normmz Tebbit). Evaluation of effect is
foregrounded in both cases, in accordance with the tendency I have
already noted in the Mirror report.
 120 Representations in documentary and news
Local coherence relations between sentences may also contribute to
the relative salience of propositions. In the case of the BBC1 report,
intersentential relations subtly contain and defuse Tebbit's
potentially explosive intervention. Notice firstly that the first of the
sentences reporting Hurd's speech is Jinked with the report of Teb­
bit's speech with the conjunction but (But the Foreign Secretary, Mr
Douglas Flurd. . .). But is often used in a reassuring way linking some
risk or threat with the means of avoiding it (e.g. we had a gas leak last
night but the Gas Board were round in less than half an hour). The BBC1
report strikes me as similar: Hurd's speech is portrayed as effectively
If-!!
■Elft rebutting Tebbit's attack. Notice also the last sentence of the news­
l ;i]jj;l reader's introduction. In this case there is no explicit linking word,
;K
but there is nevertheless a coherent meaning relation between the
NR
sentence and those preceding it, which might have been explicitly
worded as something like after the debate or in the end or finally. This
sentence imposes a closure, a conclusion, on the report, in a form
which supports the report of Hurd's speech in defusing Tebbit's
attack: since the motion is calling upon the government to continue to
build an open and outward-looking community, that implies - in
contrast to Tebbit's Speech - support for what the government has
been doing. The BBC report appears on the face of it to treat Tebbit
and Hurd equally, but they are subtly differentiated into antagonist
and protagonist respectively (Martin 1986), with the latter's position
hpine given more weight.
The informational structuring of clauses is anolher significant
factor in foregrounding. The element at the beginning of a clause is
fit ■
p called its 'theme' (Fairclough 1992a, Halliday 1985); the theme is the
topic of the clause, what it is 'about', so theme is in a prominent posi­
tion informationally, ft is often the subject of the clause, but not
always (for example, the theme is in the same breath, a 'circumstantial'
element [Halliday 1985] in the first clause of the last paragraph of the
Daily Mirror report). A significant contrast between the BBC1 and
Daily M inor reports is that whereas the theme in the former mainly
alternates between Tebbit and Hurd, the two main speakers in the
debate, in the latter the theme alternates between Major and Tebbit.
The BBC1 report represents the debate as a division about policies,
the Daily Mirror report represents it as about the misfortunes, and
survival, of John Major.
The final position in a clause, or what is sometimes called the
'information focus' position, is also prominent, especially if it comes at
the end of a sentence. In spoken language - and so in radio and tele­
vision reports - this position is usually prominent in terms of intonation
 Combination and sequencing of clauses 121
- it is where the intonational 'nucleus', the main movement in pitch,
tends to occur (Haliiday 1985). In the lead paragraph of the Daily Mir­
ror report both Major (theme) and Tebbit (information focus) are
informationally prominent: John Major is in the initial theme position
(and also subject of the clause, as is often the case), while when one
reads the sentence out loud the nucleus most naturally falls on the
final word Tebbit. Another position which is worth watching out for is
initial position in a paragraph, which is again prominent, though this
is not particularly interesting in the Daily Mirror, where paragraphs
tend to correspond to sentences.
The example of the coverage of Tebbit's speech shows that the rela­
tive foregrounding or backgrounding of aspects of represented social
practices is an important part of their representation. The analyst
does not only need to know what is represented, but what relative
weight or importance is attached to different elements within a repre­
sentation.

Local coherence relations and ideology

I would like to develop what I have said about local coherence. Local
coherence relations between clauses, clause complexes or sentences
are of three main types: elaboration, extension, and enhancement (Halli­
day 1985: 202-27). In elaboration, one clause elaborates on another by
describing it or making it more specific - by rewording it, exempli-
fvHTrrif) nr HpHfvino-
■j - -Q '■! - ' -■ - ' - •J
H' Tn pyl^ntiinn
O '" ------------
/inf
• w • •’
'oyTpr)i'i--:
' v -------
rhr> ni^hnirifr
----------- “ 'O
of another by adding something new to it'. This may be straight addi­
tion (marked with and, moreover, etc.), adversative or contrastive
(marked with but, yet, however, etc.), or variation (marked with or,
alternatively, etc.). In enhancement, 'one clause enhances the
meaning of another by qualifying it', in terms of time (e.g. A then B, A
after B, A while B - where A and B are clauses or sentences), place
(e.g. A where B), cause (e.g. A because B, A so B) or condition (e.g. if
A then B).
Local coherence relations between clauses or sentences (clause
complexes) may be explicit or implicit; they may be, but are not neces­
sarily, signalled by markers of cohesion such as conjunctions or
sentence-linking adverbiais (but, accordingly, nevertheless, afterwards,
for example). Halliday and Hasan (1976) distinguish four main types
of cohesion: conjunction, lexical cohesion, reference, and ellipsis. The
markers of cohesion Í have referred to so far are cases of conjunction,
including both what traditional grammar calls conjunctions (e.g. and,
but, because) and 'sentence adverbiais' like however and nevertheless.
122 Representations in documentary and news
Lexica] cohesion is cohesion through vocabulary - through repetition
of words, and words that are linked in meaning occurring across
clause and sentence boundaries. Reference includes referring back or
forward in a text with pronouns, the definite article or demon­
stratives (this, that). Ellipsis is missing out repeated material or
replacing it with a substitute word (e.g. Tebbit failed to mention
Thatcher. Hurd on the other hand did. i.e,, did mention Thatcher).
Relations of coherence between clauses and sentences of a text are
not objective properties of the text, they are relations that have to be
established by people interpreting it. This is clear in the case of impli­
cit relations, where there is nothing in the text to show how one sen­
tence links to its predecessor. Even where there are markers of
cohesion, these may have to be interpreted, interpreters need to
decide what a pronoun refers back to, or which items of vocabulary
are cohesively linked together. In doing so, interpreters need more
than the propositions which are explicit in the text. They draw upon
other propositions which allow them to make inferential leaps
between the clauses and sentences of the text. But texts are structured
in ways which predispose interpreters to leap one way rather than
another - which is no guarantee that they will do so, because inter­
preters may read texts in different ways. However, there will be
certain predominant readings, and we can say that the non-explicit
propositions which they depend on are part of the implicit meaning
of the text What that means is that a text addresses a sort of 'ideal
interpreter' who will bring to bear just the propositions (the implicit
meanings) needed to give the text what has been called its 'preferred
reading' (Hall et at. 1978).
The upshot of all this is that local coherence relations are very signi-
ficant indeed in the ways in which texts position people as subjects
and cumulatively shape identities (see also Chapter 7), and how texts
work ideologically. Let us take as an example the last of the three
extracts from 'A New Green Revolution?' on page 111.
Cohesion relations are largely implicit in this sequence. For example, I
interpret the clauses of the first three sentences as in relations of
enhancement, and more specifically cause-effect relations, but they
lack markers of causal cohesion. In sentence 1, the two clauses are
linked by the ail-purpose conjunction and, which leaves implicit the
cause-effect relation (poor people flock to the city because life in rural
areas gets harder). I also see an unmarked cause-effect relation between
the second clause of sentence 1 and sentence 2 (the urban poor get
poorer because so many people flock to the city). Again, although the
first clause of sentence 3 is marked with a temporal conjunction (when),
 Combination and sequencing of clauses 223
there is an implicit causal relation between the two clauses (hunger
and unrest grow because rice prices go up). It takes quite an inferen­
tial leap on the interpreter's part to establish a coherent meaning
relation between sentences 3 and 4. I interpret this as an extension-
type relation of an adversative type (unrest grows, but the people
can usually be kept in their place; or, although unrest grows, the
people can usually be kept in their place). The connection between
these sentences rests upon a 'bridging assumption' (Brown and
Yule 1983, Fairclough 1992a): that popular unrest gives rise to a
problem of order, and the need for official action to try to contain
it.
Overall, this part of the extract addresses an ideal interpreter who
is familiar with a particular preconstructed 'script' (Montgomery et al.
1989) that is being evoked here: a predictable sequence of events
leading from rural poverty to urban squalor and unrest and conse­
quential problems of order. The ideal interpreter is relied upon to fill
in the gaps, make explicit what is left implicit, and construct a
coherent, preferred, meaning for the text.
But this is not just a matter of textual economy, not bothering to
spell out what can be taken for granted. It is a moot point how many
real audience members might, if asked, actually agree with the
stereotypical narrative of Third World urban problems which consti­
tutes the script. But the text takes the script as universally given for its
audience, and so positions audience members that they are induced
tO ivpfMT it- a r r i Vv 1 V a r r x l n o r - o p k i f l t 0 t l ChP ílo C C lll -’llGO H lV

earlier observation that processes affecting the poor are represented

as processes without agents: life getting harder, or poverty increas­
ing, are things that happen (like thunderstorms) rather than effects
created by people. The script represents social processes on a natural
scientific model, as a sequence of causally linked events devoid of
human agency and outside human control. This is a representation
which (depending on context) has a potential ideological function, in
the sense that eliding responsibility for social processes and the
possibility of intervention to change them is conducive to a fatalism
which can help sustain existing relations of domination. Local coher­
ence relations in cases of this sort can therefore contribute signifi­
cantly to textual processes of ideological interpolation (Althusser
1971): audience members are, so to speak, called upon to acknow­
ledge the framework of ideological common sense (in this case, the
Third World script) within which they are positioned. Such texts can
cumulatively shape the knowledge, beliefs and values of audience
members.
124 Representations in documentary and news
The interpretative activity of the audience member in arriving at
coherent meanings is even clearer in the case of the visual images of
television, represented on the left of the transcription of the extract.
Documentary producers depend upon audiences drawing upon
visual scripts in interpreting sequences of images coherently. For
example, not only are we assumed to recognize images of a child and
a mother and child as images of poverty, we are also assumed to
know what follows - in terms of 'keeping people in their place' -
upon soldiers marching through city streets. As with the language,
such inferential work on the viewer's part can contribute to the ideo­
logical work of texts.
The connection between the question which closes this sequence
(Have the scientists' new techniques helped to increase or to decrease this
violence and tension?) and the sentences which precede if is worth
commenting on. There is a cohesive link o f the reference type (this in
the phrase this violence and tension refers back to earlier sentences),
and there is also lexical cohesion (violence echoes violent, and tension is
semantically close to unrest). But notice that the phrase presupposes
the existence of violence and tension, even though no actual violence
had been referred to or visually depicted. As so often, the way iri
which the text formulates or paraphrases earlier parts of the text turns
out to be significant (Heritage and Watson 1979): the New People's
Army's planning for 'violent revolt' is transformed into 'violence'.
This can again been seen in terms of ideological interpolation, in that
it calls upon audience members to draw upon, m interpreting a text, a
stereotype which associates especially left-wing oppositional groups
with violence, rather than governments. It is also noteworthy that the
information focus and the intonational nucleus fall on decrease rather
than this violence and tension, in accordance with the presupposed,
taken-for-granted status of the latter. Notice also that the impact of
this contentious presupposition is increased by its positioning in the
activity-type structure associated with the genre here. This is an
investigative reporting genre where the activity type - the argu­
mentation - is quite standardly structured as a sequence of: authori­
tative orientation for the report -!- question for investigation +
evidence bearing on the question. The question element, which the
presupposition is built into, is an informationally salient hinge
between the orientation and evidential stages of the report. This
shows how local coherence relations, global text structure, and the
information structure of the clause can work together in texts.
 7

IDENTITY AND SOCIAL RELATIONS IN

MEDIA TEXTS

Chapter 6 was concerned with representation m texts, and this

chapter deals with the other two processes which are always simul­
taneously going on in texts: the construction of identities, and the
construction of relations. I suggested in Chapter 3 (page 39) that
media discourse can usefully be seen in terms of the interaction
between three major categories of participants, and this can provide a
starting point here. The categories are: reporters (a term I use for all
categories of media personnel, including announcers, presenters,
anchorpersons); audiences; and various categories of 'other parti­
cipants', mainly from the public domain (politicians, trade unionists,
religious leaders, scientists and other experts, etc.).
Notice that the concern here is not with how 'others' are referred to
and represented (part of the business of Chapter 6), but with others as
direct participants in media output, for instance as interviewees in
broadcasting. We need to include amongst 'other participants' people
who participate as in some sense representatives of the audience -
members of the public' who are asked for their views on a shift in
government policy, for example, or participants in audience discussion
 126 Identity and social relations in media texts
programmes (discussed in Chapter 9). One important concern is with
relationships between categories of 'others': which are most salient,
which are similarly constructed and which are contrastively con­
structed. Another issue is absences: for instance, whereas politicians
are a well-defined and salient category, capitalists - those who dom­
inate the economy - significantly are not. There are interesting
slippages between these three main categories: for instance, the right
to directly address audiences on television, usually the prerogative of
broadcasters, is extended in certain cases to categories of 'others' -
such as politicians in election broadcasts and, in open access types of
programme, to p eo p le seen as representative of audiences.
The chapter will be constructed around these three broad cate­
gories of participant. 1 shall be concerned on the one hand with ques­
tions of identity: the sorts of social and personal identity that are set
up in media output for reporters, for audiences, and for the various
categories of 'other' participant. On the other hand, I shall be con­
cerned with relations: between reporters and audiences, between
various categories of 'others' and audiences, and between reporters
and politicians, experts, and other categories of 'others'. Although it
is analytically useful to distinguish questions of identity from ques­
tions of relations, the two are, in practice, inseparable: how a report­
er's identity is constructed is in part a question of how a reporter
relates to an audience.
Analysis the construction of relations and identities in media
texts is, I suggest, a significant constituent in addressing a range of
important sociocultural questions. This is so because of the uniquely
influential and formative position of the media in contemporary soci­
eties. Understanding how relations are constructed in the media
between audiences and those who dominate the economy, politics
and culture, is an important part of a general understanding of rela­
tions of power and domination in contemporary societies. And there
is a key question about contemporary changes in the media construc­
tion of relations between reporters and audiences and between
politicians and audiences: do they constitute a substantive
democratization, or do they primarily have a legitimizing role in
respect of existing power relations? There are also questions about
whether and to what extent the media, in the ways in which they con­
struct audience and reporter identities, operate as an agency for pro­
jecting cultural values - individualism, entrepreneurialism,
consumerism - and whose values these might be.
ji v|.
*U It is important to ensure that a critical perspective is applied in the
31 analysis of relational and identity dimensions of texts, as well as in
 Identity and social relations in media texts 127
the analysis of representations. Given the questions I have just raised
about how relations of domination in the wider society underlie the
media construction of relations and identities, how these processes
take place in texts is a major concern in ideological analysis. Indeed,
perhaps a shift has taken place over the past two or three decades in
the relative importance, ideologically, of representational aspects of
texts and interpersonal (identity and relational) aspects of texts. Most
accounts of ideology in the media stress representational issues. Yet
perhaps relatively stable constructions of social and personal identity
and relations which have become naturalized as facets of familiar
media genres and formats (the news, magazine programmes, soaps)
are now more ideologically significant in the implicit messages they
convey about people and relationships than the variable representa­
tional contents that these programmes may accommodate.
I want to emphasize the diversity, multiplicity, and variability of
identities and relations in the media. I am referring here not to the
obvious variation between different types of programme (current
affairs versus soaps versus television drama, for example), but to
diversity within a programme type such as news. This is partly
because sub-categories of a particular category of participant may be
involved in different sorts of activity within the one programme.
There is, for instance, a division of labour between newsreaders,
reporters, and correspondents in news programmes, which may well
involve different constructions of identity and different relationships
with the audience or other participants. Alsu, a single participant
may have a complex identity in the sense that s/he may individually
be having to negotiate a number of activities successively or simul­
taneously within a programme, involving perhaps the sort of
changes in 'footing' described by Goffman (1981). But even a parti­
cular single activity may involve a multiplicity of simultaneous social
purposes (van Leeuwen 1987 - see Chapter 5) that entail complex and
potentially contradictory identities and relations. A presenter may,
for instance, be trying to simultaneously manage the roles of pur­
veyor of authoritative information and entertainer, while also trying
to project herself or himself as an 'ordinary person', like the audience
(see below for examples). Furthermore, the practices of a particular
type of output such as news or documentary may be variable, so that
a presenter, for instance, has a range of models to choose between.
An added complication is that this complex picture is changing
through time. The diversity of available models at a particular point
m time may be seen as the synchronic effect of a longer-term process
of change. I referred for instance in Chapter 3 (page 46) to Kumar's
128 Identity and social relations in media texts
work (1977) on changes in the institutional voice of the BBC, which
shows how presenters have shifted in a more populist direction,
claiming common identity with audiences. There has been a shift in
the relative weighting of broadly collective and institutional and
broadly personal aspects of identity in favour of the latter, with
adherence to defined role on the part of presenters becoming rela­
tively less important, and projection of an attractive personality
becoming relatively more important.
The focus on the construction of relations and identities in this
chapter, as opposed to the focus on representation in the last chapter,
entails a concern with a different set of linguistic features of texts,
namely those which are associated with the interpersonal function in
texts. These include the linguistic systems of mood and modality,
which are concerned respectively with clause and sentence types
(choice between declarative, interrogative, and imperative clauses
and sentences), and the stance of speaker or writer to 'message' - the
degree of affinity with or commitment to a proposition expressed by a
speaker or writer, for instance. They also include what I broadly refer
to as 'interactional control features', including turn-taking (the way
in which talking turns are distributed in, say, an interview), exchange
system (organization of, for instance, interviews in terms of
question-answer sequences), control of topics and topic change, and
formulation (ways in which earlier parts of a text or interaction are
paraphrased). And they include features of texts which are relevant
to 'politeness' in the sen se in w hich that term is used within prag­
matics (Brown and Levinson 1978, Leech 1983) - features, for
instance, that may mitigate a particularly challenging question in an
interview.
In this chapter I shall explore some of these issues through analysis
of extracts from four broadcast programmes: the regular BBC Radio 4
medical programme Medicine Now; High Resolution, a six-part series
broadcast on BBC Radio 4 in 1992, described by Radio Times as 'explor­
ing the popular side of science'; a popular television programme pro­
duced in the USA, the Oprah Winfrey Show, broadcast in the UK
originally on Channel 4 and now on BBC2; and the Today programme
which is broadcast every weekday morning on BBC Radio 4.

Medicine Now
The edition of Medicine Now which i refer to here was broadcast on 12
August 1992.1 shall focus on one item in the programme which deals
 Medicine Now 129
with the influence of patients' mental states on the incidence of epilep­
tic fits, and possibilities for controlling mental state to avoid fits. The
item is structured as follows: (1 ) an opening account of these new
developments in the treatment of epilepsy by the programme's
presenter, (2 ) an interview between the presenterand an expertin this
form of treatment, identified as Dr Peter Fennick, (3) an interview
between the presenter and one of Dr Fennick's patients, identified as
Kathleen Baker, on her particular case and treatment, (4) further inter­
view between the producer and Dr Fennick, (5) a wind-up of the issue
from the presenter. There is actually some slippage between stages (3)
and (4), in that there is for a short period a three-way interaction with
the presenter questioning Dr Fennick and Kathleen Baker in turn. The
item is quite a long one, lasting approximately seven minutes, so I shall
use a number o f short extracts from different stages.
The first extract is the opening of the item. I have transcribed the
extract as normal orthographic sentences because the presenter is evi ­
dently reading from a written script, with sentence boundaries clearly
marked. In the first three sentences I have italicized all the words
which are madeprosodically prominent in the delivery, either because
they are heavily or contrastively stressed, or because they carry an
international nucleus (main pitch movement). I have also marked with
a slash the boundaries of tone units except where they come before a
full stop, question mark, or colon - these punctuation marks also corre­
spond to tone unit boundaries. (A tone unit is a stretch of talk contain-
InCT 3 loCMS of nitrh OTOrn]H?‘HC^ UpV/^fCL" Ci

downwards or a combination of the two - also called a 'nucleus'.)

E x tra ct 1

presenter: Most illness/ is affected to some degree/ by the sufferer's

state of mind. In the case of certain attacks of epilepsy! it may even be
what's in the patient's mind/ that sets them off! and this raisesI an intri­
guing question: might conscious attempts to avoid/certain states of mind!
help to prevent! epileptic seizures? Sometimes dramatic symptoms of
epilepsy/ are caused by a wave! of abnormal electrical activityI sweeping
through/ the brain. Exactly what then happens, loss of consciousness,
loss of sensations, convulsions or whatever, depends un which parts
of the brain are affected. There's more than one kind of epilepsy, but in
some patients it's possible to detect abnormal brain activity between
seizures as well as during them. This often takes the form of small elec­
trical discharges referred to as 'spikes'. If patients could learn to
suppress these discharges, or at least reduce their frequency, they
might prevent a full-blown seizure from beginning.
130 Identity and social relations in media texts
The first issue I want to take up is how the identity of the presenter
is constructed. My summary of the structuring of the item into five
stages indicates that the presenter is engaged in different activities
within the programme. In particular, he is engaged in exposition (in
the extract above from stage 1 ), interviewing, and the managerial role
implied by the term 'wind-up' in stage 5 (which consists solely of: Dr
Peter Fennick and a more thoughtful way of fending off epilepsy). In fact,
managerial work is also being done in stage 1 , as well as in the inter­
viewing: the first sentence of the extract above, apart from prefacing
the account of new approaches to epilepsy, manages the transition
between this and the previous item, which dealt in more general
terms with the influence of mind on body. The picture is actually
rather more complex, because (as I show below) the interviewing
involves not just asking questions but also comment and evaluation.
The range of these activities in itself entails a multiple and complex
presenter identity, and there are further complicating factors I deal
with below.
Having grossly identified the range of activities, the next question
is how each activity is actually handled. As my discussion of the High
Resolution programme wall showy there are options available to
presenters in respect of the construction of identity and the construc­
tion of relations. In the extract above, representing stage 1, the
presenter adopts the position of announcer, delivering an orienting
account of the new research on epilepsy which he does not claim as
his own. Cuntpcu isuu with the High RcsolnOou extract shows that the
option exists for a presenter to take more personal responsibility for
such an account, but in this case the option was not taken up. The
comparison also shows that the presenter can address and engage
with the audience as an entertainer, even a comic, and can in various
ways claim solidarity and co-membership with the audience, but
again these options are not taken up here. The audience is con­
structed as seeking information rather than entertainment, and the
presenter does not go beyond the conventional announcer's role of
facilitating the informational and educative process. A fundamental
contrast between the Medicine Now and High Resolution programmes
is that the presenter's institutional identity (role) is foregrounded in
the former, whereas the presenter's personal identity is foregroun­
ded in the latter.
Let us look in more detail at the first extract. The words on the
page, as they appear in the transcription, could have come from a
lecture, but when the delivery is taken into account it becomes dear
that the presenter is not straightforwardly lecturing. Apart from one
 Medicine Now 131
question which is there for expositional purposes (the text goes on to
answer it), the extract consists of clauses which are declarative in
mood (on mood, see Halliday 1985, Quirk ct al. 1972) and make asser­
tions, though the last sentence differs from the others in being condi­
tional (conditional clause + declarative clause). These assertions are
authoritative. They are not attributed to others, they are made on the
author's authority. The modality is not always categorical, i.e. these
are not just black-or-white assertions of truth or falsity (though one or
two are, e.g. there's more than one kind of epilepsy). Some are quite
extensively modalized - these are assertions about degrees of prob­
ability and usuality, as is indicated by the number of modal verbs
(may, might), temporal adverbs (sometimes, often), quantifiers (most,
some), and the modal adjective possible. These modalizations do not
diminish the text's authoritativeness, for they evoke the cautious
(and authoritative) discourse of science and other academic disci­
plines in their careful specification of probabilities (on modality, see
Halliday 1985). The vocabulary is not severely technical, and ordin­
ary language expressions are used in preference to professional ones
in some cases (e.g. what’s in the patient's mind, sets them off); notice also
the acceptable vagueness of ordinary language in convulsions or
zvhatever. But quite a number of semi-technical expressions are used
(e.g. epileptic seizures, electrical discharges) which also contribute to
authoritativeness, as indeed does the evaluation and this raises an
intriguing question: such an evaluation implies the author's expertise.
Taking together these features of mood (declarative elausts do asser­
tions), modality, and vocabulary, the audience addressed is a well-
informed but not specialist one (though sufficiently in tune with
scientific curiosity to be 'intrigued' by the question), and audience
members are positioned as learners.
However, the extract is read out with an announcer's delivery
rather than a lecturer's delivery. There are a number of markers of
this delivery style. First, the text is divided into a large number of tone
units. This has the effect of dividing the information here into a large
number of small information units, a typical feature of announcer
talk. Second, an unusually large number of words are stressed - all
the underlined words in the transcription. Third, quite a few of these
have exceptionally heavy stress or contrastive stress. Fourth, the
presenter uses a greater pitch range than one would expect in conver­
sation or lecturing. On these and other features of announcers'
deliveries, see van Leeuwen (1984).
The contrast between the announcer's delivery and the authori­
tative lecturing style of the language points to an ambivalence in the
 il
Mis
132 Identity and social relations in media texts
presenter's identity. As a reporter with specialized interests in
medical journalism, the presenter is in a marginal position between
medicine and the media: he is in possession of a certain amount of
what normally counts as expert knowledge, but at the same time he
lacks the credentials of expertise and is in the business of mediating
experts to radio audiences. We can see this in terms of an ambi­
valence of ownership of these assertions. From the transcript, it
might appear that the presenter is claiming to be author (the person
whose words these are) and principal (the person(s) whose position
is represented in the account) and not merely animator (the person
who says the words on this occasion), yet the deliver}' belies this
appearance - see Goffman (1981) for these distinctions.
In stages 1 and 5, the presenter is addressing the audience directly,
whereas in stages 2 -4 he is addressing the other participants directly.
The next extract follows immediately upon extract 1, completing
stage 1 and initiating stage 2 (the interview with Dr Fennick).

E x tra ct 2

presenter : Dr Peter Fennick of London's the Institute of Psychiatry

identifies various strategies by means of which patients may be able to
control their brain activity one he says is bio-feedback
dr fennick : that involves taking these abnormal . electrical dis­
charges . converting them into a form that the patient can see . and
then asking them to suppress them . for example you can put a couple
or electrode:, on the head that'* just hnw vou measure the electrical
activity . amplified . in an amplifier and then turn it into . a
pulse . which triggers a counter . so every time a spike
occurs . then the counter advances itself by one and you say . stop it
counting . and they have to find some way of stopping it counting
presenter: you say they have to find some way of stopping it counting
you can't really give any instructions about how they set about con­
trolling the electrical discharges inside their own heads
dr fennick: well the interesting thing is that if you do this many pieople
have strategies of their own anyway . and they finally . sometimes
say . well I've been doing this for many years . in other words
there's something that they know they do to stop these di- dischar­
ges . but they didn't know they were doing it
Pi
presenter : how effective is it [ mean in in people in whom you can see
these kinds of e abnormal e: patterns of electrical activity between
fits . how many are actually able to control them in this way
Dr Fennick's first speaking turn is the only one in the programme that is
not a response to a question from the presenter. In that sense, the con­
tribution of the medical expert to the programme is heavily mediated
 Medicine Now 133
(m anaged, framed and controlled) by the presenter, and there is a
strong managerial element in the interviewing stages generally.
Notice, however, that the presenter's two questions do not have the
same character. The first is not strictly a question at all, though it
shares with ordinary questions the property of requiring a reply. It is
a comment on what the expert has said, and might be taken as a
negative evaluation. It shows that the presenter is not entirely limited
to being a facilitator. So too does a later contribution:

presenter: the the conditioning you talked about just now sounds
rather like Pavlov's dogs except instead of the dogs salivating when
they hear the bell ring in this case e: the stimulus whatever it happens
to be that was originally paired with the the fit causes a seizure
dr fennick: that's absolutely right

Correspondingly, the doctor's response here counts as agreement

with the presenter's comment rather than answer to a question.
By contrast, the second question is a proper question, which asks
the expert to develop the account he has given so far of his approach
to treatment. In the third stage, where the presenter interviews the
patient, the presenter uses only such information-eliciting questions.
Their role in the development of the patient's account of her own par­
ticular case is interesting. The next extract is an abbreviated version of
part of this stage, (p stands for Presenter, kb for Kathleen Baker, the
patient, and a string of dots indicate where I have omitted material,
fur reasons of space;:

Extract 3
v: what Kathleen . ar-is the situation on the circumstances or
the thoughts which tend to bring on a seizure in your case
kb: in my case guilt
p: . when did it start
kb: e: when I was quite young.......... attacks started occurring
f epU-
r- (.epileptic attacks
kb: y e s..........

p: and did this pattern [continue . for for [years

kb: [yes [yes
p: after
kb: there are certain attacks that I know were induced by guilt
p: what about e more recent times
kb: on two very important occasions . . .
134 Identity and social relations in media texts
A feature of this sequence is the way in which the patient's narrative
has been divided up into topical chunks for presentational purposes,
with the presenter's questions controlling the topical development,
moving from what causes her attacks, to when they started, to how
long they went on, to what's happening now'. The mediating and
managerial work of the presenter as interviewer is particularly clear.
These differences in the exchange systems and in the turn-taking
between stages 2 and 3 mark a contrast between the presenter's
relationship to the doctor and his relationship to the patient. (An
exchange system is the system operative within a particular discourse
type for distributing turns at talking between participants. Exchange
systems and turn-taking have been extensively documented by con­
versation analysts. See Sacks ct al. 1974.) Given the ambivalence in
the identity of the presenter as a specialist medical journalist, the
former relationship approximates a relationship between equals. The
presenter, as we have seen, comments on and evaluates what the
doctor says on the basis of his own knowledge of the field, and his
questions (e.g. the second question in extract 2 ) also display a
knowledge and understanding of the subject-matter. It is noteworthy
that the doctor seems to make (as I show below,’) more accommo­
dations towards the lay audience than the presenter does: the latter's
questions are rather complex and som etim es difficult to understand.
By contrast, the relationship between presenter and patient is an un­
equal one. The marked degree of presenter control over the present­
ation of the patient's story might be seen as motivated by the need to
make it digestible for the audience. If so, why is there no such accom­
modation to the audience when the presenter interviews the doctor?
It may, on the other hand, show that the presenter sees the patient as
needing more guidance. Notice also that whereas the doctor is not
addressed, the patient is - by her first name. All this points again to
the complexity of presenter identity.
What about the identity of the doctor? As with the presenter, there
are alternative models available. In fact the doctor adopts a voice
which is quite conventional amongst professionals of this sort, it is a
markedly pedagogical voice, which combines an authoritative
account of the research with considerable accommodation to the lay
audience. The delivery is slow and measured, showing considerable
and indeed exaggerated care to ensure clarity and comprehensibility.
It is also, at times, an expressive delivery. For example, one accom­
modative feature of the pedagogical style in extract 2 is direct quota­
tion (recall the discussion of discourse representation in Chapter 5),
not of authoritative sources, but of patients or the doctor himself (you
 High Resolution 135
say stop it counting, they finally sometimes say well I've been doing tins for
many years), and where this occurs the doctor dramatizes it by simu­
lating the way it was actually said. Or again, he turns the conclusion of
his second turn (but they didn't know they were doing it) into a sort of
punchline by a marked reduction in loudness, a very slow/ tempo, and
a rhythmical delivery which divides it into three feet each with a
stressed syllable (in capitals): but they DIDn't/KNOW they were/ DOing
it. Similarly, later in the interview, when the doctor declares seizures
can be learnt, again as a punchline at the end of the turn, he delivers it in
a slow, quiet and breathy voice. The cumulative effect of these features
is a heavily marked teacher-learner relationship between expert and
audience which in the context of other contemporary constructions of
expertise (see the discussion of High Resolution below) sounds
patronizing. Part of the pedagogical accommodation is an accommo­
dation towards certain features of informal conversational language -
notice the use of the indefinite pronoun you and the non-specific
quantifier a couple (Quirk cl al. 1972).
Both presenter and doctor are constructed in this programme in
conventional wavs, as traditional professionals. The presenter is con­
structed as a professional broadcaster whose primary concerns are
facilitation and mediation, though he is in a somewhat ambivalent
position with respect to medical knowledge as a specialist in ibis area
of broadcasting. The doctor is constructed in conventional terms as
an expert with pedagogical skills. Notice that the accommodation to
audience in the doctor's style is not specifically an acacmmodation to
the media: one can imagine something very like this as a first-year
university lecture. This is an accommodation made from within the
profession on the professional's terms.

High Resolution
High Resolution is a popu lar scien ce radio program m e. T he topic o f the
particular edition I am using, broadcast on 8 September 1992, is
bones: what different fields of science can tell us about the way
people lived in the past through analysis of bones found in archaeo­
logical excavations. The extract begins quite early in the programme
with the presenter working on an archaeological site, and in the pro­
cess of uncovering a skeleton.
presenter (sounds of scraping of trowel on soil): 1 just want to have a
look in the soil . surrounding the skeleton . to see if i can find out if
136 Identity and social relations in media texts
he was buried in a coffin . . . what I'm looking for is a line of iron nails
or a . thin dark line in the soil that might be left by the side of the coffin
(song: 'Dig dig dig from early morn to night') now this is interesting . .
I've found a small brass pin . in the rib cage . (sounds of seagulls) it's
about the size of those tailor's pins you get in men's shirts when they're
new . but this one's a shroud pin . so there we are . this was a
shroud burial . no coffin . just laid in the grave in his shroud . now
as an archaeologist I can work out details like these . but I can't say
much more about the skeleton itself . in fact 1 can't even say what sex
it is . and when we take the bones to anatomists like Margaret
Bruce . that's one of the first questions we ask
Margaret bruce: what we do is to look for the bits where you would
expect the best difference it's in the pelvis . what we've got in a female
pelvis . is a big and roomy one for a baby . the male doesn't need that
so his pelvis is much narrower
presenter: well that's alright for the skeletons of people who actually
grew up to be adults . but how do anatomists sex the more junior
members of the cemetery
Margaret bruce: we can't really with any reliability
tell . boys . from girls . it's only at puberty . that the real differen­
ces in the sexes emerge
presenter: now believe it or not . although it's easier to sex an adult
than a child . it's the other way round when it comes to answering
another common question . how old were people when they died
Margaret bruce: it's much easier for us to tell how old a child was . at
death . because we've g u t p e r h a p s f i f t e e n y e a r s of de'^'m-ment with
nice clearly demarcated stages along dental and . skeletal develop­
ment . once we get older . we're really looking at the process of
degeneration . and as we all know just looking at people sometimes
the grey hair comes in quickly . sometimes the . aches and pains
come in quickly . so we age at different . rates . and it's very diffi­
cult to put a precise age . on an adult skeleton . easier . on an
immature one
(bone song)
The extract is in two stages which are roughly comparable to stages 1
and 2 in the Medicine Now item - direct address of the audience by the
presenter, followed by interview with an expert. But there are major
differences. In the first stage, the presenter is actually engaged in
research rather than just giving an account of it. And in the second
stage, the presenter is framing and commenting on the expert's contri­
butions but not directly addressing the expert, not asking questions.
In the first stage, the division between presenter and expert breaks
down, for the presenter is a working archaeologist. Correspond-
ino-iv in Coffman's terms (see page 132) he is constructed as author
 High Resolution 137
and principal of what he says, not just as animator, i.e. it is he whose
words these are and whose position this is, not just he who says the
words. Also, this programme differs from Medicine Now in combining
information with entertainment, the most obvious marker of which is
the periodic insertion of songs about bones, digging, and so forth.
The presenter is also constructed as an entertainer, an actor, engaged
in a simulation of archaeological work - this is obviously a dramatic
reconstruction of a dig, not a recording of a real dig. But he is also
constructed as an ordinary person, a co-member of the world of
common experience, the 'lifeworld' (Habermas 1984) which audience
members are positioned within. We thus have a configuration of pur­
poses, informing and educating, and an associated configuration of
identities in the person of the presenter - giver of information, enter­
tainer, and fellow member of the lifeworld, which is characteristic of
the contemporary media, and which is widely displacing the more
traditional model of Medicine Now. A further consequential aspect of
presenter identity is that the presenter's personal identity and perso­
nality are foregrounded, in contrast to the foregrounding of institu­
tional role in Medicine Now.
How are these properties o f presenter identity realized in the lan­
guage? Delivery is again at least as important as the words on the
page. One significant difference between High Resolution and Medi­
cine Now is that whereas both presenter and expert in the latter are
men with middle-class 'received pri>rmnriaHnn' a'-'-ents (Trudgill
1986), the presenter in the former has a Mancunian accent, and
several of the experts are women with regional accents. This in itself
indicates the bond with ordinary life implicitly claimed by the High
Resolution programme. More generally, both the words on the page
and the delivery are reminiscent of an operational form of ordinary
conversation - the sort of language you might use face-to-face to talk
someone through a job you were doing. This is relevant to both the
entertainer and the 'ordinary bloke' elements of the presenter's
identity- it is an entertaining simulation of an ordinary-life scenario.
As air example, take the beginning of the extract, I just want to have a
look in the soil . surrounding the skeleton . to see if 1 can find out if he was
buried in a coffin, and the sequence that follows the song, now this is
interesting . . I've found a small brass pin . in the rib cage . it's about the
size of those tailor's pins you get in men’s shirts when they’re new. H ere, I
just want to and now this is interesting are conversational formulae, the
former recognizable as operational language; have a look and find out
are conversational-style verbs; referring to a skeleton by gender (he),
The use of the demonstrative determine!’ (those) to refer to an item of
138 Identity and social relations in media texts
common experience, use of about for approximation, and use o f you as
an indefinite pronoun (equivalent to the middie-class pronoun one)
are all features of informal conversation. The delivery is effective not
only in making this plausible as a simulation, but also in projecting a
particular, attractive, personality for the presenter. Personality is
expressed through a configuration of linguistic choices, including
vocabulary, accent, pace of talk (slow and measured in this case),
expressiveness (for instance, I just want to have a look is delivered in a
breathy voice, and in now this is interesting, pitch range is used to
express interestingness), and so forth. At the same time, the
presenter is giving archaeological information in stage 1 , though in
an experiential way rather than through abstract theoretical
discourse, by way of the presuppositions which are drawn upon in
the account of the particular find. For example, instead of asserting
corpses were either buried in coffins or in shrouds, or a line of nails or a thin
dark line in the soil provide evidence of coffin burials, these propositions
are implicitly present as presuppositions.
As I have already indicated, in the second stage the presenter does
not directly address or question Margaret Bruce, but his interven­
tions in her account do frame and control what she says in the sense
of orchestrating change of topic, commenting on what she says (well
that's all right for skeletons of people who actually grew up to be adults), and
formulating (Heritage and Watson 1979) what she says (his last con-
i-rihntion summarizes hers in advance). In terms of presenter-
audience relations, the mediating role or the presenter rs accentuated
through maintaining the audience as addressee throughout; the
presenter is talking to 'us' about 'them', extending his implicitly
claimed co-membership of the audience lifeworld to a claim to repre­
sent the audience point of view in commenting on the experts. This
could be seen as moving towards the more thorough-going demysti­
fication of experts which we find in audience discussion programmes
like Kilroy (see the discussion of the Oprah Winfrey Show below, and
Chapter 9). At the same time, the presenter continues to entertain
and inform - the comment containing the joke about the 'more junior
members of the cemetery' is also informative in indicating a problem
with sexing skeletons of children.
The identity of the expert in High Resolution differs from that of the
doctor in the Medicine Now extract in that while the latter is overtly
constructed as a pedagogue, the pedagogic function of the former's
talk is implicit; and whereas the latter comes across as patronizing,
the former does not. The expert here seems to be a knowledgeable
person talking in a conversational way, rather than a professional
 The Oprah Winfrey Show 139
making obvious efforts to talk to lay people. Both, of course, are pro­
fessionals, but the difference is between more traditional and more
current models of professional behaviour. Currently, professionals
are widely induced to regard talking in an ordinary conversational
way in public contexts (commonly referred to in advertisements for
professional posts as an ability to 'communicate') as a part of their
repertoire. This is a matter of accommodating professional practices
to the demands of external agencies, and most notably the media.
The differences between the two experts are again partly matters of
delivery and accent (the High Resolution expert has a Scottish accent),
but notice also the use of pronouns. The expert in High Resolution
makes extensive use of we. This personalizes the anatomists, in con­
trast with the impersonal construction of scientific procedures in
some of the doctor's formulations in Medicine Noiv (e.g. that involves
taking these abnormal electrical discharges . converting them into a form the
patient can see). But there is also a slippage between 'exclusive we
('we the anatomists' - e.g. easier for us to tell how old a child ions) and
'inclusive we' ('we human beings' or 'we members of this society' -
including the audience - e.g. once we get older), marking a certain
ambivalence in Margaret Bruce's identity between the expert and the
ordinary person.
To sum up the contrasts in presenter and expert identities and rela­
tions with audiences between Medicine Noio and High Resolution: the
emphasis in Medicine Now is upon institutional roles and relations,
with the presenter as facilitator and mediator and the doctor as pro­
fessional expert and pedagogue; the emphasis in High Resolution is
upon personal identity and personality and a simulation of lifeworld,
conversational relations with the audience, as well as (in the case of
the presenter) entertainment of the audience.

The O prah W in frey S h o w

The Oprah Winfrey Show is a popular television show produced in the

USA but widely distributed in other countries including the UK. Each
show addresses a topic of concern to people in their social or personal
lives, with a panel of invited guests including ordinary people talking
about their own experiences, and contributions from experts and a
studio audience. Oprah Winfrey orchestrates the various contri­
butions, positioning herself with a hand-held microphone in front of
or within the studio audience. I shall refer to a programme broadcast
in the UK on 9 September 1992 which dealt with women who were
140 Identity and social relations in media texts
'dumped' by boyfriends when they were young (but not, inci­
dentally , with men dumped by girlfriends). For legal reasons, I am
unable to quote directly from the transcript.
One striking feature is the diversity of the voices that are given
space in the programme. These include expert voices, but are pre­
dominantly the voices of ordinary people. And in contrast to some
media output which gives access to ordinary people, these people
sometimes give opinions and even speak authoritatively on the
issues, as well as recounting their experiences. The diversity of voices
arises partly from the selection of guests to represent a range of
opposing perspectives, partly from the relationship between lay and
expert voices, and partly from the variety of perspectives voiced from
the audience. One interesting feature of these programmes with
respect to relationships between participants is the management of
diversity: how an Oprah Winfrey Show orders these various voices in
relation to each other, how certain voices frame others, how voices
are hierarchized, and in particular how the voices of ordinary people
are given space yet at the same time contained and managed.
The relationship between the voice of the expert and other voices in
the panel and the audience is particularly noteworthy. The expert is
introduced as a 'therapist' and author of two books on love and the
breakdown of relationships. In one part of the programme an opinion
is expressed, and forcefully and eloquently expressed, by an audi­
ence member. But drat opinion is, interestingly, referred by Oprah
Winfrey to the expert, who in this case endorses it. the same dung
happens a t the end of the programme - in fact the programme actu­
ally closes with the therapist still speaking, which underscores the
way in which other voices are framed by the therapist's. Audience
and panel members may be allowed opinions as well as experiences,
but their opinions are given value through being endorsed by the
expert.
The vvay in which the therapist is constructed as expert is in con­
trast with both Medicine Now and High Resolution. The expert here is
herself an accomplished media professional, a good 'communicator'
in a common contemporary sense of that term, in contrast with the
medical professional of Medicine Now, and the anatomist expert of
High Resolution (who accommodates to media requirements for a con­
versational mode of expert talk, but is still far from the therapist's
media-designed performance). One aspect of this is the fluency and
flow of her talk. This comes across in the relative absence of pauses,
and of hesitations and other disfluencies, but also in the high degree
of structuring and verbal planning evident in her contributions. The
 The Oprah Winfrey Show 141
fluency and the flow depend, however, upon the freedom that
Oprah Winfrey gives to the therapist to hold the floor and develop
her arguments at some considerable length, in contrast with the more
heavily mediated and controlled presence of experts in the Medicine
Now and High Resolution programmes. Fluency and flow are also part
of a contrast in purpose: the therapist here aims to persuade the panel
and audience to see their own experience in relationships in parti­
cular ways, whereas there is no such persuasive objective in the other
programmes.
The difference in medium (television versus radio) is, not sur­
prisingly, important. Therapist identity and therapist-audrence rela­
tions are constructed visually as well as in language. The audience
reactions shown on camera while the therapist is talking construct
the therapist as an authority whose pronouncements the audience is
prepared to accept. Audience members are shown listening more
carefully and concentrating more seriously and intently than during
other contributions. The authoritative role of the therapist within the
configuration of voices is also symbolized in her physical positioning,
on the panel facing the audience (and Oprah Winfrey), with four
other panel members - two sets of former partners on each side of
her. A significant part of the performance is the expert's non-verbal
communication. Her talk is accompanied by almost continuous
expressive hand movements (a great deal more than any other contri­
butor's), and she also uses her body more than others in alternating
address - addressing the couple on the left, the couple on the right,
particular sections of the audience, or Oprah Winfrey. Linguistically,
her authoritativeness com es across in the modality, in the categorical
assertions she makes not only about relationships in general, but also
about the particular experiences of the ex-partners on the panel. The
therapist is making the extremely powerful implicit claim of a
capacity to interpret people's experiences for them. Significantly, the
only hint of a challenge to this claim to power comes not from the
panel or the audience, but from Oprah Winfrey, in the form of inter­
ruptions. Oprah, as presenter, claims precedence even overthe pro­
fessional mystique of the therapist. It is Oprah herself who
orchestrates the hierarchization of voices in the final analysis. Also,
the therapist is faded out in full flow at the end of the programme,
which we must read as a little covert undermining of her. There is
perhaps a hint of ambivalence about the authority of the expert here
which points to the more explicit challenging of experts in other
audience discussion programmes (Livingstone and Lunt 1994; see
also Chapter 9 ).
142 Identity and social relations in media texts
Most of the voices are, as I have indicated, voices of ordinary
people, but the fact that the programme is called a 'show' is not, I
think, irrelevant in assessing the import of what appears to be a
substantive elevation of audience to the status of participant. Ordin­
ary people, and especially those on the panel, are partly there for
their entertainment value. In so far as the programme generates con­
troversy, as it standardly does, it is partly because controlled con­
troversy makes for good television. But there is often a tension
between the serious social or emotional aspects of the issues and
experiences dealt with and the search for entertainment.
The identity of Oprah Winfrey herself is very complex. In her
capacity as manager of the hierarchization of voices in the show, the
one who holds and controls the microphone, she sometimes evaluates
contributions as well as controlling their length and order, and some­
times also seeks to reconcile conflicting voices. Oprah's identity
includes also her roles as a serious social investigator questioning the
panel, and as a chat-show hostess. In the latter capacity she is an
accomplished performer, witty, humorous, with a winning smile, a
contagious laugh and a generally attractive personality. In addition,
she is also at times a moralist and educator, directly addressing
viewers on the themes of the programme. So, like the presenter of High
Resolution, she issimultaneouslyeducatorandentertainer, though the
particular realization and configuration o f these elements is quite
different - these are different types of programme, emanating from
different cultures, and we have the contrast between a black
woman and a white British man. Also, like the presenter in High Resolu­
tion, Oprah Winfrey is constructed as an ordinary person sharing the
lifeworld of people in the studio and home audiences, though again
the realizations of this ordinariness in her talk are quite different.

Today
The Today programme is broadcast every weekday morning on BBC
Radio 4. It is particularly noteworthy for the populist, common-sense.
style of its presenters, which is all I shall discuss here. The following
extracts are from an edition of the programme broadcast on 8 April
1992 during the UK general election, involving two presenters, John
Humphrys (jh ) and Brian Redhead ( b h ). The first extract is from an
interview with an expert on elections about the possibility of a coalik
fion government. Because my focus is on presenters, I have omitted
most of the expert's contributions, 1
 Today 143
Extract 1
jh : twenty to nine so what are the odds on a hung Parliament I shall be
asking the man from Ladbroke's in a moment but first someone who's
seen it all before many times before Dr David Butler . of Nuffield
College Oxford who was studying elections when Messrs Kinnock,
Major and Ashdown were still in their prams I suspect . perhaps I'm
dating you too much there David am i
d b : 1 was first election was well forty-five I was in the army fifty 1 was it

was the first one 1 watched

jh : w e ll th ere y o u a re th e y w e r e in th e ir p r a m s a t th a t
(db: yes absolutely) so are we seeing anything different this time
d b : oh y e s ( p a r t o f a n s w e r o m i t t e d ) a n d i t ' s t h e f i r s t e l e c t i o n in w h ic h a
hung Parliament has seemed the likeliest outcome
jh : a n d y o u b e l i e v e th a t is t h e c a s e th is t im e
(a n s w e r fro m db o m itted )
jh : so b e c a u s e w e h a v e n ' t s e e n t h i s b e f o r e in q u i t e t h i s w a y i n r e c e n t
h is to r y w h a t's g o in g to h a p p e n
(a n s w e r fro m db o m itted )
jh: c o z a lot of people have been saying oh well an October election you
don't think that's likely or at least
(answer from d b omitted)
jh : a n a w f u l lot o f p e o p l e w h e n y o u a s k t h e m s a y o h 1d o n 't m u c h c a r e
for a n y o f e m w h y d o n 't w e h a v e co a litio n th a t w o u l d b e th e b e s t th in g
w h a t's y o u rview of that
(a n sw e r fro m d b omitted)

j h : but i f you ask the politicians of course they will say as you say abso­

lutely toif - oi Ndoi some ul them will say absolutely terrible idea e- it's
simply unworkable . e: is is that . purely self-interest
(answer from d b omitted)
The second extract is from a panel interview with representatives of
the three main political parties (Conservative, Labour, and Liberal
Democrat) about how they would answer questions from an imagin­
ary floating voter. I have included only Brian Redhead's questions
and omitted the politicians' answers. (Recall that 1 used part of this
extract as an example in Chapter 1.)
Extract 2
br: right now thank you all three first of all for being positive and not
being negative and knocking the others now imagine this floating
voter is very thoughtful and serious and he comes back to you he
comes back to you Chris Patten and he says well that's all very well but
■ you know you've been in power so long we're not a one-party country
we do need changes occasionally maybe you lot are looking tired and
perplexed don't you need a spell in opposition just to rethink
: . (answer from Chris Patten omitted)
144 identity and social relations in media texts
br: well now Chris Patten thank you now our floating voter turns to
you Brian Gould and he says look (bg: yeah) f don't really fancy' another
Conservative government i think we've had enough of that but 1can't
really bring myself to vote for you because you've been out of office for
so long you haven't got the experience if you get in the City might say
do this lot know enough to run the country I'm nervous that a vote for
you would mean a vote for some kind of flight from the pound
(answer from Brian Gould, question from br to Des Wilson, and
answer from Des Wilson omitted)
b r : Des Wilson thank you now . imagine this floating voter actually is
a mate of all three of you . knows you personally . and has saf up
he's a differentbloke altogether this one's been here through the whole
election he's listened to every blooming broadcast (one of panel: lucky
chap) he's fed up to the back teeth (one of panel: haven't we all) . and
he rings you up and he says the same question to each of vou and i just
want a quick answer from each if you would . he says . hey
Chris . e:m . your campaign has been dreadful . 1 mean you've just
underestimated the intelligence of the electorate and particularly of
m e . what would you why did vou get it wrong
As in earlier examples in this chapter, delivery is an important
element in the construction of presenter identify and presenter-
audience relations. But I shall focus for this final example just on the
language of the transcriptions.
The populism of the style manifests itself in a high degree of con-
versationalization. One aspect of this is the direct representation of
the discourse cf other?, inHudine an attempt to simulate the voice of
the (real or imaginary) original. Recall that the expert in Medicine Now
did the same, but the two cases differ considerably, not only in the $
sorts of voices that are simulated (roughly, they are m iddle-class
voices in Medicine Now but working-class voices here), bu t also in how m
these simulations are framed. Whereas the doctor's own voice in n
■-ft'--
Medicine Now is that of a traditional professional, the presenters' own
voices here are very similar to the voices that are simulated - both are T
the voices of ordinary experience. In other words, even when the
presenters do not signal the representation of the discourse of others,
they are still in a sense speaking in the voices of others, those of vvhal
are taken to be typical audience members, and there is nothing in the
represented discourse within these extracts that could not be the
presenters' own words. In the second extract, with Brian Redhead,
i!
the whole item is built around the simulated voice of the floating
voter, but there are also simulations in the latter pari of the first I
extract (e.g. oh l don't much cure for am/ of cm why don't we have coniitiott
that would be the beet thing).

fir
 Today 145
The conversationalization of the presenters' talk is realized in a
variety of linguistic features. Most obvious are items of colloquial
vocabulary such as knocking, fancy, mate, bloke, and so forth. Notice
also the colloquial way of quantifying inn/; azuful lot of people. There are
also cases of vocabulary items which are not per se colloquial but
which are used in colloquial senses, such as dating (meaning showing
how old someone is, as in expressions like that dates you) and negative
(meaning critical). Certain expressions evoke the specific know­
ledges, wisdom or preoccupations of the culture of the lifeworld: the
man from Ladbroke's, who's seen it all before, when (they) were still in their
prams, and so forth. Pronouns (as always) are worth noting, especi­
ally the use of inclusive we (e.g. are we seeing anything different, we’re
not a one-party country) and the colloquial use of you as an indefinite
pronoun (e.g. an awful lot of people when you ask them say). Notice also in
the second extract the colloquial use of the demonstrative pronoun
this to refer to someone (e.g. imagine this floating voter actually is a mate
of all three of you). The presenters also use discourse markers (words
like oh, well, and right) in ways that are typical of conversation, and
conversational ways of using conjunctions (such as so and coz in so
because we haven't seen this before . . . and coz a lot of people have been say­
ing . . . in extract 1 ). There are also features of conversational nar­
rative and argumentation, such as the use of present tense in
narrative in extract 2 (e.g. he comes back to you Chris Patten and he says),
and the formula for prefacing a disagreement that's all very well but.
. . . Finally, we get something of the compression and high level of
implicitness that is common in ordinary conversation in (a lot of people
. have been saying) oh well an October election in extract 1, which might be
more explicitly formulated as something like 'it seems from what's
happening that there is likely to be an October election'.
On the Today programme, as with the other programmes I have
discussed, presenter identity is more complex than I have indicated
so far. Part of this complexity derives from the duality of address: the
presenter is for much of the programme addressing an interviewee or
another reporter as well as (indirectly) the audience. In extract 1, for
example, the formulation and you believe that is the case this time would
pass perfectly well within academic discourse, and indicates an
accommodation to the expert which is generally lacking here. As with
the presenter in Medicine Now, there is a certain ambivalence - if a less
clearly expressed one - in the presenter's relationship to the domain
of specialist knowledge, academic analysis of politics in this case, so
that on occasion the presenter displays his credentials as a part-
specialist.
146 identity and social relations in media texts
This chapter demonstrates some of the diversity in the construction
o f identities and relations that I referred to on page 127. Medicine Now
is the most conservative of the four programmes I have discussed and
contrasts with the others in several ways: informing is not mixed with
entertaining, institutional distance is maintained, and institutional
roles are foregrounded over personalities. The other programmes, in
differing degrees and ways, shift towards entertainment, claimed co­
membership of the audience lifeworld, and the foregrounding of per­
sonality, though entertainment is not as salient in the Today pro­
gramme as in the other two. I have shown how these developments
variously affect the construction of presenters in the three program­
mes, and the experts in High Resolution and the Oprah Winfrey Show.
These developments also entail important differences in how audi­
ences are constructed: the audience of Medicine Now is constructed as
a group of citizens, people who have an intelligent interest in keeping
up with the advancement of knowledge. The audiences of the other
programmes (particularly High Resolution and the Oprah Winfrey
Show) are in addition constructed as consumers, as people for whom
listening to radio or watching television are leisure activities, involv­
ing the expectation of entertainment. There are also, as 1 have indi­
cated, differences in terms of the ways in which ordinary people
make the transition from audience to participant. In Medicine Now,
the patient is present as a tightly managed witness who is limited to
recounting a relevant experience, as an illustration. In the Oprah
Winfrey Show by contrast, ordinary people are the main participants,
giving opinions as well as recounting experiences, though they are
partly there as spectacles for audience entertainment, and so posi­
tioned within the presenter's orchestration and hierarchization of
voices that their opinions and experiences are subordinated to the
legitimated opinion (counting as knowledge) of the expert.
I have referred at times to the options or choices available in the
construction of presenters or experts. The notion of choice, as I indi-,
cated in Chapter 1, is a helpful one on one level but can be mislead--
ing, especially since market values and ideologies which centre;,
around choice are very .much at issue in these constructions. When If
refer to choices or options, I mean that practices are variable, that the;
order of discourse includes alternative sets of practices. This vari-'
ability, as it appears at a particular point in time, can be seen from a
historical perspective as change in progress. In this case, Medicine
Noio is representative of older media practices whereas the other pro­
grammes are more representative of newer practices which have
come to be dominant. Within the newer practices there are various
 Today 247
degrees fo which the shifts towards entertainment, personality, and
commonsensicalness can be taken. There are also various ways in
which, for instance, being 'conversational' can be realized linguistic­
ally. These are choices, but that does not necessarily imply freedom
of choice for individuals, since many of them are determined by
factors such as the type of programme, the target audience, pressures
for audience maximization, and so forth.
The concept of 'personality' is salient in contemporary media
(Tolson 1990, 1991), but it is important to clarify what is meant by
the suggestion of a shift away from institutional roles towards a
foregrounding of personality. This is often perceived and
portrayed in a way which harmonizes with the core contemporary
cultural value of individualism, in terms of a foregrounding of the
unique and individual personalities of, especially, different
presenters. The shift in the way identities are constructed can
indeed be seen as significant in projecting and spreading ideas and
values of individualism within contemporary society. Yet how
individual are these various presenter identities? Of course, there
are dear differences between individuals such as Brian Redhead
and John Humphrys on the Today programme, but then there were
dear differences between different announcers in the earlier
history of broadcasting when, institutional role was more in focus.
Even fairly rigid roles allow considerable space for individual style,
in addition to the individualizing effects of differences in voice
quality and accent. On the other hand, contemporary presenter
personalities fail into fairly dear types. The Today programme is
interesting in this regard, because it was built around the person­
ality of the late Brian Redhead over a number of years, but John
Humphrys, having taken over as the main presenter, projects the
same type of personality as Redhead through a similar style of talk.
This shows how personality can become transformed into product
image, and in that sense become depersonalized, on the leisure
market. Or again, there is the aggressive school of interviewers,
including Sir Robin Day, more recently Jeremy Paxman, and
- various others. The shift that has taken place has a personalizing
nature in the sense that aspects of identity traditionally associated
with the private domain (e.g. aggression) are being appropriated
,by the media, but that does not necessarily entail a substantive
increase in individuality. This strikes me as an ideological repre­
sentation of the shift in the sense that it reads it in terms of indi­
vidualist ideologies which obfuscate ways in which the media
reflect and help shape changes of a social and cultural order.
148 Identity and social relations in media texts
Changes in construction of presenter-audience relations, also,
may have ideological import. Consider the following sequence from
extract 2 of the Today programme:
you haven't got the experience if you get in the City might say do this
lot know enough to run the country I'm nervous that a vote for you
would mean a vote for some kind of flight from the pound

One question to ask is how particular constructions of relations (and

identities) relate to particular representations, the concerns of Chapter
6 . In this case, notice the presupposition that a flight from the pound is
a City judgement on the competence of a government. This particular
piece of preconstructed knowledge may work ideologically in repre­
senting the City in terms of a preoccupation with competence which
may obfuscate its preoccupation with profits. The construction of
presenter-audience relations in terms of shared membership of a
common-sense lifeworld, the 'communicative ethos' (Scannell 1992)
that is realized in a conversationalized, public-colloquial style, can be
seen as instrumental in legitimizing such ideological aspects of repre­
sentations (see also the discussion of Fowler 1991 in Chapter 1, page
13). It is significant, too, that the City is attributed with the same
conversationalized communicative style as the presenter and the
floating voter (notice this lot). But the communicative style also has a
potential ideological import in itself, in the implicit claims it makes
about the validity of social relations (Habermas 1984): when reporters
address audiences in such a communicative style, they implicitly ?
'ifi':-
claim, as 1 have suggested, shared membership of the same lifeworld,
i -
and in so doing implicitly negate the differences of position, perspec­
tive and interest that are implied by practices in which contrasts of k
institutional role are explicit. The same is true when politicians or £ ‘
senior managers or archbishops or (as we have seen) various cate­
gories of expert address the public in a conversationalized commun­
icative style. What I am suggesting is that the style can help in an
imaginary ideological construction of social relations whose real
nature is less symmetrical and benign (Bernstein 1990).
But the communicative ethos of broadcasting is more ambivalent
politically than this suggests. It can be read at the same time as for­
ming part of a substantive democratization of cultural life and cul­
tural relations which has given value to popular culture and ordinary
practices within the wider culture. I see relational and identity
aspects of media texts as a complex field of negotiation and contes­
tation where complex, overlapping and contradictory cultural ten­
dencies meet and are enacted, are given texture. Cultural
 Today 149
democratization takes place in a society which is built upon relations
of domination and exploitation which must be reproduced and legiti­
mized, and where the market and market-consumer relations are
colonizing new domains of social life including 'leisure' and the
media. The communicative style of broadcasting lies at the intersec­
tion of these democratizing, legitimizing and marketizing pressures,
and its ambivalence follows from that.
 8

CRiMEWATCH UK

In this chapter and Chapter 9, I shall be applying the analytical

framework, presented in previous chapters, in two case-studies. My
focus in this chapter is upon a programme which has been appearing
regularly on BBC1 television for a number of years, Crimexvatch UK.
The programme enlists public help for the police in the solving of
crimes. It includes re-enactments of crimes, interviews with the
police, and public appeals for information which could help solve
crimes. I shall also refer for comparison, and more briefly, to a some­
what similar programme also regularly shown on BBC1 over the
same period, 999, which is described in the weekly TV and radio pro­
gramme guide Radio Times as 'dramatic stories of real-life rescues,
reconstructed by the emergency services, actors and those people
involved in the actual rescues'.
My main concern will be with how television in these instances
contributes to a particular construction of the relationship between,
the public and the state - more precisely, between the public and the
police in the case of Cr'unewatch UK, and between the public and
the emergency services (fire brigade, police, ambulance service) in the
 The C rim ew atch format 151
case of 999. 1shall suggest that we can use discourse analysis to throw
some light on this general issue by focusing upon the nature of the
discourse practice in these programmes - upon how genres, voices,
and discourses are articulated together. I shall look at the mix of
public information and entertainment, and of official discourse and
the discourse of ordinary experience; and how relationships between
the voices of the police, emergency services, ordinary people, and
media journalists are structured. Wha 1 1 think is going on in both pro­
grammes is intervention to shore up the crumbling public legitimacy
of the state.

The C rim ew a tch format

Crimewatch UK programmes are put together according to a relatively

fixed formula involving combinations of a small number of regular
types of feature: 'Photocall', 'Incident desk' and 'Aladdin's cave'
-(unclaimed and stolen property) are regular slots which are intersper­
sed with a number of re-enactments of crimes. Reports and updates
on cases treated in previous programmes or earlier in the programme
-information about responses from the audience and what the police
have done on the basis of them - are another element. Here, for
example, is how one edition of Crimewatch (broadcast on IS February
1993) was structured:
Opening sequence
James Bulger case: latest information
. Re-enactment: Claire Tiltman murder
Report: previous Photocall
Photocall
~ __ Re-enactment: Muriel Harvey rape
Update: James Bulger feature
Incident desk
Re-enactment: Doncaster robbery
Aladdin's cave: unclaimed and stolen property
- Update: James Bulger and Claire Tiltman features
Closing
There is, additionally, a separately scheduled Crimewatch Update one-
and-a-half hours after the end of the programme.
Let us look more closely at the elements in this structure. The
opening sequence is a series of shots showing members of the public
witnessing criminal activity (mainly getaw ays)-one watches from an
upstairs window, another from a passing car, another is a pedestrian
152 Crimewatch UK The Crimewatch format 153

with a good view of a number plate. The last shot in the sequence the police, presenters in voice-over, while images on the screen
shows a disembodied hand lifting a telephone receiver and dialling - change according to the following sequence: photograph of suspect;
presumably - the police. The shots highlight the potentially danger­ pictures of location of crime, vehicles used, etc.; 'mug shot' of suspect
ous consequences of crime for ordinary people not directly involved: with personal details shown on screen (age, height, build, hair-
they show some near misses - a pedestrian with a pram narrowly colour); original photograph of suspect, with a voice-over appeal for
missed by a speeding vehicle, a passing motorist almost in collision help. 'Photocall' ends with a number for people with information to
with a getaway car. The programme's assumptions about relation­ phone. 'Incident desk' differs from 'Photocall' in that the police
ships between criminals, the public and the police are dearly presenters are on-camera addressing the audience.
signalled at the outset. The dramatic percussion music is reminiscent The 'Aladdin's cave' of unclaimed and stolen property, the last
of the theme tunes of police drama series. substantial feature before the end, provides light relief after the cata­
The featured re-enactments have a regular and predictable struc­ logue of crimes. This feature evokes a popular television programme,
ture: an on-camera introduction by the presenter, a filmed re­ The Antiques Roadshow, in which members of the public bring their
enactment of the crime, an interview by the presenter with a police prized possessions for identification and valuation by experts in
officer, an on-camera dosing by the presenter. Appeals for help from antiques. The items of stolen and unclaimed property shown are
the audience sometimes occur within the filmed re-enactment and in exclusively (apparent) antiques, and they are displayed as they might
the interview, but they are particularly the focus of the presenter's be in an antique shop. This feature has its own presenter, who sets
dosing which always includes a telephone number for people to call. himself off from the others both in his appearance (a bow-tie and
Appeals for information are directed at possible witnesses of the , moustache are antique-dealer style rather than police or journalist
crime, but suspects and the families and friends of possible suspects style), and by giving what might pass as specialist descriptions of
are also urged to come forward. Constantly repeated motifs are that . some of the items. For instance:
there is someone out there who knows who committed the crime,
well I'm quite taken by these e this pair of glass bon-bon dishes
and that the criminal may strike again. d well . not so much the glass dishes but more the silver frames I love
The first feature, on the James Bulger case, has the substance and; . this cast swan . and he rests upon a little lion's paw foot . now
some of the structure of a re-enactment, but is actually not a typical : they've got e: English hallmarks but they they were imported if you're
feature. James Buiger w a s a sm a ll child who la ter proved to have been:: --- the owner . you ii know where they were imported from
murdered by two older boys. The case attracted a great deal of public;;
The discourse is sometimes mixed, in that collector's descriptions are
attention. The item here is a summary of the latest information on the.
sometimes merged with the police identification of distinguishing
case. There is no filmed re-enactment, but instead an extended marks.
presenter account of new security video pictures of the two boys, the
The closing of the programme brings in another common Crime-
exact route they took with James (with precise timings), what James;; match motif:
was wearing, and a forensic psychologist's report about the likely
identity and behaviour of the two boys. This is followed by the usual | must be said that e: after the grim litany of crimes we've shown tonight
sequel to a re-enactment: an interview with a police officer, plus a | itmay be worthwhile pointing out the obvious that . most people are
closing appeal for help, on-camera, from fhe presenter. h decent and . in their hundreds if not thousands . are now calling in
In both the 'Photocall' and 'Incident desk' features, police officers j ■; to help so if you can't stop up till Crimewatch Update please . don't
have nightmares do sleep well . goodnight
themselves take over the job of presenting. The police presenter team j
mirrors the team of journalists who orchestrate the programme as a j It.is commonly claimed that 'moral panics' about crime are a char­
whole - in both cases there is a team of two, a man and a woman, and j acteristic of contemporary society - that people's fear of crime is in
they alternate in presenting items. 'Photocall' and Incident desk'” many cases disproportionate given the actual idsks they are subject
give brief accounts of crimes, without filmed re-enactments, and j to. Programmes like Crimewatch might be regarded as contributing to
again appeal for audience help. 'Photocall' is built around a set of four, moral panics, and such attempts to reassure people indicate sensi-
photographs of suspects. Each case is briefly summed up by one of tivity on that score.
254 Crimewatch UK
What I have said so far about'the structure of this edition of Crime-
watch gives a preliminary sense of the articulation of voices in the pro­
gramme, which is, as I have indicated, one of my main focuses. It
indicates that voices are not neatly associated with roles - in parti­
cular, that the presentation of features is shared between media
presenters and police officers, and the presenters are doing police
work in, for instance, appealing for witnesses. This fudging of the
difference between mediators and public officials is one element in
the process I referred to earlier of restructuring of police-public rela­
tions in a way that helps legitimize the police, as also is the closing
feature which brings to police work the benign and popular reson­
ances of The Antiques Roadshow.
I n ow w ant to take the analysis further by focusing in on one of the
re-enactment features, the Claire Tiltman murder. In terms of the
critical discourse analysis framework presented in earlier chapters,
the emphasis will be on discourse practice and intertextuality, sup­
ported with selective textual analysis. Comments on sociocultural
practice will follow later. I shall begin looking at the articulation of
genres in the re-enactment, then look at how voices are combined
together, and finally at the main discourses that are drawn upon.

Generic analysis

1 have indicated that re-enactments in Crimewatch are internally

complex, involving a diverse set of elements in a predictable
sequence:
Presenter's introduction
Film ('re-enactment proper')
Interview between presenter and police officer
Appeal for audience's help by the presenter
These are indeed the main elements of the re-enactment under;
scrutiny, but this gives only a rough and inadequate idea of its gen­
eric complexity, because the film is itself internally complex. So I shall
begin with a fairly detailed summary of it.
The first point to make is that it is not a simple narrative of what
happened to Claire Tillman. The narrative element is preceded and
followed by non-narrative elements:
Fire-fighting sequence
Narrative of Claire's last day
Pool-game sequence
 Generic analysis 155
The first of these shows Claire (played by an actress) learning about
fire-fighting with the fire brigade as part of a Duke of Edinburgh
Award scheme, actually using a fire hose on a burning building. The
sequence is dramatized, with some dialogue between Claire and a
fireman, and a lot of squeals and laughter. Claire comes across as a
pleasant and lively teenager. The sequence is accompanied by voice­
over from Claire's (real) father and mother - for part of this, the fire­
fighting film is interrupted by a close-up photograph of Claire and film
of Claire's parents sitting together on a couch - and from the fireman
who is a 'character' in the film. The final pool-game sequence is similar
-it shows Claire playing very successfully at pool, with an accompany­
ing voice-over from her parents talking about her in teres t in and skill at
the game, during which the picture shifts to Claire's parents again
seated together on the couch. The sequence (and the film) closes with
an appeal by Claire's mother to the wife, girlfriend or mother of the
murderer to come forward. The central and longest part of the film is
the narrative of Claire's last day. The main episodes are:
Claire preparing to go to school
Claire's journey to school (she arranges to visit her friend that evening)
Claire taking an examination at school
Claire back home
Claire walking to her friend's house
A witness driving past where Claire was killed seeing a suspect car and
a girl (Claire) running out of an allev
The discovery of Claire's body, attempts to revive her
Apart from the examination, Claire's walk and the witness in the car,
these episodes involve dialogue, and most episodes also have
elements of 'commentary', mainly voice-over.

Voice-over: generic complexity

I want to focus on the generically diverse roles of voice-over as a way
into the generic complexity of re-enactments. Voice-over serves a
variety of functions in the film and draws upon a number of genres.
The main genres are what I shall call narrative, biography, and public
Appeal. Some of this generic diversity is exemplified in the following
example of the presenter's voice-over:
Claire had a mile to walk to Vicky's home . were you driving down
the A226 . here called the London Road through Greenhithe Kent
between Dartford and Gravesend , it's six o'clock on Monday a
month ago Monday the eighteenth of January
156 Crimewatch UK
The first sentence shows the presenter as narrator, working with the
dramatized film narrative to tell the story, in this case providing
information important to the narrative which cannot easily be shown
in the film. The question which follows (up to Gravesend) illustrates ;
voice-over as public appeal for information. Notice the official
(police) discourse manifest here in the formal description of location, «
which is 'person-neutral' in the sense that it makes no assumptions
about audience knowledge, and no claims of shared knowledge or t
co-membership between presenter and audience. The third element y
of the extract (from it's six o'clock) is again narrative, but in this case it
has a dramatizing function in marking the onset of the climactic epi-
sode of the narrative which culminates in Claire's m urder- notice the
use here of an 'historic present' tense (it's) instead of a past tense. In
its specification of date and time, it also shares with the second
element the official function of producing a precise person-neutral
description of events, though notice that a month ago is an imprecise =
lifeworld temporal specification rather than an official one.
There are several other instances of generically narrative voice- !
over from the presenter. The following example of voice-over is ;
interesting:
the exams was the last day . she'd done quite well in them actually
coz . she wasn't a . particularly a scholar but . she did work l
hard . she done very well d

it has a partly narrative role, contextualizing the him or the examin­

ation hall with the information - actually already given - that this was.
the last day of the exams. What is noteworthy is that Claire's father is
here taking on what is otherwise the presenter's role. But it is also,
partly what I'm calling biography, it is telling us about Claire rather;
than about the crime.
In the last three episodes of the narrative of Claire's last day, from
Claire walking to her friend's house, we find generically narrative
voice-over of a special kind, from three witnesses - the jogger, the
driver, and a person identified as Michael Godfrey. I want to draw a-
contrast between these as instances of testimony and the external nar-’
rative which 1 have referred to as 'narrative' so far. The difference is
one of point of view (Toolan 1988). External narrative tells the story _
from the point of view of a narrator external to the events themselves
who is able to take an overview. The 'narrator' in this case is a pro-#'--
duction team including a film team, but mainly personalized as we 4
have seen as the presenter. Testimony is the story from the point of -
view of one of those involved, indirectly involved in this case as 3 3*
 Generic analysis 157
witness of a part of the events. The use of testimony echoes common
recent practices in broadcast documentary, allowing ordinary people
to apparently 'speak for themselves' (whereas in interview genre
their voices are mediated by the presenter) and foregrounding per­
sons and personalities (Tolson 1991).
As the jogger describes what he saw, wha t he describes is shown in
the film. Testimony is typically not just a narrative of what happened,
in this case Claire's walk as witnessed by the jogger. The focus is par­
tly upon the personality of the witness, and his/her responses to and
evaluation of events. The jogger's account is prefaced with biogra­
phical contextualization (f often go jogging along the London Road), and
ends with a report of what went through his mind (oh thank you for
letting me run in the road). We might say that this testimony is trans­
formed into the external narrative of the film which accompanies it,
with quite a long objectifying shot of the jogger running down the
road, and Claire shown actually crossing the road rather than just (as
in the jogger's testimony) behaving as if she might. This sets up a
tension between testimony and external narrative. But there is also a
parallel tension within the testimony, a contrast of styles between the
part which describes what the jogger saw', and the reported thought.
The latter is conversational and colloquial in using the rhetorical form
'I thought to m yself + direct speech (1 thought to myself oh . thank you
•for . letting me run in the road). The former bears traces of written lan­
guage (e.g. use of thus and positioning of again before noticed - ! again
rnoticed Claire in the distance), and these features plus the attempt at
•..precise factual description of place and action (e.g. she walked out
towards the main road as if she was going to cross) evoke the style of formal
official (including police) report. Official report is normally associated
with external narrative - hence my allusion above to a 'parallel
tension'.
Voice-over works in a similar way in the next episode (witness driv­
ing past the scene of the crime). Again, this is testimony, and it fore­
grounds the witness's affective responses. His disgust at the way the
car he noticed was parked is expressed both lexically (e.g. stupid in
this car that's parked in a very stupid place) and through expressive
delivery. As with the jogger's testimony, the grounding of the
account in ordinary experience and sensibility is underscored by
colloquial rhetorical forms - 'I remember thinking' + direct speech,
and 'it suddenly dawned on me'. While testimony is a public genre in
the sense that it is associated with public functions - notably evidence
m courts of law - it is essentially the public appropriation of personal
experience and of the private-domain genre for recounting personal
 158 Crimewatch UK
experience, conversational narrative. Which is why it is normally
colloquial and conversational in style. Testimony is a form for linking
personal experience and public accountability.
The voice-over from the witness Michael Godfrey is in two parts.
The first is another instance of testimony, but the second, while
having a narrative function, is different from anything so far:

f " I'm sure that . Claire would have sensed . that somebody was with
| 1 her . in that last moment . I think that's quite important . for her
t family to know that . that she wasn't on her own

It is not part of the story, part of the account of what happened, but a
comment on and evaluation of the story, drawing some comfort from
it, and linking the story as past events to the present and Claire's :
family's continuing grief. Notice the deictic forms here, particularly
the verb tenses. The past-tense verbs, but also the demonstrative that |
(in that last moment) contrast with the present tenses which anchor
these words in the moment of speaking, after the event - I’m sure, 1
think, that's quite important. Although, like testimony, this is the per­
ception and point of view of one witness, it does also have a role in i
the main external narrative, or perhaps rather it serves to bring the ;
two narrative perspectives together, providing a closure, the sense of
an ending.
The biography genre is mainly present in the voice-overs in the ~
firefighting and pool-game sequences, which are primarily building a f§
sympathetic and positive picture of Claire. The voice-over from her i
parents in the opening sequence is a jointly developed account of her ]
interests:
is®
FATHER: she was in the fire brigade . the Duke of Edinburgh ;y.f|f
Scheme and [e: lyffi;
mother: 11 don't think she wanted a nine-to-five job
she wanted to go in the fire brigade
FATHER: that that was her intention to go in the fire brigade . no
one would change it (laughs) she was a lovely kid you
know s it's just such a shame
Notice how her mother cuts into and continues her father's account, r
and how he then picks up and thematizes 'the fire brigade' (with that) .j.
from her turn. Notice also how her father's account shifts from |
talking about Claire in the third-person to talking in her voice, so to {
speak, in a clause (no one would change it) that is marked as free indirect j
speech by its modal verb (would). This is followed by an affectionate
laugh and makes a transition from the third-person account of j-
 Generic analysis 159
Claire's interests to her father's explicit wording of the evaluative and
moral conclusions the account and the film point to - she was a lovely
kid, and it’s just such a shame. The voice-over from the fireman in the
opening sequence is generically different: it's a testimonial (different
from the 'testimony' I referred to earlier), an evaluation of Claire's
performance and potential. The parents' voice-over in the final pool-
game sequence is generically similar to their earlier one, a conversa­
tional account.
To sum up, then, voice-over indicates that the film is generically
complex in being part narrative of events leading up to and following
the crime, part biographical portrait of Claire, part public appeal. The
biographical element is itself generically complex, including a life-
world part, conversational reminiscence, and a more public/official
part, testimonial. The narrative too is generically complex, bringing
together external narration of events and the personal testimony of
witnesses. I now want to carry on investigating the nature of this gen­
eric complexity by taking a closer look at the narrative.

Narrative
Police work conventionally makes use of narrative in appeals for help
and information from the public. Police officers are sometimes inter­
viewed on news programmes, and give narrative accounts of crimes
as a basis for appeals fo r help. Sometimes the police use a dramatized
. form of narrative, the reconstruction, which has traditionally
involved police officers playing the parts of those involved in crimes.
" Crimeiuatch UK sees itself as an extension of this: it sees what it is
doing primarily in terms of eliciting help from the public in the
solving of crimes.
This construction of the programme places it in the domain of fact
and information, suggesting a version of external narrative which is
focused upon events of and around the crime, and their time and
, place, and linked to public appeal. But as I have already indicated,
j what we actually find in re-enactments is part external narrative but
■ also part testimony, and a focus upon personality and character - in
this instance upon the personality of Claire herself, her parents, and
the witnesses. Films are referred to in Crimewatch UK sometimes as
.re-enactments' and sometimes as 'reconstructions', and this dual
identification perhaps points to their ambivalence: they do have in
part a factual and informational character (as reconstructions), but
they also have a dramatized and fictionalized entertainment char­
acter (as re-enactments). We have in another form the tension
160 Crimewatch UK
between information and entertainment which has been referred to
through the book, and it is also in this case a tension between factual
and fictional, between public information and drama. It is worth not­
ing in this connection that the sort of crimes that Crimewatch UK
covers are mainly those which are most obviously open to dramati­
zation - corporate fraud, for instance, is low on the agenda.
Even in conventional police reconstructions, events are
dramatized in the sense that people are playing the parts of those
involved in the crime. But in Crimewatch, these include professional
actors as well as some of the people actually involved, but not police.
These are professionally produced dramas, even if they are less than
distinguished instances. They include extensive dialogue. This,
together with other generic elements I have indicated - the biography
genres of the opening and closing sequences, and the testimony -
contributes to the focusing of character and personality. In fact, one
might argue that all of the narrative of Claire's last day before her
walk to her friend's house is primarily there to develop character, first
and foremost the character of Claire. There is no question of eliciting
public information about the events of the earlier part of the day, and
a traditional police reconstruction would cover only Claire's walk to .
her friend's house up to the discovery of her body. This focus upon s
character and personality which shapes so much of the re-enactment
is central to its character as fictionalized entertainment.
A great temptation for journalists with a story of this sort is to
piay up us sensational potential, the violent, horrific and wanton;
nature of the attack. Crimewatch UK is very restrained in this regard,
as it has to be if it is to maintain any credibility for its claim that the
programme is primarily about public help for the police. But the
actual ambivalence of the programme just referred to does show
itself in certain relatively muted ways in which the presenter creates
an atmosphere of suspense. Notice, firstly, how the crime is for­
mulated in the presenter's introduction in the three expressions:
another rather grim case that's made national headlines, a killing for the__
sake of it, she was apparently picked at random. This is very tame com:-,
pared with what the tabloid press might make of it, but nevertheleis
it establishes (in this viewer, at least) an anticipation of horrors ton.
come. Notice, though, that the presenter actually says another case
I'm afraid that seems to be a killing for the sake of it: this is double-voiced, ,
for it brings together the muted foregrounding of the sensational
nature of the story with the programme's commitment (realized in.
I'm afraid, seems) not to contribute to moral panics or sensational /f
crime. Later in the presenter's commentary, the background
 Voices 161
information refers to the evening without pressure which Claire was
looking forward to. This familiar anticipatory narrative device again
builds up suspense for viewers, ironically juxtaposing what Claire
hoped for and what we all know is going to happen. A sequence
where Claire is offered a lift but refuses it has a similar ironic effect,
which is reinforced shortly afterwards by Claire's mother's explana­
tion of why she didn't drive her to her friend's house. I have already
referred to the voice-over which uses the historic present to dramatic-
,ally register the onset of the climactic episode leading to Claire's
death.

Voices

I now want to begin shifting the analysis of the discourse practice and
intertextuaiity of Crimewatch UK away from genres and towards
discourses. I shall do so by looking at the range of voices that are
included in the programme, how those voices are distributed within
the complex configuration of genres 1 have pointed to above, and
thus what relationships are set up between voices. 1 will then in the
-next section look at the relationship between voices and discourses.
The main voices, the main types of social agents, that figure in this
programme are the police and other officials (e.g. firemen), ordinary
«■people, and media presenters. Ordinarv people include in. some
-cases the victim of the crime (Claire obviously cannot speak for her­
self, but Muriel Harvey for instance, a rape victim, does), the victim's
family and friends, and witnesses.
The feature on the Claire Tiltman murder shows a distribution of
voices with respect to genres which is typical for Crimewatch UK. The
police do not figure in the film; the main account of the crime is medi­
ated by the presenters and by the testimony of ordinary people, but
not by the police. The police figure only in the person of Inspector
‘Owen Taylor who is interviewee in the interview which follows the
film. This is an instance of wrhat we might call the 'expert interview',
in which the interviewee is constructed as having expert knowledge
-which the interviewer is eliciting on behalf of the audience. The aim
of the interviewer is not to probe or challenge, but rather to facilitate
communication between expert and audience. The content of the
police officer's answers is more or less fully anticipated - and in some
‘ cases closely preformulated - in the questions, suggesting that inter­
viewer and interviewee are collaborating in covering an agreed
agenda. (Interview genre is presumably preferred to a monologue
162 Crime watch UK
from the police officer because it breaks blocks of information into
more easily digestible chunks.) For example, at the beginning of the
interview, the police officer's answer reformulates and slightly elabo­
rates the question:
presenter : Owen Taylor that must be your fear . this guy's going to
do it again
Owen t .: yes it is this was a particularly savage attack . Claire was
stabbed . several times with a . a large knife and e: I think there's
every likelihood this person could strike again

This guy's going to do it again and this person could strike again are parallel
formulations of the same proposition but in different discourses (a
contrast I return to). And again the police officer's it's not the sort oj
thing you would normally carry around closely echoes the presenter's it's -
not the sort of knife you would sort of carry, and an unusual weapon echoes
a very unusually large knife.
Another feature of the questions is the presenter's deference -
towards the police officer, which is marked in the modality. Two
instances use subjective modality to foreground the police officer's •
judgem ent-his fears in that must be your fear, his opinion in does he live T
in the area do you think. The must in the first of these, as well as presum- I
ably elsewhere, both in declarative questions where the question is in \
the form of a statement, mitigate the presenter's claim to knowledge J
by marking it as a presumption which the police officer ratifies in each |
TV»#3 nnlirp offirpr^S ^nciwprc nrp 3 HllxtllfC of C3 f0<’rOriC3 lly’ 7a-
modalized statements (such as yes it is, this was a particularly savage i
attack) where the basis for speaking authoritatively is implicit, and^
statements which foreground the police officer's own judgement“!
through use of subjective modalities (such as I think there's every likeli- ,|'
hood this person could strike again). r
The voice of the police here is mainly limited to giving information \
about the crime and the investigation, but there is an instance of ;
public appeal embedded within one of the answers (can you honestly
live with Claire Tiltman's murder on your conscience . and can you liv e^
with the thought that this person may well strike again). This is not a direcbtjp
appeal to the audience-the police officer does not look at the camera
and speak directly to the viewers, he tells the interviewer what he,jf,
would 'simply say' to someone who knew who the murderer was.
Nevertheless, this indirect appeal is in the form of direct speech, «V ||
the form of a direct question to the potential witness (compare thejy
option of an indirect question: l would simply ask that person whethetgff
they could honestly live with Claire Tiltman's murder on their conscience
 Voices 163
But elsewhere in the programme, in the 'Photocall' and 'Inquiry
desk' features, police officers take on the role of presenters. The fol­
lowing item from 'Inquiry desk' is presented by a police super­
intendent:
first tonight in Incident Desk we need your help to identify this
man . who sexually assaulted a motorist on Saturday the seventh of
November . the twenty-four-year-old woman was driving along the
A323 near Guildford when her car overheated - she was looking under
the bonnet of her white Vauxhall Astra when a man stopped . and
offered to help her . however his kindness didn't last . he grabbed
the woman and dragged her into bushes . she was knocked uncon­
scious and it was an hour and a half later before she managed to raise
the alarm . the victim remembered her attacker was smartly
dressed . and wore a shiny dark quilted jacket . he was in his early
thirties . around six foot . with a strong athletic build . perhaps you
saw him driving away in a dark-coloured car . possibly a Sierra or
- Granada . remember it was the A323 Aldershot Road on Satur­
day . the seventh of November . if you can help . please call Guild-
ford police station on 0483 31 111 . that's 0483 the code for Guildford
31111
Apart from interview, we have here the same range of activities
which 1 described for the presenter in the re-enactment of Claire
Tiltman's murder: introduction to the item; narrative of the crime;
. and public appeal. In addition there is a description of the attacker.
- Notice that the public appeal is prefaced by two cues for potential
: witnesses (perhaps you saio him, remember it was the A323), both, like the
appeal itself, directly addressing the viewer (with you and the
imperative verb remember). Police officers taking on the role of
presenter adopt a presentational style and delivery which are very
similar to those of professional presenters, though perhaps not as
fluent. One might also detect in this example some of the relatively
muted journalistic exploitation of the story which I discussed for the
re-enactment: why else does the narrative have he grabbed the woman
Jfwd dragged her into bushes rather than just he attacked the woman?
In Crimewatch UK, the relationship between voices and genres is an
Open and flexible one. In the case of presenters and police, the flexi­
bility works in both directions. One might see the Claire Tiltman re­
enactment in terms of a television presenter taking over a traditional
police genre of eliciting help with solving crimes. Actually it is not
just the presenter, for while the presenter controls the overall struc­
ture of the feature, the main narrative of the crime is jointly
developed by the presenter and a variety of ordinary people - the
764 Crime watch UK Discourses: official and lifeworld 165
parents, and the witnesses. The police are excluded. Also, it is the i1 official discourse are discernible in the contributions of ordinary-
presenter who makes the direct appeal to viewers for help. In j people, while there are elements of lifeworld discourse in contri­
'Inquiry desk' and 'Photocall', however, the police not only (as it ? butions from the police (and the fireman in the Claire Tiltman re­
were) repossess their traditional functions in appeals for public help, r enactment). The mapping of voices on to discourses shows some of
they also take over the role of the professional presenter. ■ : the flexibility which characterizes the mapping of voices on to genres.
A striking feature of the programme is the prominence given to * An interesting feature of this programme is the way in which it struc­
various categories of ordinary people - victims of crime, victims' tures the relationships between official and lifeworld discourses,
families and friends, and witnesses. The re-enactments dramatize which is linked of course to the way it structures relationships
events in the lives of ordinary people, and the real people figure or r between voices and the categories of social agent they are associated
are represented by actors in the film and dialogue. The potential rele- [ with.
vance of stories to everyday life is a vital factor in their appeal to audi- ; i The discourse of Claire's parents in the firefighting sequence (page
ences - the sense of 'it could happen to me' which is accentuated by i 154) is predominantly a lifeworld discourse. Textual indicators of life-
rooting the stories in ordinary life and experience. Ordinary people j world discourse include vocabulary (e.g. a nine-to-five job, go in rather
also have a major part in the development of narratives in the form of than, for instance, join [the fire brigade]), conversational formulae
what I've called testimony, and in the biographical parts of the film. I ' (it's such a shame, she toas a lovely kid), and the implicit nature of coher­
also suggested earlier that Claire's father briefly contributes to the i t ence relations which presupposes that audience members share
external narrative of events, as well as the biographical element. And , membership of the same lifeworld, and are therefore able to fill the
ordinary people are also involved in the public appeals, if indirectly f gaps. Examples are: coherence relations between the two parts o f she
as in the case of Claire's mother. So here again, the flexibility of the j urns in the fire brigade . the Duke of Edinburgh Scheme; and coherence
relationship between voices and genres is striking, with ordinary ; relations between the account of Claire's interest in the fire brigade,
people overlapping with presenters in narrating crimes, and with • and she was a lovely kid, and it's just such a shame. My feeling is,
presenters and police in appealing for help. Notice, however, that -£ : however, that that urns her intention is a trace of official, public
there are limits to the flexibility: apart from the o n e rather marginal * discourse. Ordinary conversation is shot through with such traces,
instance of Claire's father, ordinary people do not contribute to exter- f ( and it is not surprising to find it in experiential talk produced for
nai narrative, only to testimony, the narrative of personal experience, r public consumption on television (though one would need to look at
They are generally excluded from the overview which the external j t how the editing process affects such traces).
narrator is endowed with. The programme also draws rather a dear f The fireman's testimonial in the firefighting sequence is an inter­
line between those who are explicitly interviewed and those whose esting mixture. A testimonial is by its nature a public and often official
voices figure without the mediation of the interviewer: it is the police 4 genre, but the discourse here is largely a lifeworld one. One obvious
who are constructed as experts through interview, but not ordinary p factor is that testimonials are usually written, but this is spoken. Con­
people. ir l sider, for instance, the beginning of the testimonial:
"^ 1
'.Claire . she was a good student . she always used to come to the ses­
1 sions . she was always the first here
Discourses: official and lifeworld
A written testimonial might cover the same ground in something like
this way:
I want to focus here on the relationship between two main types of f'
discourses: official discourses, discourses of public life and especially * - Claire Tiltman was a good student who always attended classes.
o f policing; and lifeworld discourses, discourses of ordinary life and -i.y;
ordinary experience. In Crimeiuatch UK the former are primarily-ty .Features of lifeworld discourse are the use of first name, the
associated with the voice of the police, the latter with the voice of.Jyty disjunctive structure which places Claire initially as a sort of thematic
ordinary people, while the voice of the presenter mediates between * label outside clause structure, the auxiliary verb used to which fore­
the two. But the picture is rather more complex in that traces J- ; grounds the fireman's memories of Claire rather than what, for
166 Crimewatch UK
instance, the course records show. I doubt whether a formal testi­
monial would include anything equivalent to she was always the first
here - implicitly conveying a sense of someone's personality in this
way is a lifeworld practice, whereas a formal testimonial might refer
to her 'enthusiasm'. But there are elements of official discourse here,
notably at the end: possibly be in charge of a fire appliance.
The dialogue in the re-enactment overwhelmingly draws upon life-
world discourses, except for the sequence near the end involving
Michael Godfrey and the ambulance woman, which includes a
stretch of technical medical talk from the former:
no pulse I've been giving her . CPR . she's not responded . and
there's no pulse
CPR (which stands for cardio-pulmonary resuscitation) is obviously a
technical expression. But the vocabulary in which treatment (giving
the patient CPR) and its effects (patients responding) are talked about
also comes from medical discourse, as does the impersonal way of
talking about Claire's pulse (there's no pulse). (Compare the
ambulance woman's question has she got a pulse -interestingly, the lay:
first-aider bilks in a more technical way than the professional medical
worker.) The testimony from the jogger, the motorist and Michael
Godfrey again overwhelmingly draws upon lifeworld discourse,
though as I pointed out earlier there are traces of official report in the
jogger’s account, in the precise description of place and action, the
use of thus, and the positioning of again before noticed.
I suggested above that the presenter mediates between the life-:
world and official discourses. There is some evidence of this in the
presenter's contributions to the external narrative during the film.
For example, a rather official formulation of the occasion the final day
of Claire's mock GCSE exams is then translated into the lifeworld for­
mulation sitting tests. I mentioned earlier the official police discourse
in the presenter's formal description of location. But the mediating
work of the presenter is most evident in the interview which follows:
the film. The police officer is mainly, though not entirely, using
official discourse, and in some cases there is a contrast between the.
official discourse of the policeman and lifeworld discourse from the.;
presenter. I referred earlier in another connection to the relationship
between their opening turns - compare the presenter's this gw/ri
going to do it again, and the policeman's there’s every likelihood this per­
son could strike again. Official features of the latter are the carefully for­
mulated probability modality (every likelihood, could) and the non-
mminatiup and vender-neutral term person. A rather different feature
 Between state and people 167
of official discourse in the police officer's next turn is conjunctions
with or which ensure that all possibilities are covered - we have two
examples here, the person either lives in the area has got connections in the
area . or has lived in the area in the past, and someone in that local com­
munity . knows who that person is . or suspects who they are. Presenter
mediation takes a different form. The police officer refers to Claire
being stabbed with a large knife. The presenter reformulates this as a
very unusually large knife it's not the sort of knife you would sort of carry like
a little penknife or anything. This on the one hand facilitates the police
work by foregrounding a detail which might prompt information
from the audience, but it also sensationalizes by exaggerating the size
of the knife in comparison with the police officer's formulation. Else­
where the presenter himself talks in the categories of official
discourse - unpredictable violence, outbursts of aggression - or more
exactly in those categories from the specialist discourse of psychology
that are appropriated within official discourse. Notice also that there
are elements of lifeworld discourse in the talk of the police officer -
particularly in his indirect appeal for people to come forward, where
he echoes the formulation of Claire's mother (I don’t think I could live
■with myself) in can you honestly live with Claire Tiltman's murder on your
conscience . and can you live with the thought that, using the formula of
lifeworld discourse I couldn't live with X.

Between state and people

I shall conclude this discussion of Crimewatch by considering how the

features of discourse practice and text described so far link to the pro­
gramme as sociocultural practice. I see the programme as an inter­
vention into the fraught relationship in contemporary society
between the state and the people. Politicians, governments and insti­
tutions of the state have lost much of their public credibility and
authority over the last two or three decades. There is a major crisis of
legitimacy. In the case of the UK police force, the erosion of a relation­
ship of trust between the public and 'the bobby on his beat' is a matter
for endless nostalgia, much of it within the media. Even contem-
: Porary police drama series reflect the loss of innocence, focusing
upon the fallibility and corruption of the police.
Crimewatch UK can be seen in this context as fighting a rearguard
action. But it is not harking back to a golden age of police-public
relations, it is reconstructing a relationship of trust and cooperation
on a new basis, through a mediatization of police work and of the
168 Crimewatch UK
police-public relationship. The work of policing is transformed in
the programme into a joint effort between police, journalists and
ordinary people - 1 noted, for instance, that the policing genre of
public appeal for information cuts across these three major voices,
so that Claire's mother and the interviewed police officer but
especially the presenter all appeal for help with the Claire Tiltman
case. This relationship between voices and genres also applies to
journalistic work, with police officers sharing the role of presenter
(specifically in the 'Inquiry desk' and Thotocall' features) with
journalists. Moreover, the ambivalent mix of reconstruction and re­
enactment, and the mixing together of dramatized narrative and
public appeal, resituate police work in the familiar and homely
world of television entertainment. And the salience of a lifeworld
discourse within the re-enactment, in the biographical elements
and the dialogue, but also in the actual recounting of events in the
testimony of witnesses, links police work with ordinary people and
ordinary experience. We have, then, a crossing of boundaries and
a merging of voices and practices which powerfully domesticates
and so legitimizes police work. Or at least appears to do so: it
would be fascinating to know what audiences make of this
programme.
Let rne link these remarks to the question of how audience
members are addressed. Audience address is complex and contra-
Gictury. 1 ilO t - C i U l c l I LL’i ui ciUiuUuit !-• bot vvtrci’i. u"i£ Capacity ifl Vvllich-
the programme purports to address audience members, and the;
capacity in which it does in effect address them. The programme;
presents itself as eliciting public help for the police. Audience
members are addressed, most obviously in the public appeal and the;
interview, as responsible citizens who may be able to help. Yet the:
programme is nationally networked, and only a tiny fragment of,
the 11 million viewers could conceivably be in a position to help. If.
the issue is primarily helping the police, why a nationally networked
programme, why not use local media?
In fact, it is the address/construction of audience members as'
spectators of an entertainment that has general relevance for a
national audience, articulated as I have suggested above with,
address/construction of audience members as co-members of the life-
world, the world of ordinary experience, to which victims, their fami­
lies and friends, and witnesses are shown as belonging.
The following entry from the Radio Times (17 February 1994) shows
in a condensed form the contradiction between real and purported
target audiences.
 999 169
Your help is needed to find an armed gang who robbed a post office
van in Burnley, Lancashire, and then shot a policeman in the leg. The
villains drove off and hijacked two cars, forcing the occupants out at
gunpoint, before finally escaping in a Ford XR3i. if you have any
information about this or any of the other crimes featured in the pro­
gramme, call the studio free on 0500 600600.
Radio limes and similar weeklies play an important part in enticing
viewers to watch programmes. We have the usual focus on public
help at the beginning and the end, but sandwiched between is a
summary of an exciting and shocking story. There is an ethical issue
here. It is a commonplace understanding that many people get pleas­
ure from watching violence and horror in films and on television. Per­
haps Crimeiuatch UK allows people to do that while comforting
themselves that they are doing something else. The ethical issue is
whether the BBC should tolerate and sustain the discrepancy- which
they surely well understand - between what is claimed to be going on
and what is actually going on.

999
'Like Crimeiuatch UK, 999 is a prime-time B8C1 programme. It is pre-
sented by a well-known television journalist and newsreader,
' Michael Buerly find, features dram a-: rescues usually involving the
emergency services. My discussion of 999 will be briefer than the
- section on Crimeiuatch. My purposes in including it are twofold: first,
to show that re-enactments of people's misfortunes using something
■like a Crimeiuatch formula have a wider appeal in television and do not
need the particular justification - public help for the police - that
Crimeiuatch claims; second, to show that the sensationalist aspect of
such stories and their entertainment value are not always as muted as
they are in Crimeioatch.
Two stretches of voice-over from the presen ter during the opening
sequence give the flavour of the programme:
all of tonight's rescues are true stories . we've sometimes used actors
or stuntmen . but everything you see and hear is based upon the
accounts of the people involved . they've helped us to reconstruct
events . ns they happened
tonight on 999 . a daring lifeboat rescue in a storm off the Isles of
Scilly . captured . as it happened on video - the three-year-old on an
adventure that could easily kill him . and the men who struggled for
T~ his life
170 Crimewatch UK
These extracts point to two essential properties of 999 stories: they
must be true and based upon authentic accounts of those involved,
and they must be good drama. The first of these is made explicit,
whereas the second is implicit in the 'headlines' of the second extract.
The reserved and muted treatment of the entertainment value of
stories in Crimewatch gives way here to a foregrounding of drama and
suspense and the deployment of various devices for heightening
them (see below for examples). The italicized words are all given par­
ticular salience in the delivery - all is heavily and contrastively
stressed, as they happened is given emphasis through intonation, loud­
ness and pace of delivery, and struggled is heavily stressed and
expressively articulated to suggest struggle. The first voice-over is
accompanied by the series of dramatic images of rescue which make
up the opening visual sequence, and the second by shots of a lifeboat
in heavy seas then a yacht capsizing, and a small child walking the
streets alone then a man struggling in soft mud to position a car tyre
in front of him (to stop himself from sinking). The loud and dramatic
theme music continues throughout.
While the programme is characterized by a strong element of
voyeuristic fascination with the misfortunes of others, this is
mitigated in various ways, so that, as in the case of Crimewatch there is
an ambivalence of intent and ethos about the programme. In some
cases, the moral implications of a story are stressed by the presenter. ■
Some stories provide the basis for educational work - for example,
the feature I focus on below concerns a heart attack, and the story is
followed by advice for the audience on how to give firs t aid in the case
of a heart attack. Throughout, the presenter's serious and austere
tone and expression hovers between highlighting the drama and
excitement and deprecating the terrible things that happen to people. >
1 shall focus on one feature which describes how fire and:;
ambulance services rescued a man who had a heart attack near the
top of a 200 -foot crane. I shall summarize the main events of the story,
indicating points at which tension is particularly built up, and then
look in more detail at short transcribed extracts.
A crane operator arrives at work on a building site and chats with his _
friend the site foreman, telling him he's not feeling too well but will 'be
alright'. He starts to climb up to his cab at the top of the crane. Shortly
afterwards the foreman tries unsuccessfully to call him on an intercom- ~
He notices a hand waving from high up on the crane, realizes there'sa
problem, and climbs up to find his friend collapsed on a platform just;
below the top of the crane. The fire brigade and ambulance service
arrive. The ambulance man is terrified at the prospect of climbing up
 999 171
the crane. He controls his fear and starts an agonized ascent. The first
drama is when part of his equipment catches on the scaffolding of the
crane and pulls him back, so that he seems to be at risk of falling. He
at last reaches the crane operator and realizes that he has had a mas­
sive heart attack, and tells the fireman and the foreman (who are on
the platform with him) that unless the man is got down quickly he
will die. Will they make it in time? The fireman proposes to lower him
down the inside of the crane in a stretcher, but the foreman says that
will take too long if it is possible at all. He suggests winching up a skip
to lower the man down in. This is agreed, but tension builds up as
vital moments are lost because the man on the ground can't shorten
the chains enough for the skip to be winched to the top of the crane.
He eventually succeeds, and the skip is winched up. But can the fore­
man (operating the crane) manage to get it close enough to the plat­
form for the transfer? He does. To lower the crane operator in the
skip, he has to be placed horizontally on the stretcher, a dangerous
position after a heart attack. Will he suffer cardiac arrest? He doesn't.
But can the ambulance man overcome his terror sufficiently to climb
from the platform into the skip? With great difficulty he manages to
do so, but there is then another drama: can the stretcher be moved
across without the skip moving and the man falling to his death?
They manage it, the skip is lowered to the ground to the applause of
those waiting below, and the man is taken to hospital in an
ambulance, where he makes a full recovery. The freal's crane operator
praises the ambulanceman and gives his own account of his ordeal,
which is interspersed with more video re-enactments. The presenter
makes a link from the story to a feature on first aid for heart attack
victims.
Here is the presenter's introduction to the feature, delivered on-
camera with a building site and a large crane in the background:

few people would relish swaying about on a crane nearly two hundred
feet up in the air . the fear of heights can make you panic . can make
you freeze . luckily most of us can avoid the experience . but for
members of the emergency services there are times when the job leaves
them . no choice they're forced to confront their personal fears and
overcome them . for the sake of the people they're trying to
help . that's exactly what happened one day on a building site here in
London

This introduction first evokes the vicarious experience of danger

which is a key feature of the programme through an alarming and
exaggerated formulation of the rescue (swaying about on a crane nearly
172 Crimewatch UK
two hundred feet up in the nir). The presenter's face is serious and
indeed grim throughout, and the toughness of the scenario evoked is
metaphorically conveyed in a tough, hard, tone - the initial conson­
ants of crane and two are given a particularly explosive release.
Toughness is also conveyed through gesture: a thumb sharply jerked
backwards to point to the crane behind the presenter. Both panic and
freeze are heavily stressed, and accompanied by both hands being
pushed sharply forward to give further salience to these evocations of
danger and fear. The alarming image is made safe (luckily most of us
am avoid the experience), and the vicarious experience of the audience
is contrasted with the real experience of people in the emergency
services. The internal struggles and courage of emergency workers (a
major focus of this feature) are again visually represented through a
gestural metaphor of physical confrontation - the presenter pushes
his two clenched fists together in front of his chest. This formulation
of the story to follow endows it with a serious purpose and point, a
moral, which (as suggested above) mitigates the voyeurism: it shows
people confronting their personal fears for the general good.
999 is similar to Crimewatch UK in constructing narratives through a
combination of re-enactment and dialogue, presenter voice-over,
and accounts by the people involved. One difference in the formula is
that whereas these accounts are in voice-over in Crimewatch, in 999
they are mainly given on-camera. This is part of a more general
enhancement of the voices of ordinary people, and an exclusion of
official voices: although some of these accounts are given by
members of the emergency services such as ambulance workers or
firefighters, they generally tend to speak personally as ordinary
people, not officially. What tends to be in focus is their reactions and
feelings.
Here is an extract at the point where the ambulanceman crosses
from the platform into the skip. The alternation between re­
enactment and the on-camera account by the ambulanceman in the
studio is shown by printing the on-camera account in italics.

presenter: once John was strapped into the stretcher . they came to
the most difficult part of the rescue
fireman : okay you go over to the skip first . and I'll pass him to you
ambulanceman: you just be sure . to hold the skip
ambulanceman (on-camera): now . I'm now an absolute nervous
wreck . I'm shaking l can't breathe properly I'm . 1 can't talk properly
ambulanceman: just be sure to hold the skip
fireman : I've got it
ambulanceman : . just hold the skip . okay . ready . okay
 999 773
ambulanceman (on-camera): fear's not the word . I mean l was absolutely
petrified
ambulanceman: o h Jesus Christ
ambulanceman (on-camera): to step off the side of a hundred-and-eighty-
foot - (shakes his head) it just sounds crazy to even think about it now . but
just to step off the side . a good gust of wind . a-huttdred-and-eighh/-
foot . believe me . it's windy up there . and you can just . be blown away
fireman : go . you're nearly there
ambulanceman: got me
fireman : yeah
ambulanceman: I'm in . I'm in
ambulanceman (on-eamera): / w- wouldn't do it again . I couldn't do
that again . not walk off the side like that

The frequency of the alternation between re-enactment and on-

camera account is striking. A relatively brief episode in the story is
stretched over quite a long section of the feature. The camera shot
of the actual step from the platform to the skip is itself stretched
from okay . ready . okay to I'm in . I'm in. The first two on-camera
accounts make explicit in some detail what is already implicit in
what we see and hear in the re-enactment: the ambulanceman's
fear. The ambulanceman's first turn [you just be sure . to hold the
skip) is delivered in a voice that is shaking with fear, and vve have a
close-up of the man's agonized face. The same is true of his next
two turns - indeed his voice breaks on the words hold the skip. The
extremity of the man's terror is signalled in his blasphemy (oh Jesus
Christ), which would, I suspect, not lightly be allowed into a script.
Notice that the first account (I'm now an absolute nervous wreck . . .)
uses present tense; it comes across as reliving the experience - the
man's voice is actually shaking on the word nervous, even though
he is sitting in the safety of the studio. The third and longest on-
camera account (to step off the side . . .) reiterates what is again
already made clear in the re-enactment: the enormity of what the
ambulanceman did. We have already seen distance shots of the
skip floating high up in the air, and of the man stretched between
the platform and the skip.
A salient property of this whole extract is repetition. The on-
camera accounts repeat the enacted terror of the ambulanceman,
and vice versa, and the terror is repetitively enacted in the one and
represented in the other. In the re-enactment, you just be sure lo
hold the skip is repeated with minor variations three times. The
enormity of the step is repeated visually in the re-enactments and
verbally in the accounts: step (or walk) off the side occurs three times,
174 Crimewatch UK £
...&■

a-hundred-and-eighty-foot occurs twice; wind is repeated as windy, and I

notice the parallelism at the end between 1 wouldn't do it again and '
l couldn't do that again. Repetition serves to elongate this one short |
incident in a way that fully exploits its potential as vicarious t
entertainment.
A lthough the real people involved and the characters depicted f
here are an ambulanceman and a fireman, they are not talking in |
official or technical ways. The dialogue is the sort of talk that forms
part of ordinary collaborative action, and the ambulanceman's p
commentary is about his personal response to the incident. The f
discourse, apart perhaps from the fireman's you go over to the skip f.
first . and I'll pass him to you which hovers between (official) instruc- j
tion and (lifeworld) suggestion, is a lifeworld discourse. Markedly so
in, for instance, the colloquial formula 'x is not the word' (fear's not the {
word), the blasphemous curse (oh jesus Christ) indicative of a loss of |
control which itself is a marker of lifeworld discourse, the implicitness of 1,
coherence relations within the ambulanceman's longest commentary j,
(to step off the side . . .), and the formula /wouldn't (couldn't) do it (that) j
again as a coda for a story told in conversation.

Crimewatch UK and 999 manifest in particular forms the shifting and

unstable boundary between information and entertainment, fact and
fiction, documentary and drama, w hich l have referred to at various
points during the book. In the case of Crimewatch UK, the element of
entertainment is, as it were, shamefaced, and hides itself under the-
claim that the programme is about getting the public to help solve
crimes. In the case of 999, the element of entertainment is more open,
though even here it is mitigated by some emphasis on the moral
implications of the stories, and on how viewers might help people in r
need.
I have suggested that Crimewatch UK is an exercise in legiti­
mation, in rebuilding a relationship of trust between police and
public at a time when trust has been seriously eroded. I think we ;
can see 999 in similar terms, but there are important differences. In
Crimewatch UK, the filmed re-enactments are embedded within
substantive studio-based activity, in which, as I argued above, the
roles of police and presenter are partly merged. But. 999 consists,
almost solely of re-enactments with linkage and continuity provided ;
only by the presenter, so such a merging of roles is not part of the
legitimation process. A key difference between the programmes
with respect to legitimation is that Crimewatch UK centres upon the
police, whereas 999 is at least as much concerned with the fire and
 W i/0
ambulance services. Whereas the public reputation of the police has
sunk sharply, fire and ambulance workers are still highly regarded.
While Crimewatch UK is actively restructuring relationships between
police and public, 999 is rather consolidating relationships which are
already fairly solid.
This contrast shows itself in the discursive practices of the two pro­
grammes in the different handling of the voices of police and
emergency workers. In Crimeiuatch UK the police present features
and are interviewed, but do not figure much in re-enactments and do
not provide personal testimony. In 999, emergency workers figure
prominently in re-enactments and dialogue, and also provide testi­
mony which contributes to the narration of rescues. The difference
goes along, of course, with a difference in topical focus: Crimewatch
UK re-enacts crimes, and obviously the police will tend to be involved
only at the margins if at all; 999 by contrast focuses upon the activities
of the emergency services by re-enacting rescues. But the salience of
emergency workers in the provision of testimony (illustrated in the
extract from 999 analysed above) is important, because it indicates
that the main merging of voices in 999 is between ordinary people
.and emergency workers. The latter are largely constructed as ordi­
nary people, as like 'us' in how they behave and how they react to
situations, and, of course, in how they talk. I suspect that this con­
struction is plausible because it corresponds to reality: in real life-
world communities, the barriers and suspicions which tend to attach
To community members who are police officers do not attach to
/ambulance or fire workers. The legitimizing agenda of 999 is less
problematic than that of Crimewatch UK. Nevertheless, both pro­
grammes strike me as intervening to legitimize state-public relations.
 3

POLITICAL DISCOURSE IN THE MEDIAI

I iV m c n H"' r-, f U ' •: fl-io -» f-iT -ili El|r->n i ?-> •> ? - » W -im -i i l r s r- £■•> r . n .

study of the analytical framework introduced earlier. But the

approach will be rather different. Kecall m y argument in Chapter 4
that analysis of media discourse needs two twin focuses - on parti­
cular instances, particular texts (communicative events), and on.
the order of discourse. In Chapter 8 the emphasis was upon parti­
cular instances, particular programmes in the Crimewatch UK and
999 series. In this chapter, the emphasis will be upon the order of
discourse. I shall give an overview of how political discourse is
structured in the media - of what I shall call the 'order of;
mediatized political discourse'. C;
The decision to place the emphasis on describing the order of
discourse in my discussion of politics in the media is partly a
response to Bourdieu's critique of a particular sort of discourse:
analysis as applied to political discourse. Especially given the influ­
ence of Bourdieu in contemporary social theory, the critique must
be taken seriously. Thompson sums up Bourdieu's position as
follows:
 Creating a new political discourse 177
it would be superficial (at best) to try to analyse political discourses or
ideologies by focusing on the utterances as such, without reference to
the constitution of the political field and the relations between this field
and the broader space of social positions and processes. This kind of
'internal analysis' is commonplace . . . as exemplified by . . . attempts
to apply some form of semiotics or 'discourse analysis' to political
speeches. . . . all such attempts . . . take for granted but fail to take
account of the sociohistorical conditions within which the object of
analysis is produced, constructed and received.
(Thompson 1991: 28-9)
The first section of the chapter gives a summary of my analysis in
an earlier work (Fairclough 1989) of the political discourse of
Margaret Thatcher. That analysis partly meets Bourdieu's critique,
and partly does not. I draw upon the Thatcher example through the
chapter. In the next section, I summarize Bourdieu's account of the
political field, as a preliminary to my main concern in this chapter: the
overview of mediatized political discourse, which provides a partial
picture of the constitution of the political field. 1 suggest that part of
the analysis of examples of political discourse in the.media needs to
be placing them within the order of mediatized political discourse, in
accordance with Bourdieu's argument. I give most attention in the
overview to the classification and articulation of voices, discourses
and genres in mediatized politics, with an example from a current
.affairs programme (on the Parent's Charter). Other facets of the order
.of discourse are more briefly discussed in the iinal section: bound­
aries between mediatized political discourse and other orders of
discourse; the production, distribution and reception of mediatized
political discourse; the diversity of practices within it; ongoing
change in practices; politicians' access to, and training in, the prac­
tices of mediatized political discourse. I conclude with a discussion of
the relationship between politics and media.

Creating a new political discourse

In an earlier book, I included a chapter on the discourse of Thatcher­

ism (Fairclough 1989, chapter 7, 169-96). This centred upon an
analysis of part of a BBC Radio 3 interview with Margaret Thatcher in
1985. It did go beyond the purely 'internal analysis' that Bourdieu is
criticizing: it attended to the sociohistorical conditions of Thatcher­
ism, it tried to locate Thatcherism and Thatcher's political discourse
within the field of politics, and it considered how Thatcherite political
#13
•Itt4i
178 Political discourse in the media
discourse contributed to the restructuring of that field. But the focus
was still mainly (in Thompson's terms) on 'the utterances as such',
ii« and I want to discuss later how the analysis could be developed to
à few
V0 x more adequately meet Bourdieu's critique.
1 shall begin with a summary of my analysis. Part of the extract
which I used in 1989 is reproduced below:

michael charlton; Prime Minister you were at Oxford in the

nineteen forties and after the war Britain would embark on a
period of relative prosperity for all the like of which it had hardly
known but today there are three and a quarter million unem­
ployed and e:m Britain's economic performance by one
measurement has fallen to the rank of that of Italy now can you
imagine yourself back at the University today what must seem
to be the chances in Britain and the prospects for all now
Margaret thatcher: they are very different worlds you're
10 talking about because the first thing that struck me very forcibly
as you were speaking of those days was that now we do enjoy a
standard of living which was undreamed of then and I can
remember Rab Butler saying after we returned to power in
about 1951-52 that if we played our cards right the standard of
15 living within twenty-five years would be twice as high as it was
then and em he was just about right and it was remarkable
because it was something that we had never thought of now I
don't think now one would necessarily think wholly in material
Lï ü b il :U.etJU A ui i i. i IL J.Ü

20 because really the kind of country you want is made up by the

strength of its people and I think we're returning to my vision of
Britain as a younger person and I was always brought up with
the idea look Britain is a country whose people think for them­
selves act for themselves can act on their own initiative they
25 don't have to be told don't like to be pushed around are self-
reliant and then over and above that they're always responsible
for their families and something else it was a kind of em I think it
was Barry who said do as you would be done by e: you act to
others as you'd like them to act towards you and so you do
30 something for the community now I think if you were looking at
another country you would say what makes a country strong it
is its people do they run their industries well are their human
relations good e: do they respect law and order are their families
strong all of those kind of things and you know it's just way
35 beyond economics

In the analysis, I suggested that Thatcherism was a

response from the right to Britain's long-term economic and po
 f Creating a new political discourse 179
I
v decline. It broke with the 'postwar consensus' in British politics after
f the Second World War which underpinned the welfare state,
s policies of full employment, and nationalization of utilities and
* services. It rejected the Conservatism of Macmillan and Heath as
: decisively as the social democracy of the Labour Party. The break
■ entailed a restructuring of the political field to carve out a space and
? a political base for Thatcherism. Following analyses of Thatcherism
T by Stuart Hall and others (Hall and Jacques 1983), I saw this in terms
i of a new articulation of elements of traditional conservatism, neo-
‘ liberalism, and political populism, constituting a political mix which
i these authors refer to as 'authoritarian populism'. The key point for
t present purposes is that this articulation is partly brought off by a
1 restructuring of the order of political discourse: a new Thatcherite
political discourse is constructed by articulating together elements
y of conservative, liberal and populist discourses. There is a conse-
j quenlial struggle for hegemony within the Conservative Party, and
then within the political order of discourse and the field of politics
’[ more generally.
\ The constitution of a Thatcherite political discourse is illustrated
j through an analysis of the discourse practice of the radio interview -
\ especially of how these discourses are mixed together - and a
, selective analysis of features of the text to show how this hybrid
t discourse practice is realized in a heterogeneous text. The analytical
E .. ..........
framework will bv- now be familiar to readers. I focused on interner- t
sonal issues: the construction of an identity (a 'subject position') for
~ Thatcher as a political leader, of an identity for the political public
('the people'), and of a relationship between leader and public. I shall
just summarize the main lines of the analysis.
For a new political tendency like Thatcherism to achieve power, it
has to carve out a political base, a sufficiently powerful constituency
:

of supporters. Such a political base is partly 'talked into' existence -

politicians construct and reconstruct the people, the political public,
in their discourse, and a measure of their success is the degree to
- which people accept, and so make real, these (often wildly imagin­
'-r*y* '~,vs*'-»r* ^

ary) constructions. In the radio interview, Thatcher frequently talked

about 'the British people' (lines 23-29, 32-34, in the extract are
. examples). Furthermore, she often did so by listing their attributes,
as in these cases. What is striking about these lists is that they draw
upon diverse discourses and condense these discourses together. In
the first extract, for example, she evokes both a liberal political
discourse of individual responsibility { t h e y d o n 't h a v e to b e to ld d o n 't like
; J j T- J ° p u s h e d a r o u n d a r e s e l f - r e l i a n t ) , and a conservative political
180 Political discourse in the media Bourdieu: the field of politics 181
discourse with themes of family, community, and law and order. The her relationship to 'the people' in the particular local form of the
links between the elements in such lists are left implicit, which means Radio 3 audience (though the particular nature of the audience may
that the important ideological work of constitutinga constituency for well affect how that relationship is constructed). I have suggested a
this marriage of conservative and neo-liberal agendas is partly done tension between ('masculine') authoritativeness and femininity, but
by members of that constituency themselves, in arriving at coherent there is also another tension between an authoritative relationship
interpretations of these lists. In saying that Thatcher 'evokes' these with the audience and 'the people' and a relationship of solidarity.
discourses, I am suggesting that audience members may bring fuller This shows itself in the mixing of traditional political discourse with
and richer versions of them to interpretation than the few phrases life world discourse, discourse of ordinary experience. Pronouns are
used here by Thatcher. Audiences can, for this reason, draw out key worth looking at in this regard. The pronoun we, used for example
meanings which are left implicit - for instance, in neo-liberal in lines 11—17, is sometimes used 'inclusively' to include the
discourse, people 'don't like being pushed around' specifically by the audience and people generally (e.g. we do enjoy a standard of living
('nanny') state, and the trade union 'bosses'. which was undreamed of then) and sometimes 'exclusively' (e.g. after
Thatcher was faced with the unprecedented problem of construct­ we returned to power in about 1951-52, where uw identifies the
ing a plausible identity fora woman political leader, and moreover a Conservative Party), and sometimes ambivalently (e.g. if we played
woman leader of a tough and resolute right-wing tendency. Avail­ our cards right - is this the Conservatives? the government? or the
able models for such a leader had strongly masculine resonances, so nation?). Inclusive uses of we are a common feature of political
she had the dilemma of needing to appropriate masculine models discourse. On the one hand they claim solidarity by placing every-
without compromising her femininity (Atkinson 1984). She has taken •one in the same boat, but on the other hand they claim authority in
a great deal of professional advice on the construction of her identity that the leader is claiming the right to speak for the people as a
over the years, for instance lowering the pitch of her voice and slow­ «whole. Vagueness about who exactly we identifies, and the con­
ing her speed of delivery to make herself sound more statesmanlike, stantly shifting reference of we, are important resources in political
but also more feminine and more sexually interesting (with her discourse. You as an indefinite pronoun (in lines 28-31 in the
'husky' voice), and to overcome being perceived as 'shrill', which has extract) - meaning people in general - also claims solidarity, but it is
dangerous stereotypical connotations of a woman emotionally out of ;:not authoritative: you is a colloquial form in contrast to the mainly
cLUitrui. Site is very much a product of the techuoiogization or; written one, it belongs to a lifeworld discourse, and its use claims
discourse, of the engineering of discursive practices to achieve lnstiy membership of a shared lifeworld. It also can be vague and shifting
tutional objectives. Her dress and hair style are also carefully man-: in who it identifies - see Fairclough 1989 for examples. (Claiming
aged to highlight her womanliness. Her language is a mixture of; solidarity with ordinary people was not, however, something that
elements which further contribute to ensuring her femininity is; .Thatcher was good at: her appearance and communicative style
beyond question, and elements which appropriate 'masculine' prac­ -were too emphatically those of the middle class and shire counties.)
tices of authoritative and even tough talk. Look again from this point: So, the mixing of zue and you here is one manifestation of the mixing
of view at the list in lines 22-29. Ostensibly this is a quotation from; of political discourse and lifeworld discourse.
some unspecified source (the idea, line 23), but Thatcher herself often; !
talks in this way. Look as an imperative verb form is interesting;
because it connotes toughness and straight-talking in ordinary life,; I Bourdieu : the field of politics
Thatcher uses it a lot. And the claims that are made about the British:
people in the list are made authoritatively, using categorical.; In the tradition of the German sociologist Max Weber, Bourdieu sees
modalities. •«the emergence of modern society in terms of a differentiation of
The question of how Thatcher is constructed of course slides oven fields: the economy, the state, the legal system, religion, culture and
into the question of what relationship is set up between her and 'the; The arts emerge as separate fields which are partly autonomous
people' - a person is always constructed in relation to others. So though intricately linked. Each is marked by its own particular form
Thatcher's authoritativeness in this extract is obviously a matter of of institutionaliza tion. The field of politics has undergone a process of
I
182 Political discourse in the media
professionalization such that 'political capital', the means for produc- s
ing political policies and programmes, is increasingly concentrated ;
amongst professionals. To become an actor in the political field, any
class or group of people must find professional politicians to repre- ;
sent them, which means, according to Bourdieu, that they must para- ;
doxically become politically dispossessed in order to be politically
represented. Professional political organizations acquire a life of their
own and become increasingly cut off from the people they claim to •
represent. A key element in the training of politicians is acquiring a
sense of the overall structure of the field of political discourses, a
sense of the range and relationships of actual and possible political
stances, a feel for the political game; a political 'habitus', in Bour-
dieu's terminology. An important insight in Bourdieu's analysis
which I develop below is that the discourse produced by professional
politicians is doubly determined. On the one hand it is internally
determined by its position in the increasingly autonomous and rare­
fied field of politics, on the other hand it is externally determined by
its relationship to the world outside politics, and particularly to the
lives and struggles of the people whom politicians represent, whose
trust and support has to be won and sustained.
Let me add to this brief sketch of the political field a few more of
Bourdieu's observations about political discourse. Political discourse
provides the clearest illustration of the constitutive power of
discourse; it reproduces or changes the social world by reproducing i
or changing people's representations of it and the principles of
classification which underlie them. It also clearly shows the insepar­
ability of ideational and interpersonal processes in discourse; it i an
reproduce or change the social world only in and through reproduc- f"
ing and changing social classes and groups - it works simultaneoi sl\ '
on representations and classifications o f reality, and representah ins }
and classifications of people. The power of political discourse L
depends upon its capacity to constitute and mobilize those social Ior-
ces that are capable of carrying into reality its promises of a new
reality, in its very formulation of this new reality.
As I indicated earlier, my analysis of the Thatcher interview did
roughly locate the discourse of Thatcherism in the field of politics and
political discourse, did comment on its conditions of production in s
terms of the state of British politics, and did indicate its effecl m 1
restructuring the field of politics and political discourse. What it did
not do was provide the systematic account of the field of political
discourse that Bourdieu's approach would require. Bourdieu's thesb i
of the double determination of discourse is a useful biwi1- l°r
 Bourdien: the field of politics 183
developing the analysis. And it can be accommodated within the
critical discourse analysis framework introduced in Chapter 4.
The concept of 'order of discourse', and the distinction I drew in
Chapter 4 between internal relations within an order of discourse and
external relations between orders of discourse, allows us to focus
upon the field of politics in its discursive aspect. On the one hand, we
might describe the order of discourse of professional politics, political
communication and struggle between professionals, as it manifests
itself in, for example, Parliament, party conferences, political meet­
ings and discussions. One concern here would be to specify the struc­
tured set of political discourses, so that one could see more clearly
how, for instance, the discourse of Thatcherism was located within
the internal relationships of the order of discourse. On the other
hand, we might describe external relations between the political
order of discourse and other orders of discourse. This would allow a
focus on communication between politicians and publics, as well as
on discursive aspects of interfaces between politics and the economy,
law, religion, and so forth.
A key external relation is between the political order of discourse
and the order of discourse of the mass media. Communication
between politicians and publics is the second of Bourdieu's twin
determinants of the discourse produced by politicians. Since it is now
so heavily concentrated in the mass media, it makes sense to fore­
ground the relation between politics and media in investigating this
second determinant, and I shall do so below in my discussion of
mediatized political discourse. Much critical work on mediatized
politics has stressed complicity between the media and politicians,
' but it is also important to be alert to tensions, contradictions and
straggles in the relationship between the political order of discourse
and the order of discourse of the media. Surprisingly, the mass media
is in fact virtually absent from Bourdieu's account of political
discourse, which is a major weakness. Note that the distinction
between an internal description of the political order of discourse per­
haps with an emphasis on parliamentary discourse, and an external
; description of the interface between political and media discourse,
should not blind us to the fact that much political discourse is now
- open to being reported and represented in the media, so that politi-
■cians even in their parliamentary discourse are partly addressing the
public in anticipation of mediatization, as well as addressing each
other. The televising of Parliament has aggravated this tendency.
Another important omission in Bourdieu's account of political
.discourse is genre. Even in internal analysis of relations between
 184 Political discourse in the media
discursive practices within the political order of discourse, genre has
to be taken into account. Politicians never articulate their various
discourses in a pure form, their talk is always situated, always shaped
by genres such as political speech making, parliamentary questions
and answers, debate, or negotiation. The point is always significant,
but becomes crucial when mediatized political discourse is in focus.
The genres of the mass media do not at all neatly correspond to the
genres of politics, and this lack of fit is a source of constant tension
,r, and difficulty for politicians. An account of the contemporary field of
1<•'! -j politics which omits this tension cannot be satisfactory. My analysis
> :r of the Thatcher interview did allude in passing to this tension, but it
did not give sufficient weight to the fact that this is Thatcherite pol­
'fpF itical discourse within a particular media genre with its own expecta­
rlSii
tions and assumptions which are not the politician's. 1 shall develop
the analysis in this direction below.
Bourdieu describes political discourse as a field of struggle, internal
struggle to produce and sustain a coherent political discourse within
the current structured set of political discourses, external struggle to
constitute a political public and a base of support and trust for that
political discourse and the institution and charismatic individuals
associated with it. These processes of struggle can only properly be
appreciated through fine-grained analysis of texts. This is where
critical discourse analysis can supplement Bourdieu's rather more
abstract analvsis. Just as Bourdieu's analysis is an analysis of discour-'
ses but not genres, it is also an analysis mainly of choice or paradigma­
tic relations between discourses rather than the chain or syntagmatic;
relations of political texts constituted within particular genres.
Although it is not the focus of his concerns, Bourdieu does allude to
the reception of political discourse by its publics, and especially to
apoliticism and antagonism towards professional politics and its per­
formances. Bourdieu attributes this phenomenon to the political,
impotence of ordinary people and the monopolization of politics by-
the professionals. As recent work in media sociology has shown, the-
analysis of reception has to be a significant element in analysis of
mediatized political discourse. (One clear limitation of my Thatcher
study was that it ignored this issue.)

The order of mediatized political discourse

My aim in this section is to begin to give a broad overview of the order

of mediatized political discourse. 1 will continue the overview in the-
 The order of mediatized political discourse 185
next main section, after discussing an example. This is one part of the
approach I suggested in the last section; the other would be a similar
overview of the (non-mediatized) order of political discourse, in
Parliament and so forth. This is an initial attempt to map out a
■ complex area of practice, and much of the detail would have to be
.... worked out through closer studies of particular parts of the area (e.g.
political interviews, or party political broadcasts). Needless to say,
the map is a tentative one, which should be regarded as a suggestion
for readers to check out, modify and argue with. I shall sometimes
refer to the Thatcher example as a point of reference, suggesting how
the analysis might be developed.
We need to specify the repertoire of voices, discourses and genres
within the order of mediatized political discourse, the relationships
f : o f choice and alternation within each of these repertoires, and how
r» particular voices, discourses and genres are articulated together in
different types of media output. Let's begin with voices. A first step
here is to identify the main categories of social agent that contribute to
> political discourse in the media. We can initially distinguish five: pol­
itical reporters (journalists, correspondents, radio and television
presenters, etc.); politicians, trade union leaders, archbishops, etc.;
experts; representatives of new social movements; and ordinary
people. There are two important questions about voices. One is to
specify in more detailed terms which voices figure in mediatized
politics, given the obvious but important point that it is not just pro-
t fessional politicians who produce media political discourse. The
\ other is to show how the various categories of voice are structured in
relation to each other in mediatized political discourse - who, for

JI
*"
example, tends to have the last word?
The politicians who feature in the media are for the most part lead­
ing members of the main parliamentary political parties (government
ministers, MPs, MEPs, etc.), but members of smaller extra-
parliamentary parties and groups, as well as local politicians, do
•sometimes figure. We can also include other groups who sometimes
; feature in their capacity of significant actors in political life - trade
p unionists, representafives of religious organizations, members of the
royal family, and so forth. Experts include political commentators

r
|
{ and analysts, and experts in various fields of social policy and so
forth, who are often academics.

Representatives of the new social movements are a significant con-

temporary addition to the range of political voices in the media.
These include green organizations like Friends of the Earth or Green-

1 Peace, Shelter (an organization which supports homeless people),

t iN'

186 Political discourse in the media

Qxfam, or various 'single-issue' organizations representing, for

instance, single parents (National Council for One-Parent Families),
deprived children (e.g. the Child Poverty Action Group), or former
prisoners. The prominence of these voices in contemporary
mediatized politics (McRobbie 1994) reflects their increasing influ­
ence and support, and the relative weakening of traditional political
parties. McRobbie points out that effective media opposition to the
government's scapegoating of weaker social groups such as single
mothers often comes from articulate, media-trained representatives
of these organizations rather than from opposition political parties.
Another major political voice is the voice of ordinary people. Ordi­
nary people have featured for a long time in news and documentary
programmes as 'vox pop', edited and circumscribed extracts from
interviews with ordinary people which incorporate an element of
popular reaction into reports on political and social issues. More
recently, ordinary people have started to play a more active part in
political conversation and debate in talk shows and especially audi­
ence discussion programmes such as the Oprah Winfrey Show or (in
the UK) Kilroy and The Time, The Place, and the audience-discussion
format seems to be having some influence on news and current
affairs broadcasting more generally (Livingstone and Lunt 1994). The ■;
hallmark of such programmes is that ordinary people are involved on
a relatively equal footing with experts (sometimes politicians) in
unrehearsed and virtually unedited discussions on topical issues,
loosely managed by a media 'host', in which the experiences of ordi­
nary people often have a higher status than the expertise of the
experts, undermining the conventional status of the latter in the
media (though recall my analysis of an extract from the Oprah Winfrey,
Show in Chapter 7).
These major categories of voice are distinguished at a high level of
generality, and each of them is internally diverse. Amongst profes­
sional politicians, for example, there is a structured complex of voices
categorized according to political party and according to tendencies
within political parties - individuals who figure in the media are
standardly identified as Conservative, Labour, etc., and more speci­
fically as, for instance, Thatcherite, 'pro-' or 'anti-European', and in
the Labour Party as 'traditionalist' or 'modernizer'. Note that media
categorizations of professional politicians may be rather different,
from categorizations within the world of professional politics itself or
within academic analysis. One issue worth attending to is the social
class, gender and ethnic distribution of the range of voices within
mediatized politics. There has been a significant increase in the
 The order of mediatized political discourse 187
salience of women's political voices in the media. In particular, many
representatives of new social movements are women, though
women are still very much in the minority overall, certainly amongst
professional politicians, political analysts and presenters. An inter­
esting question is how those women who figure prominently in the
media, for instance, Virginia Bottomley in the Major government, are
constructed - sheer presence is not in itself a straightforward
measure of greater equity. Virginia Bottomley, for instance, who is
very good at taking and holding the floor in interviews, is commonly
referred to as 'unstoppable' or something equivalent, evoking
stereotypes of women as tending to talk too much. Black and brown
faces and voices, on the other hand, are still very unusual in tele­
vision politics (the television newsreader Trevor Macdonald is the
noiable exception). Working-class voices are quite common amongst
ordinary people who figure in programmes, though hardly in pro­
portion to the overall composition of the population, but are rela­
tively rare otherwise. Again, consti'uction is a key issue: Ken
Livingstone and Dennis Skinner are examples of working-class politi­
cians who have been demonized in the media.
The relationship between voices and discourses is often far from
simple, and cannot be taken for granted. For instance, the relation­
ship between discourses, political parties, and positions and ten­
dencies within political parties, is a variable and shifting one. The link
between a voice in the political field and a political discourse is not as
inherent and essential as it might appear. Groups and individuals
change their discourses in response to changes in the political field,
one consideration being to sustain their relationships of similarity to
and difference from other groups and individuals. The political
discourse of a political voice is in this sense always relational and rela­
tive (Bourdieu 1991). For example, in the mid-1990s leading Labour
politicians draw upon discourses which would have been 'Thatcher-
ite' fifteen years ago, and with the election of Tony Blair as Labour
leader, a preoccupation in the Conservative Party is establishing
'clear water' between Conservative and Labour.
A major opposition in mediatized political discourse is between the
professional political discourses which derive from the field of
politics, and lifeworld discourses which are based in ordinary experi­
ence. While professional political discourses are mainly drawn upon
by politicians and lifeworld discourses mainly by ordinary people,
the picture is rather complex. On the one hand, ordinary people may
draw upon professional political discourses to varying degrees. On
the other hand, it is a striking feature of contemporary mediatized
188 Political discourse in the media
politics that lifeworld discourses are appropriated, again to varying
degrees, by politicians and media reporters. Another issue is what
discourses are brought into the political domain by the new social
movements - ecological and feminist discourses are obvious
examples. Discourses may be 'drawn upon' in various ways. They
can simply be unselfconsciously used, they can be self-consciously
deployed for rhetorical purposes, or they can be contested, under­
mined and struggled against. See below for an example of the latter.
In the mapping of voices on to discourses, it is sometimes difficult
to determine what is collective and what is individual. There is, for
instance, an unclarity in my analysis of the Thatcher interview: is the
discourse that I describe the discourse of a collective voice (the
Thatcherite political tendency), or of an individual voice (Margaret
Thatcher)? It seems to be a bit of both. The merging of conservative
and neo-liberal discourses is a general feature of Thatcherism,
whereas the way in which Thatcher manages her self-positioning as a
woman political leader is obviously a matter of her individual style.
But I am not sure where exactly to draw the line between general and
individual features in the analysis. To do so would require a fuller
investigation of the discourse of Thatcherism and a comparison
between Thatcher and other leading figures within the tendency
(such as Sir Keith Joseph in the earlier part of the Thatcher period, or
John Major at the time of writing). Be that as it may, individual
idenucy and cnunsiua p-iUv-ui^on lAoT; is an essoTTtial part or politics,
and the analysis therefore needs to attend to distinctive individual
voices and styles of political discourse as well as to the discourse
associated with collective voices at various levels of generality.
Much of contemporary political discourse is mediatized political;
discourse. Its major genres are no longer just the traditional genres of
politics, they are also the genres of the media. Traditional political
activities and their genres - parliamentary debates, party conferen­
ces, international conferences - carry on, but they too are represented
in the media. And they are represented within the formats and gen­
res of the media - news, documentary, and so forth - so that their
representation is always a selective recontextualization (see Chapter
6 ) according to the requirements of these formats and genres. At the
same time, genres for political discourse that the media themselves
generate are increasingly important for politicians - most notably the
political interview, but also, for instance, phone-in programmes.
Also, the boundaries between professional politicians and other,
media voices are increasingly blurred, with politicians appearing, for'
instance, on chat shows along with other celebrities such as pop
 The order of mediatized political discourse 1S9
stars, or presenting or 'hosting' programmes like journalists (e.g.
Robert Kilroy-Silk, the host o f K/lroij, isa former Labour MP - see Liv­
ingstone and Lunt 1994). An example is a two-part television docu­
mentary in 1993 called Tomorrow's Socialism which was presented by
the former Labour Party leader, Neil Kinnock. An account of the
order of mediatized political discourse would need to specify the
; range of genres where political discourse appears, and the roles that
various political voices have within them.
Genre is the point of intersection between choice relations and
chain relations. Genres have structure, constraints on syntagmatic
| organization which I discussed in Chapter 5 in terms of 'activity
j; types'. A fruitful line of analysis is to look at the ordering which these
| activity types impose upon political voices and political discourses.
L For example, the genre of political interview can be seen as a device
|, for articulating together the voices of the professional politician and
| the radio or television reporter. Political interviews are not, of course,
I private conversations, they are interactions with audiences. The ori-
I entation to audience is evident in the discourses that are drawn upon.
| Reporters justify the adversarial nature of many contemporary pol-
f itical interviews and their challenging and even aggressive ques-
l tioning of politicians (Bell and van Leeuwen 1994) by claiming to
| speak on behalf of the ordinary people in their audiences. Sometimes
I’ they draw upon lifeworld discourses, simulating the talk of ordinary
\ people. But politicians also often claim their credentials as repre-
.I sentatives of ordinary people by drawing upon lifeworld discourses,
, so that many political interviews are an amalgam of the discourses
of professional politics and ordinary life. Recall, for instance, the
discussion of extracts from the BBC Radio 4 Today programme in
Chapter 7.
While much media output draws upon established genres in rather
conventional ways, some is more creative, generating novel genres
through innovative combinations of established ones. Livingstone
! and Lunt (1994) see audience discussion programmes as creative in
j' this sense, evolving a genre which combines three established genres
) of debate, romance and therapy, with the debate element itself
involving a configuration of different genres of dialogue - quarrel,
debate proper, critical discussion, inquiry and negotiation. An
important consequence is that the main voices that figure in audience
‘ discussion programmes - television reporters, experts, politicians,
and ordinary people - are themselves complex configurations of
roles, given that each different genre which contributes to the mix is
likely to ascribe a particular role to these voices For instance
190 Political discourse in the media
the generic ambiguity is clearly seen in the role of the host: is he or she
the chair of a debate, the adored hero of a talk show, a referee, a conci­
liator, a judge, the compere of a game show, a therapist, the host of a
dinner party conversation, a manager or a spokesperson? At times, the
host plays any one o£ these roles, thus altering the roles of other parti­
cipants and listeners.
(Livingstone and Lunt 1994: 56)
Combinations of genres may be simultaneous in the sense that even a
short sequence from a programme may be generically hybrid. But
programmes also combine together different genres and multiple
instances of the same genres sequentially, producing complex articu­
lations and orderings of voices. I shall illustrate this in the next main
section.
Although the focus in my analysis of the Thatcher interview was on
Thatcher's political discourse, I did give some peripheral attention to
questions of genre which might fruitfully be developed. The media
genre here is a form of celebrity interview: the programme was one of
a series of in-depth interviews with prominent figures in various
walks of public life. This genre is associated with particular expecta­
tions about the construction of the interviewer-interviewee relation­
ship and the interviewee-audience relationship, and about the
nature of and relationship between questions and answers. Ques­
tions are designed to probe the personality and outlook of the inter­
viewee, and answers are expected to be at least to some degree frank
and self-revelatory. Audience members are constructed as over­
hearers listening in on a potentially quite intense interaction between
interviewer and interviewee. The programme should at once be
educative and entertaining.
Thatcher, however, handles the interview in part as if it were a pol- ;
itical interview and also an occasion for political speech making, ■
treating the audience rather than the interviewer as addressee, and(
constructing the audience as a mass audience rather than purporting;;
to address audience members individually as broadcasters normally:
do. The speech-making element is evident, for instance, in the listing,
structures I referred to earlier, such as lines 23-29 in the extract (page
178). The tension between the two participants' assumptions about
genre - the tension between the practices of political discourse and
the practices of media discourse - is evident in the uneasy relation­
ship between the interviewer's questions and Thatcher's answers.-
Thatcher behaves as she might in a political interview, using the
questions as opportunities to say what she wants to say without
trying to compliantly answer them. Thatcher does not, for instance,.
 The order of mediatized political discourse 191
really answer the question in the extract. Moreover the questions, in
accordance with the interviewer's assumptions about genre, are
asking for a level of reflection and self-analysis which the answers do
not give; the discourse of the questions is studiously avoided in the
answers. This is particularly clear when Thatcher is asked (not in the
extract) about her 'vision of Britain', 'what is it that inspires your
action'; Thatcher's answer avoids the self-analytical discourse of the
question. While the interviewer is trying to get Thatcher the person to
reveal herself, Thatcher the politician is intent on constructing herself,
and her public. A later question tries unsuccessfully to engage
Thatcher in theoretical debate, referring to Thatcherism as 'radical'
and 'populist'; Thatcher says she 'wouldn't call it populist' but then
talks (in a populist way) of 'striking a chord in the hearts of ordinary
people', and uses the word 'radical' in a common-sense way but not
the semi-technical political sense intended by the interviewer.
The tension indicated here between traditional practices of political
discourse (including practices of political interview which originate
in the media but have now become naturalized as part of the tradi­
tion), and the media practices which constitute the frames within
which politicians are now required to operate, are I suggest a rather
general feature of mediatized political discourse. From a media per­
spective, Thatcher's performance in this instance is not awfully
successful, and indeed failure on the part of politicians to successfully
negotiate the demands of the new and for many of them unfamiliar
field of politics is quite common, despite increasing attention to
media training (Franklin 1994). How audiences might react to a per­
formance of this sort requires separate analysis. My analysis of the
Thatcher interview in effect abstracted the political discourse of
Thatcherism from the media genre in which it was located. I would
now argue that the question of what sort of political discourse politi­
cians produce in mediatized politics cannot and should not be
divorced from the question of how politicians reconcile their tradi­
tions with the unfamiliar and shifting demands of media practices
and genres.

An example: the Parent's Charter

I want now to show how complex sequences of genres within a pro­
gramme can impose orderings upon voices and discourses, and to
illustrate again the tensions that can arise between political discourse
and media formats, i shall refer to an edition of the BBC1 current
affairs programme Panorama which was devoted to the Citizen's
'iH f l

i , ..qt j
■' --C- 192 Political discourse in the media
Charter, a government initiative associated especially with Prime
Minister John Major and designed to give people redress in cases
where public services fail to meet designated standards. One of the
reports which made up the programme dealt with the Parent's
Charter, the part of the Citizen's Charter that applies to schools. My
focus will be on how professional politicians are positioned and
isolated in this part of the programme.
The report has three main stages - introduction, story, and
analysis; an introduction by the presenter, which also effects a transi­
tion between this report and the previous one; a story, about children
who have been left without places in secondary schools in the outer
London borough of Bromley, and in particular about one such child,
Helen; an analysis stage, in which the Parent's Charter and especially
its claims about parental choice are analysed. A variety of source
■materials are deployed including several interviews and films. These
are heavily edited, with short sequences being spliced together, and
different sequences from the same interview or film being placed at
different points in the report. Several different genres are drawn
upon: interview, presenter narrative, political commentary and
analysis, political speechifying, expert opinion, and ordinary conver­
sation and conversational narrative.
I shall begin with a summary which indicates some of the articulations
of genres, voices and discourses. This report immediately follows an
exrracc ia m i a n u d a view with. William Wakkgrave, Cabinet Minister .
with responsibility for the Citizen's Charter in which he gives an account
of its objectives. In the introduction, the presenter, in making the transi­
tion from this to the Parent's Charter report, gives a contrary and critical
account of the Citizen's Charter as part of a Tory strategy to pursue private-
sector methods while holding the public purse strings tightly.
The story opens with film of a public meeting and the angry voices
of parents whose children are without a place at secondary school
because the government's strategy of 'parental choice' - a focus of the
Parent's Charter - has led to local schools being over-subscribed. :
Voice-over commentary from the presenter formulates the situation
as one person's choice has become another's denial. The presenter then
quotes from and critically analyses the Parent's Charter, and narrates
the case of Helen, who has no secondary school place to go to. Like
most of the presenter's contributions, this is in voice-over. There foly
low extracts from interviews with her mother, then her father, then.
Helen. The presenter goes on to talk about the school Helen wanted
to go to. There is an interview with the headteacher, the presenter
describes Helen's 'painful' position of being excluded while most of
 The order of mediatized political discourse 193
her friends were given places, and there is then more interview with
Helen. The next sequence deals with Helen's second-choice school.
The presenter describes the school, there are further extracts from the
Parent's Charter, and an interview with the headteacher. There is
then a sequence focusing on the local education authority, opening
with film of the town hall and voice-over from the presenter which
refers to officials' apprehension about the public meeting, and
shifting to more shots of the meeting with the voices of angry parents
and the headteacher of Helen's first-choice school. The presenter
announces that Helen has now at last been allocated a place, and
there is an interview with her mother which closes off the story stage.
The analysis stage opens with extracts from an interview with a
local Conservative councillor who resigned her position as chair of
the Education Committee in disgust at the chaos in secondary school
admissions. This is juxtaposed with further extracts from the inter­
view with William Waldegrave. The presenter then comments that
the Citizen's Charter is not about rights but about remedies, and the fea­
ture closes with a snatch of interview with the Treasury Minister
Francis Maude, which appears to confirm that conclusion.
The main voices, then, are those of the presenter, ordinary people
(Helen, her parents, people at the meeting), the headteachers
(experts of a particular sort), and the politicians (the local councillor,
the government ministers). Particular voices are associated with par­
ticular genres. The presenter narrates, analyses and comments. The
headteachers give expert opinion. The ordinary people in the reports
converse, but so too does the local councillor. Waldegrave, the
government minister, by contrast makes a political speech. The
/report is organized in a way which sets up an opposition between the
■generally discredited genre of political speechifying used by Wal­
degrave, and the other genres. Particularly striking, given that they
■are both Conservatives, is the juxtaposition of the conversational lan­
guage of the local councillor and the political rhetoric of the minister:

conservative COUNCILLOR: I blame the government . 1 . 1 feel .

that if . I - I very much support this Citizen's Charter I think it's about
time . the ordinary citizen had the right to take public authorities to
task . if they don't deliver what they what they ought to 1 - I very
much support that 1 think it's a great idea . a lot of people who are in
favour of market market forces say no that's wrong market forces must
be an absolute thing . but . in the case of choice it can't be absolute
can it it can't be . absolute for everybody . 1 mean these parents who
are . milling about . so unhappily now . they won't feel that
they've had wonderful choice
194 Political discourse in the media
The order of mediatized political discourse 195
william waldegrave: now we're enabling parents to make a . a real Charter, and it is noteworthy that all categories of voices - ordinary
fuss . a fuss that means something . saying we've compared schools people, experts (headteachers) and the presenter as well as politi­
we actually like . these ones we want more of these now I don't I'm of
cians - speak politically and engage in a struggle over the discourse of
course not welcoming the fact that not all getting their first choice .
choice. Let me list in order some of the formulations used. Presenter:
now . but
presenter : it's not that they're not getting their first choice (Wal­
the question is whether all this choice will really make the difference in hospi­
degrave: right) they're saying you can't go to any school (in the tals schools. Parent in meeting: it strikes me that it's the schools that have
borough) the choice and not the parents. Presenter: in Bromley one person's choice has
waldegrave: well they want to g- they want to go to schools like those become another's denial. First headteacher: if you have the . a vast amount
provided in Bromley . they want more schools . like that . and of choice you've also got the responsibility o f coping with that choke and that
they're going to push to get em . now in the old days they would have may mean encouraging a child to understand that she's got to wait.
just had to take what they were given . by local education authorities Presenter: like the Parent's Charter's promise o f choice . the music depart­
following guidelines from the department . but that's what's turning ment Helen fancied was just wishful thinking. Second headteacher: Brom­
round it's rather a good example of how we want to . get the energies ley parents have less choice now than they did before. Parent in meeting: i
of citizens . to to work f- for themselves but also . for the standards feel you've taken my decision-making away from -me. Presenter: a
which will then spread across the country government that has promised them a right to choose whichever school they
Panorama, 18 May 1992
want to go to and yet the Council can't deliver. Conservative councillor: in
the case of choice . it can't be absolute can it . it can't be absolute for every­
There is a widely noted ambivalence about the Citizen's Charter body . I mean those parents who are milling about so unhappily now . the]/
which allows it to be read in terms of either a discourse of citizenship won't feel they've had a meaningful choice. Waldegrave's only mention of
or a discourse of consumption. The issue is whether it seeks to choice differs significantly from all these in limiting and relativizing
enhance citizens' rights, or is solely concerned to subject public choice: he refers to parents' first choice.
services to market pressures from consumers. The formulation ot the In addition to these various formulations, justifications and contes­
Citizen's Charter given by William Waldegrave immediately before tations of the political discourse of the Charter, a number of other pol­
the Parent's Charter report (getting quality into public service) draws itical discourses briefly surface in the feature. The presenter draws
upon the discourse of consumption, as does his first contribution in: upon a left oppositional discourse early on, and actually gives this
the extract above in foregrounding parental (in effect, consumer) discourse the status of truth: the Charter has to be seen for what it is .
pressure as a mechanism for raising standards. There is a market- part of a Tory strategy to pursue private-sector methods while holding the
oriented metanarrative underlying the government's political public purse strings tightly. There is also a snatch of cynical lifeworid
discourse around the Citizen's Charter: the Charter will give people' discourse about politicians from a parent in the film of the meeting:
rights of redress for poor service and will thus stimulate consumer, -s- tit's because of this meeting I reckon they got them out quick. And the last
pressure which will push up standards. Maude, at the end of the tea-; contribution from Helen's mother draws upon a discourse of grass­
ture, explicitly denies that the Citizen's Charter has to do with citi-l roots political campaigning:
dii:
zens' rights, and formulates it in consumer terms as making sure the! HI
citizen gets a good deal for the money that the government spends on the citi­ is presenter: are you going to continue this process to see it doesn't hap­
zen's behalf. But notice that although he is referring to people as con-: pen again
ST: . mother: oh absolutely until every child this year is settled into school
sumers he uses the word citizen, articulating together the discourses 1
mm u and then we shall continue until we make sure it never . ever .
of consumption and citizenship. . happens again
Apart from the voices o f government ministers, a discourse at
m
citizenship predominates. The focus is upon the right to parental The mother's contribution is striking in its delivery, drawing upon a
choice which the Parent's Charter seems to embody: You can choose the \ non-professional political rhetoric which is manifested in a slow
school you would like your child to go to. The report is a problematic iln>n emphatic delivery of absolutely and the last four words (never . ever .
and elaboration of this aspect of the political discourse of the Citi 'en 5 W- happens again).
196 Political discourse in the media
The isolation of the government ministers (Waldegrave in parti­
cular) is terms of genre and discourses is underscored by the posi­
tioning of their voices in the report. There is a broad movement up a
chain of responsibility within the report, from the voices of ordinary
people (the meeting and the case of Helen), to the experts (head­
teachers), to the local councillor, and finally to the government
ministers. The voices of the two government ministers are cut off
from the others in being located peripherally at the end of the feature
(there is also the Waldegrave interview extract that immediately pre­
cedes the feature). There is an element of drama here, with the
ministers being confronted with an alliance of voices, discourses and
genres which is built up in the main body of the report, especially in
the story. The sense of Waldegrave being confronted by this chorus
of critical voices is accentuated by the juxtaposition already noted
between his contribution and that of the highly critical councillor, the
fact that she is also a Conservative underscoring the dash of
positions.
The reading of this report which 1 have been foregrounding is one
which emphasizes tension and struggle between the media and pro­
fessional politicians, seeing the presenter in this case as orchestrating
and contributing to an oppositional alliance drawing together voices
from the lifeworld, the professions, local politics, and journalism.
The professional politicians are edited into a corner, as it were.
According to this account, using the media as a way of communicat-
•mv
x- ‘ o "
IVic * rnnhli
.... —
.''1ic tv> c.iniiMp
- — ........... r ...... . •
fn r rrinf'prrinrirpjrv
.......- - -. - j
T^oHHci^nS
* ■......- ■■ ’■
The terms on which the media can be so used are often demanding,
and in some cases the media seem to be able to dictate them (though
less so for a powerful politician like Thatcher).
However, a very different reading of reports of this sort is possible,
according to which the difficulties and challenges facing politicians
are mainly there to give an impression of the autonomy of the media
and the answerability of politicians, while at a deeper level there is
complicity and cooperation between journalists and politicians to
sustain the status quo. Waldegrave and Maude may be isolated and
confronted at the end of the report, but the emphasis they clearly
want to put on quality and value for money in their account of the
Citizen's Charter is allowed to stand as the last word, and the political
discourse of the professional politicians is given the legitimacy of pro--
viding closure for the feature. Moreover, the oppositional voices and
discourses with which the ministers are confronted are very limited,
with notable absences. For the most part, it is accepted that citizens,
rights in the domain of education are appropriately formulated m
 Further considerations 197,

terms of the consumerist discourse of choice; people do not challenge

'choice', they merely argue that 'choice' isn't working. This is linked
to the programme's selection of the white, middle-class, Conser­
vative borough of Bromley; the ordinary people who figure do not
include people of colour, or working-class people. Their predomin­
antly middle-class nature is visually obvious, as well as indicated by
the language they use. If the programme had set out to seriously chal­
lenge and corner ministers, it could have brought together far more
damaging voices and discourses. Voices and discourses which,
especially since the retreat of the official opposition Labour Party to
safe middle ground, are rarely seen or heard at all in the mainstream
media. These two alternative readings suggests an ambivalence in
the relationship between politics and the media which I return to in
the final section of the chapter. In certain cases, it may seem that the
media control the politicians, in others that politicians manipulate the
media in complicity with journalists. In many cases, as here, the
relationship is ambivalent.

Mediatized political discourse: further considerations

So far, my overview of the order of mediatized political discourse has

focused upon the identification of repertoires of voices, discourses
and genres, and the shifting configurations/articuiations which they
enter into. But there are a num ber of other considcratiorib which are
relevant in characterizing the order of mediatized political discourse.
I shall discuss some of these rather briefly in the final section. The
issues I take up here are: boundaries between mediatized political
discourse and other orders of discourse; the production, distribution
and reception of mediatized political discourse; the diversity of prac­
tices within it; ongoing change in practices; politicians' access to, and
training in, the practices of mediatized political discourse.
The issue of boundaries or frontiers between orders of discourse
has in fact already been quite extensively discussed above, in that
mediatized political discourse has itself been presented as an area of
intersection and tension between the orders of discourse of profes­
sional politics and of the media. However, there are other frontiers,
also areas of intersection and tension, which need to be attended to.
There is, for instance, the ill-defined frontier between political
discourse and the discourse of government (the state, administra­
tion, bureaucracy), which is of current concern given the dramatic
mcrease in government and local government advertising in the
198 Political discourse in the media
media and accusations of the politicization of public information (in,
for instance, campaigns to encourage people to buy shares in pri­
vatized public industries such as Gas and Water). There is also an
important frontier area between mediatized political discourse and
academic discourse, the discourse of various types of experts in and
around the political arena, including the discourse of the 'think tanks'
which have become a major element in the contemporary political
process. These and other examples underline the view that the field
of political discourse must be seen as an open field, where frontiers
with a range of other fields are constantly being negotiated.
An adequate account of the order of mediatized political discourse
would need to incorporate accounts of its key processes: the pro­
duction, distribution, and consumption/reception of mediatized pol­
itical texts. In Chapter 3 I discussed in general terms the importance
of attending to all three processes in analysing media discourse, and I
shall not repeat that argument here. The discussion above of voices,
discourses and genres has focused upon the processes of production,
though without going systematically into the practices and routines
of production. Political discourse gives rise to particular issues of dis­
tribution and reception. With regard to distribution, one issue is what
one might call the 'trajectories' of different types of political
discourse: their varying distributions across discursive practices and
domains within the order of discourse of the mass media and within
other orders of discourse (such as those of government, or
education), the mreriexiual chains (set- Chapter 4) they enter into,.
and the transformations they undergo as they move along these
chains. Some types of political discourse-for example a major speech j
by a leading politician - have highly complex trajectories, entering'
into many domains of reception, which will to a degree be anticipated
in the way in which they are produced. Part of what is involved here,:
is a complex process of recycling within the media: an interview with
a politician on a breakfast television show can become a lead item in
both newspapers and broadcast news programmes, and a topic for
editorials, current affairs programmes, and phone-in programmes.
Mediatized political events often themselves constitute the main pol­
itical news.
The consumption/reception of media discourse raises a number of-
specific issues. One is the general question of how political discourse
impacts upon people's lives - what wider influences and ramifi­
cations it has beyond the portions of people's lives that they devote to
watching, reading or listening to the news and other political mater­
ial. One way into this difficult question is through discourse analysis
j Further considerations 199

j7 of the ways in which people talk about mediatized politics, focusing

j specifically upon the question of what parts of their discursive reper-
! toires they draw upon in doing so, and how the political discourse of
' the media is recontextualized and transformed in their talk, and ar-
! ticulated with elements from other discoursal sources. For example,
l in an article in Discourse and Society (Fairclough 1992b) 1 suggested that
( people taking part in a reception study of Israeli political television
I were drawing upon the discursive practices of ordinary life ('the life-
! world') in talking about mediatized politics - recontextualizing that
t" ■ political discourse within everyday experience. A related issue
discussed briefly in that paper is whether and to what extent audien-
i ., ces critically analyse and 'deconstruct' mediatized political discourse
( (see Liebes and Ribak 1991, Livingstone and Lunt 1994). A more
■ general issue is the worrying alienation of people from party politics
{ s in a number of western democracies. Analysis of consumption/
reception of mediatized political discourse should be able to throw
t: some light upon this very significant development. (See Livingstone
and Lunt 1994 for a discussion of the related issue of increasingly
* critical responses by audiences to experts on television.)
I A danger in giving an overview of the complex field of mediatized
political discourse is that the diversity of practices will be lost sight of.
A fuller description wou Id need to go into differences between types of
| media (television, radio, press), differences between outlets in each
i type (between television stations, between radio stations, between
j newspapers), and diversity within particular outlets. For example,
practices of political interview even within a single television station
; such as BBC1 are not homogeneous - they vary between programmes,
] but even within programmes according to the models preferred by
^ particular interviewers and editors. The picture is one of considerable
diversity, instability and change. Practices like political interview are
( sensitive barometers of wider processes of social and cultural change,
j showing subtle shifts in, for instance, the construction of the identities
of both politicians and journalists, and in social relationships between
: j~~". them, and between them and audiences (Fairclough 1995). Relation­
s' ships between diverse practices may be relationships of struggle, with
f*"' particular practices coming to symbolize wider positions and interests
within media institutions which in turn may be linked to wider social
; struggles. The order of mediatized political discourse, like other ord-
j ers of discourse, can usefully be regarded as a domain of cultural hege-
. . rnony which is constantly open to hegemonic struggle, a struggle for
: power within media institutions which will relate, if in possibly
, - complex and indirect ways, tostrugglesforpowerinthewidersociety.
200 Political discourse in the media
The picture, then, is a very complex one, and vve need to beware of
easy simplifications.
Another issue is the issue of access to mediatized political
discourse. This partly overlaps with the discussion earlier of the
range of voices within the order of discourse, and the relationships
that are constructed between the voices (e.g. between professional
politicians and ordinary people) in particular types of output. There
is also the increasingly important matter of the apparatuses which
political parties have developed to train their members in using the
media, to prepare and groom them for media appearances (often
radically changing their appearance, clothes, and communicative
style), to set the agenda of political news, and to optimize the media
exposure of their members.

The emergence of mediatized politics is sometimes seen as the colon­

ization of politics by the media, and sometimes seen as the colon­
ization of the media by politics. Certainly the energy and resources
that political parties and national and local government are now put­
ting into their information and communication departments indi­
cates a major effort on their part to control their relationship with the
media - to 'package politics' as the title of a recent book on the subject
puts it (Franklin 1994). Yet if we consider the relationship between
politics and media from the perspective of discursive practices, it is
clear that it has required more concessions and adaptations from
politicians than trout the media. One indicator ui this is what has hap­
pened recently to politicians who are exceptionally gifted in the tradi­
tional discursive practices of politics- the great political orators. They
have become marginalized, have lost their public visibility, and have
even become figures of ridicule. Michael Foot, a former leader of the
Labour Party, is a good example. If the political apparatuses do
largely dictate the agenda of mediatized politics, they do so only at
the price of a radical mediatization of the internal practices of politics
which has profound implications for the viability and legitimacy of
the political public sphere. But the settlement that has been arrived at
between politicians and the media is not a stable one. It is a relation­
ship of complicity and mutual dependence which is constantly unset­
tled by its contradictions, for the agendas of politics and media are,
not in the end the same. Oscillation between harmony and tension,;,
trust and suspicion, are inherent. The order of mediatized political?
discourse is itself, therefore, an essentially unstable one.
 CRITICAL MEDIA LITERACY

in tiie course or this book, i have emphasized the importance of the media
and of media discourse in wider processes of social and cultural change,
and in wider power relations and ideological processes in society. The
media, and media discourse, are clearly a powerful presence in contem­
porary social life, particularly since it is a feature of late modernity that
.cultural facets of society are increasingly salient in the social order and
social change. If culture is becoming more salient, by the same token so
too are language and discourse. It follows that it is becoming essential for
effective citizenship that people should be critically aware of culture,
discourse and language, including the discourse and language of the
media. As a conclusion, I want to draw together some of the issues and
analytical methods dealt with in this book in the form of questions that
someone who is critically li terate in the language of the media ough t to be
able to answer about a media text - a newspaper article, a programme on
television, or a radio programme. Critical media literacy is not just a
matter of awareness of media discourse (Luke 1994)- i t also includes, for
instance, awareness of the economics of media and production processes
within the media - but critical awareness of language and discourse is an
202 Critical media literacy
important part of it (Fairclough 1992c). What follows can be regarded
as a tentative agenda for teachers.
1 suggest that it ought to be an objective of media and language
education to ensure that students can answer four questions about
any media text:
1. How is the text designed, why is it designed in this way, and how
else could it have been designed?
2. How are texts of this sort produced, and in what ways are they
likely to be interpreted and used?
3. What does the text indicate about the media order of discourse?
4. What wider sociocultural processes is this text a part of, what are
its wider social conditions, and what are its likely effects?
These are of course very general questions, which can be developed,
into more specific ones. Notice that the book has not dealt equally
with all of them: there is more material appertaining to 1 and 3 than to
2 and 4. I shall discuss them in turn.

1. How is the text designed, win/ is it designed in this way, and how else
could it have been designed?
This question highlights the idea that texts are based upon choices,
and that alternative choices might always have been made. Some­
times the question will direct attention to the variation that currently
exists in media practices - for instance, the sort of variation in radio
science that was brought out in Chapter 7 in the comparison between
Medicine Now and High Resolution. Sometimes the question will sug­
gest that current practices are shaped by (and help shape) current
social and cultural circumstances - and that things might be (and per­
haps once were, and will be) different.
The question of how texts are designed has received more a ttention in .
the book than any other. I have provided a 'metalanguage' for talking
about the language and intertextuality of texts. Such a metalanguage is
essential for a critical literacy of media language, but developing a
metalanguage which can be made generally accessible through the;,
educational system is a formidably difficult problem which I have not
addressed in this book. Let me summarize as a further and more specific
series of questions the main forms of analysis introduced in the book. I
also include some of the types of linguistic and textual analysis used in
connection with each group of questions.
(a) Intertextuality
« What genres, voices and discourses are drawn upon, and how arc
they articulated together?
 Critical media literaaj 203
- direct and indirect speech, generic structure or 'staging', narrative
analysis (story, presentation), conjunctions, collocations

(b) Language
1. Representations
• What presences and absences, foregrounding and background­
ing, characterize the text?
• What process and participant types are there? How are processes
and participants categorized and metaphorized?
« What relationships are set up between propositions (clauses) in
texts?
- presupposition, process and participant types, nominalization,
agency and voice (active and passive), categorization and wording,
metaphor, main and subordinate clauses, theme, local and global
coherence relations

ii. Relations and identities

What are the participants (voices) in the text, and how are they con­
structed?
• What relationships are set up between participants - specifically
between:
- media personnel (journalists, presenters) and audiences/
readerships
- 'others' (e.g. experts, politicians) and audiences/readerships
~ media personnel and 'others'
• Are constructions of participants and relationships simple, or
complex/ambivalent?
• What relative salience do institutional and personal identities have
in the construction of participants?
- oral delivery, body movement, key (serious or humorous), con-
versationalization, vocabulary, mood, modality, interactional control
features, lists

iii. Image and text

• In the case of television, how are visual images constructed, and
what relationships (e.g. of tension) are set up between language
and image?

2. How are texts o f this sort produced, and in what ways are they likely to be
interpreted and used?
This question relates to some of the issues dealt with in Chapter 3, but of
course neither processes of production nor processes of consumption
have been major concerns of the book. With respect to production, it is
204 Critical media literacy
important to be aware that what we read in a newspaper or see on the
television screen is not a simple and transparent representation of the
world, but the outcome of specific professional practices and tech­
niques, which could be and can be quite different with quite different
results. It is also important to be aware that the practices which
underlie texts are based in particular social relations, and particular
relations of power. With respect to consumption, important issues
are the diversity of practices of reading, listening and viewing (and
their social conditions), and the potential for divergent interpreta­
tions and uses of any given text by different sections of a readership
or audience.

3. What does the text indicate about the media order of discourse?
Part of critical media literacy is an overall sense of the practices of
media and of the media order of discourse, and a sensitivity to signifi­
cant tendencies of change. This question assumes that any given
media text will shed some light upon these issues, in that it will be a
product of a particular state and evolution of the order of discourse.
Particular questions here include:
• Is the text indicative of stable or unstable relationships, fixed or
shifting boundaries, between discursive practices within the order
of discourse, and between the media order of discourse and
socially adjacent orders of discourse?
a What particular choices (indusions/evohrsions, of genres or
discourses) is this text associated with?
• What chain relationships across the media order of discourse
and/or socially adjacent orders of discourse is this text situated
within?
• What particular tendencies of change (e.g. commodification or
conversationalization of media discourse) does this text
exemplify?

4. YJhat wider sociocultural processes is this text a part of, what are its
wider social conditions, and what are its likely effects?
This question brings into the picture wider social conditions (includ­
ing economic and political ones) which constrain media discourse
and media texts, and their social effects - in terms of systems of
knowledge and beliefs (and ideologies), social relations of power,
and the positioning of people as social subjects. It also draws atten­
tion to changes in society and culture which frame the sort of
changes in the media order of discourse alluded to in question 3
above. The three-dimensional framework for critical discourse
 Critical media literacy 205
analysis introduced in Chapter 4 is relevant here: the analysis of any
media event links together statements about:
- the text and its linguistic properties
- the discourse practice - processes of text production and con­
sumption (recall that intertextual analysis links text to discourse
practice)
- the sociocultural practice which the discourse practice and the
text are embedded within

We might add to the four questions so far a fifth question suggested

in Luke et al. (1994):
5. What cati be done about this text?
The point of this question is to highlight the status of media texts as a
form of social action which can be responded to with other forms of
social action. These may be other texts - letters of congratulation or
complaint, reviews, discussions - or nontextual forms of action.
Some media texts, for instance, can stimulate public campaigns,
meetings and demonstrations. One example is the widely noted
influence in 1989 of media representations of struggles within the
former socialist countries of Eastern Europe upon the development
and spread of those struggles. Another is a powerful documentary
produced by John Pilger for Channel 4 in 1994 on the genocide
practised by the Indonesian government against the people of East
l'imor. One effect of this question, within a programme of critical
literacy awareness, may be to encourage people to move beyond
reception of media texts to action in response to those communicative
events.
2OS Bibliography
Fairclough, N. 1994: Conversationalizaticm of public discourse and the
authority of the consumer. In ICeat, R., Whiteley, N. and Abercrombie, N.
(eds), T h e a u t h o r it y o f the c o n s u m e r , Routledge.
------ . 1995: Ideology and identity change in political television. In
Fairclough, N ., C ritic a l d is c o u r s e a n a ly s is , Longman.
Featherstone, M. 1991: C o n s u m e r c u ltu re a n d p o s tm o d e r n is m , Sage Pub­
lications.
Fisk.e, ]. and Hartley, J. 1978: R e a d in g tele v is io n , Methuen.
Forgacs, D. 1988: A G r a m s c i re a d e r, Lawrence & Wishart.
Foucault, M. 1972: T h e a r c h a e o lo g y o f k n o w le d g e , trans. Shericlan-Smith, A.
M., Tavistock Publications.
------ . 1978: D is c ip lin e a n d p u n is h : th e b irth o f th e p r is o n , trans. Sheridan-Smith,
A. M., Penguin Books.
Fowler, R. 1987: Notes on critical linguistics. In Threadgoid, T. and Steele, R.
(eds), L a n g u a g e to p ics : e s sa y s in h o n o u r o f M ic h a e l H alliclay, John Benjamins.
Fowler, R. 1991: L a n g u a g e in the N e w s : d isc o u r se a n d id eo lo g y in th e p re ss,
Routledge.
Fowler, R., c t a l . (eds) 1979: L a n g u a g e a n d c o n tr o l, Routledge.
Franklin, B. 1994: P a c k a g in g P o lit ic s : p o litic a l c o m m u n ic a tio n in B rita in 's m ed ia
d em o c ra c y , Edward Arnold.
Fraser, N. 1989: U n ru ly p r a c tic e : p o w e r , d is c o u r s e a n d g e n d e r in c o n te m p o r a r y
s o cia l th eo ry , Polity Press.
Freire, P. 1972: P ed a g o g y o f th e o p p r e s s e d , Penguin.
Garfinkel, H. 1967: S tu d ie s in e t h n o m e t h o d o lo g y , Prentice Hail.
Garnham, N. 1986: The media and the public sphere. In Golding, P., et al.
(eds), C o m m u n ic a tin g p o litic s , Leicester University Press.
v_jCe, j. ivc(j: cjo ciu t u n g u i ^ i a s i m u ■..., ■ a p o l o g y w L ■o : . The ■hnlmer
Press.
Giddens, A. 1991: M o d e r n ity a n d s e lf- id e n tit y , Polity Press.
Goffman, E. 1969: T h e p re sen ta tio n o f s e l f in e v e r y d a y life, Penguin.
------ . 1981: F o rm s o f ta lk , Basil Blackwell.
Gramsci, A. 1971: S e le ctio n s fr o m th e p r is o n n o teb o o k s , trans. Hoare, Q. and
Mowell-Smith, G., Lawrence & Wishart.
Greatbatch, D. 1986: Aspects of topical organisation in news interviews: the
use of agenda-shifting procedures by news interviewees. M e d ia C u ltu re
a n d S o c ie ty , 8 (4): 441-55.
Habermas,). 1984: T h e o r y o f C o m m u n ic a tiv e a c tio n vol. 1, trans. McCarthy, T.,.
Heinemann.
------. 1989: T h e S tr u ctu ra l T r a n s fo r m a tio n o f th e P u b lic S p h e r e , trans. Burger, T .,;
Polity Press.
Hall, S. 1977: Culture the media and the ideological effect, in Curran, J., et:
al. (eds), M a s s c o m m u n ic a tio n a n d s o c ie ty , Edward Arnold and the Open.
University P r e s s . .i
Hall, S. and Jacques, M. 1983: T h e p o litic s o f T h a tc h e r is m , Lawrence &
Wishart.
| Bibliography 209
'H all, S., el al. 1978: Policing the crisis: mugging, the state, and law and order,
I Methuen. '
I Halliday, M: 1978: Language as social semiotic, Edward Arnold.
L: ------ . 1985: Introduction to functional grammar, Edward Arnold.
!. ------ . 1989: Spoken and written language. Oxford University Press.
Halliday, M. and Hasan, R. 1976: Cohesion in English. Longman,
i Hartley, ]. 1982: Understanding news. Routledge.
Heritage, J. 1985: Analyzing news interviews: aspects of the production of
f talk for overhearing audiences. In van Dijk, T. (ed.), H a n d b o o k o f d is c o u r s e
■ a n a ly s is , 3, London: Academic Press.
Heritage, J. and Watson, D. 1979: Formulations as conversational objects. In
: Psathas, G. (ed.), Everyday language: studies in ethnomethodology, Irvington.
Herman, E. and Chomsky, N. 1988: Manufacturing consent: the political
] ■': economy of the mass media, Pantheon Books.
: Hodge, R. and Kress, G. 1979: Language as ideology, Routledge.
------. 1988: Social semiotics, Polity Press.
fy ------. 1992: language as ideology, 2nd edn, Routledge.
It ' Hutchby, 1 .1991: The organization of talk on talk radio. In Scanned, P. (ed.),
Broadcast talk, Sage Publications,
i Inglis, F. 1990: Media theory: an introduction, Blackwell.
! Kellner, D. 1992: The Persian Gulf television war, Westview Press.
| Kress, G. 1989: History and language: towards a social account of language
j change. Journal of Pragmatics, 13: 445-66.
J Kress, G. and Threadgold, T. 1988: Towards a social theory of genre.
I Southern Revieiu, 21: 215—43.
| Kress, G. and van Leeuwen, T. 1990: Reading images. Deakin University Press.
^ Knsceva, j. 198t>: Word, dialogue and novel. In Moi, T. (ed.), The K r is te v a
J_ Reader, Basil Blackwell.
Kumar, K. 1977: Holding the middle ground. In Curran, ]., etal. (eds), Mass
j- communication and society, Edward Arnold and the Open University Press.
1 Labov, W. 1972: Language in the inner city, University of Pennsylvania Press,
i Labov, W. and Waletzky, ]. 1967: Narrative analysis: oral versions of perso-
if nal experience. In Helms, J. (ed.), Essays on the verbal and visual arts, Uni-
^ versify of Washington Press.
17 Lakoff, G. and Johnson, M. 1980: Metaphors we live by, University of Chicago
f’7 Press.
Lash, S. 1990: Sociology of postmodernism, Routledge.
Leech, G. N. 1966: English in advertising, Longman.
----- . 1983: Principles of politeness, Longman.
Levinson, S. 1979: Activity types and language. Linguistics, 17: 365-99.
L-'” ----- . 1983: Pragmatics. Cambridge University Press.
W ~ Liebes, T. and Ribak, R. 1991: A mother's battle against TV news: a case
study in political socialization. Discourse and Society 2 (2); 203-22.
_ , L ' Livingstone, S. and Lunt, P. 1994: Talk on television: audience participation and
-u * public debate, Routledge.
210 Bibliography
Luke, A., O'Brien, J. and Comber, B. 1994: Making community texts objects
of study. T h e A u stra lia n Jo u rn a l o f L a n g u a g e a n d L ite r a c y , 17 (2): 139-49.
Luke, C. 1994: Feminist pedagogy and critical media literacy. J o u r n a l o f
C o m m u n ic a tio n In q u iry , 18 (2): 27-44.
McRobbie, A. 1994: Folk devils fight back. N ew L e ft R e v ie w , 203: 107-16.
Mancini, P. 1988: Simulated interaction: how the television journalist
speaks. E u ro p ea n J o u r n a l o f C o m m u n ic a tio n , 3 (2): 151-66.
Mardh, I. 1980: H e a d lin e s e : o n the g r a m m a r o f E n g lis h f r o n t p a g e h e a d lin e s , CVVK
Gleerup.
Martin, J. R. 1986: Grammaticalising ecology: the politics of baby seals and
kangaroos. In Threadgold, T., et a l. (eds), L a n g u a g e s e m io tic s id e o lo g y , Syd­
ney Association for Studies in Society and Culture.
Media Literacy Report 1993 (draft): WIR Industry Research, Sydney.
Mishler, E. 1984: T h e d is c o u r s e o f m e d ic in e : d ia le c tic s o f m e d ic a l in te r v ie w s , Ablex
Publishing Corporation.
Montgomery, M. 1988: DJ talk. In Coupland, N. (ed.), D is c o u r s e S ty listics,
Croom Helm.
------ . 1990: M e a n in g s a n d th e m ed ia : s tu d ie s in th e d is c o u r s e a n a ly s is o f m ed ia texts.
PhD thesis, University of Strathclyde.
------ . 1991: 'Our Tune': a study of a discourse genre. In Scannell, P. (ed.),
B ro a d c a st ta lk , Sage Publications.
Montgomery, M ., Tolson, A. and Garton, G. 1989: Media discourse in the
1987 general election: ideology, scripts and metaphors. E n g lis h L a n g u a g e
R e s e a r c h , 3: 173-204.
Norris, C. 1992: U n critica l th eo ry , Lawrence & Wishart.
Pêcheux, M. 1982: L a n g u a g e , se m a n tic s a n d id e o lo g y : s ta tin g the o b v io u s , trans.
Nagpal, H., Macmillan.
Pilger, J. 1 9 9 2 : D is ta n t v o ices. Vintage.
Postman, N. 1987: A m u s in g o u rs elv es to d ea th : p u b lic d is c o u r s e in th e a g e o f show
b u sin e ss, Methuen.
Quirk, R., Greenbaum, S., Leech, G. and Svartvik, J. 1972: A g r a m m a r o f con ­
te m p o r a r y E n g lish , Longman.
Richardson, K. 1987: Critical linguistics and textual diagnosis. T ex t, 7 (2):
145-63.
Sacks, H., Schegloff, E. and Jefferson, G. 1974: A simplest systematics for
the organisation of turn-taking in conversation. L a n g u a g e , 50: 696-735.
Scannell, P. (ed.) 1991: B ro a d ca st ta lk , Sage Publications.
Scannell, P. 1992: Public service broadcasting and modern public life. In
Scannell, F ., et al. (eds), C u ltu re a n d p o w e r , Sage Publications.
Silverstone, R. 1985: F r a m in g sc ie n c e: th e m a k in g o f a B B C d o c u m e n ta r y , British
Film Institute Publications.
Sinclair, J. 1992: Trust the text. In Davies, M. and Ravelli, L. (eds), A dvances
in s y s te m ic lin g u is tic s , Pinter Publishers.
Straumann, H. 1935: N ew s p a p e r h e a d lin e s , Heinemann.
Stubbs, M. 1983: D is co u rs e a n a ly s is , Basil Blackwell.
 Bibliography 211

Swales, J. 1990: G e n re a n a ly s is : E n g lis h in a c a d e m ic a n d re sea rc h settings Cam

bridge University Press.
Talbot, M. 1990: L a n g u a g e , in te r te x tu a lity a n d s u b je c tiv ity : v o ice s a n d th e c o n ­
s tr u c tio n o f c o n s u m e r fe m in in it y . PhD thesis, Lancaster University
Thompson, J. B. 1984: S tu d ie s in th e th e o r y o f id eo lo g y , Polity Press.
------ . 1990: Id eo log y a n d m o d e rn c u ltu r e , Polity Press.
------ . 1991: Editor's introduction to Bourdieu 1991.
Threadgold, T. 1989: Talking about genre: ideologies and incompatible
discourses. C u ltu ra l S tu d ie s, 3 (1): 101-27.
Tolson, A. 1990: S p e a k in g f r o m e x p e r ie n c e : in te r v ie w d is c o u r s e a n d fo r m s o f su b je c ­
tiv ity . PhD thesis, University of Birmingham.
------ . 1991: Televised chat and the synthetic personality. In Scanneil, P.
(ed.), B ro a d c a st ta lk , Sage Publications.
Toolan, M. 1988: N a r ra tiv e : a c ritica l lin g u is tic in tr o d u c tio n , Routledge,
Trew, T. 1979a: Theory and ideology at work. In Fowler, R., e t a l . (eds), L a n ­
g u a g e a n d c o n tro l, Routledge.
------. 1979b: 'What the papers say': linguistic variation and ideological
difference. In Fowler, R., et a l. (eds), L a n g u a g e a n d c o n tr o l. Routledge &
Kegan Paul.
Trudgill, P. 1986: D ia lects in c o n ta c t, Basil Blackwell.
Tuchman, G. 1978: M a k in g n e w s: a s t u d y in th e c o n s tr u c tio n o f re a lity , New
York: Free Press.
van Dijk, T. (ed.) 1985: H a n d b o o k o f d is c o u r s e a n a ly s is , 4 vols. Academic Press.
van Dijk, T. 1988a: N ew s a s d is c o u r s e , Erlbaum.
------. 1988b: N eros a n a ly s is , Erlbaum.
------. 1991: R a cism a n d the p re ss, Routledge
van Leeuwen, T. 1984: Impartial speech: observations on the intonation of
radio newsreaders. A u s tr a lia n Jo u r n a l o f C u ltu ra l S tu d ie s, 2 (1).
------. 1987: Generic strategies in press journalism, A u s tr a lia n R ev iew o f
A p p lie d L in g u istic s, 10 (2): 199-220.
------. 1991: Conjunctive structure in documentary film and television.
C o n tin u u m , 5 (1): 76-114.
------. 1993: Genre and field in critical discourse analysis. D is co u rs e a n d
. S ociety , 4(2): 193-223.
Wernick, A. 1991: P r o m o tio n a l c u ltu r e , Sage Publications.
Williams, G. forthcoming: F ren ch d is c o u r s e a n a ly s is , Routledge.
Williams, R. 1975: T e le v is io n : te c h n o lo g y a n d c u ltu r a l fo r m , Shocken Books.
■----- . 1981: M a r x is m a n d lite r a tu r e , Oxford University Press.
Wodak, R. 1990: Wir sin d a lle U n sc h u ld ig e T ä te r. D is k u r s h is to r is c h e S tu d ien z u m
: N a c h k r ie g s a n tis e m itis m u s , Frankfurt-am-Main: Suhrkamp.