Exploring Silence and Absence in Discourse: Empirical Approaches

Exploring Silence and
STUDIES IN DISCOURSE
POSTDISCIPLINARY
Absence in Discourse
Empirical Approaches
Edited by
Melani Schröter and
Charlotte Taylor
Postdisciplinary Studies in Discourse
Series editor
Johannes Angermuller
University of Warwick
Coventry, UK
Postdisciplinary Studies in Discourse engages in the exchange between
discourse theory and analysis while putting emphasis on the intellectual
challenges in discourse research. Moving beyond disciplinary divisions in
today’s social sciences, the contributions deal with critical issues at the
intersections between language and society.
More information about this series at

http://www.palgrave.com/gp/series/14534
Melani Schröter • Charlotte Taylor
Editors
Exploring Silence and

Absence in Discourse
Empirical Approaches
Editors
Melani Schröter Charlotte Taylor
Modern Languages and European Studies School of English
University of Reading University of Sussex
Reading, UK Falmer, UK
Postdisciplinary Studies in Discourse

ISBN 978-3-319-64579-7 ISBN 978-3-319-64580-3 (eBook)
https://doi.org/10.1007/978-3-319-64580-3
Library of Congress Control Number: 2017962073
© The Editor(s) (if applicable) and The Author(s) 2018

This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether
the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of
illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and trans-
mission or information storage and retrieval, electronic adaptation, computer software, or by similar or
dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this book
are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or
the editors give a warranty, express or implied, with respect to the material contained herein or for any
errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional
claims in published maps and institutional affiliations.
Cover illustration: © fanjianhua / Getty Images
Printed on acid-free paper
This Palgrave Macmillan imprint is published by Springer Nature

The registered company is Springer International Publishing AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Contents
1 Introduction 1
Melani Schröter and Charlotte Taylor
Part I Comparison as Means to Identify Silence and Absence 23
2 Not for Twitter: Migration as a Silenced Topic in the 2015

Spanish General Election 25
Manuel Alcántara-Plá and Ana Ruiz-Sánchez
3 Absence in Visual Narratives: The Story of Iran and

Pakistan across Time 65
Sameera Durrani
4 Intimations of ‘Spring’? What Got Said and What Didn’t

Get Said about the Start of the Middle Eastern/North
African Uprisings: A Corpus-assisted Discourse Study
of a Historical Event 95
Alan Partington
v
vi Contents
5 Cross-media Studies as a Method to Uncover Patterns of

Silence and Linguistic Discrimination of Sexual Minorities
in Ugandan Print Media 125
Cecilia Strand
6 Critically Illuminating Relevant Absences in Public

Sphere Arguments via Digital Mining of Their
Weblinks: A Software-based Pedagogy 159
Kieran O’Halloran
7 Silence and Absence in Chinese Smog Discourses 191

Jiayi Wang and Dániel Z. Kádár
Part II Exploring Means that Produce Silence and Absence 213
8 Theoretical and Methodological Challenges in

Identifying Meaningful Absences in Discourse 215
Patricia von Münchow
9 What’s Not in a Frame? Analysis of Media

Representations of the Environmental Refugee 241
Nina Venkataraman
10 A Discourse Analysis of Absence in Nigerian News Media 281

Taiwo Oluwaseun Ehineni
Part III Analysing Surface Indicators of Silence and Absence 303
11 What the f#@$!: Policing and Performing the

Unmentionable in the News 305
Crispin Thurlow and Jamie Moshin
Contents
vii
12 The Use of No Comment by Suspects in Police Interviews 329

Joanna Garbutt
13 Conspicuous by Presence: The Empty Signifier

‘Interdisciplinarity’ and the Representation of Absence 359
Dorte Madsen
I ndex 391
Notes on Contributors
Manuel Alcántara-Plá is an Associate Professor in the Department of Linguistics

and Modern Languages at the Universidad Autónoma de Madrid, Spain. He is
interested in Corpus Linguistics and Digital Communication. More specifically,
his current work examines the linguistic characteristics of the New Media using
Corpus-Assisted Discourse Studies. He is the PI of the project ‘Framing and
Articulation Strategies in the Political Discourse on Twitter’ (2015–2017), and
co-editor-in-chief of the international journal, CHIMERA: Romance Corpora
and Linguistic Studies.
Sameera Durrani has a PhD in Media, Film and Theatre from UNSW,
Australia, and an M.Phil in Communication Research, from the University of
Punjab, Pakistan. She has published previously on the coverage given to Pakistan
in the international press before and post 9/11, and on the cross-cultural com-
parison of film narratives in Iran and China, with reference to gender roles. Her
research interests include visual communication, semiotics, and political
communication.
Taiwo Oluwaseun Ehineni currently teaches as an Associate Instructor at
Indiana University, USA. He has degrees in Linguistics from Indiana University
and in English from the University of Ibadan in Nigeria. He was a former
Fulbright Scholar. His research areas include pragmatics and (critical) discourse
analysis, syntax and morpho-phonology. He has presented related papers in con-
ferences in the United States, Slovakia, and Spain. He previously published a
paper on the linguistic and rhetorical features of Nigerian news media where he
ix
x Notes on Contributors
also discussed linguistic devices used in Nigerian media discourse to project to

or conceal information from the audience.
Joanna Garbutt completed her PhD in Applied Linguistics at Birkbeck,
University of London. Her research was concerned with the use of discourse
markers in police suspect interviews, constructing a detailed analysis of the pro-
cess by which officers and suspects create evidential accounts for the legal pro-
cess. Her research interests include the account creation process of police
interviews generally with regards to the fulfilment of institutional objectives and
the management of interpersonal interaction between participants.
Dániel Z. Kádár is Professor of English Language and Linguistics and Director
of the Centre for Intercultural Politeness Research at the University of
Huddersfield, UK. Dániel has published 17 monographs and edited volumes,
some with Cambridge University Press and Palgrave Macmillan. He has also
published many papers in peer-reviewed journals such as Journal of Pragmatics,
Multilingua and Journal of Politeness Research. His recent works include
Understanding Politeness (with Michael Haugh, Cambridge University Press
2013) and Relational Rituals and Communication (Palgrave Macmillan 2013).
His main research interests include metapragmatics, linguistic politeness and
impoliteness, rituals and intercultural communication.
Dorte Madsen is Associate Professor in the Department of Management,
Society and Communication at Copenhagen Business School, Denmark. She
has a MA in International Business Communication, and a PhD in translation
and specialist communication. She lectures in philosophy of the social sciences,
information science and communication. Her current research interests focus
on interdisciplinary communication in research groups, and discourses on inter-
disciplinarity, including linguistic and non-linguistic articulations of epistemic
authority in academic discourse.
Jamie Moshin is a Lecturer in Communication Studies and Liberal Arts at
the University of Michigan, USA. He is a Critical Rhetorician, whose scholarly
interests lie primarily at the intersection of identity and discourse. In particu-
lar, he focuses on American Jewish identity, and what its constructions and
representations tell us about liminal Whiteness, appropriation and authentic-
ity, and unusual identity performances. His work has appeared in many schol-
arly venues, and has addressed such issues as ‘new’ performances of Jewishness
that resist Whiteness, the repression of ‘taboo’ language in the media, the rep-
resentation of marginalized masculine identities, and the intersection of tragedy
and humor.
xi
Kieran O’Halloran is a Reader in Applied Linguistics at King’s College,

University of London. He researches and teaches in the following areas: posthu-
manist approaches to critical discourse studies and critical thinking; stylistics,
poetry and film; using digital text analysis tools and corpus linguistics in relation
to these foci. Recent publications include Digital Literary Studies: Corpus
Approaches to Poetry, Prose, and Drama (with David Hoover and Jonathan
Culpeper, Routledge 2014) and Posthumanism and Deconstructing Arguments:
Corpora and Digitally-Driven Critical Analysis (Routledge 2017).
Alan Partington is Associate Professor of English Linguistics at Bologna
University, Italy. His research interests include corpus research methodology,
corpus-assisted discourse studies, lexical grammar, modern diachronic language
studies, pragmatics, evaluation and evaluative prosody, and irony studies. He is
the author of Patterns and Meanings (Benjamins 1998), The Linguistics of
Political Argument (Routledge 2002), Persuasion in Politics (with Charlotte
Taylor, LED 2010), The Linguistics of Laughter: A Corpus-Assisted Study of
Laughter-talk (Routledge 2006), Patterns and Meanings in Discourse (with
Alison Duguid and Charlotte Taylor, Benjamins 2013). His paper, ‘Mind the
gaps’, published in the International Journal of Corpus Linguistics is the first
article-length study to investigate methods of locating, tracking and evaluating
absences from corpora.
Ana Ruiz-Sánchez is Associate Professor in the Department of Linguistics
and Modern Languages at the Universidad Autónoma de Madrid, Spain, and
she has a European PhD in German Studies. She is a researcher in the project
‘Framing and Articulation Strategies in the Political Discourse on Twitter’
(2015–2017). Her research interests are the analysis of intercultural discourse
in Europe, and she works as a consultant on Human Rights and Minorities.
She is the co-author of Interkulturelle Literatur in Deutschland (Metzler 2000)
and Bewegte Sprache: Vom ‘Gastarbeiterdeutsch’ zum interkulturellen Schreiben
(Thelem 2014).
Melani Schröter is Associate Professor in German Linguistics at the University
of Reading, UK. Her research interests include political discourse analysis,
silence and absence in discourse and communication, comparative analyses of
European migration discourses and discursive resistance (in particular, subcul-
tural/punk discourse). She has published on these aspects in German as well
as English. based on studies of German as well as British political and/or
media discourse. She is the author of Silence and Concealment in Political
xii Notes on Contributors
Discourse (Benjamins 2013) and has published in international journals and

edited volumes on political discourse analysis, silence and punk.
Cecilia Strand worked as program officer for multilateral and bilateral devel-
opment partners in Lesotho, Namibia and Uganda between the years 2003 and
2011. She defended her PhD thesis entitled, ‘Perilous Silences and
Counterproductive Narratives Pertaining to HIV/AIDS in the Ugandan,
Lesotho and Namibian Press’ in 2011, and has since worked as a senior lecturer
in the Department of Informatics and Media at Uppsala University, Sweden.
Her research interests primarily revolve around minorities’ media representa-
tions, absent voices and silenced narratives in media spaces.
Charlotte Taylor is Senior Lecturer in English Language and Linguistics at the
University of Sussex, UK, and is the editor of CADAAD Journal. Her research
interests include impoliteness implicatures and discourses of migration, and she
has a long-standing interest in methodological issues and the ways in which
these choices affect the research. In this vein, she has published on intra-
researcher variation (CADAAD Journal) and the importance of looking at simi-
larity as well as absence (Corpora). Charlotte is a co-author of Patterns and
Meaning in Discourse: Theory and Practice in Corpus-Assisted Discourse Studies
(with Alan Partington and Alison Duguid, Benjamins 2013) and is the author
of Mock Politeness in English and Italian: A Corpus-Assisted Metalanguage Analysis
(Benjamins 2016). She has published in International Journal of Corpus
Linguistics, Journal of Pragmatics, Intercultural Pragmatics, Journal of Politeness
Research and Gender & Language.
Crispin Thurlow is Professor of Language and Communication in the
Department of English at the University of Bern, Switzerland. More informa-
tion about his research and teaching can be found at www.crispinthurlow.net.
Nina Venkataraman is a doctoral student in the Language and Literature
Department at the National University of Singapore. The focus of her doc-
toral study is the representations of environmental refugees in selected elite
newspapers. As an ecolinguist, she proposes that victims of climate change
need a voice too.
Patricia von Münchow is Professor of Linguistics and Director of the Master’s
programme in Linguistics at Paris Descartes University, Sorbonne Paris Cité,
France. She has specialized in Contrastive Discourse Analysis and is the author
of Les journaux télévisés en France et en Allemagne. Plaisir de voir ou devoir de
s’informer (Presses Sorbonne Nouvelle, 2004, third edition 2009) and Lorsque
xiii
l’enfant paraît… Le discours des guides parentaux en France et en Allemagne (Presses

universitaires du Mirail, 2011). In her recent research she concentrates on
absences and silence in discourse.
Jiayi Wang is a Subject Leader in Chinese and an Acting Course Leader in
Modern Languages at the University of Central Lancashire, UK. Her current
research interests include environmental discourse, corpus-assisted discourse
analysis, and im/politeness, all of which are facets of her overarching interest in
intercultural communication. Jiayi earned a PhD in Applied Linguistics from
the University of Warwick, UK, where her research focused on professional
intercultural communication. Prior to her PhD, she was an international project
manager at the Chinese Ministry of Justice, and she also worked as a conference
interpreter/translator for a wide range of organizations, such as the Supreme
Court, Deutsche Bank, and Fortune magazine. She has published research arti-
cles on comparative law, intercultural pragmatics, and foreign language and
second-language education.
List of Figures
Fig. 3.1 Diachronic patterns for Gaze (Eye Contact): Iranian

Women (1981–2010) 76
Fig. 4.1 The White House press room seating chart 104
Fig. 4.2 How the Libyan administration is referred to by the
podium in the first three months of White House press
briefings in 2011 105
Fig. 4.3 How the Syrian administration is referred to by the
podium in the first three months of White House press
briefings in 2011 105
Fig. 4.4 CNN Libyan government/regime/Gadhafi regime 107
Fig. 4.5 CNN Syrian/Assad government/regime 108
Fig. 4.6 CNN Syria: Government or regime 108
Fig. 4.7 Mentions of the items Middle East* and North Africa*
quarter-yearly in 2010 in the Guardian and the Telegraph111
Fig. 5.1 SMUG tweets distribution over time Nov 2015 to
Feb 2016 136
Fig. 5.2 Exclusion and invisibility as enablers of other types of
discursive discrimination 140
Fig. 5.3 Daily Monitor silencing of local sexual minorities during
the 2016 election period 141
Fig. 5.4 New Vision silencing of local sexual minorities during the
2016 election period 142
xv
xvi List of Figures
Fig. 6.1 Major lexical cohesive chains in Monbiot’s argument 169

Fig. 6.2 How the cohesive structure of ‘moral’/‘humanitarian’ and
thus coherence of the argument are disturbed by
relevant absences 180
Fig. 6.3 Showing how the cohesive structure of ‘bomb’ and thus
the coherence of the argument are disturbed by relevant
absences183
Fig. 6.4 Concordance for ‘Muslim’ in Monbiot’s argument 185
Fig. 7.1 Reporting trend in the first half of 2016 197
Fig. 9.1 Overview of the conceptual tools 248
Fig. 13.1 Analytical distinction between a signifier’s form, its function
and absence as an ontological category 362
Fig. 13.2 Signification and the logic of hegemony 364
List of Tables
Table 2.1 Migration in the election manifestos 39

Table 2.2 Tweets with the root migra- by newspaper 44
Table 4.1 The main corpora employed in this study 100
Table 4.2 How the two newspapers referred to the Libyan
administration and its leader throughout 2010 116
Table 6.1 Frequencies for lemmas in Monbiot’s argument with a
threshold of four using a stopword list 167
Table 6.2 The most frequent lexical words in the parliamentary
debate which are absent from Monbiot’s argument 175
Table 6.3 The most frequent two-word expressions in the
parliamentary debate which are absent from Monbiot’s
argument176
Table 7.1 The smog corpus 195
Table 7.2 Newspaper breakdown (1 January 2016–30 June 2016) 196
Table 7.3 Subtopics of causes mentioned in the corpus of 415
news articles published in major Chinese newspapers,
1 January–31 June 2016 (440,266 Chinese words) 202
Table 7.4 Key lexical words 203
Table 8.1 Types of social representations, linguistic marking and
analytical procedures 225
xvii
xviii List of Tables
Table 9.1 Overview of four issue-frames 251

Table 9.2 Sample list of the use of the pronoun ‘we’ in The Times
and The Guardian261
Table 12.1 Interview data 354
Table 13.1 Typologies of interdisciplinary research (Aboelela et al.,
2007, p. 337) 377
1
Introduction
Melani Schröter and Charlotte Taylor
This edited collection aims to fill a gap in the field of discourse studies by
addressing the issue of silence and absence in discourses and by introduc-
ing routes into the empirical analysis of what is absent in discourses.
While (critical) discourse analysis has been interested in the phenomena
of absence (for example, hiding agency through the use of the passive
voice), little attention has been devoted to how we can systematically
identify and analyse absences more broadly. How do we come to notice
absences? How can we argue the existence of absences, what shapes they
take, where and why they occur? What sense can we make of them, how
do they determine what is present? Are they entailed in what is semioti-
cally perceptible to us, or how are absences determined by what is semi-
otically present? Can we analyse them empirically in a way that is
systematic and methodologically sound?
M. Schröter (*)
University of Reading, Reading, UK
C. Taylor (*)
University of Sussex, Falmer, UK
© The Author(s) 2018 1

M. Schröter, C. Taylor (eds.), Exploring Silence and Absence in Discourse,
Postdisciplinary Studies in Discourse, https://doi.org/10.1007/978-3-319-64580-3_1
2 M. Schröter and C. Taylor
What arises from questions like these is, first of all, the need for more
conceptualisation of how and why silence and absence in discourse can
be meaningful and, second, how we can identify and analyse what is
absent in discourse. By assembling contributions that use different meth-
odological approaches to analysing silence and absence, the volume seeks
to promote the empirical study of phenomena of discursive absences and
to incorporate silence and absence as a line of enquiry in discourse
studies.
The contributions in this volume therefore do not so much pursue the
conceptualisation or theorisation of silence and absence, but suggest
approaches to the empirical analysis of absences. Taken together, they
contribute to the aim of this volume to provide an initial toolkit for any-
one who wishes to pursue the study of silence and absence in discourse.
There is still a lack of empiricism when it comes to the study of silence
and absence especially from a (critical) discourse analysis viewpoint.
This lack first of all prevents a better understanding of phenomena of
absence in discourse and communication and, second, it prevents a better
understanding of discourse itself.
Regarding the first point, until now many pragmatic, sociolinguistic or
discourse-oriented studies on silence have either mapped out different
types, meanings and functions of silence on the basis of either constructed
or context-isolated examples (e.g. Bruneau, 1973; Ephratt, 2008;
Jaworski, 1993; Kurzon, 2007; Tannen & Troike, 1985)—contributing
more to conceptualising, classifying and theorising silence—or they aim
to situate the phenomena of silence within the study of language/dis-
course (e.g. Achino-Loeb, 2006; Glenn, 2004), again discussing various
aspects of silence illustrated with examples, rather than letting them
emerge from an analysis of silence in particular discourse contexts. Where
there is a focus on specific discourse contexts, the focus on silence/absence
can be partly lost, and methodologies of tackling absence are not deliber-
ately explored, discussed or explicated. However, the ambiguity and
context-dependency of silence have often been noted (Bergmann, 1982;
Clair, 1998; Jensen, 1973; Sifianou, 1997). It therefore seems all the
more important to build a pool of empirical studies of silence and absence
in specific contexts. Bergmann (1982) argues that within an ethnographic
framework, the “context and placing of stretches of silence need to be
Introduction 3
understood as resources for interpretation for the interactants themselves,

and need to be analysed as such” (145, translated MS). This is in line with
van Dijk’s (2008) socio-cognitive conceptualisation of context, as well as
with Blommaert’s (2005) premise that when analysing language in social
contexts, “the focus should be on what language [and the absence thereof,
MS/CT] means to its users” (14; italics in the original).
Regarding the second point, the focus on discourse in this volume brings
with it a focus on socio-political contexts, on patterns of and resources for
social interaction, on representation (including the notion of foreground-
ing and backgrounding) and on power/hegemony. Important questions
have been raised as to how silence and absence relate to these, but have yet
to be addressed. Regarding contexts that structure discourse, and patterns
and resources available for interaction, Blommaert remarks that
[t]he emphasis on linguistic analysis implies an emphasis on available dis-

course, discourse which is there. There is no way in which we can linguisti-
cally investigate discourses that are absent, even if such analyses would tell
us an enormous amount about the conditions under which discourses are
being produced (by whom? When? For what purpose?) and circulated
(who has access to them and who doesn’t?). It also means that discourse
analysis starts from the moment that there is linguistically encoded dis-
course, bypassing the ways in which society operates on language users and
influences what they can accomplish in language long before they open
their mouths, so to speak. (2005, p. 35)
The question of power and hegemony is closely related to this since

“[t]he road to overt ideological domination rests on a bedrock of silence
running through different layers of suppression that […] begin at selec-
tive perception of significance and end in the consensus that […] is the
necessary condition for the effective wielding of power” (Achino-Loeb,
2006, p. 13f.). The conditions for accessing, producing, receiving and
participating in discourse are not afforded randomly, but interact with
social status, resulting in “differential access to forms, to linguistic/com-
municative resources, resulting in differential capacities to accomplish
certain functions” and in “differential access to contextual spaces, i.e.
spaces of meaning ratification where specific forms conventionally receive
specific functions” (Blommaert, 2005, p. 76; italics in the original).
This is relevant for silence and absence since “[s]ignificance involves

something other than mechanical registering; it involves a selection of
sorts […]. Therefore, at the heart of our meaning construction process
is an act of suppression: hence the need to look at agency in such a pro-
cess” (Achino-Loeb, 2006, p. 38). Once established, “hegemonic dis-
course can be at its most powerful when it does not have to be invoked,
because it is just taken for granted” (Baker, 2006, p. 19). Unless social
contexts change, it can be very difficult to resist the suppression of alter-
native versions of social reality, of the viewpoints of marginalised groups,
of tabooed narratives.
Regarding representation, the discursive construction of reality in “dis-
course which is there”, critical discourse analysis has often considered the
question of “which elements of events or events in a chain of events are
present/absent, prominent/backgrounded” (Fairclough, 2003, p. 139).
Systemic functional linguistics (Halliday, 1994), social actor analysis (van
Leeuwen, 2008) and the notion of conceptual metaphor since Lakoff and
Johnson (1980) all help identify which aspects are foregrounded and
backgrounded aspects in discourses, for example, hiding individual agency
by use of the passive voice, by vague characterisation of social actors or by
metaphorically conceptualising events which involve human agency as
natural catastrophes. However, tools like these are hardly ever employed to
decidedly shed a light on silence, and they are hardly ever drawn together
in order to decidedly carve out what is arguably absent in any given text.
In the following, the Introduction will serve to propose some concep-
tual clarification, not least a differentiation between silence and absence
in so far as they can be regarded as relevant for linguistic and discourse
analysis (Sect. 1.1). This differentiation, however, focuses on the scope
and aims of this volume and does not aim to provide an all-purpose defi-
nition of discursive absence. We will also discuss a variety of manifesta-
tions of silence that are relevant to discourse analysts and which have
been noted in previous literature on the subject (Sect. 1.2) as it helps to
develop our proposal of how to differentiate between silence and absence
and how both can be meaningful in discourse. Following this, empirical
approaches to the study of silence and absence as presented in this vol-
ume will be outlined, with reference to similar approaches in selected
previous studies (Sect. 1.3). The aim of the following is to point out the
Introduction 5
contribution made with this volume to (1) the study of silence from a
discourse analysis angle (Sects. 1.1 and 1.2); and (2) the development of
a methodological toolkit to analyse silence and absence which this vol-
ume aims to inaugurate (Sect. 1.3).
1.1 C
onceptualising Meaningful Silence
and Absence in Discourse
The heading for this sub-section already suggests a delimitation of the
phenomena that we seek to specify in the following, and it is an impor-
tant one to begin with. For the purposes of linguistic and discourse ana-
lytic enquiry, what we are concerned with are signs that carry meaning.
On this basis, we enquire in various ways how such signs are structured
and how meaning is assigned to them. If a sign did not carry any meaning
whatsoever, it would not even only be a meaningless sign—it would not
be a signifier that we are concerned with, it would be uninterpretable. We
do not have to understand signs (e.g. foreign languages or scriptures) to
acknowledge their signifying potential and interpretability—even if we
are unable to ‘read’ their meaning. When we perceive something as a
sign, we take into account that ‘it means’, even if a given sign does not
mean anything to us at a given point in time. Conversely, we will only
perceive absences when there is a potential for them to be significant and,
therefore, meaningful. Where this is not the case, we seem to be dealing
with what Dieckmann describes as follows:
“[...][A] kind of inexpressive, ‘concealed’ silence that does not want to say
anything and for which—because of its paradoxical nature—we do not
even seem to have a name, or not have a name anymore: silence at the
border and as the border of language, to which we cannot get any closer by
reading from reality […]” (Dieckmann, 1992, translated MS).
It seems that only when we can hold non-occurrence of speech against

the possibility of occurring, and only when we can hold something that
gets not said against the possibility of saying it, are we dealing with epis-
temologically salient cases of absence. The contributions in this book are
limited to such cases. Silence and absence are of interest to us in that they
can be interpreted, and this is only possible if they are relatable to an
alternative presence that can be spelled out. “In general, for each possibly
relevant thing, there is a corresponding anti-thing, an absence. In its par-
ticularity, this anti-thing is not at all equivalent to no-thing, although it
may look the same” (Bilmes, 1994, p. 73). Whatever signs we are exposed
to will always be a fraction of what is possible, so there is a plethora of
absences around whatever is given, and we do not usually notice any
absences that can be related to what is phenomenologically manifest to
us. Only when a thinkable alternative occurs to us will we begin to per-
ceive its absence as meaningful. Wherever this is not the case, absences
remain unnoticed because they are meaningless, and meaningless absences
are outside of human perception. We would not notice the existence of
an absence outside of a perceptive framework that renders them mean-
ingful on the basis of an imaginable alternative of presence.
This is a very provisional and much too brief claim for the broad ques-
tion about human perception lurking behind it. However, the point that
we are trying to make is that as linguists and discourse analysts, we only
need to be concerned with meaningful absences and that for absences to
be meaningful, they require an arguable alternative of presence.
To help distinguish between absence and silence, it is useful to refer to
the notion of discourse that is most prevalent in Critical Discourse
Studies, i.e. discourse as ways of speaking that are determined by and
which at the same time reflect social, political, historical and cultural
contexts (cf. Fairclough, Mulderrig, & Wodak, 2011). Following this line
of thinking, which is broadly based on Foucault’s works, such contexts
determine what is thinkable and speakable, and they determine what, out
of the speakable, is considered more or less salient and which ways of say-
ing are more or less socially acceptable at a given time and place. In this
view, choices of individual speakers are predetermined by the discursive
contexts in which they find themselves, which make certain topics, prop-
ositions and perspectives more likely than others. Following this line of
thought, the focus moves from the agency of individual speakers to the
discursive constellations around them, even though one line of enquiry
might be to what extent individual speakers reproduce or resist preformed
patterns especially of hegemonic discourse. It would also be important to
Introduction 7
investigate development and emergences of discourse and to look at pro-

cesses of narrowing down and ordering of discourse, considering the
exclusions involved in this process. When it comes to agency in the estab-
lishment of hegemony, it would be more suitable to assume, rather than
intentionality, constellations of self-interests (cf. Achino-Loeb, 2006,
p. 13) that are pursued ‘naturally’ and not with the conscious aim of pro-
ducing a discourse that is shaped in a certain way.
We consider absence in two ways: first, as an umbrella term to contain
all forms of perceptible and meaningful absences in discourse and com-
munication, including the various forms of silences discussed in the fol-
lowing. Second, and more specifically, absence is what arises from discursive
constellations as sketched above. Studying absence in discourse therefore
requires a framework that allows us to reconstruct, to re-think what is
given (cf. von Münchow in Chap. 8 in this volume) with a view on alter-
natives afforded by, but possibly also beyond the determining contexts
(e.g. when using comparisons to different contexts). Achino-Loeb (2006)
points out the role of selectivity and salience in producing structural, dis-
cursive absence. Choices as to what is considered to merit perception and
communication are usually not made with the deliberate intention to
exclude others, but absences result from a process of choosing, they are the
other side of the coin that results in the presence of the chosen. Most
speakers’ choices are not made with a view to producing a certain order of
discourse, while at the same time contributing to it, which might most
adequately be imagined as an invisible hand process (Keller, 1994).
In contrast to absence, silence can be ascribed to individual speakers
when they make a more conscious and intentional choice about what
(not) to say—when they choose to say nothing, but instead could have
said something. (Please note that ‘saying or speech’ in the following is
meant to include other modes of communication without having to spell
out each possible mode in every instance.) Schröter (2013) discusses at
length how silence becomes meaningful when there is (1) an intention to
be silent, which determines the logical existence of a silence from the
point of production; (2) a disappointed expectation of speech, which
determines the phenomenological existence of a silence from the point of
perception; and (3) that what is not said is relevant to the context or situ-
ation at hand.
Concealment and omission can also be assigned to individual agency.

We regard silence then as a term to also include concealment and omis-
sion because both of these link with intentionality. As explained in
Schröter (2013), concealment can be considered as “silence about X”,
which might go along with talking about Y instead, but the identification
of concealment still relies on intentionality, disappointed expectation to
talk about X and relevance of X, just as with silence above. Omission is a
term that can sometimes be found in the literature on silence, and while
it, like concealment, seems to entail a specific X, it appears to focus on
removal (possibly of something that was present before or elsewhere) (cf.
Jaworski & Galasiński, 2000) and, unlike concealment, less on being
silent by hiding X from the start or putting forward something else
instead. There is a link to secrecy here (cf. Nippert-Eng, 2010; Roberts
2006; Black 2006) as well as other constellations that have been noted in
the existing literature on silence; professional obligations of silence
(Bellebaum, 1992, pp. 81–128; Ulsamer, 2002, pp. 225–236) and rights
to silence (Cotterill, 2005; Kurzon, 1995, 1998; Garbutt, Chap. 12 in
this volume), all of which relate to individual agency, even if in part
afforded by rules and regulations. Such rules and regulations need to be
consciously adhered to, maintained and reinforced by individuals sub-
scribing to them when they become part of the relevant professions. This
angle also helps characterise conventional silence, i.e. silence that is related
to certain situations and assigned a certain meaning; minutes of silence or
silence at a funeral which is meant to signify commemoration or mourn-
ing. Similar to professional secrecy, speakers make conscious decisions to
adhere to silences predetermined by specific cultural conventions of
which they are conscious, and it would be possible, if they bear the social
cost, for them to act out of line.
Of course, we can also investigate discourses (and silences) of individu-
als or how they link in to (absences in) discourse as above, but then we
are moving between the level of intentional choice and contextual
predetermination. The interplay between the two levels is perhaps most
interesting when looking at taboo and self-censorship. Taboos are collec-
tive, arising from discursive predetermination beyond the decisions of
individuals (cf. Zerubavel, 2006). To adhere to taboos, we do not nor-
mally have to actively suppress that which would challenge a taboo
Introduction 9
because we normally adhere to norms of sayability either through self-

censorship or because it does not occur to us to say something that would
cross the boundaries of what is socially acceptable to say. However, indi-
viduals might face a situation in which a conscious decision is required as
to whether or not to break a taboo. Similarly, adherence to professional
obligations of silence involves less of a conscious decision than breaching
such rules and, for instance, becoming a whistle-blower.
The notion of contextual predetermination does not deny that power-
ful individuals can (and intentionally so) shape discourse and thereby
also determine what remains unsaid, but in order to find evidence for
this, we would have to refer back to discourse as above, and moreover,
these would be exceptionally powerful speakers who can steer, but not
provide, a whole body of communication that sustains a discourse which
shapes what can (or cannot) be said for the speakers involved in it.
Censorship (cf. Anthonissen, 2003, 2008; Galasiński, 2003) needs to
be regarded as a form of silencing, but these two terms again point towards
the level of discourse and predetermined ways of speaking versus indi-
vidual choice and agency. Censorship involves powerful actors’ attempts
to control, restrict and suppress speech by others in specific ways.
Silencing can be achieved through censorship, but certain speakers or
groups or points of view can also be silenced as a consequence of their
marginalisation determined by the order of discourse that neither affords
salience to certain points of view, nor resonance for voices from groups
who are not perceived to be proper, or entitled, or participating
speakers.
A major function of silencing is to contain […] opposition by identifying

categories of persons and ideas about which speech and texts will be unac-
ceptable, that is, categories of forbidden speech and ‘forbidden reading’.
This process is complemented by the circulation of acceptable speech and
texts that express some things at the expense of others; it is thus a discursive
displacement. (Thiesmeyer, 2003, p. 9)
What is unsaid, or, in nominalised form, the unsaid (cf. Jalbert, 1994),
is not used specifically as a term here, nor has it been defined in existing
literature on silence. We use it as a synonym to loosely describe any of the
absence phenomena above in order to vary expression, but it does not

designate a specific form or occurrence of absence.
As we aimed to show above, we think that intentionality and individual
agency provide a useful angle for distinguishing and characterising differ-
ent phenomena of absence and silence. We could envisage a scale between
low and high intentionality. We argue that absence in discourse arises
from orders of discourse that are usually brought about without the
intention of producing it in the shape that it takes. Conspiracies of
silence, taboos and silencing (e.g. of marginalised groups and points of
view) are related phenomena; we might speak in ways that contribute to
establishing these phenomena without wanting to establish them, hence
our intention to produce such absences is low, and they can only be pro-
duced by collective, rather than individual agency. Given the discourse
analytical orientation of this volume, most of the contributions deal with
absence. This is particularly true for the contributions that are based on
media discourse.
Self-censorship can result in absences as an unconscious by-product of
conspiracies of silence, but it can also be the result of a more deliberate
decision to remain silent in order not to rock the boat. Similarly, adher-
ence to silence required by professional roles involves intentional choices,
but these are pre-shaped, in a recognised way, by rules and regulations.
Adhering to conventional silences equally requires a degree of intention-
ality, but less so than deliberate silences. Perhaps due to the lesser depen-
dency of these types of silences on discourse contexts, none of the
contributions in this volume deal with them.
More deliberate silences can include concealment, censorship, omis-
sion, evasion, lying and deception or metalinguistic comments such as
announcements to not say anything. They require individual and inten-
tional agency. These may occur in various discourse contexts, but from a
discourse analytical point of view, they are less of interest as occasional,
individual occurrences, but more so when they are related to discourse
contexts or social norms (e.g. avoiding taboo words, see Thurlow and
Moshin’s Chap. 11 in this volume) or genres (e.g. police interviews, see
Garbutt’s Chap. 12 in this volume).
Introduction 11
1.2 Manifestations of Silence and Absence

Absences can be as multi-layered as discourses themselves, in terms of the
levels of language use at which they can manifest and include phenomena
dealt with by different approaches to linguistic description. They can man-
ifest, for example, as vague terms, ellipsis, implication and presupposition.
They can arise from the interplay of highlighting and hiding or fore-
grounding and backgrounding in the framing of topics, use of the passive
voice or in metaphorical conceptualisation. They can arise from the struc-
ture of conversations, e.g. a missing second move in an adjacency pair, or
from deictic expressions with a contextually unclear reference. They can
involve strategies of evasion (e.g. Bull, 2003) and deception (Galasiński,
2000) and such strategies can involve a range of means to achieve the
intended silence or concealment. Silences can also be symbolically repre-
sented, e.g. as *** or through verbalising an intention to remain silent.
It is neither our aim here to establish a full list of possible phenomena
of absence, nor to be prescriptive about what phenomena can constitute
absences. On the one hand, it is open to dispute whether presupposition
and implicature are by default forms of absence, when it would be an
illusion to think that direct and all-explicating speech was possible. On
the other hand, they can be means used to omit and conceal. What is
therefore important in any discourse analyses of absence, is to deliver, by
means of contextualisation, description and interpretation, a convincing
argument as to how and why phenomena such as the above constitute,
produce or indicate absences in the discourses under investigation.
The three previous verbs (constitute, produce, indicate) have been
piled up deliberately, because they indicate different angles under which
phenomena can be seen to relate to absences in discourse. Rather than
trying to provide a list of possible absence phenomena, especially given
that, like most signs, they might have varying functions and meanings
according to context, it seems more useful to find a wider angle.
Looking at a phenomenon that constitutes an absence means establishing
what can be regarded as an absence (of something else) in itself, i.e. this is
about identifying or locating absences by means of analysis. Part I of this
volume, ‘Comparison as Means to Identify Silence and Absence’, contains
contributions that propose comparison as a way to identify absence in dis-

course. Manuel Alcántara-Plá and Ana Ruiz-Sánchez look in Chap. 2 at
missing statements relating to the immigration of refugees in the main par-
ties’ election manifestoes and in candidates’ tweets by comparing these
against the omnipresence of the topic in national newspapers. In Chap. 4,
Alan Partington looks at the absence of the topic of the early Arab Spring
uprisings in British newspaper discourse by comparing their references to
Middle Eastern and North African countries to White House press briefings
and CNN coverage. In Chap. 6, Kieran O’Halloran analyses a newspaper
opinion piece and shows how it builds up a straw man argument against an
aspect of a parliamentary debate that was largely absent, by comparing it to
the latter. Cecilia Strand, in Chap. 5, identifies the voices of sexual minori-
ties as absent from Ugandan mainstream media by contrasting discrimina-
tory media reporting with tweets by the Sexual Minorities Uganda network.
In Chap. 3, Sameera Durrani compares visual representations of Iran and
Pakistan in Time magazine and uses comparison of the representations of
these two countries to identify absences in each of them. Jiayi Wang and
Dániel Kádár deal with smog in Chinese news media in Chap. 7 and show
how there is little smog reporting during the times of greatest pollution, but
an increase when triggered by official political announcements. Bridging
over to Part II of this volume, ‘Exploring Means that Produce Silence and
Absence’, they also show that vagueness and backgrounding of causes and
agency contribute to silences about smog in Chinese media discourse.
Looking at phenomena that produce absences means that certain aspects
of language use can result in absences (e.g. framing, metaphorical concep-
tualisation). Part II of this volume contains contributions that investigate
how the interplay of a variety of means can produce discursive absences.
Patricia von Münchow analyses in Chap. 8 how dominant and obvious
representations in parenting discourse constitute an absence of fathers as
active caretakers of their children. Nina Venkatamaran looks at textual
silences in discourses about environmental refugees in Chap. 9 and shows
how presupposition, implicature, metaphor, nominalisation and transitiv-
ity patterns produce the effect of silencing aspects of climate change and,
as a consequence, environmental refugees. In Chap. 10. Taiwo Oluwaseun
Ehineni analyses headlines of Nigerian newspaper articles about the abduc-
tion of more than 200 girls by the radical Islam sect Boko Haram. He
Introduction 13
illustrates how topicalisation, omission, ellipsis and the use of deixis, rhe-
torical questions, acronyms and numbers contribute to foregrounding a
small number of girls who were released while backgrounding the large
number of girls who were still held captive by Boko Haram.
Looking at phenomena that indicate absences means looking at how
given phenomena relate to either a largely absent signified (e.g. vagueness)
or signifier, often metalinguistically. Part III of this volume, ‘Analysing
Surface Indicators of Silence and Absence’ contains contributions that look
at such surface indicators of silence and absence. In Chap. 13, Dorte Madsen
uses a discourse theoretical approach to look at ‘interdisciplinarity’ as an
empty signifier, arguing that the signified remains largely absent through
lack of specification. In Chap. 11, Crispin Thurlow and Jamie Moshin illus-
trate various ways of omitting and replacing swear words in newspaper dis-
course—and thereby also note how these work to highlight the ostentatiously
absent. Joanna Garbutt looks in Chap. 12 at the use of ‘no comment’ as a
reply that indicates absence of the expected answer in police interviews.
1.3 Methodological Approaches

to Meaningful Silence and Absence
in Discourse
Referring back to Blommaert’s remark quoted above, we agree that “[t]he
emphasis on linguistic analysis implies an emphasis on available dis-
course, discourse which is there” (2005, p. 35). What is given offers itself
to analysis, whereas what is absent often remains unnoticed and opaque.
Zerubavel further asserts the difficulty of studying silence empirically:
As one might expect, what we ignore or avoid socially is often also ignored
or avoided academically, and conspiracies of silence are therefore still a
somewhat undertheorized as well as understudied phenomenon.
Furthermore, they typically consist of nonoccurrences, which, by defini-
tion, are rather difficult to observe. After all, it is much easier to study what
people do discuss than what they do not (not to mention the difficulty of
telling the difference between simply not talking about something and spe-
cifically avoiding it). (Zerubavel, 2006, p. 13)
The previous work on silence is somewhat eclectic. While a large body

of work in Critical Discourse Analysis involves notions and means of
highlighting/hiding, or foregrounding/backgrounding, as stated above, it
is hardly devoted to furthering the study of silence and absence. While
there is a body of work that helps classifying, conceptualising, and theo-
rising absences in discourse, there are only a few studies that analyse such
absences empirically, in discourse contexts, and even fewer with the aim
of providing a methodology for tackling silence and absence in
discourse.
Strictly speaking, we also have to agree with Blommaert that “[t]here
is no way in which we can linguistically investigate discourses that are
absent” (2005, p. 35). However, this should not lead us to erase absence
from any research agenda when we can still use what is present to look at
what constitutes, produces or indicates absence. In particular, through
this volume we find that comparisons help to identify absences, that a
range of means can be employed to produce absences as can be shown in
and through specific discourse contexts, that metadiscourse can indicate
absence, and that analysing metalinguistic references to absence can tell
us something about how they are perceived, interpreted and evaluated
by discourse participants.
1.3.1 Comparison as a Means to Identify Absence
If we accept that meaningful absences require a possible presence against

which the absence can be identified, then comparison seems to be a good
way to locate absences. We can compare chronologically by looking at
what is present or absent at different stages of developments of texts (e.g.
successive drafts) and discourses. Shenhav (2007) looks at a set of differ-
ent texts produced successively by the same (group of ) speakers in differ-
ent situations and for different addressees. He compares confidential
discussions among Israeli government officials, semi-public discussions
about the same matter, e.g. in parliamentary committees in which these
government officials took part, and public political speeches and news
interviews. Shenhav is mostly interested in traces of information that was
intended to remain behind closed doors which can be found in statements
Introduction 15
from the semi-public and public sphere. Following these traces elicits
some interesting indicators of where information hidden from the public
still looms in the background. His study allows an insight into attempts
to control the flow of information in parallel to increasing publicity; his
access to verbatim stenographic records of confidential discussions offers
an insight into the kind and amount of information that was not meant
for the public.
Diachronic comparisons would also provide scope for presence and
absence, with a view on inclusions and exclusions of issues and perspec-
tives in the development of a discourse over time. This is partly involved
in Partington’s analysis in Chap. 4 in this volume.
Contributions in this volume use comparisons between different
media and/or social actors in order to locate absences, mostly via cross-
media studies. Such comparisons help in arguing, for instance, that if an
aspect is salient in public discourse at a certain point in time, then its
absence in parts of it at the same time becomes relevant and interpretable
(Chap. 2 by Alcántara-Plá and Ruiz-Sánchez, Chap. 4 by Partington).
Absences identified this way might be indicative of denying voice to a
social group (Chap. 5 by Strand) or of suppressing an issue that is argu-
ably relevant to society (Chap. 7 by Wang and Kádár). They might also
be indicative of attempts to build up a straw man argument and fight a
phantom enemy (Chap. 6 by O’Halloran). They might also reflect and
influence representations of social reality, e.g. in the portrayal of coun-
tries, cultures and politics (Chap. 3 by Durrani). Especially Durrani’s and
Strand’s contributions show that such studies can at the same time help
to bring some of these absences into presence—voices of sexual minori-
ties in Uganda, when from an outside perspective the repression would
seem total, as well as issues in Iranian and Pakistani politics that get high-
lighted at the expense of others which remain unportrayed in Time maga-
zine, as well. The latter contribution also clearly demonstrates that the
question of presence and absence need not be limited to verbal discourse,
but can also be extended to multimodal discourse analyses.
Most of the contributions that use comparisons also make use of cor-
pora and of corpus analysis tools, which seem to provide a good entry
point into contrasting presence and absence on an empirically reliable
basis (cf. Taylor, 2013; Partington, 2014).
1.3.2 Exploring Means That Produce Absence
As stated above, means that produce absences (such as passive voice, met-
aphorical conceptualisation, framing of topics, reference to social actors)
have played a role in Critical Discourse Analysis, but not often with a
view on silence and absence. Previous literature which includes a certain
focus on silence includes Jalbert (1994), Huckin (2002), and Felton-
Rosulek (2008). Jalbert (1994), taking examples from news reports, but
not looking at a specific discourse context, aims to illuminate “[s]truc-
tures of the ‘Unsaid’” by looking at presuppositions, the non-elidability
principle, opacity and transparency elisions and the use of the passive
voice. The non-elidability principle refers to morally significant factors of
statements and the attempts to hide morally problematic aspects by
under-attributing aspects of cause and consequence. Opacity and trans-
parency elisions refer to the use of contested terms to describe ‘reality’,
e.g. the use of contested and therefore inherently opaque terms, as though
they were transparent and neutral, uncontested, not bound to a specific
perspective. Huckin (2002) distinguishes different forms of textual
silences (speech-act silences, presuppositional silences, discreet silences,
genre-based silences and manipulative silences). After further characteris-
ing manipulative silences, he suggests a methodology for identifying
these that relates to the way in which a topic is framed and how knowl-
edge about an issue is constructed. He identifies, first, from scholarly lit-
erature about the topic, the main categories of topics regarding
homelessness (causes, effects, responses, demographic data) and then
from a corpus of newspaper articles about homelessness in the United
States, subtopics relating to the issue. In this way he shows how the topic
of homelessness was framed in the news media and how in particular
causes for homelessness were attributed to the affected individuals while
omitting other, structural reasons that would have highlighted political
responsibilities. Felton-Rosulek (2008) combines this approach with an
analysis of representations of social actors as well as the syntactic roles
given to actors in the closing arguments of a child sexual abuse case. Her
analysis also involves comparison (of the defence lawyer’s and the prose-
cution’s statements), as she aims to show how, by including and silencing
aspects of the crime as well as of the defendant and victim, two different
versions are constructed of the case at hand.
Introduction 17
Contributions in this volume explore means that can have the effect of
producing absence in similar ways. They are mostly concerned with how
such absences relate to the representation of certain topics or social actors
in news media (see chapters by Venkatamaran, Ehineni, Partington,
Wang and Kádár, O’Halloran, Thurlow and Moshin, Durrani, and
Strand) as well as social media (Chap. 2 by Alcántara-Plá and Ruiz-
Sánchez), parental guidance and history textbooks (Chap. 8 by von
Münchow). Means of producing absence that are discussed are also
related to strategies and how they interact; Ehineni shows in Chap. 10
how topicalisation, focus on detail, and omission can work together in
creating absence; Venkatamaran distinguishes in Chap. 9 between traces
(an absence that can be traced within what is present, e.g. implicature)
and masks (which replace an omitted aspect with a distorted version of
it). Von Münchow in Chap. 8 discusses different degrees of marked-ness
of social representations in relation to how dominant and self-evident
they are within a given speech community, and according to these, sug-
gests different routes for the researcher into the analysis of the silences
associated with them.
1.3.3 A
nalysing Metadiscursive References
as Indicators of Silence and Absence
Verbalised references to silence and absence by discourse participants

might in particular link to an interest in language ideology and in what
language—and its absence—mean to language users, how they note,
make sense of and evaluate silence and absence. It also allows the study of
expectations of speech that are the basis for thematising silences, or
attempts to alter an existing order of discourse by highlighting absence.
Verschueren in his study, What People Say They Do with Words (1985),
devotes a chapter to silence and collates an inventory of reference to
silences. While this is not related to discourse contexts, it is interesting to
see how this study allows conclusions about different types of silence (to
say nothing, to keep secret, to button one’s lip, to fall silent, to refuse to
comment, to withhold). From this metalinguistic material, Verschueren
is able to elicit characteristics of silence such as codes, reasons and motives
for more or less deliberate silence. His chapter goes to show the depth
and breadth of conclusions that can be drawn about perceptions of silence

only by looking at metalinguistic expressions referring to it. Schröter
(2013) investigates metadiscourse about silence in political discourse
contexts, using three case studies that provide a more sustained metadis-
course about instances of silence by politicians. She shows that silence
can get noted, debated and interpreted in public discourse, and that a
variety of meanings can be assigned to it. She also discusses how expecta-
tions of speech in political discourse relate to language ideologies that
value transparency and openness and talk as a means to solve problems.
Contributions in this volume are concerned with how metadiscourse
about silence can be indicative of perceived absences and how a verbalised
refusal to provide requested information triggers subsequent evaluations
of such silences (Chap. 12 by Garbutt). They also show interplays between
absence and presence in a different perspective, relating to the relationship
between signifier and signified. On the one hand, despite attempts at fix-
ing meaning by definition, a signifier can remain empty, i.e. existing and
used words can refer to a signified that remains absent (Chap. 13 by
Madsen). On the other hand, when a taboo signifier is omitted and made
absent and is replaced with various ways of reference to indicate the pres-
ence of the taboo word while maintaining its absence, its presence is actu-
ally highlighted through its avoidance (Chap. 11 by Thurlow and Moshin).
1.4 Conclusion
On the basis of the differentiations made above, in particular between
silence and absence along the lines of intentionality and individual
agency, and between constituting, producing and indicating absences,
this volume aims to show routes into the empirical analysis of silence and
absence in discourse.
The contributions in this volume approach absences in several ways,
involving a broad variety of different discourse contexts and communi-
ties, but three approaches become discernible: Comparing in order to
identify discursive absences, exploring means that produce discursive
absences and analysing metadiscursive reference or surface indicators of
silence and absence in discourses.
Introduction 19
As all these studies illustrate, successful methods of identifying absence

and silence require a rigorous integration of knowledge of the context of
production. Furthermore, the challenges posed by the task can be met by
a variety of existing concepts and methodological approaches such as
framing (Chap. 9 by Venkatameran), conversation analysis (Chap. 12 by
Garbutt), and discourse theory (Chap. 13 by Madsen). They might also
require a multi-methods approach in response, such as the integration of
(critical) discourse studies with corpus linguistics (chapters by Alcántara-
Plá and Ruiz-Sánchez, Partington, O’Halloran, and Wang and Kádár), or
the combination of content analysis with semiotic analysis (Chap. 3 by
Durrani).
We hope that this volume will be useful for anyone who wishes to look
into phenomena of absence in discourse contexts and finds themselves
wondering how to tackle a topic like this within an empirical
framework.
References
Achino-Loeb, M.-L. (Ed.). (2006). Silence. The currency of power. New York;
Oxford: Berghahn.
Anthonissen, C. (2003). Challenging media censoring. In J. Martin & R. Wodak
(Eds.), Re/reading the past. Critical and functional perspectives on time and
value (pp. 91–111). Amsterdam; Philadelphia: John Benjamins.
Anthonissen, C. (2008). The sounds of silence in the media. Censorship and
self-censorship. In R. Wodak & V. Koller (Eds.), Handbook of communication
in the public sphere (pp. 401–428). Berlin; New York: Mouton de Gruyter.
Baker, P. (2006). Using Corpora in discourse analysis. London: Continuum.
Bellebaum, A. (1992). Schweigen und Verschweigen. Bedeutungen und
Erscheinungsvielfalt einer Kommunikationsform. Opladen: Westdeutscher
Verlag.
Bergmann, J. (1982). Schweigephasen im Gespräch—Aspekte ihrer interaktiven
Organisation. In H. G. Soeffner (Ed.), Beiträge zu einer empirischen
Sprachsoziologie (pp. 143–184). Tübingen: Narr.
Bilmes, J. (1994). Constituting silence. Life in the world of total meaning.
Semiotica, 98(1–2), 73–87.
Black, A. (2006). Blacked out. Government secrecy in the information age.
Cambridge: Cambridge University Press.
Blommaert, J. (2005). Discourse: A critical introduction. Cambridge: Cambridge

University Press.
Bruneau, T. J. (1973). Communicative silences. Forms and functions. Journal of
Communication, 23, 17–46.
Bull, P. (2003). The microanalysis of political communication. London: Routledge.
Clair, R. P. (1998). Organizing silence: A world of possibilities. Albany: State
University of New York Press.
Cotterill, J. (2005). ‘You do not have to say anything…’: Instructing the jury on
the defendant’s right to silence in the English criminal justice system.
Multilingua, 24(1–2), 7–24.
Dieckmann, B. (1992). Das Schweigen: Moment des Sprechens und Grenze der
Sprache. In D. Kamper (Ed.), Schweigen. Unterbrechung und Grenze der men-
schlichen Wirklichkeit (pp. 65–74). Berlin: Reimer.
Ephratt, M. (2008). The functions of silence. Journal of Pragmatics, 40, 1909–1938.
Fairclough, N. (2003). Analysing discourse: Textual analysis for social research.
London: Routledge.
Fairclough, N., Mulderrig, J., & Wodak, R. (2011). Critical discourse analysis.
In T. van Dijk (Ed.), Discourse studies. A multidisciplinary introduction. Los
Angeles; London: Sage.
Felton-Rosulek, L. (2008). Manipulative silence and social representation in the
closing arguments of a child sexual abuse case. Text & Talk, 4, 529–550.
Galasiński, D. (2000). The language of deception: A discourse analytical study.
Thousand Oaks, CA: Sage.
Galasiński, D. (2003). Silencing by law. The 1981 Polish ‘performances and
publications control act’. In L. Thiesmeyer (Ed.), Discourse and silencing.
Representation and the language of displacement (pp. 211–232). Amsterdam;
Philadelphia: John Benjamins.
Glenn, C. (2004). Unspoken. A rhetoric of silence. Carbondale: Southern Illinois
University Press.
Halliday, M. A. K. (1994). An introduction to functional grammar. Baltimore,
MD: Edward Arnold.
Huckin, T. (2002). Textual silence and the discourse of homelessness. Discourse
& Society, 13, 347–372.
Jalbert, P. L. (1994). Structures of the ‘unsaid’. Theory, Culture & Society, 11,
127–160.
Jaworski, A. (1993). The power of silence. Social and pragmatic perspectives.
Newbury Park, CA: Sage.
Jaworski, A., & Galasiński, D. (2000). Strategies of silence. Omission and ambi-
guity in The Black Book of Polish Censorship. Semiotica, 131(1–2), 185–200.
Introduction 21
Jensen, V. (1973). Communicative functions of silence. ETC: A Review of

General Semantics, 30, 249–257.
Keller, R. (1994). On language change: The invisible hand in language. London;
New York: Routledge.
Kurzon, D. (1995). The right of silence. A socio-pragmatic model of interpreta-
tion. Journal of Pragmatics, 23, 55–69.
Kurzon, D. (1998). Discourse of silence. Amsterdam; Philadelphia: John Benjamins.
Kurzon, D. (2007). Towards a typology of silence. Journal of Pragmatics, 39,
1673–1688.
Lakoff, G., & Johnson, M. (1980). Metaphors we live by. Chicago; London:
University of Chicago Press.
Nippert-Eng, C. (2010). Islands of privacy. Chicago; London: University of
Chicago Press.
Partington, A. (2014). Mind the gaps: The role of corpus linguistics in research-
ing absences. Journal of Corpus Linguistics, 19, 118–146.
Roberts, A. (2006). Blacked out. Government secrecy in the information age.
Schröter, M. (2013). Silence and concealment in political discourse. Amsterdam;
Philadelphia: John Benjamins.
Shenhav, S. R. (2007). Detecting stories. Revealing hidden ‘voices’ in public
political discourse. Journal of Language and Politics, 6(2), 177–200.
Sifianou, M. (1997). Silence and politeness. In A. Jaworski (Ed.), Silence.
Interdisciplinary perspectives (pp. 63–84). Berlin; New York: Mouton de Gruyter.
Tannen, D., & Troike, M. S. (Eds.). (1985). Perspectives on silence. Norwood,
NJ: Ablex.
Taylor, C. (2013). Searching for similarity using corpus-assisted discourse stud-
ies. Corpora, 8(1), 81–113.
Thiesmeyer, L. (Ed.). (2003). Discourse and silencing. Representation and the lan-
guage of displacement. Amsterdam; Philadelphia: John Benjamins.
Ulsamer, F. (2002). Linguistik des Schweigens. Eine Kulturgeschichte des kommu-
nikativen Schweigens. Frankfurt am Main: Peter Lang.
van Dijk, T. A. (2008). Discourse and context. A Sociocognitive Approach.
van Leeuwen, T. (2008). Discourse and practice. New tools for critical analysis.
Oxford: Oxford University Press.
Verschueren, J. (1985). What people say they do with words. Norwood, NJ: Ablex.
Zerubavel, E. (2006). The Elephant in the room: Silence and denial in everyday
life. Oxford: Oxford University Press.
Part I
Comparison as Means to Identify
Silence and Absence
2
Not for Twitter: Migration as a Silenced
Topic in the 2015 Spanish General
Election
Manuel Alcántara-Plá and Ana Ruiz-Sánchez
2.1 Silencing in the Digital Debate

2.1.1 Silenced Topics
This work is part of a project focused on the political discourse in the

Twitter campaign of the 2015 Spanish General Election. The general
election of 20 December 2015, in Spain was marked by the emergence of
new parties claiming new ways of understanding politics. For these par-
ties the use of social media is a key element of democratic regeneration
(Fuchs, 2014; Mancera & Pano, 2013). With respect to Twitter, it was
used as a major tool for political communication by all the parties, even
the most traditional ones.
Our corpus comprises the messages sent by the five main parties and
candidates, a total of 16,305 texts. Twitter is a very popular microblog-
ging platform for publishing short texts of up to 140 characters. It is a
free service where users can broadcast messages (called tweets) and read
M. Alcántara-Plá • A. Ruiz-Sánchez (*)

Universidad Autónoma de Madrid, Madrid, Spain

26 M. Alcántara-Plá and A. Ruiz-Sánchez
those published by the users whom they choose to follow. At the time of
writing, Twitter was the fourth-most-visited website in Spain (http://
www.alexa.com/siteinfo/twitter.com), just below Google, YouTube, and
Facebook. Tweets have a specific structure and allow much more than
text: URLs, references to users’ names (prefixed with @), hashtags or key-
words (prefixed with #), and embedded media (videos, pictures, sound).
In this chapter, we focus on how migration to Spain is (under-)repre-
sented in the Twitter accounts of both the political parties and the candi-
dates. Our initial aim was to investigate whether politicians communicate
in a different way when they are using Twitter. We chose this particular
social network hoping that its quite specific characteristics would help us
to find innovative strategies. In this preliminary stage, we used frequen-
cies in order to choose the most relevant issues, but soon it was very clear
to us that some key topics were missing. Hot issues of that period did not
show up within the most frequent words. We found it particularly sur-
prising that refugees were not a frequent subject. It was December 2015
and the news all around the world was focusing on the Syrian War and
on the migration phenomenon it was causing. Thousands of Syrian refu-
gees were drowning in the Mediterranean Sea trying to reach Europe
illegally by boat. Migration across the Mediterranean has always been an
important topic for Spain. The Strait of Gibraltar, which separates Spain
from Morocco, is only 7.7 nautical miles at the strait’s narrowest point.
This is a natural route for migration from Africa.
We decided to look into the corpus for other issues which were also very
relevant to the Spanish political context and were controversial for the
political parties: feminism, sexuality, religion, racism, and linguistic minor-
ities. The results showed a clear pattern where topics conspicuous in the
press, in everyday discussions, and even in the election manifestos, were
missing from our Twitter corpus. This situation compelled us to foreground
silences in our research, and to try to answer the questions of why those
topics were silenced in the digital discourse and how this was carried out.
A crucial point in any research about silence is to prove that the silenced
topic should be there. In other words, it has to be shown that there is an
absence since all sentences and discourses lack some information. It
would not be possible to make explicit all the details involved in any
event or situation (Alcántara-Plá, 2014). However, missing information
Not for Twitter: Migration as a Silenced Topic in the 2015… 27
should not be crucial or relevant to the intended communication. When

we say that a topic has been silenced, we are implying two different
things. On the one hand, it means that the missing topic should have
been made explicit. In order to do this, we have to provide convincing
proof of its relevance. On the other hand, we imply that there is an inten-
tion in the silencing of the topic. Silence can be a communicative act
when there is a reasonable expectation of something because it is relevant,
but it is not mentioned (Schröter, 2013). In these cases, silence can be
interpreted as concealment (Jaworski, 1993): the topic is present in the
communicative context, but not in the discourse.
Political discourses are “primarily seen as a form of political action, and
as pan of the political process” (van Dijk, 1997, p. 20). Fairclough (2010,
p. 241) uses the concept of “depoliticisation” to refer to “the exclusion of
issues and/or of people from processes of political deliberation and deci-
sion—placing them outside politics”. In order to show the relevance of
migration in 2015 and the depoliticisation of migrants and of their issues
through the digital debate, we will review the national and international
political context of the electoral campaign, the presence of migration in
the Spanish newspapers in 2015, and the proposals published by the par-
ties in their electoral manifestos. We take these corpora for comparison
with what we (or do not) find in Twitter.
The approach we follow in this project is based on both articulation
(Howarth, 2005) and frame semantics (Fillmore, 1982; Huckin, 2002;
Langacker, 1991; Lakoff, 2004). In these theories special relevance is
given to lexical selection and framing strategies. We follow a quantitative/
qualitative methodology (Partington, Duguid, & Taylor, 2013) where we
take as reference both statistics of words related to the topic and the
frames they convey. Following a Corpus Assisted Discourse Studies
(CADS) approach (Baker, 2006; Partington et al., 2013), our work dif-
fers from both Corpus Linguistics (CL) and Discourse Analysis (DA).
We obtain the most basic information using CL methods (frequencies
and concordances), but we use it as a basis for further analysis. Regarding
DA, the most significant difference is the number of texts we take into
account. We try to work with a corpus as large and complete as possible.
In this case, the main corpus is made up of all the tweets published dur-
ing the official campaign for the election.
Both statistics and qualitative analysis help us to describe the way

migration is misrepresented and silenced, and the strategies used in social
media to leave social issues out of the agenda (see Sect. 2.4).
2.1.2 Silence and Power
This chapter follows a critical discourse analysis perspective, where “dis-

course is socially constitutive as well as socially conditioned” (Fairclough
& Wodak, 1997, p. 258). Focusing on the relationship between language
and power (Fairclough, 1989; Wodak, 1996) in digital discourse, we see
silence as a strategy for using covert power, ignoring existing problems
and persons (Lukes, 2005). Brummett (1980, p. 289) defines political
“strategic silence” as “the refusal of a public figure to communicate ver-
bally when that refusal (1) violates expectations, (2) draws public attribu-
tion of fairly predictable meanings, and (3) seems intentional and directed
at an audience”. The unsaid can give us clues about how power relations
are being maintained or challenged in the same way that the presence of
keywords does. Silencing certain issues “often is a major political tool for
control and imposing the status quo” (Jaworski, 1993, p. 110).
What we have in our corpus is not the silence of those who cannot
speak because they are powerless (Thiesmeyer, 2003), but the silence of
those who should be speaking for the powerless, but choose to remain
silent instead. We approach the unsaid as an intentional communicative
act. Silencing means that one discourse disables another. In this regard, it
is important to point out that communication in Twitter is not horizon-
tal. As pointed out by Castells (2011, p. 43)
Network Gatekeeping Theory has investigated the various processes by

which nodes are included or excluded in the network, showing the key role
of the network’s gatekeeping capacity to be the enforcement of the collec-
tive power of some networks over others, or of a given network over dis-
connected social units.
The accounts of political parties and their candidates are privileged

nodes in Twitter. All of them have thousands of followers and what they
publish is taken into consideration not only by them, but also by journal-
ists and the mass media. It means that it is possible for them to set topics
for discussion not only in the social network, but also on a much broader
scale. We agree with van Dijk (1997, p. 44) that “much of political power
may safely be operationalized in terms of the means and patterns of access
and control of politicians, parties or political movements over public dis-
course”. Today much of the political agenda is set on Twitter (Conway,
Kenski, & Wang, 2015).
2.1.3 Digital Silence
Digital silence is particularly relevant in current politics, framed in a soci-

ety with what Hine has called the E3 Internet: “Embedded, Embodied,
Everyday Internet”, and where “people turn to the Internet not as an
escape from everyday existence, but in order to inform and enrich their
understanding of events happening in their lives” (Hine, 2015, p. 44).
The popular belief that “if it is not on Internet, it does not exist”, fre-
quently used for marketing reasons, might have a dark side in politics.
Digital media have their own strategies for censorship. Silence is one of
the most powerful means of censorship in the digital era. Similarly, as with
tabooed ideas, silence censures topics with no need for laws or interdicts. To
take an example, social networking services have their own policies regard-
ing what can be published, but they are rarely explicitly stated. Their most
powerful means of policing the content on their network are their users and
their habits. It does not mean that the platform is not controlling us. We
know that if one user flags our content as inappropriate, the network author-
ities may investigate it, issue a warning or even disable our account. The
social network Instagram replied to a controversy about censorship of breast-
feeding pictures that they “try hard to find a good balance between allowing
people to express themselves creatively and having policies in place to pro-
vide a comfortable experience for our global community” (Brown, 2014).
What is not admissible is not pointed out because it does not even have a
presence in the social n
etwork. The alleged “global community” is in charge
of deciding what is within the limits and therefore deserves to be talked
about. The rest (what might not be comfortable for the majority) is silenced.
These implicit rules of what is permissible have strong social and polit-
ical consequences. As Fuchs has pointed out, we should not confuse
eParticipation with eDomination (Fuchs, 2008), but the difference is not

always that obvious. There was a time when the Internet belonged to a
different––virtual––world. That is no longer the case. “We do not watch
the Internet as we watch television. We live with the Internet and in the
Internet” (Castells, 2009, p. 64). We have moved from virtual reality to
augmented reality, and the digital information shapes our worldview. In
this new context, there are topics that we will never find in some social
networks and we will not be alerted either that they have actually been
suppressed. They just remain silenced and, as such, not discussed.
2.2 The Corpus

In order to empirically prove the relevance of migration in 2015 and its
absence from Twitter, we compiled three corpora: (1) tweets; (2) electoral
manifestos; and (3) articles from digital newspapers. We briefly describe
them in the following sub-sections.
Some linguistic pre-processing was carried out on the three corpora.
They were converted into text files, which were tokenised, tagged with
Part-of-Speech (POS) features using the Freeling software (Padró &
Stanilovsky, 2012), and cleaned. POS tagging was used in this work in
order to disambiguate words. As we used the electoral manifestos and the
digital newspapers only for frequency and co-occurrences analysis of rel-
evant terms, stop lists were used to exclude empty words such as articles
and pronouns. Frequencies and collocations were obtained using the
Antconc software (Anthony, 2014) and NLTK scripts (Bird, Loper, &
Klein, 2009).
2.2.1 Tweets
We compiled a corpus of 16,305 tweets from the Twitter accounts of the

five main political parties and their candidates. These parties are Partido
Popular (right-wing), Ciudadanos (right-wing), Partido Socialista (tradi-
tional social democrat), Podemos (left-wing), and Izquierda Unida
(left-wing). Messages are publicly accessible and toll-free on this digital
platform.
The corpus was collected during the 2015 official campaign, from 4
December to 20 December. It comprises 244,346 tokens (words) and
28,700 types (different forms). Therefore, we have 15 words per tweet on
average.
We used a bot connected to the Twitter API that downloaded every
message when it was published. This means that our corpus also contains
those messages that may have been removed afterwards since any user can
delete their own texts. The accounts published tweets with different fre-
quencies. Izquierda Unida was the party that most frequently published
(5348 tweets), followed by Podemos (4552), Partido Socialista (1905),
Ciudadanos (1659), and Partido Popular (1048). The first two parties
(Podemos and Izquierda Unida) are those with younger voters, which
makes social media the preferred tool for communication. It should be
pointed out, however, that only 8.2% of the voters in this election were
under 25 years old.
Regarding their five candidates, the figures are quite different with
Pedro Sánchez (Partido Socialista, traditional social democrat) the one
with the most tweets (523), closely followed by the 488 tweets of Mariano
Rajoy (Partido Popular, right-wing). Albert Rivera (Ciudadanos, right-
wing) published 318 tweets, Alberto Garzón (Izquierda Unida, left-wing)
published 298 tweets, and Pablo Iglesias (Podemos, left-wing) 166.
2.2.2 Electoral Manifestos
We also compiled a corpus from the electoral manifestos of the five par-
ties for the 2015 campaign. Izquierda Unida is represented in the mani-
festo of the group Unidad Popular, a coalition for the 2015 election
whose main party was Izquierda Unida, but also included other smaller
regional parties such as Chunta Aragonesista or Batzarre.
Electoral manifestos are the main traditional tool for conveying politi-
cal programmes in elections. Since the main purpose of a manifesto is
always to garner votes, they are written to appeal to the general public,
but manifestos are also supposed to reflect the set of principal goals which
are supported by the party and its candidate. For this reason, they are an
essential point of reference if we want to discover what is distinctive in
the discourse in Twitter.
The manifestos in our corpus all have clean and neutral designs. Their
structure is always a list of problems and proposals, divided into chapters
and sections. The document of the Partido Popular has some large photos
of smiling citizens, which is the only multimodality that we find in the
corpus except typography.
The manifestos range in size from the 103 pages of Unidad Popular to
the 338 pages of Ciudadanos, the 332 pages of Podemos, the 274 pages
of Partido Socialista, to the 214 pages of Partido Popular.
All the manifestos were published in PDF files which made it easy for
us to convert them to text files. Once the stop words were removed, the
five documents add up to 205,723 tokens and 18,309 types.
2.2.3 Newspapers
We compiled a corpus of over a million and a half words taken from the
digital versions of the main Spanish newspapers, El País, El Mundo, La
Vanguardia and La Voz de Galicia in 2015, which were the four news-
papers most widely read in Spain in 2015 (EGM, 2015). The corpus
consists of news automatically chosen using the BootCat software
(Baroni & Bernardini, 2004). In order to retrieve relevant documents,
at least three of the following keywords had to be present in a text to be
included: inmigrantes (immigrants), frontera (border), Mediterráneo
(Mediterranean), humanitario (humanitarian), guerra (war), refugiados
(refugees), sirios (Syrians), terrorismo (terrorism), and Siria (Syria). We
think the political context (see below) justifies why any news contain-
ing at least three of these words was considered relevant to our study.
Though some of these concepts are not necessarily connected to migra-
tion, they were published together in the same news in Spain in 2015.
To take an example, the Spanish Minister of Home Affairs declared in
September 2015 that the presence of terrorists among refugees was
plausible. It was his excuse for the existing delay in the processing of
asylum requests, but it shows how clearly terror and refugees were
related, even in the official discourse. The amount of news we found
with these combinations shows how frequently they were related
together in the newspapers.
Although beyond the scope of this chapter, we should point out that
the importance of the Syrian crisis in the news overshadowed other issues
typically related to migration in Spain, such as the relationship between
Spain and Morocco, and the migration from American countries. These
topics were set aside by the urgency of the Syrian tragedy.
The number of articles in our corpus from each newspaper varies from
237 (La Voz de Galicia) to 581 (El País) depending on the availability of
relevant texts. For some of the frequency analysis that we will describe
below, we used a stop list in order to remove all words that did not pro-
vide information (articles, conjunction, empty verbs). After applying the
stop list, the corpus has 880,593 tokens and 47,542 types.
2.3 Immigration as an Expected Topic

As we have explained, we decided to study the discourse about immigra-
tion because we felt that it was missing from our corpus. In order to test
this intuition, we investigated the topic in three different contexts. First
of all, we will give a brief description of what migration means for Spain
as a social and historical phenomenon. Second, we will prove that it was
a hot topic for the newspapers in 2015. Finally, we want to see whether
the political parties were aware of its importance and wanted to give it
prominence. In order to do so, we will summarise the sections focused on
migration of each party’s electoral manifesto.
This three-pronged approach will help us determine whether the topic
was expected in the discussion in Twitter.
2.3.1 Political Context
A press release of the Instituto Nacional de Estadística, the official source

of population statistics in Spain, clearly shows how important migration
was to the country during 2015. On 1 January 2016, Spain had a
population of 46,438,442. In 2016, 291,387 persons migrated to Spain
while 253,060 emigrated from Spain to other countries. Immigration
had grown 12.5% with respect to the previous year of 2015.
Almost 4.5 million people living in Spain in 2016 were not born in

that country. Romania and Morocco are the most frequent countries of
origin, with around 700,000 people coming from each of these countries.
Other frequent nationalities are the United Kingdom (297,000), Italy
(192,053), China (171,508), Ecuador (158,967), Germany (142,316),
Colombia (135,954), Bulgaria (130,506), Portugal (102,318), and
France (101,336).
Regarding emigration, this has been an important phenomenon in
Spain at least since the Spanish Civil War (1936–39). This trend only
changed between 1997 and 2011. After that, the economic crisis and a
high rate of unemployment once more encouraged the young population
to look for a better life in other countries.
Migration was an especially important issue in 2015 because of Syria. As
we will see below, the mass media reported daily on the Syrian refugee crisis
during the period of time analysed in our study. It is the largest migration
process in Europe since the Second World War. The International
Organization for Migration (IOM) estimated that more than 1,011,700
migrants arrived in Europe by sea in 2015, and almost 34,900 by land.
Spain had the third largest migrant quota of the EU for refugees in 2015,
only behind Germany and France. News was often tragic and shocking. To
take an example, in April alone more than 1200 people drowned in the
Mediterranean Sea trying to reach the European coastline.
For Spaniards, this crisis has clear resonances of what had happened
here with its own civil war. During the Spanish Civil War (1936–39),
more than 35,000 children were evacuated to other countries such as
Mexico, the UK, and the USSR. After the victory of the nationalists,
thousands of leftist militants fled to refugee camps in France.
2.3.2 Migration in the News
As we said in the corpus description, it was built only with articles that
included at least three of the following words: inmigración (immigration),
frontera (border), Mediterráneo (Mediterranean), humanitario (humanitar-
ian), refugiados (refugees), terrorismo (terrorism), guerra (war), Siria (Syria),
and sirios (Syrians). We did not want to carry out only a frequency analysis,
but to try to show how these concepts are most commonly presented. In
order to define the frames conveyed in the news, we looked for the most
frequent collocates for every keyword, using mutual information (MI) as
collocation measure and a window span of 10 words.
When we look at collocates with words from the stem migra-, we see
that migration issues appear frequently in the corpus, not only as a refer-
ence to inmigrantes (immigrants), but also migrantes (migrants), migración
(migration), and migratorio (migratory). The most frequent word they are
framed with is refugiado/s (refugee/s), followed by (in order of impor-
tance): crisis, Europa (Europe), Mediterráneo (Mediterranean), asilo (asy-
lum), organización (organisation), frontera (border), personas (persons), flujo
(flow), drama (tragedy), presión (pressure), número (number), costas (coasts),
cientos (hundreds), comisario (commissioner), entrada (entrance), mar (sea),
and barco (boat). Irregular, ilegal, masivo (massive), and rescatado (rescued)
are adjectives that we also find in the texts with high frequencies. Italy,
Greece, Libya, Turkey, Germany, and Hungary are the countries most
mentioned in these news. Regarding the verbs, llegar (to arrive), compartir
(to share), and proceder (to come from) stand out.
This vocabulary clearly shows two different frames for migration. First of
all, we find a geographical frame with countries, borders, flow of persons,
and movement verbs (accomplished by means of the Mediterranean Sea by
boat). This is the frame we also find as the definition for “migration” in the
Oxford Dictionary: “Movement of people to a new area or country in order
to find work or better living conditions”. Interestingly, goals and reasons
for the migration are not within the most frequent words in our corpus.
“Crisis” is the only vague reference to what was happening in Syria. It
shows that the chosen perspective is probably not that of the migrants.
A second frame we find for migration is one of official control. The texts
focus on illegal and irregular persons who cross borders without permission.
This frame includes many references to numbers: it is not only newsworthy
that there are “illegal migrants, but also that there are too many. The massive
arrival of hundreds of persons intensifies this frame’s negativity. The semantic
preference for quantification when referring to refugees and migration had
already been noted in the literature” (Baker et al., 2008; Taylor, 2014).
Europe and the EU commissioner represent the official power that
should control the entrance through the European borders. Refugee and
asylum are also official terms that set the limits of the migration these
documents are dealing with. It is not general migration, but one acknowl-
edged as a political priority since it has violent causes.
The other keywords we used to collect the corpus give us more infor-
mation about the same frames. We find the root refug- mostly in the
plural form of the noun refugiados (refugees). Refugiados is the most fre-
quent word in our corpus with 7110 occurrences. The use of the singular
totals only 5% and the feminine forms are marginal (around 0.5%).
Specific issues of women are not dealt with. We do not find verbs either
(refugiarse or buscar refugio could have been expected). We do find quite
frequently both the acronym ACNUR (UNHCR) and the complete name
of the United Nations High Commission for Refugees. When the refugees’
nationality is stated, it is mostly sirios (Syrians), though Palestinians,
Iraqis, and Afghans are also present. Again, numbers are important and it
is frequent to find words such as millones (millions), miles (thousands),
cientos (hundreds), masivo (massive), and muchos (many). Most of the
actions do not have the refugees as their agents. The most frequent actions
are reparto (distribution), ayuda (help), and acogida (reception). Verbs with
the refugees as subjects of the sentence are huir (to flee), llegar (to arrive),
entrar (to enter), recibir (to get), esperar (to wait), and aceptar (to agree to).
As we can see, the subject has no choice but to do or suffer what these
verbs mean. Nouns such as campos de refugiados (refugee camps), centros de
internamiento (detention centres), cuotas (quotas), and asilo (asylum)
complete the mentioned frame of official control. As we saw above,
migrants are referred to as persons, refugees are also referred to as children,
acknowledging the high proportion of children among those who are
seeking refuge. In contrast with what we found before for migration,
guerra (war) is a frequent co-occurrence of refugee. We cannot forget that
war is a defining feature of their status. The Oxford Dictionary defines
refugee as “a person who has been forced to leave their country in order to
escape war, persecution, or natural disaster”. Unlike the migration frames,
the cause of fleeing is explicitly stated in these news.
Guerra (war) is precisely another keyword. Syrian and civil are the
two concepts most frequently used with war in the corpus. Other fre-
quent words are años (years), mundial (global), país/es (country/coun-
tries), II Guerra Mundial (Second World War), terrorismo (terrorism),
crímenes (crimes), fin (end), Europa (Europe), France (Francia), and

Estado Islámico (Islamic State). Some of these words may need some
comment. Crimes and terrorism are concepts which are used both to
refer to what was happening in Syria and to what was happening in
Europe. The latter is another keyword and its analysis shows this
dichotomy. As we have already pointed out, José Manuel García-
Margallo, Minister of Home Affairs, declared in September 2015 that
the presence of terrorists among refugees was plausible. Terrorismo
appears with lucha (fight), yihadista (jihadist), internacional (interna-
tional), amenaza (threat), François Hollande, Francia (France), and aten-
tado (terrorist attack). The references to France and its president show
that it is not only about terrorism in Syria, but it is terrorism related to
Syria (and the Syrian refugees) that Europe is suffering.
Frontera/s (border/s) is another keyword in our study. In the corpus it
appears related to the Syrian border most of the time, though mentions
of the Serbian border are also frequent. Turkey, Hungary, Greece,
Macedonia, and Croatia, in this order, are other countries linked to bor-
der issues in the news. The refugees are the concept with the highest co-
occurrence frequency with borders. The rest of frequent words complete
both the geographical and the control frames. For the former, we find
país (country), kilómetros (kilometres), sur (south), Líbano (Lebanon), and
Alemania (Germany). For the latter, we find control, cierre (closing),
FRONTEX, and Schengen. Regarding the verbs, cruzar (to cross), cerrar (to
close), reforzar (to reinforce), and proteger (to protect) show a very strong
frame of control where the crossing of the borders by the refugees has to
be avoided by reinforcing and protecting the entry points. Both the
Schengen Agreement and the FRONTEX Agency are key elements in
migration control in the EU.
We chose humanitario (humanitarian) as a keyword because we wanted
to search for articles which were looking into the topic of migration from a
civil rights perspective. After all, that had been the reason we had decided to
study the silencing of this phenomenon with a CDS approach in the first
place. We searched the corpus for words with the root “human-”. Though
the frequencies we found were much less important than those of the previ-
ous terms, we did find humanitaria (humanitarian) as an adjective for ayuda
humanitaria (humanitarian aid) and crisis humanitaria (humanitarian crisis).
Other co-occurrences for humanitaria are situación (situation), organizacio-

nes (organisations), catástrofe (disaster), agencias (agencies), drama (tragedy),
and emergencia (emergency). We also found other words starting with human-
: derechos humanos (human rights), humanidad (humanity), seres humanos
(human beings), and vidas humanas (human lives). As we see, they are in a
minority and a completely different frame.
Finally, we looked for words with the root siri- (Syri-). Syria is the sec-
ond most frequent word in the corpus. It appears 6560 times (to com-
pare: Syrian appears 4839 times, war 2780, border 2496, immigrants
2477, and Islamic State 2149). Both Syria and Syrian appear with
refugiado/s (refugee/s), guerra (war), frontera (border), Irak (Iraq), (Bashar
al-)Asad, conflicto (conflict), Turquía (Turkey), islámico (islamic), and mil-
lones (millions). The relationship of the war in Syria to the refugee crisis is
clear. Also its consideration in Spain as a migration issue. When newspa-
pers report on refugees, they talk about borders and about how many
millions of these refugees there are.
The importance of the Syrian crisis and its migration phenomenon in
the news is clear. Rather than a silenced topic, what we find is an omni-
present one. More than 7000 repetitions of the word refugiados is an
irrefutable proof. Results show the two proposed frames with great detail
and complexity. In the geographical frame, countries which are neigh-
bours of Syria or the destination of the refugees are clearly identified. The
Mediterranean Sea has a tragic prominence. In the borders and control
frame, refugees are seen as a massive movement of people that needs to be
controlled. They are considered dangerous because they are too many of
them and because they are linked to terrorism.
2.3.3 Migration in the Electoral Manifestos
As we have pointed out, electoral manifestos are the main point of refer-
ence since they are the documents where parties must make explicit what
they consider national problems and which measures they propose to
solve them. Finding (or not finding) migration issues in them can help us
determine if it is a topic not considered at all by the parties or if it is a
topic they find important, but they preferred not to discuss it in the social
networks. With this in mind, we will briefly summarise in Table 2.1,
Table 2.1 Migration in the election manifestos

Political party Mention in manifesto
Ciudadanos Chapter: “Social Policies: Dignity without Exclusion”
(new right- Section: “Towards new migration policies: flexible and
wing party) realistic”
Length: 5 pages (1.48% of the document)
Proposals for immigrants:
• It is acknowledged that border control is not working
properly, and that it has to be improved, particularly in
“sensitive zones” such as the cities of Ceuta and Melilla.
• To give more information in the borders to prevent
“delinquency, terrorism, and human trafficking”.
• Immigration is considered a European Union responsibility,
which should implement better regulations.
• Immigration phenomena do not affect all the regions in
the same way. New policies should be put in place in order
to “integrate” those who already live in Spain.
Proposals for emigrants:
A reform of the consulates, to adapt them to a new reality
with many Spaniards living in other countries.
IU Chapter: “Human Rights: Celebrating Diversity”
(traditional Sections: “Migration Policies” divided into “Policies for
left-wing (Spanish) Emigrants” and “Foreigners in Spain: From
party) Immigrants to Citizens”
• Human rights must be universal.
• To avoid all institutional discourses that criminalise
foreigners.
• To promote employment policies for immigrants.
• To develop a plan against human trafficking.
• To facilitate the process for obtaining an EU residence
permit.
• To develop an education plan for the integration of the
immigrant children.
• To adapt Spanish laws so that immigrants can have the
right to vote.
To guarantee medical coverage and to promote bilateral
agreements with countries of destination; to guarantee a
decent return for those living abroad and to open Spanish
schools abroad; to make the voting process from abroad
easier; to protect Spanish workers and to facilitate the
cooperation with unions; and to help the emigrants in their
integration in the country of destination.
(continued)
Table 2.1 (continued)

Podemos Chapter: “International Democracy”
(new left-wing Sections: “Migration and Asylum” and “Foreign Policy and
party) International Cooperation”
• To guarantee the right to vote for immigrants in Spain.
• To simplify the nationalisation process.
• Creation of a Secretary of State for Migration.
• New processes for legally accepting immigrants and for
making it easier for reunification.
• New law against racism and discrimination.
• Detention centres should be closed.
• To guarantee the right of asylum in the EU.
• To guarantee the human rights in the border with
Morocco.
Universal medical coverage; the creation of a new “Emigration
Office” dedicated to help those Spaniards living abroad; a
new regulation for Spanish language and culture education
abroad (for the second generation); measures to guarantee
the emigrants’ pension; and measures to ensure that
emigrants can come back if they want to.
PP (Partido Chapters: “Committed to Welfare” and “European Policy and
Popular) Outreach”
(traditional Sections: “Integration: Same Rights and Same Duties” and
right-wing “Spaniards Abroad”
party) Length: 3 pages (2.34% of the document)
(continued)

Partido Popular was in charge at the moment and its
document starts with a surprisingly optimistic sentence: “The
most important fact about migration in our country is the
outstanding integration we have achieved.” It points out
that “work, sacrifice, and social mobility” frequently are “the
immigrant’s values”, that the European Union agreements of
2008 are a good basis for future policies, and that
employment is the best way for integration.
• To favour legal immigration, fighting against mafias and
human trafficking.
• Abolition of geographical restrictions for work permits.
• New law for underage persons.
• Faster integration of minors into the school system.
• To make the recognition of individual awards easier.
• “The knowledge of the values of the Spanish Constitution”
and of the Spanish language, history and culture will be
required for immigrants.
• Measures for strengthening control at airports and
consulates to avoid illegal immigration.
• Measures to strengthen controls at the European borders.
To modernise the administration abroad; to encourage voting
in consulates; to facilitate the return of those emigrants
living abroad; to implement a new certification for those
who have come back; to guarantee the social rights
(education, health, culture) abroad; to facilitate the
recognition of individual awards; and to promote the
internationalisation of Spanish business.
(continued)

PSOE Chapter: “Migration Policies” and “Rights of Citizens Abroad”
(traditional Length: 10 pages (3.65% of the document)
centre-left Proposals for immigrants:
party) • The last government reforms and laws should be abolished
in order to give back their rights to the immigrants.
• To write a report about the current situation so that the
European Union can improve its policies.
• To develop a structure in the government dedicated to
migration.
• To recover the cooperation with immigrants’ countries of
origin.
• To abolish the right to vote of foreigners in municipal
elections.
• To create a fund for integration policies.
• To promote full integration of second generations.
• To improve the process for nationalisation.
• To remove the barbed-wire border fences.
• To improve the detention centres.
• To give priority to American and African countries.
• To promote actions in favour of immigrant women.
• To improve processes for family reunification.
• To approve by law new measures against human
trafficking.
• To help as soon as possible refugees both in other countries
of destination and in Spain.
• To improve the offices for asylum collaborating with
UNHCR.
• To coordinate actions with the Spanish regions for a better
integration of refugees.
To approve a new nationality law to guarantee the rights of
those who live abroad; to develop a strategic plan about
their needs for 2016–20; to change the process of voting
from abroad, making it easier; to recover the universal
medical coverage; to make bilateral agreements with
countries of destination; to adapt the eGovernment to the
emigrants’ needs (including the voting process); and to
promote cultural activities abroad for the emigrants.
those sections of the manifestos related to the migration phenomena.

Since it is the topic we are focusing on, proposals for immigration will be
enumerated in more detail.
As we see in Table 2.1, the five parties took into consideration the
migration phenomena when writing their political manifestos, both of
Spaniards abroad and of immigrants into Spain. They dedicated sections
to these topics and included an important number of proposals. We see
also that there is no agreement on the policies. While Partido Popular
seems proud of the situation, the other parties want to change it. There
are also different perspectives of how to do this. While Ciudadanos asks
for “flexible and realistic” migration policies, presupposing that other
candidates may propose unrealistic ideas, Izquierda Unida prefers a more
emotional approach “celebrating diversity”. While Partido Popular
explains that integration has been achieved, Podemos and Izquierda
Unida point out that discrimination must be eradicated. The control
frame is very present. It appears linked to complex issues such as human
trafficking and the right to vote. Some parties (Ciudadanos and Partido
Popular) propose measures for more control, while the others propose
more flexible approaches.
The main picture we get from the reading of these documents is that
migration is a controversial topic, with many variables and different
opinions that should be further debated. The only agreement that we find
is that this is a relevant issue, important enough to have its own sections
in these documents, though the low percentages already show a reduction
in comparison with its presence in the press.
2.4 Migration as a Silenced Topic in Twitter

The Twitter results are particularly striking in light of the importance that
has been given to migration in both the newspapers and the manifestos,
and considering the international context.
In a corpus of 16,305 tweets, we find only 57 messages with the root
migra-, words being emigrantes, emigración, emigrar, emigrados, inmigran-
tes, and inmigración (Table 2.2).
Table 2.2 Tweets with the root migra- by newspaper

Party Tweets with migra- % of all their tweets
Izquierda Unida 22 0.41
Podemos 18 0.39
Partido Socialista 7 0.36
Partido Popular 2 0.19
Ciudadanos 1 0.06
Mariano Rajoy 4 0.82
Alberto Garzón 2 0.67
Pedro Sánchez 1 0.19
Pablo Iglesias – –
Albert Rivera – –
These figures already show two trends: (1) left-wing parties have
included this topic more than right-wing parties; and (2) candidates have
barely discussed the topic in their personal accounts. The four messages
of Mariano Rajoy are the only ones written in English and published
information from the EU institutions. We will see below that the rele-
vance of Izquierda Unida is even greater when considering only immi-
grants in Spain.
2.4.1 Emigration
Some 42 of these 57 tweets are focused on Spanish emigration (0.26% of

all the tweets). This predominance, also present in the manifestos, con-
trast with what we have found in the media.
Migration to other countries is seen by opposition parties as a conse-
quence of wrong political decisions taken by the Partido Popular
Government. Some examples are:
1. [Izquierda Unida] Nos cerraron las puertas, pero las abriremos y los
emigrantes vamos a #VolverConGrazón para construir un #NuevoPaís
[They closed the doors, but we will open them and we migrants are
going to #ComeBackWithGarzón so that we can build a #NewCountry]
2. [Partido Socialista] Rajoy es el maestro del eufemismo. Ha llamado a
la emigración de nuestros jóvenes, movilidad exterior @sanchezcaste-
jon #VotaPSOE
[Rajoy is the master of the euphemism. He calls the migration of our

young people “external mobility” @sanchezcastejon #VotePSOE]
These messages give a very specific image of migration. First of all, they
presuppose that migration was imposed on migrants. It seems as if they
had no choice. This is what Partido Socialista reproaches Partido Popular
for in tweet 2). “External mobility” sounds like an option, but they
understand migration as an obligation. Besides they also presuppose that
migrants want to come back. It is as if they were waiting for an opportu-
nity to live in Spain again. This is a logical outcome if we think that they
left against their own will, which is the idea tweet 2 tries to convey. This
is made explicit in several tweets:
3. [Ciudadanos] Por un País en el que los jovenes no tengan que emigrar

y puedan realizarse en España #JovenesconAlbert #Ciudadanos
[For a country where the young people do not have to emigrate and
can develop in Spain #YoungPeoplewithAlbert #Ciudadanos]
4. [Podemos] Hoy es el Día Internacional del Migrante. Por un país
donde emigrar sea una decisión y no una necesidad, proponemos:
https://t.co/V8IheEgm1r
[Today is the Migration International Day. We want a country where
emigrating should be a decision and not a necessity. We propose:
https://t.co/V8IheEgm1r]
The idea of “wanting them back home” shows that migrants do not
have agency. They were expelled and they will be brought back. This
return is seen as a migrants’ right, but also as a necessity for the country
(which is the agent):
5. [Partido Socialista] “Necesitamos recuperar a los investigadores/as

emigrados”. Artículo de @sanchezcastejon vía @Hipertextual #Ciencia
[“We need to get back the researchers who migrated”. Article of @
sanchezcastejon via @Hipertextual #Science]
6. [Podemos] Es necesario recuperar a toda la gente joven altamente for-
mada que ha tenido que emigrar #elDBT
[We need to get back all the young and highly educated people who
had to migrate #theDbt]
There are four tweets by both Izquierda Unida and Podemos remind-
ing the reader of the historical processes of emigration in a very explicit
gesture aimed at the elderly:
7. [Izquierda Unida] Porque no tuvieron niñez. Porque sufrieron la emi-

gración. Porque lucharon contra el franquismo. #AbuelasConGarzón
[Because they didn’t have a childhood. Because they suffered emigra-
tion. Because they fought Francoism. #GrandmasWithGarzón]
8. [Podemos] Emigraste para darnos un futuro mejor que hoy nos están
robando. Votemos para no tener que irnos. #ConMiAbuPodemos
[You migrated in order to give us a better future that they are stealing
from us today. Let’s vote so that we don’t have to leave
#WithMyGrandmaPodemos]
Izquierda Unida talks about the migrants’ vote in two ways. First, they
assume that emigrants would vote for Izquierda Unida if they were able
to. Second, they complain about a new law that makes voting more dif-
ficult for migrants.
9. [Izquierda Unida] Ya van saliendo votos desde Francia para @

UnidadPopularHU Emigrantes con @agarzon @Unidadpopular
#DerechoAVolver
[Votes are already leaving France supporting @UnidadPopularHU
Migrants with @agarzon @Unidadpopular #RightToComeBack]
10. [Izquierda Unida] El censo electoral de emigrados españoles es de
1,880,026. Sólo podrán votar 115,055 el 20D. Estas elecciones son
un pucherazo.
[The electoral register of Spanish migrants comes to 1,880,026. Only
115,055 will be able to vote on 20D. This is a rigged election]
Podemos makes reference to the same law:
11. [Podemos] Sánchez dice que derogará el voto rogado que dejó sin
voto a tantos jóvenes emigrados. Rajoy: lo aprobó el PSOE. Y así
todo #CaraACaraL6
[Sánchez says that he will abrogate the voting law that left so many
young migrants without the right to vote. Rajoy: it was passed by the
Partido Socialista. And so on #FaceToFaceL6]
As we see in these examples, emigration is used as electoral propaganda

more than as an issue for discussion. The topic is portrayed in a simplified
manner and without reference to the real problems of the emigrants,
though some of them had been stated in the electoral manifestos. Those
documents talked about the rights of the emigrants in their countries of
destination, about the administrative problems they may find abroad,
about the difficulties of maintaining Spanish culture and language, etc.
None of these issues are discussed in the digital campaign. The only real
problem we find in this sub- corpus is the right of emigrants to vote,
which is in fact a shared problem with the political parties that are trying
to garner their votes.
Seven tweets talk about the European Union. One of them was sent by
Izquierda Unida, quoting a member of the party strongly questioning the
EU policies:
12. [Izquierda Unida] .@MarinaAlbiol: “Las políticas criminales de la

UE sobre inmigración son el caldo de cultivo perfecto para la trata”
#NuevoPaísFeminista
[.@MarinaAlbiol: “The criminal policies of the EU on immigration
are the perfect breeding ground for trafficking of women”
#AFeministNewCountry]
The other six tweets about the EU are references of the Partido Popular
to its participation in the EU institutions. All these tweets are similar:
they are written in English, they are linked to media from the EU, and do
not give much information on their own.
13. #EUCO—National briefing: Spain @marianorajoy #migrationEU

#refugeecrisis #UKinEU #Syria https://t.co/nHUZP8tycU [link to a
video in the European Council Newsroom webpage]
14. European Council family photo. More about this meeting: https://t.
co/oHZYs2d0oM. #EUCO #migrationEU #UKinEU https://t.
co/88PyI5LjhL [link to a photo of the European Council’s members]
Specific characteristic of these tweets are that they include many

hashtags and that these keywords are not easily understandable by Spanish
citizens. It is, first of all, because they are written in English. Furthermore,
they appear without enough context. We do not think that “UKinEU”,
“EMU”, or “migrationEU” alone have a clear meaning for most of the
Partido Popular followers on Twitter.
2.4.2 Immigrants
We focus in this chapter on the parties’ and candidates’ apparent lack of

interest in the problems in Spain of migrants coming from other coun-
tries. Comparing it with the media, it is inevitable to think that this topic
is being silenced in Twitter, where there are only eight tweets about immi-
grants (0.05% of all the tweets of the corpus), one of them published
twice. They are:
15. [Izquierda Unida] @agarzon denuncia la exclusión d inmigrantes d

la sanidad y los precios impagables de los posgrados
#ServiciosPúblicosUniversales #elDBT
[@agarzon denounces the migrants’ exclusion from medical care and
the expensive costs of postgraduate studies #UniversalPublicServices
#theDebate]
16. [Izquierda Unida] “Nos expulsan de la sanidad por ser inmigrantes,
de la universidad por ser hijos de la clase trabajadora.” @agarzon
#GarzónEnZaragoza
[“They throw us out from medical care because we are immigrants,
from the university because we are children of the working class” @
agarzon #GarzónInZaragoza]
17.
[Izquierda Unida] #JóvenesconGarzón @ma_bustamante84
“Queremos sanidad de calidad, pública y para todos, incluidos los
inmigrantes, aunque eso no de votos”
[#YoungPeopleWithGarzón @ma_bustamante84 “We want quality

medical care, public and for everyone, including immigrants, even
though it doesn’t give us votes”]
18. [Izquierda Unida] Rajoy habla de inmigración: recordemos a los 16
muertos en Ceuta [enlace a sección del Eldiario.es titulada “Muertes
en la frontera de Ceuta”] #CaraACaraL6 [enlace a imagen]
[Rajoy talks about immigration: let’s remind him of the 16 deaths in
Ceuta [link to a feature in the newspaper Eldiario.es titled “Deaths
in Ceuta’s border”] #FaceToFaceL6 [link to a photo]]
19. [Izquierda Unida] Rajoy habla de inmigración: recordemos a los 16
muertos en Ceuta [enlace a un reportaje en Eldiario.es titulado “Las
muertes de Ceuta”] #CaraACaraL6 [enlace a imagen de las vallas de
Ceuta]
[Rajoy talks about immigration: let’s remind him the 16 deaths in
Ceuta [link to a feature in Eldiario.es titled “The deaths in Ceuta”]
#FaceToFaceL6 [link to a picture of the barbed-wire wall in Ceuta]]
20. [Izquierda Unida] @marianorajoy, ¿hablamos de migrantes? ¿De

concertinas, de las devoluciones en caliente, de los CIE? ¿O de Ceuta?
[link to a feature in Eldiario.es titled “The deaths in Ceuta”]
[@marianorajoy, should we talk about immigrants? About barbed-
wire walls, pushbacks, and detention centres? About Ceuta? [link to
a feature in Eldiario.es entitled “The deaths in Ceuta”]]
21. [Izquierda Unida] Hay que abrir Europa a refugiados, inmigrantes y
que cierre las puertas a Merkel y sus políticas @MarinaAlbiol
#GarzónEnMadrid aplausos
[Europe has to be open for refugees and immigrants, and with the
doors closed for Merkel and her policies @MarinaAlbiol
#GarzónInMadrid applause]
22. [Partido Socialista] Rajoy ha recortado los derechos de las personas emi-
grantes @sanchezcastejon #GanaPedroGanasTú #CARAaCARA2015
[Rajoy has cut the rights of migrant persons @sanchezcastejon
#PedroWinsYouWin #FaceToFace2015]
We see that only the left-wing Izquierda Unida (7 tweets) and the tra-
ditional social democratic Partido Socialista (1 tweet) talk about immi-
gration. Two messages are repeated with minor differences (only the
media they link to). It is also interesting to notice that four messages have
a direct reference to TV debates (“elDBT”, CaraACaraL6”, and
“CARAaCARA2015”), one was also published quoting what was being
said at a debate (“@marianorajoy, ¿hablamos de…”), and the rest quoted
what was being said at different rallies. One of them even represents the
“applause” of the rally audience. Therefore, none of them is a digital dis-
cussion in origin since they are actually echoing communications from
outside the Internet.
Half of these messages deal with more than one topic. The first three
tweets of Izquierda Unida mix migration with other claims that are not
clearly related with it: postgraduate studies fees, access to university for
the working-class students, better medical care for everyone (including
migrants), and strong disagreement with Angela Merkel’s policies. In the
last case, the current Chancellor of Germany symbolises the European
Union, even though she had made a clear stand in favour of the refugees’
integration in Europe. Having more than one issue per tweet, and being
them so unrelated to migration, makes the discussion difficult. It seems
hard to reply to all the references in a coherent way.
Three tweets by Izquierda Unida are reactions against something
Mariano Rajoy had said in a TV show. All of them reproach the strong
policies he had undertaken against those who were trying to reach Spain
from Africa. Barbed-wire walls and pushbacks are two measures that had
been severely criticised by society and the media during the last
government.
The only tweet by the Partido Socialista, also from a TV debate, is a
very general statement. However, it uses a particular way of referring to
emigrants, “personas emigrantes”, emphasising that migrants are people
(hence they should have rights).
As we did with news and electoral manifestos, we looked the Twitter
corpus for the other keywords related to immigration in 2015: frontera
(border), Mediterráneo (Mediterranean), Siria (Syria), humanitario
(humanitarian), refugiados (refugees), guerra (war), and terrorismo (terror-
ism). Since they are semantically related to immigration, these words help
us to understand the presence (or absence) of the topic.
2.4.3 Borders and the Mediterranean
There are only two messages about literal borders in the whole corpus
(0.01%). Both tweets were published by Partido Popular and made refer-
ence to the European Union. None of them includes a proposal or a clear
statement. The first is an invitation to watch a video of a speech at the
European Parliament, and the second is a complaint about the control
the EU wants to have over the Spanish borders. This last example is inter-
esting because, as we will see below, Partido Popular usually refers to the
EU as the institution responsible for international affairs.
23. [Partido Popular] Mi intervención en Parlamento Europeo de hoy

sobre propuesta de guardia de fronteras europea. @EPPGroup @
ppegrupo https://t.co/9EiCGFNz7K
[My speech at the European Parliament of today about a proposal for
a European border guard. @EPPGroup @ppegrupo [link to a video
of the European Parliament]]
24. [Partido Popular] Que haya imposición por parte de la UE para con-
trolar la frontera de España no me parece bien @marianorajoy
[I don’t like the EU trying to impose its control over the Spanish
borders @marianorajoy]
There is only one tweet mentioning the Mediterranean. It was pub-

lished by Partido Socialista showing gratitude to ACNUR (UNHCR) for
being “there”. It includes a link to a video where ACNUR explains the
Syrian crisis.
25. [Partido Socialista] ¡Gracias por estar allí! https://t.co/P4aAYFXnaK

#refugiados #Siria #Mediterráneo
[Thank you for being there! [link to the video] #refugees #Syria
#Mediterranean]
2.4.4 Syria
We find only four other tweets about Syria and the Syrians. The first one
was sent by Izquierda Unida and proposed an agreement with the Syrian
government. Two synonyms, “strategic” and “tactical”, are used to define
the kind of agreement that should be reached. Both adjectives imply that
the agreement is temporary and circumstantial, making clear that there is
disagreement about other issues. Daesh, the common enemy, is the rele-
vant connection here.
26. [Izquierda Unida] “Hay que tener un acuerdo táctico y estratégico

con el gobierno sirio que también combate al Daesh @agarzon
#AndalucíaConGarzón
[We need to reach a strategic and tactical agreement with the Syrian
government, which is also fighting Daesh @agarzon
#AndalusiaWithGarzón]
The second tweet was published by Mariano Rajoy and is a good

example of the above-mentioned strategy regarding the EU: The Spanish
government does not need to make a proposal because the EU is the
institution in charge. It should be pointed out that the hashtag “#BCN…”
means that it is a statement made in Barcelona and therefore in a national
context.
27. [Mariano Rajoy] Resolver la situación de los refugiados sirios es de

humanidad. Debe actuarse desde el origen es el gran reto de la #UE
#BCNTribunaRajoy
[To solve the Syrian refugees situation is a matter of humanity.
Actions should be taken at the source of the problem; this is the great
challenge of the #EU #BCNTribuneRajoy]
One tweet was sent twice, by Partido Popular and Mariano Rajoy, in
English, from the European Council, and linking to the national brief-
ing. As already mentioned about these messages in English from the EU
institutions, they do not convey much information by themselves, though

they are full of hashtags.
28. [Partido Popular] #EUCO—National briefing: Spain @marianora-

joy #migrationEU #refugeecrisis #UKinEU #Syria https://t.co/
nHUZP8tycU
Finally, Izquierda Unida mentions Syria in a list of conflicts that should

have been avoided.
29. [Izquierda Unida] No más intervenciones militares como las de Iraq,

Afganistán, Libia o Siria que son una amenaza para la paz y la seguri-
dad #VotaUPVotaPaz
[No more military interventions such as Iraq, Afghanistan, Libya or
Syria, which are a threat to peace and security #VoteUPVotePeace]
2.4.5 Humanitarian
We find 44 tweets with the root “human-” in the corpus (0.27%).

However, only two (0.001%) are related to migration. Besides the one of
Mariano Rajoy in (tweet no. 27), we find a quote by Monica Oltra, a
well-known member of Podemos in the region of Valencia.
30. [Podemos] Nuestro gobierno es insensible ante la crisis humanitaria

más grave de las últimas décadas @monicaoltra #VotaPodemos20D
[Our government is insensible to the most important humanitarian
crisis of the last decades @monicaoltra #VotePodemos20D]
We classified the rest of tweets with the root “human-” under five top-
ics. They are (in order of frequency): human rights (12), human capital
(9), humanity (7), candidates’ humanity (2), and humanitarian crisis (1).
The latter, which is the only one related to migration, was sent by the
Partido Popular and referred to the situation in Venezuela.
2.4.6 Refugees
Some 16 tweets include the word refugiado/s. None of them was sent by
Ciudadanos. Izquierda Unida (4 tweets) uses the hashtag “#refugeeswel-
come” (in English), an international campaign to receive refugees in
European countries, and demands an EU that is open to the refugees.
Partido Socialista (3 tweets) criticises the government policies, considered
to conflict with the citizens’ ethics.
31. [Partido Socialista] Rajoy ha planteado una política europea con-

traria a los valores de los españoles en materia de refugiados @san-
chezcastejon #CARAaCARA2015
[Rajoy has proposed a European policy on refugees contrary to the
values of the Spaniards @sanchezcastejon # FaceToFace2015]
Podemos (3 tweets) makes explicit its position, guaranteeing the refu-

gees’ rights.
32. [Podemos] Vamos a ser referente en Europa de los derechos, la paz y

la acogida de refugiados q huyen de la guerra @OviedoTaboada
#PodemosEnAsturies
[We are going to be an example in Europe for rights, peace and asylum
for refugees fleeing from war @OviedoTaboada #PodemosInAsturies]
Partido Popular (1) and Mariano Rajoy (3) sent tweets from the
EU and in English, while Pedro Sánchez published a link to an inter-
view for UNHCR about refugees and asylum policies. In that inter-
view, when asked about what his message to his voters is, Pedro
Sánchez answers that: “We the socialists keep in mind the memories
of the exile and also the memories of reconciliation. The elderly know
first-hand what it means to suffer persecution and I think that is the
reason why we have a special sensitivity to the current issues of
refugees.”
2.4.7 War
Refugees are fleeing from war. As we saw in the newspapers, the Syrian
War was a hot topic in 2015, especially because of the migration phe-
nomenon that it caused. We found 29 tweets with the word “war” (0.18%
of the corpus). Only four (0.02%) are related to the Syrian conflict. We
classified the other tweets into three frames: (1) war as a general concept
(20); (2) the Spanish Civil War (4); and (3) metaphorical uses as an
important conflict (1). We did not find any direct reference to the current
war in Syria.
Some 19 of the 29 messages were published by Izquierda Unida,
mostly of war as a general concept and showing a pacifist point of view.
33. [Izquierda Unida] #PorUnNuevoPaís pacifista que se oponga a la

guerra como forma de resolución de conflictos https://t.co/
Ot2zekE1Nn #VenimosDeMuyLejos
[#ForANewCountry pacifist and against the war as a way of conflict
resolution [link to a campaign TV spot] #WeComeFromFarAway]
34. [Izquierda Unida] “La doctrina de la guerra que va desde Bush hasta
Hollande no ha hecho más que empeorar las cosas” @agarzon
#JóvenesConGarzón
[“The war doctrine that comes from Bush to Hollande has only wors-
ened the situation” @agarzon #YoungPeopleWithGarzon]
References to the civil war are part of a historical frame that Izquierda
Unida uses to claim a tradition that includes the Second Spanish Republic.
In fact, there is a direct reference to the Constitution of 1931, which was
the formal start of the republic.
35. [Izquierda Unida] “La Constitución del 31 renunciaba a la guerra

como modo de hacer política. Queremos recuperar ese espíritu” @
agarzon #GarzónEnZaragoza
[“The Constitution of 31 gave up on war as a way of doing politics.
We want to bring back that spirit” @agarzon #GarzónInZaragoza]
2.4.8 Terrorism
We have seen that the Syrian civil war is closely related to terrorism, both
in Syria and in Europe. Terrorism is a recurring topic in Spanish election
campaigns because Spain has a long history of terrorist attacks, in the last
decades mostly by the Basque terrorist group ETA, but also by radical
Islamist groups. There are 75 tweets with the root terror- in our corpus
(terrorismo, terrorista), which are 0.46% of the corpus. Therefore, this is
the most frequent root of the keywords we have selected. Most of them,
45, refer to the fight against terrorism. In this frame, we find concepts
close to the situation in Syria and the refugees’ migration such as musul-
mán (Muslim), jihad, Turquía (Turkey), bombardeo de países (bombing of
countries), and coordinación entre miembros de la UE (coordination between
EU members). However, we find important differences between the
accounts we analysed. Izquierda Unida points out that the “Western
World” is not only a victim, but also responsible for the terrorism. They
want also to make clear that the link between terrorism and Muslims is
unfair, and that terrorist threats are used in the national context as an
excuse for the deprivation of social rights.
36. [Izquierda Unida] Reconocer que el aumento de terrorismo se debe

en parte a acciones militares de Occidente bordeando legalidad inter-
nacional #VotaUPVotaPaz
[We have to acknowledge that the rise of terrorism has partly to do
with the military actions of the Western World which were almost
illegal #VoteUPVotePeace]
37. [Izquierda Unida] “El terrorismo no puede servir de excusa para que
se nos impida ejercer nuestros derechos y libertades” @Rsixtoiglesias
#GarzónEnValència
[“Terrorism cannot be used as an excuse for preventing us from using
our rights and freedom” @Rsixtoiglesias #GarzónInValència]
For the right-wing parties, Ciudadanos and Partido Popular, terrorism

is a common enemy that should unite all the parties. For Ciudadanos, it
means that it is a shared responsibility and it must be shown as such.
38. [Ciudadanos] .@Albert_Rivera “Contra el terrorismo internacional

lucharemos juntos y con inteligencia” #MurciaCiudadana https://t.
co/9NxX6AaN99
[.@Albert_Rivera Against international terrorism we will fight
together and with astuteness” #MurciaCiudadana [link to a photo of
Albert Rivera]]
39. [Partido Popular] Tenemos que estar todos unidos en la lucha contra
el terrorismo @pablocasado #LMElecciones https://t.co/Km2KiAzvAx
[We must be together fighting terrorism @pablocasado #LMElections
[link to a campaign picture with a list of proposals against terrorism]]
However, some Partido Popular tweets are ambiguous regarding the

frame they are in. As we can see below, they claim that the fight against
terrorism is “part of our DNA” and also reclaim its memory. Therefore,
they are talking about the past, and any Spaniard would identify terrorism
in the Spanish history with ETA. This ambiguity is relevant because,
though unity against jihad terrorism is clear for every party, unity against
ETA is not an easy issue. It has a long history that starts with the Franco
dictatorship. So when Pablo Casado states in (39) that “we must be
together fighting terrorism”, the interpretations are very different if we
read it within a frame of the current situation (Syria, the jihad) or within
a frame of Basque politics.
40. [Partido Popular] .@isanseba Las víctimas del terrorismo han sido,
son y serán parte del ADN del Partido Popular https://t.co/
dTcCYAQX36
[.@isanseba Victims of terrorism have been, are and will be part of
the Partido Popular DNA [link to a photo of Pablo Casado, Partido
Popular politician]]
41. [Partido Popular] Con @marianorajoy el @PPopular va a reivindicar
siempre memoria, dignidad y justicia para las víctimas del terrorismo
#CaraACaraL6 #YoVotoPP
[With @marianorajoy @PPopular will always vindicate memory, dig-
nity, and justice for the victims of terrorism #FaceToFaceL6 #IVotePP]
Podemos has only one tweet on terrorism and it is also to show its
support for and loyalty to the government. Rajoy is the only candi-
date with messages on this topic. He sent 12 tweets, mostly referring
to the fight against terrorism as a priority and related to national
security.
Some 11 tweets from the 75 with the root “terror-” are reactions to the
terrorist attack of Kabul on 12 December against the Spanish Embassy,
where two police agents were killed. Parties and candidates showed their
condolences in Twitter. Another 11 tweets are direct references to ETA
terrorism, all of them published by Partido Popular and Ciudadanos.
Eight tweets sent by Izquierda Unida use the term “terrorismo machista” as
a way of referring to violence against women.
There is one only tweet that makes the connection between refugees
and terrorism. It was sent twice, both in the account of Partido Popular
and in the account of its candidate Mariano Rajoy, and includes the
hashtag of the European Council (#EUCO).
42. [Partido Popular/Mariano Rajoy] #EUCO Nuestra solidaridad con

las personas que huyen de la guerra, de la barbarie terrorista y de la
vulneración de sus elementales DDHH
[#EUCO Our solidarity with those persons fleeing from war, terror-
ist cruelty and violation of their fundamental rights]
2.5 Aunque eso no dé votos

In electoral campaigns politicians are communicating in order to garner
votes. What they choose to discuss and what they choose to silence should
be read from that perspective. However, it can tell us a lot about which
issues are relevant to the society we live in and which topics are too contro-
versial to guarantee votes. This is especially clear when the issues we are
analysing are as tragic and urgent as the ones we have chosen for our study.
Regarding the methodology, we find corpus linguistics the most coher-
ent framework for the empirical analysis of what is absent in discourse.
The title of this chapter, “Not for Twitter”, only makes sense if we find
the topic in other contexts. We have followed a three-pronged approach

in our study that combines both an objective review of the historical
context and data from corpora that show us how this context was being
read by the society. The main idea behind our methodology is to test
whether immigration is so present in other discourses that we should also
expect it in the one we are analysing.
Key to this is choosing relevant corpora for comparison, which must
be important enough so that we can draw clear conclusions. Being
focused on an electoral context, we chose two different corpora. On the
one hand, the electoral manifestos of the five political parties whose
Twitter accounts were included in the main corpus. On the other, articles
devoted to immigration in 2015, the time frame of the main corpus,
which were taken from the four most widely read newspapers. These
complementary corpora have enabled us to assess how relevant migration
was in that political context.
We complemented the findings from corpora with a review of the his-
torical context. In order to do so, we reviewed official statistics relevant to
the presence of the topic in Spanish society. Examples are the proportion
of the Spanish population who are immigrants (around 10%), the
number of refugees who arrived to Europe in 2015 (more than 1 mil-
lion), and the migrant quota of Spain for refugees within the EU (the
third largest quota in the EU).
The historical context, the impact on the media, and the electoral
manifestos have shown that migration phenomena were a very important
issue when the 2015 election took place. Furthermore, they were being
presented mixed with other important topics such as terrorism, human
rights, and control. Between migrants who were seeking jobs and better
lives abroad (what governments call “economic migrants”) and those flee-
ing war-torn countries such as Syria (who are likely to be granted refugee
status), the latter were the ones who took up most of the media reports in
2015. The former were important in electoral manifestos, primarily
focused on the emigration of Spaniards.
The conclusion we reach from the news analysis is that migration was
being intensely discussed and, being a political issue, we would have
expected some debate on it by the election candidates. However, we have
seen in our research that, though migration is clearly a historical element
of 2015, very present in the news because of the Syrian crisis, it is barely
present in the electoral manifestos (around 3% of the pages, but with
very explicit proposals), and almost completely missing from Twitter.
We have also seen that parties and candidates did publish a lot in their
Twitter accounts during the campaign. They tweeted several times every
hour, but migration barely got space in the discussion. Only 8 tweets out
of 16,305 were dedicated to non-Spanish migrants. All but one were
published by Izquierda Unida quoting or discussing what was being said
on TV debates and televised rallies.
Because of their situation, immigrants have no voice on the Internet.
Because of our politicians’ decision, immigrants do not get talked about
on Twitter either. New digital media have adopted strategies from alterna-
tive and community media, reformulating the original citizen journalism
practice of gate-keeping into a collective exercise of gate-watching (Bruns,
2015). This strategy of gate-watching has been taken on by traditional
media due to the huge amount of information and sources they must deal
with, with their new task to highlight that information which they believe
is of most relevance to their audiences. This is not the case with politi-
cians, who keep behaving as gate-keepers: filtering which topics are rele-
vant according to internal selection policies. It shows us that the use of
new technologies does not imply new ways of communication. As tradi-
tional media used to do following their own idealised image of what their
audiences were interested in, politicians are publishing themselves now
following their idea of which topics can garner votes (Gallardo, 2016).
We saw that it was made explicit in a tweet by Izquierda Unida:
“Queremos sanidad de calidad, pública y para todos, incluidos los inmi-
grantes, aunque eso no dé votos” [“We want quality medical care, public
and for everyone, including immigrants, even though it doesn’t give us
votes”]. To include immigrants in the equation “doesn’t give us votes”.
The data show that every party has different messages, and their figures
are not exactly the same. However, differences are not important enough
to consider that the silencing of migration was not general. We saw that
Izquierda Unida published much more tweets on migration that the oth-
ers, but it is still an insignificant 0.15% of its tweets. Results for Podemos
and Ciudadanos are surprising since they are parties that have stated their
intention to pursue new ways of doing politics and of communicating

with its electorate.
When immigrants’ issues did appear in our corpora, it was through
specific frames. We have shown that newspapers conveyed two different
frames for refugees: one geographical and one of control. It means that,
even though immigrants and refugees were being taken into consider-
ation, their discourses were not. Their thoughts, feelings and needs were
not represented. It was also the case with the Spanish emigrants, appar-
ently better represented in the Twitter corpus. We have seen that mes-
sages referred to them did not give them any agency. It was not their voice
or their wishes, but ones imposed on them. In fact, these tweets seem
clearly electioneering when emigrants are not the addressee, as we have
seen with messages about them addressed to their “grandmas”.
One of the definitions of silence we gave in the Introduction was that
it is a way of disabling discourses. Our analysis has corroborated this.
Glenn writes that, depending on the context, silence can deploy power or
can defer to power (Glenn, 2004). What we have in the Twitter campaign
is clearly the first option. Parties and candidates were able to send 57
tweets about migration in 2015 without mentioning the refugee crisis.
They could also publish 29 messages with the word “war” without refer-
ring to the frame everyone had in their minds: the tragic news coming
from Syria.
The use of digital platforms has transformed communication in many
sectors, frequently towards more collaborative scenarios. It does not seem
to be the case with electoral communication, where the most powerful
nodes keep using their position to silence what they do not find relevant
or comfortable. As Coleman writes, “The Internet, like the rest of the
social world, is replete with inequalities and frustrating injustices”
(Coleman, 2013, p. 384). Twitter is an interactive service that allows
digital conversations using hashtags and references to user names. The
tweets we have analysed do not try to start an open conversation. They do
not ask for suggestions or comments. They convey very specific frames
about selected topics, both chosen to garner votes. Everything else
remains silenced.
Acknowledgements We thank the editors for the insightful comments on ear-

lier versions of this chapter. Corpus linguistics approaches always call for many
helping hands, and we are grateful to the members of our research group “Wor(l)
ds Lab” at the Universidad Autónoma de Madrid. The study reported in this
chapter is part of the project “Estrategias de encuadre y articulación del discurso
politico en 140 caracteres”, funded by the Spanish Ministry of Economic Affairs
(FFI-2014-53958-P).
References
Alcántara-Plá, M. (2014). El concepto de oración para el español oral. Oralia,
17, 63–84.
Anthony, L. (2014). AntConc (Version 3.4.4) [Computer Software]. Tokyo,
Japan: Waseda University. Retrieved from http://www.antlab.sci.waseda.ac.jp
Baker, P., Gabrielatos, C., Khosravinik, M., Krzyzanowski, M., McEnery, T., &
Wodak, R. (2008). A useful methodological synergy? Combining critical dis-
course analysis and corpus linguistics to examine discourses of refugees and
asylum seekers in the UK press. Discourse & Society, 19(3), 273–305.
Baroni, M., & Bernardini, S. (2004). BootCaT: Bootstrapping corpora and
terms from the web. Proceedings of LREC 2004.
Bird, S., Loper, E., & Klein, E. (2009). Natural language processing with python.
O’Reilly Media.
Brown, K. V. (2014, July 17). Instagram’s ‘mommy wars’: The right to bare chest.
SFGate. Retrieved from http://www.sfgate.com/news/article/Instagram-s-
mommy-wars-the-right-to-bare-chest-5626780.php#photo-6604616
Brummett, B. (1980). Towards a theory of silence as a political strategy. The
Quaterly Journal of Speech, 66, 289–303.
Bruns, A. (2015). Working the story. In C. Atton (Ed.), The Routledge compan-
ion to alternative and community media (p. 384). New York: Routledge.
Castells, M. (2009). Comunicación y poder. Madrid: Alianza.
Castells, M. (2011). A network theory of power. International Journal of
Communication, 5, 773–787.
Coleman, S. (2013). The Internet and the opening up of political space. In
J. Hartley, J. Burgess, & A. Bruns (Eds.), A companion to new media dynamics
(p. 384). Somerset: Wiley Blackwell.
Conway, B. A., Kenski, K., & Wang, D. (2015). The rise of Twitter in the politi-
cal campaign: Searching for intermedia agenda-setting effects in the presiden-
tial primary. Journal of Computer-Mediated Communication, 20, 363–380.
EGM: Encuesta General de Medios. (2015). Official report. Retrieved from

http://www.aimc.es/-Datos-EGM-Resumen-General-.html
Fairclough, N. (1989). Language and power. London: Longman.
Fairclough, N. (2010). Critical discourse analysis: The critical study of language
(2nd ed.). London: Routdledge.
Fairclough, N., & Wodak, R. (1997). Critical discourse analysis. In T. A. van
Dijk (Ed.), Discourse studies. A multidisciplinary introduction. Vol. 2. Discourse
as social interaction. London: Sage.
Fillmore, C. J. (1982). “Frame semantics”. Linguistics in the morning calm. Seoul:
Hanshin Publishing.
Fuchs, C. (2008). Internet and society: Social theory in the information age.
Fuchs, C. (2014). OccupyMedia! Zero Books.
Gallardo, B. (2016). Pseudopolítica: el discurso político en las redes sociales.
Universidad de Valencia.
Glenn, C. (2004). Unspoken: A rhetoric of silence. Southern Illinois University
Press.
Hine, C. (2015). Ethnography for the internet: Embedded, embodied and everyday.
Bloomsbury Publishing.
Howarth, D. (2005). Applying discourse theory: The method of articulation. In
Co-edited book with J. Torfing, Discourse theory and European politics:
Identities, policy and governance. Palgrave.
Jaworski, A. (1993). The power of silence: Social and pragmatic perspectives. Sage.
Lakoff, G. (2004). Don’t think of an Elephant. Hartford: Chelsea Green
Publishing.
Langacker, R. W. (1991). Foundations of cognitive grammar. Descriptive applica-
tion. Stanford: Stanford University Press.
Lukes, S. (2005 [1974]). Power: A radical view. New York: Palgrave Macmillan.
Mancera, A., & Pano, A. (2013). El discurso político en Twitter. Barcelona:
Anthropos.
Padró, L., & Stanilovsky, E. (2012). FreeLing 3.0: Towards wider multilingual-
ity. Proceedings of the Language Resources and Evaluation Conference (LREC).
Partington, A., Duguid, A., & Taylor, C. (2013). Patterns and meanings in dis-
course. Amsterdam: John Benjamins.
Schröter, M. (2013). Silence and concealment in political discourse. Discourse
approaches to politics, society and culture, 48. Amsterdam: John Benjamins.
Squires, L. (2016). Twitter: Design, discourse, and the implications of public
text. In A. Georgakopoulou & T. Spilioti (Eds.), The Routledge handbook of
language and digital communication. Devon: Routledge.
Taylor, C. (2014). Investigating the representation of migrants in the UK and

Italian press: A cross-linguistic corpus-assisted discourse analysis. International
Journal of Corpus Linguistics, 19(3), 368–400.
Thiesmeyer, L. (Ed.). (2003). Discourse and silencing: Representation and the lan-
guage of displacement (Vol. 5). John Benjamins Publishing.
van Dijk, T. A. (1997). What is political discourse analysis? Belgian Journal of
Linguistics, 11(1), 11–52.
Wodak, R. (1996). Disorders of discourse. London: Longman.
3
Absence in Visual Narratives: The Story
of Iran and Pakistan across Time
Sameera Durrani
3.1 Introduction
The way in which the story of a country is told in the elite western news
media significantly influences the manner in which it is perceived and
treated by members of the international community. The image of a
country, as constructed by the media, impacts both economic opportuni-
ties available to it (Gertner & Kotler, 2004), and the foreign policy for-
mulated towards it (Gilboa, 2002). Understandably, therefore, it
constitutes an area of interest for social science researchers from different
domains.
The concept of a national image encompasses a diverse range of epis-
temic concerns, and has therefore been investigated by researchers from
different disciplines, such as Public Diplomacy (McPhail, 2010) and
Nation Branding (Anholt, 2004). Researchers from these domains exam-
ine this issue from a diplomatic and advertising/marketing perspective,
and concentrate more on the policy and public agenda aspects of the
S. Durrani (*)
University of New South Wales, Sydney, NSW, Australia

66 S. Durrani
debate. News media agendas are investigated by analysts from the domain
of communication and journalism; these, however, are dominated by
studies of textual coverage (see Sect. 3.2 for a more detailed discussion)
and coverage of short-term events such as armed conflicts (for example,
see Fahmy, 2004; Griffin, 2004; King & Lester, 2005). This leaves a gap
in the field: the examination of diachronic visual news narratives of
national images.
This chapter is part of a larger research project that aims to address this
research gap. It examines the visual treatment of Iran and Pakistan in
Time over the course of 30 years (1981–2010). It does so with the help of
a triangulated methodology that employs both quantitative and qualita-
tive techniques: content and semiotic analysis. The interpretive potential
of the findings with the help of the works of theorists that include Kress
and van Leeuwen (2006), van Leeuwen (2008), and Homi Bhabha
(1991). Historical contextualisation is employed where relevant.
The content is sampled from an elite western news magazine, Time,1
chosen because of three key reasons: its prestigious status, the priori-
tisation of the visual in its content, and its longevity—the magazine
has stayed in publication for the entire duration of this study, and as
a brand name, still enjoys the status of an intermedia agenda setter
(Vliegenthart & Walgrave, 2008) as was witnessed recently in the cov-
erage given to its decision to award its prestigious ‘Person of the Year’
title to U.S. President-elect Donald Trump.
As discussed in Chap. 1 in this volume, and reflected in the research
presented in the other chapters, research into absences tends to focus on
analysing linguistic resources. Given that visuals increasingly dominate
news discourse in the digital age, it is important to craft methodologies
that analyse absences within visual news discourse. This is what this study
sets out to accomplish. The aim of the analysis here is to demonstrate
how absences in visual discourse can be identified using systematic,
empirical analysis. It provides an insight into what a methodology for
analysing semiotically invisible elements of photojournalism can look
like. To that end, it combines tools from media analysis literature with
theoretical lenses from semiotics, cultural studies, and post-colonial the-
ory. Quantitative techniques such as content analysis are an efficient
means for documenting what is present, thus opening a door to the
Absence in Visual Narratives: The Story of Iran and Pakistan... 67
a nalysis of what might be missing. Qualitative lenses provide an effective

mode of discussing the significance of these unavailable semiotic choices,
with reference to how such absences serve to structure specific types of
discursive constellations, and their interplay with broader issues of power
The data for this study was analysed with the help of a relational data-
base, MS Access, which allows intensive data analysis at both a macro and
micro level. Statistical analysis allowed the documentation of diachronic
trends with reference to various variables. The abundance of data in cer-
tain categories, and its scarcity in others, were one indication of potential
discursive absences. In other cases, the scarcity of the data made it neces-
sary to investigate the presence of certain themes in adjacency pairs, as
the alternative presence of a theme in the news coverage given to one half
of a pair may direct attention to its absence in the other half. For this
reason, the study focuses on a pair of neighbouring countries that are
similar in terms of several key dimensions, thus making them an apt
combination for the purposes of comparison. This will be discussed in
greater detail in Sect. 3.4.
The analysis in this chapter focuses on the coverage given by the maga-
zine to two countries, Pakistan and Iran. Both are two geo-strategically
significant nations that have experienced significant social and political
upheavals within the selected time periods. Iran has gone from an Islamic
revolution in 1979 to a post-revolution generation at odds with its legacy,
as seen in the protests of the Green Movement (2009). Meanwhile,
Pakistan has experienced two proxy wars, the Afghan Jihad in the 1980s,
the War on Terror in the 2000s, and remains a geo-politically significant
nuclear power. Given their turbulent history, and their centrality to myr-
iad political issues and conflicts, and their consistently negative coverage
in the press (see Sect. 3.2), the two represent interesting, complex case
studies as diachronic narratives of national images in the news media.
The quantitative tracking of semiotic and thematic patterns facilitates
the documentation of the manner in which the story of a country is told
over a generational time span (30 years). Like most media content stud-
ies, this research focuses primarily on documenting what the stories about
Pakistan and Iran in Time say: themes and ideas present within them.
However, the time span of the study facilitates retroactive historical anal-
ysis, making it easier to identify what is not there: absences.
68 S. Durrani
This chapter examines, therefore, visual absences in the narratives of

Iran and Pakistan in Time. It looks at how absences arise from, and struc-
ture, discursive constellations. In this context, it identifies the presence of
three different types of absence within the examined sample: minus in
origin (Bhabha, 1991), exclusion (van Leeuwen, 2008), and semiotic,
interactive absences (Kress & van Leeuwen, 2006). These will be dis-
cussed in detail in Sect. 3.4, alongside relevant examples. It also examines
the impact of time in mediating absences. It provides an insight into the
manner in which the consequences of events absent from a discourse can
find their way into news narratives decades later, and it examines how
certain absences can dissipate, under the influence of changing geo-
political events.
3.2 P
akistan and Iran, Through the Eyes
of the Global News Media: General
Trends
Both Iran and Pakistan have received their share of unfavourable media
coverage over the past three decades. This has been documented in vari-
ous research studies, discussed in this section, although more research
exists with reference to Iran than Pakistan. This uniformity of negative
coverage exists despite differences in the way these two countries are posi-
tioned in terms of geo political alliances. Since the 1977 Islamic revolu-
tion, Iran is seen an adversary (Semati, 2008), while Pakistan has been an
official ally of the United States in two proxy wars: the Afghan Jihad in
the 1980s (Talbot, 1998), and the post 9/11 War on Terror (Hanif,
2011).
Officially, Pakistan has been a western ally through most of its history.2
In terms of media coverage, though, the alliance is perceived as a compli-
cated one. This trend goes back many decades. While the American gov-
ernment’s official stance towards Pakistan during the 1971 Indo-Pak war
was supportive, the same was not true of representatives of the elite
American press, such as The New York Times (Becker, 1977). Research
studies that look at the coverage given to the regional dispute between
India and Pakistan over Kashmir find similar trends—Pakistan’s role in

the conflict is evaluated negatively (Mughees, 1995). Mughees (1997)
goes on to note in another study notes that The New York Times tends to
be critical of the policies of economic and military aid towards Pakistan.3
Negative opinions towards Pakistan have gained further traction post
9/11—the country, as Jalal (2011) puts it, is now seen by many ‘as the
world’s largest assembly line of terrorists’ (p. 7). Studies on news coverage
of the country bear out this assertion. Khan (2002) notes that after 9/11,
news photographs from Pakistan in news magazines like Time and
Newsweek primarily focused on documenting stereotypical themes, such
as religious extremism and oppression of women. Durrani and Mughees
(2010) observe similar trends, while adding that Pakistan’s then President,
General Musharraf did receive more favourable coverage post 9/11, while
Khan and Irtaza (2010) found that the country’s official alliance with the
U.S. did not result in any softening of editorial and op-ed stances in elite
newspapers like The Washington Post.
Iran’s coverage in western media is influenced strongly by one event:
the Islamic revolution of 1977. Tadayon (1980) found the pre-revolution
era newspaper coverage for the country revolved around support for the
ruling monarch, Shah Reza Pahlavi. Tadyaon (1980) noted one critical
absence during this period: reports of opposition to the Shah were virtu-
ally non-existent.
After the 1977 Islamic revolution, media coverage of Iran became hos-
tile, and focused extensively on issues such as the government-sanctioned
hostage-taking in Tehran in 1980, nuclear and arms build-up, and eco-
nomic sanctions imposed by the West on Iran (Naficy, 2008, p. 77). Iran’s
nuclear capability has consistently received negative coverage over the past
three decades in the elite American press (Izadi & Saghaye-Biria, 2007), as
does the issue of compulsory veiling (Chan-Malik, 2011), and the status
of women in Iran (Roushanzamir, 2004). Holistically, since the 1979
Iranian revolution, ‘the official narrative of the United States and many
other Westerners regarding the Islamic republic often depicts Iran as a state
that is simply ruled by a handful of mad mullahs’ (Semati, 2008, p. 2).
Despite the general hostility, there have been periods of relative détente,
a process facilitated by cultural and political developments within Iran
itself. Since the 1990s, Iranian cinema has won considerable acclaim at
70 S. Durrani
international film festivals. This has served to moderate ‘the barbaric and
backward impressions of Iran as a nation’ (Naficy, 2008, p. 189). The
tone of the international press softened further after the victory of the
‘moderate’, reformist clergyman, Mohammed Khatami, in the May 1997
presidential elections, (Khiabanay, 2008). Sporting events, such as foot-
ball matches, have also generated positive coverage of ordinary Iranians,
although Iranian government officials continue to be depicted negatively
(Delgado, 2003). The hostility towards the Iranian government within
the media was compounded further by the inclusion of Iran in the ‘Axis
of Evil’ by President George Bush (Kafala, 2002), and the election of
hardliner Mahmoud Ahmadinejad as President in 2005 (Semati, 2008).
Another shift in the global perceptions of Iran occurred in 2009, with
the Green Movement, which erupted following outrage over charges of
electoral fraud levelled at Ahmadinejad (Mortensen, 2011). Online
forums, the content of which was available to both local and global audi-
ences, played a crucial role in organising and expressing dissent against
the government, an instance of which can be found in the coverage given
to Neda Agha Sultan, an Iranian woman shot dead during anti-
government protests. The video of her final moments went viral, and her
face became the global icon of the Green Movement (Mortensen, 2011).
Absences, as noted in the beginning of this volume, may arise from
contextual predetermination. Briefly, as noted in Chap. 1 in this volume,
this means that what is present in a discourse is governed by the norms of
sayability, and adherence to professional obligations. The discussion in
this section provided an overview of what prevalent norms of sayability
about Pakistan and Iran look like, as accepted by professional journalists
working for prestigious global news outlets; this has been done in order
to better contextualise the findings reported in this chapter.
3.3 Methodology
The study examines a purposive sample of 840 images (Iran: 376 images,
Pakistan: 454 images). This is a type of sample chosen to be representa-
tive of a population (Wimmer & Dominick, 2003, p. 465). Researchers
who use this type of sampling purposely choose subjects relevant to the
research topic (Sarantakos, 1998, p. 152). This research study limits itself
to relevant material by focusing on images taken from sections of Time
that emphasise the visual in their layout, and which are consistently
recurrent across the three decades: 1980s, 1990s, and the 2000s.4 These
include: photo features, photo essays, interviews, Time Person of the Year,
Time’s selection of the most memorable photographs of the year, and
feature articles. Feature article subtypes include news features, profile fea-
tures, lifestyle features, issue-based features, and supplementary/explana-
tory features, which are defined as small articles on supplementary themes
appended to the main feature article. Types of images included in the
sample are: cover images, secondary cover images (thumbnail images
which sometimes appear on the cover of Time) and photos that appear
inside the abovementioned sections. These were analysed with the help of
an MS Access database, an electronic relational database that allows for
data and relationships among data to be stored in the form of tables
(Caple, 2013). This data was then coded into categories that draw on
Kress and van Leeuwen’s (2006) framework for analysing images, specifi-
cally, their system for analysing visual grammar with the help of three
different metafunctions: Representation, Interaction and Composition.
This chapter focuses on notable absences in the data trends which
characterise the first two metafunctions, Representation and Interaction.5
These terms are defined and explained in the next two paragraphs. The
MS Access database contains provisions for quantitative coding, as well as
qualitative notes. The first section of analysis demonstrates absences pri-
marily through quantitative analysis, while the second section relies
mainly on a qualitative, historical approach. As demonstrated in Sects.
3.4 and 3.5, these choices flow from the nature of absences analysed.
Section 3.4 focuses on empirical analysis of absent interactive cues, by
looking at data that dominates the coverage, while Sect. 3.5 examines
specific representational cues, scarce to the point of absence, and engages
in qualitative contextualisation of why this may be so.
To Kress and van Leeuwen (2006), Representation, or the ideational
metafunction, refers to ‘the ability of semiotic systems to represent objects
and their relations in a world outside the representational system or in
the semiotic systems of a culture’ (p. 47). In the context of news photo-
graphs, this may refer to themes, people and activities portrayed in the
72 S. Durrani
images. It is pertinent to mention that representational meanings are

determined here with the help of a verbal context unit (Huang & Fahmy,
2011), a common practice in visual research (see, for instance, Parry,
2010, 2011; Fahmy and Kim, 2008) verbal context unit (VCU) consists
of three elements: the headline, the stand-first (Economou, 2010), opera-
tionally defined as the short paragraph that accompanies the headline,
and the caption.
Interaction refers to the process whereby a semiotic mode projects the
relations between the producer of a sign, and the receiver/reproducer of
that sign. Kress and van Leeuwen (2006) define interactive participants
as the people who communicate with each other through images, i.e. the
producers and viewers of images. With reference to visual communica-
tion, this translates into the use of visual cues strategies such as eye con-
tact (e.g. direct or indirect), or power relations (conveyed with the help
of camera angles). The specific cue analysed in this chapter is Gaze (eye
contact). The subcategories are: direct (when a person looks directly at the
camera), indirect (when person looks away from the camera, though the
face is still visible, none (when a person’s back is turned to the camera),
and mixed (when some people look at the camera, and some look away).
The systemacity of the coding system allows the documentation of
empirical patterns with reference to both representational and interactive
cues, and this creates the possibility of systematically identifying the scarcity
or absence of specific types of semiotic cues. The analysis in Sect. 3.4 focuses
on patterns of interactive cues employed to represent Iranian women over
30 years. Sections 3.5.1 and 3.5.2 focus on a specific decade, the 1980s.
These sections show comparisons of compare patterns of similar represen-
tational cues for coverage of educational reforms in Pakistan and Iran.
3.4 Interactive Metafunction: Iranian Women

and the Absent Gaze
Representation of Muslim women in the media is an issue that is fre-
quently discussed in communication and linguistics literature. For
instance, Al-Hejin (2015) combines critical discourse analysis with cor-
pus linguistics to investigate semantic macrostructures in the BBC’s
c overage of Muslim women. In a similar vein, this study combines semi-

otics with content analysis in order to analyse visual representations of a
specific group of Muslim women (Iranian women), and investigates how
these patterns have evolved over the years.
The interactive dimension of images refers to the ‘writing of what is
usually called non-verbal communication’, a language shared by produc-
ers and viewers alike (Kress & van Leeuwen, 2006, p. 116). The method-
ological design of this project enables the documentation of patterns of
interactive analysis across three decades for both Iran and Pakistan. The
categories for interaction analysis used for this study draw on the frame-
work devised by Kress and van Leeuwen (2006), and consist of the fol-
lowing: gaze (eye contact), social distance (the illusion of distance created
by camera shots) and power relations (camera angle). The absence dis-
cussed here occurs with reference to the first category: gaze. The specific
section in the sample to which it applies is the coverage given to Iranian
women.
The presence or absence of eye contact within a photograph is an
important arbiter of meaning potentials. There is a fundamental differ-
ence between pictures in which people look directly into the camera, and
into the viewer’s eyes, and pictures in which this is not the case (Kress &
van Leeuwen, 2006). A direct gaze demands that the viewer enter into
some kind of relationship with the viewer. These are referred to as ‘demand
images’, in which represented participants are allowed, as subjects, to
‘demand’ a social response of some kind from the viewer (Kress & van
Leeuwen, 2006). In contrast, images which depict an indirect gaze, where
people look away the camera, are much less interactive—the person
depicted within it is offered to the viewer almost as an object for contem-
plation. A real or imaginary barrier is erected between the represented
participant and the viewer, a sense of disengagement. These may there-
fore be described as ‘offer’ images (Kress & van Leeuwen, 2006, p. 120).
Based on trends derived from the data set, this research adds two more
categories to Kress and van Leeuwen’s system. The first is ‘no eye contact’
(when the subject’s back is turned towards the camera, and the face is not
visible). I argue that ‘no eye contact’ offers an even more intense sense of
disengagement, than indirect eye contact. This is another additional
dimension of the ‘offer’ category. The second category is ‘mixed’ where
74 S. Durrani
some people look towards the camera, and some do not. This is a mixture
of ‘demand’, and ‘offer’.
To begin with, it should be noted that the norm for the sample is indi-
rect eye contact, for both Iranian men and women. From a total of 189
photographs, 75.66% of Iranian men look away from the camera. Direct
eye contact is established in only 3.70 % of the images (ten photographs),
while the remaining images fall within the categories of ‘none’ (face
turned away completely from camera), or ‘mixed’ (some people look at
the camera, while some look away). For Iranian women, though, the
number is even lower—four photographs feature direct eye contact,
across 30 years, out of a total of 41 photographs (one from the 1990s,
and the rest from the 2000s).6 In terms of percentage, 9.75% of the pho-
tographs that depict Iranian women feature photos with direct eye con-
tact. A commutation test reveals similar patterns for Pakistan: most of the
represented participants from Pakistan are depicted with indirect eye
contact, men (75.60%) as well as women (80.26%).7 A total of 7.2% of
Pakistani men make direct eye contact with the camera, while the same is
true of 11.84 % of women.
The first Pakistani woman to make direct eye contact in the sample is
the politician Benazir Bhutto, in a photograph published on November
14, 1988. This, as the following analysis demonstrates, is a much earlier
occurrence as compared to the appearance of the same category in the
data subset for Iranian women. There is a general absence in the dataset
of photographs that contain direct eye contact. This may be indicative of
the general norms that govern the genre of photojournalism. The absence
is, however, further compounded when the aspect of gender is taken into
account, and becomes even more pronounced when nationality is taken
into account. The most compounded absence occurs in the subsection
where nationality and gender intersect: Iranian women.
The first photograph featuring an Iranian man who makes direct eye
contact with the camera appears the very first year of the data set, 1981.
It features an Iranian journalist, Time’s Raji Samghabadi (February 16,
1981; see section 3.5.3 for a more detailed analysis of this photo, in the
context of political dissent). The first photograph featuring an Iranian
woman who makes direct eye contact with the camera occurs 17 years
into the sample (December 15, 1997). The picture shows a young girl in
a black headscarf, with a sombre expression, draped over another older

women’s shoulder. The other women have their back turned towards the
camera. Their figures form an implacable black wall of chadors. The lit-
tle girl’s pale face, and her light turquoise dress, serve to draw the view-
er’s eye straight towards her. The headline reads ‘A Remembrance of
Things Past’. The text identifies the women as attendees of a ceremony
that commemorates and mourns a fallen soldier. The image is more
about the context, or the situation, rather than the actors; the little girl
is not named.
The first named Iranian woman who looks directly into the camera is
the Nobel Laureate Shirin Ebadi, in a photograph that appears in the
sample 26 years into the coverage (May 15, 2006). It is a close-up shot
that shows Ebadi, minus the traditional head scarf, looking directly at the
camera. This photo represents a break from the stereotypes of the 1980s
and 1990s, and is the first of a trio of photographs which depict pioneer-
ing, independent women. The second one depicts a female footballer
flanked by her fellow players (June 12, 2006), while the third one shows
a female health worker, her team standing behind her in the distant back-
ground (March 23, 2009).
The absence of Iranian women who look directly at the camera is,
then, a trend with three key dimensions. The first, and the most obvi-
ous one, is the scarcity of data. The second is the time lapse—the
sheer amount of time that elapses before such a photo appears in the
sample. A third aspect worth examining here is the diachronic pattern
of gaze trends that helps demonstrate how this absence gradually
dissipates.
The data in this study examines trends from three different decades.
While the 1980s and the 1990s are both characterised by indirect gaze
(see Fig. 3.1), it is a very different kind of indirect gaze. Until 1989, pho-
tographs of Iranian women in Time show them clad in dark cloaks, with
their backs turned towards the camera. Rather than indirect gaze, it is
better characterised as no gaze. Photographs from the 1990s, however,
contain examples of what may be seen as a more accessible indirect gaze.
This may be attributed partly to changes in the conventions that govern
the photojournalism genre. Photographs that appeared in Time during
the 1980s were usually black and white, and sometimes grainy, in a way
76 S. Durrani
Fig. 3.1 Diachronic patterns for Gaze (Eye Contact): Iranian Women (1981–2010)
that signifies documentary-style authenticity. Photographs in the 1990s

were always in colour, and favoured clarity. The trend applies to this par-
ticular subset of the data as well. Iranian women photographed in the
1990s were presented in colour photos that clearly showed their faces.
Most of the photos show young women in public spaces. A typical exam-
ple would be the photo of a smiling customer in a restaurant, who looks
over her shoulder into the distance, published on March 22, 1993. While
the women in these photographs also look away from the camera, the
relative visibility of their faces creates a more interactive relationship
between the viewer and the person represented in the photographs. In
terms of historical context, this trend may partially be attributable also to
a general atmosphere of détente, facilitated by favourable perceptions of
the men who ruled Iran during this decade—President Rafsanjani, and
then President Khatami, both of whom are portrayed within the news
magazine’s narrative as progressive.
This trend is reinforced further in the 2000s, which features three pho-
tographs where women make direct eye contact with the camera, with
relatively closer camera angles. For a quantitative overview of the manner
in which the patterns of eye contact change across Time, see Fig. 3.1.
3.5 Absences: Representational

Metafunction
The data for this study extend across three decades. However, in terms of
the representational metafunction, two visual absences noted within the
data set pertain to the coverage given to Pakistan within one specific
decade—the 1980s. These absences are highlighted with the help of com-
parison conducted against the second half of an adjacent pair: Iran. It is
noted here that the analysis is conducted with the help of an analytical
binary division which contrasts scarcity with absence. The absent themes
and actors are not abundantly present in the data for Iran, but they do
exist, while they are absent altogether in the data for Pakistan.
The study from which these results are taken codes data into a set of
themes, as well as categories of social actors, which allowed the researcher
to track which themes were more prevalent, and which were not, and
which categories of social actors were given coverage, and which were
not. Politics, predictably, was the predominant theme, while certain other
themes, such as education, were given much less space. Politicians and
military men dominated the coverage for Pakistan, while religious leaders
and politicians dominated the coverage of Iran (Durrani, 2016). Avenues
of scarcity provided inspiration for investigation into potential absences.
The analysis in this section looks at two of these: one instance of thematic
absence/scarcity, and one instance of absent/scarce social actors.
Section 3.5.1 provides relevant historical context. The analysis in Sect.
3.5.2 focuses on thematic absence/scarcity: education, specifically, educa-
tional reforms. This section of the analysis relies heavily on historical con-
textualisation, since in both countries, the scarce/absent reforms were
carried out in one specific decade: the 1980s. At the time, Pakistan and
Iran were ruled by heavily authoritarian systems of government, dominated
78 S. Durrani
by strong individuals: General Zia and Ayatollah Khomeini respectively.

Political dissent is a theme that exists in the coverage given to both.
However, certain types of dissidents present in the coverage for Iran are
absent in the coverage given to Pakistan. This trend, and possible expla-
nations for the discursive constellation that prioritises certain actors over
others, is discussed in Sect. 3.5.3.
3.5.1 H
istorical Context, and Key Strategies
for Identifying Absences
Before venturing into an analysis of data for this decade, it is relevant to

first provide some historical, geo-political context. As a state, during
1977–1988, Pakistan was ‘authoritarian in political structure’ and
‘aspired to be an ideological state’ (Talbot, 1998, p. 245). The country
was ruled by General Zia-ul-Haq, who took power in a military coup
on the 5th of July, 1977, following political unrest which preceded the
general elections (Hevisi, 1988). The coup, and the consequent disrup-
tion of democracy, initially led to the country being sidelined by the
international community. This changed as a result of two events that
transpired in the same year, 1979: the overthrow of the pro-American
Shah of Iran, and the Soviet occupation of Afghanistan (Talbot, 1998).
The outbreak of this conflict forced the Americans to reappraise their
security arrangements with Pakistan (Talbot, 1998, p. 267). General
Zia went from being an international pariah to America’s frontline ally
in the fight against communism (Talbot, 1998, p. 246). General Zia
ruled Pakistan for 11 years, until his death in an airplane crash on
August 17, 1988 (Talbot, 1998). Internationally, Zia was seen as a reli-
able ally, but domestically, he was seen as a deeply divisive figure, and
the legacy of his political reforms continues to be debated in Pakistan
(see, for instance, Rehman, 1988; Nasir, 2012; Murtaza, 2015). These
reforms are seen to have shaped what Pakistan is today. Their absence in
international news discourse has a significant impact on how Pakistan is
understood within the global community. It is the salient absence of
some of these reforms within the selected sample, therefore, that is dis-
cussed in this section.
Before an absence can be documented, it is useful to have an idea of

what is present. Pakistan received 110 photos as coverage over the course
of ten years (between 1981 and 1990), and 43% of all images published
in this decade focus on two key political elite actors, President Zia, and
his chief political opponent, Benazir Bhutto. Zia featured in 22 images,
while Benazir featured in 29. Until his death in 1988, both the Pakistani
state and its structure were influenced strongly by the man who ruled it
during this period, and that was Zia. The analysis in this section therefore
focuses on absences in coverage relating to the policies enacted by his
government.
During the 1980s, Pakistan, as seen through the lens of Time, was seen
as a country ruled by a reliable dictator who, while facing dissent, steered
the ship in a stable manner. A small illustration of this frame of coverage
comes from the eulogy given to him after his death (August 29, 1988).
The article which documents his death in a plane crash is headlined as
‘Vacuum’. The stand-first reads ‘A suspicious airplane crash kills President
Zia and plunges Pakistan––and Southwest Asia––into uncertainty’. A
photo gallery in the same article is accompanied by the following text ‘His
enemies described Mohammed Zia-ul-Haq as tough, uncompromising,
even brutal. But those who got to know the late President privately dis-
covered a devout, often charming man with a strong sense of mission.’
Iran’s theological shift towards conservatism is well documented in the
media and international consciousness via two iconic historical events,
embedded in public consciousness—the 1979 Islamic revolution, and
the Hostage Crisis that followed. The crisis was instigated on November
4, 1979, when Islamist students climbed up the walls of the U.S. Embassy
and took 66 diplomats hostage. It was an event that changed U.S.-Iranian
relations for a generation (Tabaar, 2014), and still continues to define
popular perceptions of Iran in the media. Since the revolution, and its
consequent events, the general stereotypical frame that dominates the
perception of Iran in the media is that of a country ruled by ‘mad mul-
lahs’ (Beeman, 2005), and these are the perceptions that dominate Iran’s
coverage in Time during the 1980s.
The analysis looks at two key types of absences within the narrative of
Pakistan in Time in the 1980s, gauged by contrasting them with Iran’s
coverage. The first absence—the lack of coverage given to critical
80 S. Durrani
institutional reforms—may be seen as an example of what Bhabha terms

a ‘minus in origin’ (Bhabha, 1991, p. 160), a narrative strategy that is
applied to deprive selected actors of agency. The second absence makes
use of the strategy of exclusion—the absence of people (in this case, cer-
tain types of dissidents) from contexts where they are present in reality
(van Leeuwen, 2008).
The presence and significance of these absences are explored with the
help of two analytical strategies. The first is a semiotic commutation test
(Lacey, 1998). This involves the replacement of one sign by another (for
example, to replace the gender of a character in a novel). Substituting
objects for other signs in the same paradigm, and decoding the new
meaning, helps isolate the contribution of the original sign to the mean-
ing of the image. The paradigm here, in Sects. 3.5.1 and 3.5.2, is educa-
tional reforms, while the signs substituted are Pakistan and Iran. This
strategy relies on the use of comparisons to spot absences. As outlined in
Chap. 1 in this volume, comparisons have been used elsewhere in this
volume to examine absences in textual coverage. Here, in the form of a
semiotic commutation test, this idea is applied to the analysis of the
visual news coverage of national images. Specific discursive constellations
come to characterise the coverage given to different countries and ethnic
groups, and this strategy is suggested here as a means of questioning these
constellations. For instance, if a terrorist attack occurs in Pakistan, and
another occurs in France in the same year, and both claim roughly the
same number of victims, how and why does the coverage vary in a pres-
tigious global news outlet like BBC? In a world increasingly defined by
filter bubbles, where specific groups are increasingly branded with polar-
ised discourses, semiotic commutation tests could be a good strategy for
challenging the preconceived suppositions that produce absences. The
second strategy used here is retrospective historical analysis. The absence
of a phenomenon in a news narrative does not mean that it ceases to exist
and impact reality. At times, the event itself may not register in the public
consciousness, but its historical consequences might, especially when
they are too dramatic to ignore. Therefore, where relevant, the analysis
draws on historical data, connecting the past with the present. These
strategies are applied to the data in order to generate questions about the
possible causes of such absences, their consequences, and their possible
presence in other kinds of news narratives.
3.5.2 Educational Reforms
Absence is a relative phenomenon, which is best illustrated, at times, with

the presence of the same things in a parallel context. Using a semiotic
test, the analysis here looks at the coverage given to educational reforms
carried out in the 1980s in two Muslim countries in Asia that underwent
significant ideological changes at the state level in this decade. The spe-
cific paradigm here is education, and broadly, the signs substituted here
have to do with Pakistani and Iranian education systems.
The Islamic revolution and the hostage crisis dominate the coverage of
Iran in Time during the 1980s. The hostage crisis received 77 images in
Time, and was a dominant point of coverage in this decade. The impact
of conservative ideology on education is, not a foregrounded issue, but it
does make an appearance. The most salient example of this theme is a
news feature published in the magazine on April 2, 1984. The article,
which takes up two-thirds of a single page in the magazine, includes a
single black and white image that shows a page taken from an Iranian
textbook. This snapshot shows a medium close-up profile view of Iran’s
Supreme Leader Ayatollah Khomeini, in his trademark turban and gown,
with his hands raised in prayer. A corner of this photograph is splattered
with a small drop of blood. The headline reads, ‘Children’s Lit: Pages
from a primer for war’. The caption adds, ‘Book of Souvenirs: Khomeini
and a blood spot’. While this research study only examines images and
verbal context units, for the sake of contextualisation, it is noted here that
the text of this article examines propaganda primers produced by the
Iranian state which were used by young Iranian soldiers. A second exam-
ple which appears on August 26, 1985, comments on the effect of politics
on university campuses. The photograph, a high angle long shot, shows a
group of men with their backs towards the camera, prostrated on the
ground, offering namaz, i.e. Muslim prayers. The caption for the photo-
graph comments, ‘At the university of Tehran politics weeds out the best
students reflecting admissions’ (August 26, 1985). These are the only two
photographs that explicitly connect with the theme of education in Iran
during the 1980s.
It is useful, here, to apply a commutation test. While ideologically
driven educational reforms are given some space within the coverage of
Iran, any mention of the occurrence of a similar phenomenon in Pakistan
82 S. Durrani
is non-existent within the narrative. This absence assumes further signifi-

cance, when contextual historical analysis of the reforms that were carried
out in Pakistan is taken into consideration here, as documented and dis-
cussed by other scholars and commentators.
During the 1980s, under Zia-ul-Haq’s government, Pakistan experi-
enced two different types of educational reforms. The first reform was
initiated with the help of international aid. During the mid to late
1980s, a USAID-funded project, in conjunction with the University of
Nebraska, Omaha, printed millions of textbooks in the northwestern
city of Peshawar, Pakistan (El-Edroos, 2011). Published in the domi-
nant Afghan languages of Dari and Pashtu, the textbooks were devel-
oped in the early 1980s under an AID grant to the University of
Nebraska, and its Center for Afghanistan Studies. The agency spent $51
million on the university’s education programmes in Afghanistan from
1984 to 1994 (Stephens & Ottaway, 2002). These books, which may be
seen as being similar to the ‘primers of war’ alluded to in the story about
Iranian educational reforms, were distributed in both Pakistan and
Afghanistan (El-Edroos, 2011; Haider, 2011). The curriculum included
books with titles such as ‘The Alphabet of Jihad Literacy’, and used
images of weapons to teach the alphabet (Crilly, 2014). Educationist
Dana Burde notes that while ‘the U.S. program ended with the collapse
of Afghanistan’s communist government, its textbooks have spawned
dozens of copies and revised editions’ (Crilly, 2014). Burde (2014)
argues that this was a ‘flawed approach intended to spur conflict in the
short run, with unpredictable long term consequences’ (p. 6). These
books continued to be used in the 1990s and 2000s; the process to
replace them did not begin till 2002 (Stephens & Ottaway, 2002).
Some media reports note that ‘even the Taliban used the American-
produced books, though the radical movement scratched out human
faces in keeping with its strict fundamentalist code’ (Tharoor, 2014; for
further discussion, see Stephens & Ottaway, 2002; El-Edroos, 2011;
Crilly, 2014; Tharoor, 2014; Burde, 2014).
The second set of reforms was initiated within Pakistan with a view to
reforming the state education system along ideological lines (see
Hoodbhoy & Nayyar, 1985; Nayyar & Salim, 2005, for a detailed dis-
cussion of these reforms). There is extensive debate within Pakistan itself
about the role of these reforms in shaping societal attitudes in ways which
continue to affect the region today (see, for instance, Nayyar & Salim,
2005; Haider, 2011; El-Edroos, 2011; Kureshi, 2016). However, as
Burde (2014) notes, while these educational reforms are remembered and
talked about in the region itself, most Americans are unaware of them
(p. 55).
Keeping in mind Burde’s (2014) observation, this chapter argues that
given that the trend of exclusion for these particular reforms is likely to
be true of other mainstream western news narratives at the time, this
qualifies as an instance of absence which may have given rise to a ‘minus
in origin’ (Bhabha, 1991, p. 160). This effect is created when certain
actors are selectively deprived of agency. As an example, Bhabha (1991)
notes that historical accounts of the Indian Mutiny of 1857 tend to deny
the rebel soldiers the existence of a controlling mind, painting them
instead as irrational, inscrutable individuals. Within a news narrative
then, it is important to identify which actors are attributed agency, and
which ones are not. The image from the data for Iran, which shows the
blood-spattered picture of Khomeini, makes visual connections between
him and the educational and political policies in effect at the time. The
educational policies effected by the Pakistani government, and the dis-
semination of war primers in Pakistan during the 1980s, do not receive
any visual coverage within the sample. The agency of the key actors
involved in the reforms—both Pakistani and foreign—is erased from the
narrative.
It is an absence within the collective consciousness of western news
audiences that has consequences for how the origins of militancy in the
regions are popularly understood. It is worth noting here that with refer-
ence to the data for Pakistan, the specific theme of the role of educational
institutions in fomenting extremism makes an appearance in the sample
for the first time after 9/11. The first photograph to make an explicit con-
nection between education and extremism appears on September 24,
2001. The picture shows a child with a Muslim prayer cap bent over a
book, while the headline and caption read respectively ‘Sacrificial
Warriors’, ‘ABC’s: Studying Islam in religious schools, like this one in
Pakistan, can be as critical to molding extremists as the training at this
Bin Laden camp in Afghanistan’.
84 S. Durrani
This minus-in-origin, with reference to the cultivation of extremist views

within educational materials and matters, is analysed here to raise several
questions for future researchers. First, researchers who look at the coverage
given to institutional reforms conducted by autocratic allies of western gov-
ernments may wish to investigate the question: are the policies of these allies
less likely to be scrutinised? This question can be used to generate research
in the current political context, with other countries and alliances.
Second, is the significance of educational reforms in general an area of
absence within news cycles? With reference to the case studies in question,
it must be acknowledged that the issue received two images only with
reference to Iran as well, and that is not a significant number. The trend
may have something to do with what is considered newsworthy—educa-
tional reforms are perhaps not at the top of the news cycle agenda, given
the lack of immediacy of outcomes. However, it is worth asking this ques-
tion here: with the rise in populism and the emergence of ideology-driven
movements across the world—in both the first and the third worlds—will
we see more ideology-driven reforms to educational curriculums across
the world? Will these reforms—and their possible, unintended genera-
tional consequences—gain any coverage within the news media?
As Cynthia Dunbar, a Christian activist who served on the Texas
Education Board puts it, ‘The philosophy of the classroom in one genera-
tion will be the philosophy of the government in the next’ (Shorto, 2010).
The absent coverage of populist curriculum reforms within the main-
stream news media may produce long-term consequences unrelatable to
causes, as it did with Pakistan. Given the current political climate of the
world, the absence of this particular issue within news discourse may well
influence the shape of the global political system for the next generation.
It is, therefore, an issue that merits further research, in various contexts
and countries across the world.
3.5.3 D
issent: ‘Valuable’ and ‘Non-Valuable’
Dissidents
A second theme that merits mention here is that of dissent, and the
relative difference in the kind of dissidents given coverage in Time, with
reference to Iran and Pakistan. This second absence makes use of the
strategy of exclusion—the absence of people (in this case, certain types of

dissidents) from contexts where they are present in reality (van Leeuwen,
2008). Within the data for Iran, dissidents are generally named individu-
als, depicted in the classic headshot format- photographs, which empha-
sise the story of one individual (Durrani, 2016). With reference to
Pakistan, the key face of dissent is Benazir Bhutto. Other photographs
that depict dissent tend to show unidentified individuals protesting in
groups, sometimes clashing with the police. The specific type of dissent
that is absent from the narrative for Pakistan is problematised with the
help of a commutation test (Lacey, 1998), which involves the replace-
ment of one sign by another. In this particular example, the paradigm
here is freedom of press; the signs substituted are Pakistani and Iranian
journalists respectively. It should be noted here that, as with education,
this is an example of the scarcity/absence binary. Press freedom in Iran is
not an extensively covered topic in Time in the 1980s, but it is covered.
Its visual presence in the narrative is what drew the researcher’s attention
to its absence in Pakistan’s data for the same decade.
The first named dissident within the data for Iran is identified as a
journalist working for Time. Raji Samghabadi (February 16, 1981). The
stand-first of the article, a two-page spread, describes how the ‘Iranians
subject a Time reporter to a mock execution’. The centre of the first page
of the article shows a reconstruction of the event in the form of a paint-
ing. The second page features a mid-shot of a sombre Samghabadi, look-
ing directly at the camera, his arms folded across his chest. The caption
reads, ‘Time’s Raji Samghabadi, safely in New York: ‘I am more saddened
than angered”. It is worth noting here that the fact that there is only one
named journalist dissident in Time’s coverage of Iran in the 1980s consti-
tutes a scarcity verging on an absence. However, in the case of Pakistan,
it is a complete absence.
General Zia’s government is remembered by press historians for the
constraints it placed on the press with the help of various measures. This
includes arrests, public floggings, newspaper closures, censorship, etc. (see,
for instance, Niazi, 1986, and Aziz, 2015, for a more detailed history).
However, violations of press freedom in Pakistan are not documented in
the 1980s. Instead, the visual documentation of dissent seems circum-
scribed to the domain of political opposition activists.8 Pakistan was, at
86 S. Durrani
that time, an ally of the West in the Afghan Jihad. Dissidents provide a
bulwark against authoritarianism, which, in the case of countries like
Pakistan, has consequences that reverberate across decades, and indeed,
geographical boundaries. In terms of holistic, diachronic representational
consequences, exiling these dissidents to the margins of mainstream news
discourse helps create the impression that (1) these countries lack such
democratically aware and active individuals altogether, and are therefore
socially and politically backwards; and (2) the autocratic regime in ques-
tion is a stable entity that is acceptable to the people of that country.
This exclusion leads to certain questions for future research. To borrow
a notion from Butler (1993), are some dissidents more ‘valuable’ than
others for the purpose of news narratives? What makes certain categories
of dissidents more likely to be mentioned and noted within news narra-
tives, as compared to others? Are dissidents less likely to be named and
valued, if they oppose a leader viewed as political ally? Or, to situate the
question within the present political scenario—are the elite western press
less likely to cover violations of press freedom committed by, say, current
autocratic allies in ongoing military conflicts in the Middle East, com-
pared to autocratic adversaries, like China? What are the long-term con-
sequences of these choices for the countries in question?
Journalism and communication research, as a field of enquiry, empha-
sise the textual over the visual. Investigations into absences in news
accounts reflect this bias. This chapter seeks to compensate for this ten-
dency, and extends the notion of absence to the idea of visual grammar,
as expressed by semiotic cues. Moreover, by documenting these diachronic
semiotic changes over the course of three decades, this chapter argues for
a paradigm shift within semiotics itself, a field that tends, as noted by van
Leeuwen (2005), to focus largely on synchronic descriptions. Therefore,
this research seeks to examine the implications of diachronic changes in
visual grammer, and the manner in which they connect with issues of
representation. Which groups of people are consistently allowed, in news
narratives, to look directly at the camera? Can these semiotic absences be
traced with reference to other cues, such as power relations (camera
angle—high, low, etc.), and social distance (camera shot—close-ups,
long shots, etc.)? Given the increasing emphasis on the visual in news
delivery processes in a predominantly online world, it is useful to focus
more research on the role of visual elements in news stories—the ones

that are present, as well as the ones that are absent.
3.6 Conclusion
The news stories the media tells about places and people are structured
not just by what is said, but also by what is not said. The news photo-
graphs that go with these stories provide an aura of authenticity and
veracity to these narratives, and it is, therefore, important to examine
what kind of themes and people merit sufficient significance to be docu-
mented visually, and which ones do not. As the analysis in this chapter
demonstrates, absences in news discourse can be mediated by different
types of concerns: technological restrictions as they exist at a certain point
in time (in this case, for news photography—the gritty black and white
documentary photography of the 1980s may have influenced the way
Iranian women are depicted), news genre and cycle requirements (issues
like education, no matter how critical, lack a sense of immediacy, and are
insufficiently sensational) and geo-strategic political concerns (at the
same point, news discourse appeared to express concern for the use of war
primers in an adversarial nation-state, while simultaneously not attend-
ing to their use in an allied country). Diachronic analysis adds another
dimension to the analysis of absence. It illustrates how the consequences
of an absent event can continue to reverberate across time in the news
discourse, as with the rise of extremism in Pakistan post 9/11, while some
of the historical causes rooted in the 1980s are rendered absent from
popular, collective consciousness. It also demonstrates how absences can
be gradually eroded, as was the case with the averted gaze of Iranian
women across Time.
As the visual grows ascendant in the delivery of news, in a world that
relies increasingly on online platforms for information, and where algo-
rithmic ‘filter bubbles’ (Parser, 2011) structured predominantly by the
absence of opposite viewpoints increasingly govern perceptions, visual
absences are likely to become a significant avenue of research within the
domain of social sciences. It is hoped that the research presented in this
chapter offers useful avenues of discussion for future researchers.
88 S. Durrani
Notes
1. It is relevant to note here that Time publishes four different editions, for
the U.S, Asia, Europe and the Middle East, and the South Pacific region.
The editorial philosophy remains essentially the same across all editions,
and all four editions share the same content with some variations, usually
with reference to cover stories. This study focuses on the Asia edition,
since the two countries included in the study are given extensive coverage
within it.
2. See www.dawn.com/news/731670/timeline-history-of-us-pakistan-relations
3. It is worth noting here that existing research literature on the news cover-
age given to Pakistan and Iran tends to focus on American media outlets.
This may be due to two reason. First, American media outlets such as The
New York Times and Time magazine have global audiences, and are there-
fore seen as influential agenda setters. Second, given the complicated his-
tory of political ties between the United States, with Pakistan and Iran
respectively, researchers from both countries are, perhaps, more likely to
focus on publications emanating from a global superpower that has con-
siderable influence over the international community.
4. It is noted here in the interests of full disclosure that a certain issues were
missing from the two archives from which this data was collected (4.68%
of the total sample). However, given the sheer volume of the data, this is
a negligible percentage.
5. For an exploration of the data trends for Composition, see Durrani and
Caple (2018).
6. It should also be noted here that with reference to quantitative coding, the
category of ‘gender’ was coded into the following subcategories: Male,
Female, Both (a photo that shows both men and women), and Unclear
(where the photo is too grainy to allow a determination of gender). The
subset analysed here refers to the categories of Male and Female, and more
specifically, only those males and females identified in the verbal context
unit as Iranian, since the sample does contain stories about Iran that con-
tains photographs of other nationalities (e.g. American).
7. It is useful here to look at real numbers, not just percentages, because the
numbers for photographs that feature direct eye contact for both genders,
with reference to Pakistan, is somewhat higher: men (18) and women (9).
Also, while the percentage of Pakistani women who make direct eye contact
with camera is higher, the number is clearly lower. For Iran, the numbers
are fewer: men (10) and women (4). Iran does, however, receive less cover-
age than Pakistan in overall terms, so this pattern is consistent with overall
trends.
8. It is interesting to note here that the first visual instance of press freedom
violations in Pakistan appears in May 1999 in the sample. This is when
Pakistan was under democratic rule, with Prime Minister Nawaz Sharif at
the helm.
References
Al-Hejin, B. (2015). Covering Muslim women: Semantic macrostructures in
BBC News. Discourse and Communication, 9(1), 19–46.
Anholt, S. (2004). Editor’s foreword to the first issue. Place Branding and Public
Diplomacy, 28(3), 253–265.
Aziz, S. (2015). A leaf from history: Four journalists flogged, two newspapers
shut. Dawn. Retrieved from http://www.dawn.com/news/1179170/newspa-
per/newspaper/newspaper/column
Becker, L. (1977). Foreign policy and press performance. Journalism Quarterly,
54(2), 364–368.
Beeman, W. (2005). The ‘Great Satan’ vs. ‘The Mad Mullahs’: How the United
States and Iran demonize each other. Westport: Praeger.
Bhabha, H. (1991). The location of culture. New York: Routledge.
Burde, D. (2014). Schools for conflict or for peace in Afghanistan. New York:
Columbia University Press.
Butler, J. (1993). Bodies that matter. London: Routledge.
Caple, H. (2013). Photojournalism: A social semiotic approach. Basingstoke:
Palgrave Macmillan.
Chan-Malik, S. (2011). Chadors, feminists, terror: The racial politics of U.S.
media representations of the 1979 Iranian women’s movement. The Annals of
the American Academy of Political and Social Science, 637, 112–140.
Crilly, R. (2014). Infidels are our enemy: Afghan fighters cherish old American
schoolbooks. Aljazeera America. Retrieved from http://america.aljazeera.
com/articles/2014/12/7/afghan-fighters-americantextbooks.html
Delgado, F. (2003). The fusing of sports and politics: Media constructions of U.S.
versus Iran at France ‘98’. Journal of Sports and Social Issues, 27(3), 293–307.
Durrani, S. (2016). Representing nations through visual narratives: Pakistan and
Iran in Time magazine. Unpublished doctoral thesis, University of New
South Wales, Australia.
90 S. Durrani
Durrani, S., & Caple, H. (2018). Stories across time: Exploring internal and
external cohesion in the visual representation of Iran and Pakistan in Time
Magazine. Manuscript submitted for publication.
Durrani, S., & Mughees, S. (2010). The pictorial image of Pakistan in Newsweek
and Time: Pre and post 9/11. Journal of the Research Society of Pakistan, 47(1),
1–31.
Fahmy, S. (2004). Picturing Afghan women: A content analysis of APWire
Photographs during the Taliban regime and after the fall of the Taliban
regime. The International Journal for Communication Studies, 66(2), 91–112.
Fahmy, S., & Kim, D. (2008). Picturing the Iraq War: Constructing the image
of war in British and U.S. press. International Communication Gazette, 70(6),
443–462.
Economou, D. (2010). Having it both ways? Images and text face off in the
broadsheet feature story’. In V. Rupar (Ed.), Journalism and meaning making:
Reading the newspaper (pp. 175–193). Cresskill, NJ: Hampton Press.
El-Edroos, S. (2011). Learn to be Taliban: K is for Kalashnikov. Express Tribune.
Retrieved from http://blogs.tribune.com.pk/story/4877/learn-to-betaliban-
k-is-for-kalashinkov/
Gertner, D., & Kotler, P. (2004). How can a place correct a negative image?
Place Branding and Public Diplomacy, 1(1), 50–57.
Gilboa, E. (2002). Global communication and foreign policy. Journal of
Communication, 52(4), 731–748.
Griffin, M. (2004). Picturing America’s ‘War on Terrorism’ in Afghanistan and
Iraq: Photographic motifs as news frames. Journalism: Theory, Practice and
Criticism, 5(4), 381–402.
Haider, Z. (2011). Ideologically adrift. In M. Lodhi (Ed.), Pakistan: Beyond the
crisis state (pp. 113–130). London: Hurst and Company.
Hanif, M. (2011). What impact did 9/11 have on the world? Pakistanis realise
that the trauma is far from over. The Guardian. Retrieved from http://www.
theguardian.com/world/2010/oct/30/pakistan-floods-aftermath
Hevisi, D. (1988). Mohammad Zia-ul-Haq: Unbending commander for era of
atom and Islam. The New York Times. Retrieved from http://www.nytimes.
com/1988/08/18/world/mohammad-zia-ul-haq-unbendingcommander-for-
era-of-atom-andislam.html?pagewanted=all
Hoodbhoy, P., & Nayyar, A. (1985). Rewriting the history of Pakistan. In
A. Khan (Ed.), Islam, politics and the state: The Pakistan experience
(pp. 164–177). London: Zed Books.
Huang, Y., & Fahmy, S. (2011). Same events, two stories: Comparing the pho-
tographic coverage of the 2008 anti-China/Olympics demonstrations in
Chinese and US newspapers. The International Communication Gazette,

73(8), 732–752.
Izadi, F., & Saghaye-Biria, H. (2007). A discourse analysis of elite American
newspaper editorials: The case of Iran’s nuclear program. Journal of
Communication Inquiry, 31(2), 140–165.
Jalal, A. (2011). The past as present. In M. Lodhi (Ed.), Pakistan: Beyond the
crisis state (pp. 1–20). London: Hurst and Company.
Kafala, T. (2002). Analysis: Iran and the axis of evil. BBC. Retrieved from http://
news.bbc.co.uk/2/hi/middleeast/1814659.stm
Khan, A. (2002). Content analysis of pictorial coverage of Muslim world in
Newsweek and time. Unpublished master’s thesis, University of Punjab.
Khan, M., & Irtaza, M. (2010). Post 9/11 scenario and image of Pakistan as
portrayed by top US newspapers. Journal of Media Studies, 25(1), 5–20.
Khiabanay, G. (2008). The Iranian press, state and civil society. In M. Semati
(Ed.), Media, culture and society in Iran: Living with globalization and the
Islamic State (pp. 17–35). New York: Routledge.
King, C., & Lester, P. (2005). Photographic coverage during the Persian Gulf
and Iraqi wars in three U.S. newspapers. Journalism and Mass Communication
Quarterly, 82(3), 623–637.
Kress, G., & van Leeuwen, T. (2006). Reading images: The Grammar of visual
design. London: Routledge.
Kureshi, S. (2016). Socio-cultural re-engineering. Daily Times. Retrieved from
www.dailytimes.com.pk/opinion/27/2016/socio-cultural-re-engineering
Lacey, N. (1998). Image and representation: Key concepts in media studies.
Basingstoke: Palgrave Macmillan.
McPhail, T. (2010). Global communication: Theories, stakeholders and trends.
West Sussex: Wiley Blackwell.
Mortensen, M. (2011). When citizen photojournalism sets the news agenda:
Neda Agha-Soltan as a Web 2.0 icon of post-election unrest in Iran. Global
Media and Communication, 7(1), 4–16.
Mughees, S. (1995). Image of Pakistan in the New York Times (1980–1990).
PakistanDefence, 7(1), 76–98.
Mughees, S. (1997). Editorial treatment of U.S Foreign policy in the New York
Times: The case of Pakistan (1980–90). National Development and Security,
6(2), 33–63.
Murtaza, N. (2015). A post Zia Pakistan? Dawn. Retrieved from http://www.
dawn.com/news/1200668
Naficy, H. (2008). Iranian Emigre cinema as a component of Iranian National
cinema. In M. Semati (Ed.), Media, culture and society in Iran: Living with
globalization and the Islamic state (pp. 167–191). New York: Routledge.
92 S. Durrani
Nasir, A. (2012). Zia’s long shadow. Dawn. Retrieved from http://www.dawn.

com/news/732329
Nayyar, A., & Salim, A. (2005). The Subtle subversion. Sustainable Development
Policy Institute.
Niazi, Z. (1986). The Press in chains. Karachi: Royal Book Company.
Parser, E. (2011). Beware online filter bubbles. [Video file]. Retrieved from
https://www.youtube.com/watch?v=B8ofWFx525s
Parry, K. (2010). A visual framing analysis of British press photography during
the 2006 Israel–Lebanon conflict. Media, War and Conflict, 3, 67–85.
Parry, K. (2011). Images of liberation? Visual framing, humanitarianism and
British press photography during the 2003 Iraq invasion. Media, Culture and
Society, 33(8), 1185–1201.
Rehman, I. (1988). Zia-ul-Haq: Master of illusion. Dawn. Retrieved from
http://herald.dawn.com/news/1153499
Roushanzamir, E. (2004). Chimera veil of “Iranian woman” and processes of
U.S. textual commodification: How U.S. print media represent Iran. Journal
of Communication Inquiry, 28(1), 9–28.
Sarantakos, S. (1998). Social research. London: Macmillan Press.
Semati, M. (2008). Living with globalization and the Islamic state: An introduc-
tion to media, culture and society in Iran. In M. Semati (Ed.), Media, culture
and society in Iran: Living with globalization and the Islamic State (pp. 1–13).
Shorto, R. (2010). How Christian were the founders? The New York Times
Magazine. Retrieved from www.nytimes.com/2010/02/14/magazine/ 14text-
books-t.html
Stephens, J., & Ottaway, D. (2002). From U.S., the ABC’s of Jihad. The
Washington Post. Retrieved from https://www.washingtonpost.com/archive/
politics/2002/03/23/from-us-the-abcs-of-jihad/d079075a-3ed3-4030-9a96-
0d48f6355e54/?utm_term=.57e529aba384
Tabaar, M. (2014). Strategic anti-Americanism in Iran from the hostage crisis to
nuclear talks. The Washington Post. Retrieved May 25, 2015, from http://
www.washingtonpost.com/blogs/monkey-cage/wp/2014/11/04/strategi-
canti-americanism-in-iran-from-the-hostage-crisis-to-nuclear-talks/
Tadayon, M. (1980). The image of Iran in the New York Times. International
Communication Gazette, 26(4), 217–233.
Talbot, I. (1998). Pakistan: A modern history. London: Hurst and Company.
Tharoor, I. (2014). The Taliban indoctrinates kids with Jihadist textbooks paid
for by the U.S. The Washington Post. Retrieved from https://www.washing-
tonpost.com/news/worldviews/wp/2014/12/08/the-taliban-indoctrinates-
kids-with-jihadist-textbooks-paid-for-by-the-u-s/?utm_term=.abad
1278a9b0
van Leeuwen, T. (2005). Introducing social semiotics. London: Routledge.
van Leeuwen, T. (2008). Representing social actors. In T. van Leeuwen (Ed.),
Discourse and practice. New York: Routledge. https://doi.org/10.1093/acprof
:oso/9780195323306.001.0001
Vliegenthart, R., & Walgrave, S. (2008). The contingency of intermedia agenda
setting: A longitudinal study in Belgium. Journalism and Mass Communication
Quarterly, 85(4), 860–877.
Wimmer, R., & Dominick, J. (2003). Mass media research: An introduction.
Belmont: Wadsworth/Thomas Learning.
4
Intimations of ‘Spring’? What Got Said
and What Didn’t Get Said about
the Start of the Middle Eastern/North
African Uprisings: A Corpus-assisted
Discourse Study of a Historical Event
Alan Partington
4.1 Introduction
4.1.1 Absence/s from a Corpus Linguistic Perspective
In a previous paper (Partington, 2014), written partly in response to the

kind of pessimism in critical discourse analysis about the possibility of
studying discourse absence/s as typically expressed by Blommaert (2005)
(see Chap. 1 in this volume), and partly because of the paucity of corpus
linguistics research into the topic at that time, I set out to break down the
notion of absence/s into a theoretical—and, from a corpus linguist’s view-
point, practical—categorisation as follows:
A. Partington (*)
Bologna University, Bologna, Italy

96 A. Partington
(i) ‘known’—or suspected, or ‘searchable’—‘absence’. You already

know which linguistic feature you are searching for and simply want
to know whether or not it is in the corpus.
(ii) ‘unknown absence’, an absence stumbled upon serendipitously in
the course of piece of research.
(iii) relative absence and absolute absence.
(iv) absence from a sizeable corpus, which may raise questions about the
representativeness of the corpus.
(v) absence from a limited set of texts, including from a specific portion
of a corpus.
(vi) absence from a position in a single text, including from a location in
a phrase.
(vii) absence defined as ‘hidden from open view’, that is, hidden

meaning.
These are not meant to be necessarily mutually exclusive categories, for

instance, one might be studying absolute and relative absences (type iii)
from a specific portion of a corpus (type v), say, in the utterances of one
particular type of speaker.
Various considerations arose in the course of this research. Since most
things are absent from most places most of the time, we need to decide
the parameters of relevant or salient or meaningful absence/s, that is, those
which:
(i) if somehow suspected, are worth searching for, or

(ii) if stumbled upon, are worthy of further investigation.
One indication could be unexpectedness, that is, discovering absence

when a presence is expected. This, however, raises the question of expected by
whom and why, especially since researchers have their own unique past prim-
ings (Hoey, 2005), which influence expectations in the present. And then,
when an absence is discovered, how does one decide whether the absence is
intentional or otherwise, especially given that, as already stated, absence is
usually quite normal? Far too often in critical discourse analysis (or CDA), it
is taken for granted that a silence or absent message or voice must have been
deliberately suppressed with little tangible evidence of intentionality. Finally,
Intimations of ‘Spring’? What Got Said and What Didn’t Get... 97
once an absence is adjudged relevant and worthy of investigation, do we

attempt to explain it? If so, what kinds of explanations are valid and interest-
ing? Which are trivial and which non-trivial, that is, are themselves non-
obvious and unexpected?
4.1.2 M
odern-diachronic Corpus-assisted Discourse
Analysis
One main reason for the current increase in interest in absence in corpus
linguistics research (Duguid & Partington, in press; Taylor, 2012, 2013),
is the recent arrival on the scene of systematically designed diachronic
corpora, which make it possible to compare and contrast both the lin-
guistic features of, and the social and political issues present or absent in,
specific discourse types of a language over different periods of time. The
Corpus of Historical American English (COHA, Davies, 2012) which
contains more than 400 million words of text from the 1810s–2000s,
appeared on-line in 2010. In the same year, in a set of papers using the
SiBol (Siena-Bologna) suite of UK broadsheet newspaper corpora from
1993 to the present, the term ‘modern diachronic corpus-assisted dis-
course studies’ (MD-CADS) was coined to describe research on corpora
from different recent time periods (Partington, 2010). UCREL (Lancaster
University) is producing versions of both the Brown family of corpora
and of the British National Corpus containing more recent texts, part of
whose raison d’être is for comparison purposes with the older corpora
(Baker, 2009).
This chapter then contains a related set of a modern diachronic corpus-
assisted discourse studies of a particular form of presence and absence,
namely, that of the mention and failure to mention in political and media
news reporting of messages relating to a particular historical event, spe-
cifically, the outbreak of protest in the Arab Middle East and North
Africa (henceforth MENA) at the beginning of 2011. The analyses will
take a ‘before and after’ contrastive approach, comparing how particular
political and media sources talked about events and people involved in
the period before the start of the protests and the period following the
outbreak of the protests.
98 A. Partington
4.1.3 The Aims of the Current Research
The overarching preliminary research aim is to discover if the discussions

changed, if so, how, and, in particular, which messages were present or absent,
appeared or disappeared, completely or in relative terms, before and after the
outbreak of the uprisings. Various sets of data are examined in order to exam-
ine three subsidiary research aims contributing to this main one.
More specifically, then, the original aims of the project described in
this chapter were the following:
1. To track the evolution of discussions of the events in the MENA

(Middle East and North Africa) area in White House press briefings,
in the 12 months from December 2010 to November 2011.
2. To see if characteristics of this evolution were reflected in the televised
press in the form of CNN news reports in 2011.
3. To analyse the representation of the Arab MENA in the UK broad-
sheet press in 2010, the year leading up to the uprisings, in particular,
to check whether there is any awareness, any inkling of what is about
to happen. The newspapers chosen were the left-leaning Guardian and
the right-leaning Daily Telegraph, hereafter the Telegraph, in case there
was a political aspect to these questions.
From a methodological perspective, it was intended as research using a

variety of corpora each containing various political and media institu-
tional voices with the aim of studying the interplay between them.
During the course of the study, however, I found myself uncovering mes-
sages from these voices that were somehow missing, that were not explic-
itly present in the data sets, and that this fact was as interesting as studying
those which were actually present. It appeared that the contrastive capa-
bilities, which corpus research facilitates, make it a good way of uncover-
ing and quantifying absences in political and media discourses (Marchi
& Taylor, 2009, p. 222). While it may be true that, in the past:
[…] corpus-based analysis tends to focus on what has been explicitly written,
rather than what could have been written but was not or what is implied,
inferred, insinuated or latently hinted at. (Baker et al., 2008, p. 296)
this certainly does not need to be the case.

More generally, MD-CADS investigations are conducted within the
methodological framework of corpus-assisted discourse studies (CADS).
These types of research are eclectic and pragmatic in the techniques they
adopt since they are goal-driven, that is, the aims of the research dictate
the methodology. Nevertheless corpus-assisted discourse studies tend to
display a number of common characteristics and tendencies (Partington,
Duguid, & Taylor, 2013). These include, first of all, the emphasis on
comparison among data sets which entails the use of multiple corpora,
sub-corpora and corpus divisions and the frequent need to compile
‘bespoke’ (or ‘ad hoc’) corpora, including concordance corpora (see Sect.
4.2). They also include the often complex interaction of statistical analy-
sis and close reading of extended stretches of text. And finally they include
the combination of corpus-generated observation with data from other
external sources. In other words, CADS—to resurrect Sinclair’s (2004)
famous dictum—‘trusts the text’, but it also exploits the corpus texts, in
particular by encouraging the serendipitous opening of new avenues of
research—following up leads the corpus data supplies, especially when
that data includes ‘outliers’, that is, unexpected findings. It also enhances
the corpus texts, through triangulation, that is, approaching the corpus
data from different directions, including those suggested by external
sources of information on the topic in hand (which may well include
non-corpus sources of data, for example, publications, reports, surveys
and interviews).
4.2 The Corpora and Methods Employed

Two corpora were used in tracking the White House press briefings dis-
cussions of the early stages of the MENA uprisings. The first, named
WH-Obama, consists of all briefings held in the 12 months from 1st
December 2010, the month of elections in Egypt and protests in Tunisia
to 30th November 2011 and consists of 1,300,000 words. For any
potential comparison, I was already in possession of a collection of brief-
ings from an earlier period, during the previous administration, entitled
WH-Bush, consisting of 3,400,000 words.1
100 A. Partington
The CNN corpus consists of all news reports from the year 2011 and
contains 61,000,000 words (but reports are often repeated, which were
not removed).2
The two main newspaper corpora are the Guardian 2010 comprising
41,600,000 words and Telegraph 2010 comprising 48,700,000 words.3
Both consist of all the articles published by each newspaper in 2010.
Reference will also be made to a publication of the Guardian’s own selec-
tion of 2011 reports on the uprisings (Manhire, 2012) and finally a brief
comparative reference will be made to Guardian 2013 (32,300,000
words) and Telegraph 2013 (37,000,000 words), parallel corpora to
Guardian 2010 and Telegraph 2010 but from the year 2013 (Table 4.1).
This configuration of corpora permitted various methodologies to be
performed. Most obviously it allowed, if and when convenient, various
key item comparison/contrasts, for instance, WH-Obama vs WH-Bush,
WH-Obama vs CNN and Guardian 2010 vs Telegraph 2010. By ‘key
item’ here is meant items found, usually by software, to be appreciably
more frequent in one text or set of texts than another. Less obviously, it
was also possible to subdivide each corpus into time periods in order to
track developments, changes, appearance and disappearance of messages
over very short time periods. For instance, it was possible to subdivide
WH-Obama month by month to see if the MENA discussions altered
over the period. I also found it useful to subdivide the newspaper corpora
into two subsections, the first containing the articles from the first nine
months of 2010, the second those from the final three months, with the
aim of ascertaining whether there was an increase in the latter part of the
year of mentions of MENA-related events and actors. This would be a
Table 4.1 The main corpora employed in this study

WH briefings
WH-Obama, Dec. 2010–Nov. 2011 1,300,000 words
WH-Bush 3,400,000 words
CNN news reports 2011 61,000,000 words
UK press
Guardian 2010 41,600,000 words
Telegraph 2010 48,700,000 words
Guardian 2013 32,300,000 words
Telegraph 2013 37,000,000 words
strong indication of an inkling, a suspicion, on the part of these news

sources that something unusual was about to happen.
Another technique employed was the concordance-corpus (Marchi,
2010; Taylor, 2010), a way of creating a corpus which is highly focussed
on a particular lexicalised topic. The terms Middle East* and North Africa*
were concordanced in each newspaper, stipulating 500 characters of co-
text for each occurrence. The resulting concordances could thencefor-
ward be employed and analysed as corpora in their own right.
Cluster or n-gram study was also useful especially regarding the White
House briefing discourse. The reason that press briefings were instituted
in the first place by the US administration was to propose the administra-
tion’s favoured view of events to the press and through the press to the
public. To this end, the podium’s discourse is often packed with repeated
phrases, sometimes with minor variation. The podium’s discourse thus
provides many clear examples of what Duguid (2007) has named ‘forced
priming’, that is, the process whereby speakers or authors frequently
repeat a certain form of words to deliberately ‘flood’ the discourse with
messages for a particular strategic purpose. This forced priming was first
noted in earlier research while both watching briefings, broadcast by
C-Span public service TV and by reading a good number of transcripts
(Partington, 2003). To uncover what repeated messages are contained in
this data set, lists of clusters, 4, 5, 6 and 7 items in length, were prepared
using the WordSmith (Scott, 2008) Wordlist Tool. Individual items
which reoccurred in these clusters could then be concordanced in the
hope of throwing light on the nature of messages being launched.
However, and as is often the case in CADS, the most frequent tech-
niques were that of, first of all, to borrow a term from Halliday (1994),
shunting back and forth between key items and close reading of both con-
cordances and sections of the original texts and, second, of ‘serendipitous
concordancing’, that is, stepping from one concordance to another to
follow up potentially interesting lexical-item observations. As an example
of the latter, from a concordance of Mubarak (the Egyptian President at
the outbreak of the uprisings) in WH-Obama, in 250 characters of co-
text, the phrase was noted: ‘we condemn the use of violence on both
sides’, uttered by the podium. This led me to concordance violence in the
co-text of all, both, either sides, and in the resulting concordance lines I
102 A. Partington
noticed the podium’s opinion that ‘all sides need to show restraint’. This
led to a further concordance of restraint with the items urge and show
within a three-word span. This yielded altogether 18 results, all of them
contained in the podium’s turns. The countries where, according to the
White House podium, both sides in a conflict—that is, the government/
regime and its opponents—need to refrain from violence and exercise
restraint are principally Bahrain (7 occurrences) and Yemen (5).4 The mes-
sage is also used three times about Egypt but just once about Syria. It is
completely absent in discourses about Libya, where only one side, the
administration, is portrayed as perpetrating violence, as the following
typical statement implies:
(1) MR CARNEY: Well, let me just say that the President strongly con-
demns […] the bloodshed perpetrated by the Libyan government in Libya.
(23/02/2011)
This is an indication (admittedly limited given the small sample) that

the administration is unwilling to take sides against the strategically
‘friendly’ governments of Bahrain and Yemen, but exhibits no such
qualms about the less pro-Western rulers in Libya and Syria. It is also the
first indication of how corpus techniques can reveal the absence as well as
presence of a particular political message and would constitute an exam-
ple of type (ii) serendipitous discovery of a previously unsuspected
absence. It can also tell us where the absence is located, in this case, in the
discussions on Libya and Syria. And, finally, although it has been claimed
that
A traditional corpus-based analysis is not sufficient to explain or interpret

the reasons why certain linguistic patterns were found (or not found).
(Baker et al., 2008, p. 293; my emphasis)
it would seem that, given a modicum of knowledge of the political

world on the part of the analyst, corpus-assisted analysis, in this particu-
lar case by allowing us to contrast what is said about various MENA
governments, can supply us with data to help us infer an interpretation
or explanation of why a certain message—support for the Libyan and
Syrian governments—might be absent from a discourse.
4.3 White House Press Briefings and MENA

4.3.1 Naming Choices
In order to study how the Arab uprisings were debated in the White
House briefings, one obvious early step was to concordance the names of
some of the countries involved, namely, Libya/Libyan(s), Syria/Syrian(s)
and Egypt/Egyptian(s), along with the names of the countries’ leaders,
Qaddafi, Assad and Mubarak. Since each briefing is contained in a sepa-
rate file named by date it is possible to track which topics are discussed
month by month or even day by day.
The first finding was that none of the MENA countries (Egypt*, Tunis*,
Libya*, Syria*, Bahrain, Yemen were all concordanced) are mentioned in
December 2010 or indeed until the actual occupation of Tahrir Square in
Cairo in late January 2011, with the exception of a solitary but fore-
sighted question on the wisdom of re-opening the US Embassy in
Damascus. We can infer quite strongly then an absence of any particular
attention or interest among the White House press corps regarding events
in the MENA region before the Egyptian protests actually occurred. This
is in spite of the fact that street protests began in Tunisia in December
and national elections were held in Egypt that month. Given how repre-
sentative the WH press corps is of the US and the Western world’s broad-
cast and printed press (see Sect. 4.4.1) there is strong evidence that there
was little inkling in the Western media of what was about to occur
(Fig. 4.1).
We can turn our attention to the two countries where the most intense
violence eventually occurred, that is, Libya and Syria. In January 2011,
there are still no mentions of Libya or Libyan. In February, both the
podium and press are comfortable in discussing the Libyan government,
which is mentioned 32 times, but in March only nine times, and after
that never at all (except a couple of times in the context of freezing Libyan
government assets). In the same February, there are six mentions of the
Libyan regime and six of the Qaddafi regime. By March 2011, however,
regime is used a total of 58 times, 37 co-occurring with Qaddafi and 21
with Libyan. In the final six months of the year, we find only Qaddafi
with regime and never Libyan. The evaluatively neutral Libyan government
104 A. Partington
Fig. 4.1 The White House press room seating chart
has rapidly been replaced in briefings discourse with the negative Qaddafi
regime in what appears to be a deliberate priming shift to create diplo-
matic distance between the White House and the Libyan administration.
Perhaps the most interesting aspect is that Libyan government becomes
absent from the journalists’ speech almost as quickly as from the podi-
um’s. They appear to acquiesce to the White House’s message and evalu-
ations on the issue of Libya (Fig. 4.2).
There is a similar process of diplomatic distancing regarding Syria in
WH-Obama, but the process is slower and not complete. In the first six
40
35
30
25
Libyan government
20
Libyan regime
15 Qaddafi regime
10
5
0
January February March
Fig. 4.2 How the Libyan administration is referred to by the podium in the first
three months of White House press briefings in 2011
45
40
35
30
25 Syrian government
20
Syrian regime
15
Assad regime
10
5
0
Dec-Feb March- June- Sept-Nov
May Aug
Fig. 4.3 How the Syrian administration is referred to by the podium in the first
three months of White House press briefings in 2011
months of 2011, we find 48 occurrences of Syrian government and only

three of Syrian regime. In the second six months, there are 34 references
to Syrian regime but it is still called government 18 times, all by the
podium. However, the main reason why the item regime is absent from
the press language is because the questions tend to contain abbreviations
such as On Syria […], where government or regime is not explicitly speci-
fied (Fig. 4.3).
As a comparison, in WH-Bush, we find seven mentions of the
Libya* + government and no mentions of either Libya* or Qaddafi + regime.
There are four mentions of Syria* + government and none of regime,
106 A. Partington
although one mention is of the terrorist government of Syria. Overall,

these figures seem to demonstrate an absence of interest in these coun-
tries on the part of the press corps during the Bush period.
The combination of cluster analysis and concordancing mentioned
above (Sect. 4.2), showed how the WH-Obama administration’s mes-
sages were very different according to the country and its leader.
Regarding Egypt, we find rather sedate, diplomatic and formulaic
expressions such as call* for a(n) (orderly) transition (67 occurrences) and
call* for free and fair elections (39). The terms Mubarak regime or Mubarak
… dictator are never used by the White House podium (in other words,
they are absent from the podium’s discourse), not even when he is asked
the direct question:
(2) Q: Do you think Mubarak is a dictator?
Realising this phrasing might afford the podium some room for foot-
ing evasion (for instance, ‘it’s not important what I think’), the journalist
switches the target or recipient (Partington, 2003, p. 51) of the question
to the President:
(3) Q: More importantly does the President think Mubarak is a dictator?
But they are again not obliged with a straight answer, and Mubarak
still deserves the title President:
(4) MR GIBBS: The administration believes that President Mubarak has a

chance to show the world exactly who he is by beginning this transition
which is so desperately needed in his country and for his people now.
(02/02/2011)
Assad is initially requested to change course and cease the violence, but
after August, when violence in Syria became more intense, he is asked to
step aside/step down (22), while the US policy is to put/enhance/ratchet up
the pressure on the Syrian regime (32).
The messages to Gadaffi are much more pressing. We find the demand
that he must remove himself from power (10) and, should he fail to be so
obliging, he must be remove* from power (13). These more forceful phrase-
ologies are never used of either Mubarak or Assad in 2011, another inter-
esting absence. Nor does the White House make any pretence of hiding
its realpolitik. The template each/every country + different occurs 40 times,
for example:
(5) MR. CARNEY: Well, Dan, as we’ve said, each country that has been
affected by this unrest is different. Each country in the region is different.
Each country has different traditions, political systems and relationships
with the United States and other countries around the world. (24/02/2011)
4.3.2 CNN
The main hypothesis I wished to test on the 2011 CNN broadcast news
reports was whether the abrupt change of appellation from government to
regime also occurred in this data set, that is, whether either term was or
became relatively absent over this time period, and so the same procedure
was followed, namely, concordancing the names of the countries involved,
that is, Libya/Libyan(s) and Syria/Syrian(s), along with the names of the
countries’ leaders, Qaddafi and Assad. Figure 4.4 shows the proportions
of the use of Libyan government, Libyan regime and Gadhafi regime over
the same time period we observed for the briefings (note that the name of
the Libyan leader is spelled in different ways in the briefings, CNN and
the UK press).
300
250
200
Libyan government
150
Libyan regime
100 Gadhafi regime
50
0
February March April
Fig. 4.4 CNN Libyan government/regime/Gadhafi regime

108 A. Partington
What is apparent is that there is a fairly even numerical balance between

the choice of government or regime, and that the proportions remain very
similar in each month, in contrast to the briefings discourse, where the
use of regime took over very rapidly.
As regards Syria, Fig. 4.5 shows that, although Syrian government
remains the most popular of the three possibilities throughout the year,
the combined use of regime—Syrian regime and Assad regime—eventu-
ally comes to outnumber government (see Fig. 4.6). It also shows a marked
falling away of interest towards the end of the year (pressing internal
financial matters dominate the briefings during this later period), a phe-
nomenon we can relate to relative absence, type (iii) in the categorisation
given in Sect. 4.1.
300
250
200
Syrian government
150
Syrian regime
100 Assad regime
50
0
1st Qtr 2nd Qtr 3rd Qtr 4th Qtr
Fig. 4.5 CNN Syrian/Assad government/regime
300
250
200
150 Government
Regime
100
50
0
1st Qtr 2nd Qtr 3rd Qtr 4th Qtr
Fig. 4.6 CNN Syria: Government or regime

As regards the clusters we noted in the briefings, concerning Gaddafi,

we find the expression remov* from power (161 occurrences) but never the
somewhat odd remov* himself from power we saw above (Sect. 3.3.1) in
the press briefings. However, the expression remov* from power is also
used on 42 occasions regarding Mubarak, a message which, as we saw,
was absent from the briefings. The CNN journalists evidently feel less
obliged to be delicate about the Egyptian leader than the White House,
for whom he was a crucial geopolitical ally.
4.3.3 The UK Newspapers
4.3.3.1 The Guardian’s MENA Representations in 2011
As stated earlier (Sect. 4.1.3), whenever relevant, CADS type research

also makes use of non-corpus sources of information. In 2012, the
Guardian published a selection of its own articles from 2011 in a book,
tracing in chronological order its own reporting of the uprisings. The
book was entitled The Arab Spring: Rebellion, Revolution and a New World
Order (Manhire, 2012).
There are four articles in the selection dedicated to Egypt. Mubarak is
mentioned four times, twice as Hosni Mubarak but never as President. He
is referred to as tyrant and his government as a gang and a regime, which
is criticised as follows:
(6) A tyrannical regime might deprive the people of their freedom, but in
return they are offered an easy life. A democratic regime might fail to beat
poverty, but the people enjoy freedom and dignity. The Egyptian regime
has deprived the people of everything including freedom and dignity, and
has failed to supply them with their daily needs. (Guardian 28/01/2011)
There are five articles in the Guardian’s selections dedicated to Libya

and Gaddafi’s fall from grace is quick and complete. He is referred to as
Colonel Gaddafi just once and never as the Libyan leader; on all other
occasions he is simply Gaddafi. He is referred to as the great dictator, the
Libyan despot and even plonker and reference is made to his supposed
madness. The Libyan administration is referred to exclusively as the
110 A. Partington
Gaddafi regime (5 occurrences), never as the Libyan government or the

Libyan regime. There are discussions in three of the articles of ‘Gaddafi’s
notorious Tripoli prisons’ (Guardian 31/10/2011) and the torture which
was practised there.
There are four articles in this selection dedicated to Syria which appear
to show that the Guardian’s attitude to Assad and the Syrian administra-
tion is more complex and changes more slowly. The administration is
referred to by Guardian columnists as regime 13 times, including murder-
ous regime and Assad’s Mafia-like regime, as well as once as a vile dictator-
ship. However, in contrast to Libya, it is still referred to by Guardian
journalists as government 11 times, along with another three occasions
where the term is attributed to external voices (twice to a Russian and
once to a CIA voice). In March, Assad is still seen with some sympathy
and good humour: ‘Bashar-al-Assad doesn’t really look like an Arab presi-
dent. Or a dictator come to that […] Seeing him reminds me of some
gangly scoutmaster’ (Brian Whitaker, Guardian 31/ 2011), whereas later
in the year he turns into ‘a chuckling and snorting Bashar-al-Assad [who]
tried to deny any responsibility for the attacks on his own people’ (Mehdi
Hasan, Guardian, 12/12/ 2011). The regime’s murder, torture and muti-
lation of prisoners are also highlighted in one of the selected articles
(Guardian, 20/09/2012).
Since the articles were selected by a Guardian journalist and the book
was published by Guardian books, we can infer that the newspaper
wishes, in 2012, to portray itself as highly critical of the leaders and
governments of these three MENA countries. But how were they depicted
by this newspaper and the Telegraph in the year before the uprisings, that
is, in 2010? Was there an equally assiduous attention to and an equally
negative evaluation of the MENA’s rulers’ misdeeds?
4.3.3.2 The Guardian’s and Telegraph’s MENA Reporting

in 2010
In this section we examine data from the two corpora Guardian 2010 and
Telegraph 2010 (see Sect. 4.2). Having found a large number of negative
evaluations of the relevant governments and leaders in the Guardian’s
2012 post-uprisings book (relating to 2011 articles), in this section I will

investigate whether we can find any such negative evaluations in various
press sources before the uprisings occurred. This corresponds to a type (i)
absence research, that is looking for ‘known’—or suspected, or ‘search-
able’—‘absence’, when you already know which linguistic feature—or, in
this case, political issues they may relate to—you are searching for and
want to find out whether or not it is in the corpus. In the present case,
then, we are searching for such negative evaluations in the two 2010
corpora.
The first thing to note in the two newspapers’ reporting is the absence
of statistical evidence of an increase in interest in the MENA region over
the course of the year 2010. Figure 4.7 is a graphic illustration tracking
mentions of the items Middle East* and North Africa* quarter-yearly in
2010 in the Guardian and the Telegraph. In line with the observation
regarding the White House briefings, there appears to be no significant
trend exhibiting any rising interest in the regions over the year, no inkling
on the part of these news outlets that something of interest is about to
break out, yet another interesting absence.
Middle East */ North Africa*

900
800
700
600
1st quarter
500 2nd quarter
400 3rd quarter

4th quarter
300
200
100
Fig. 4.7 Mentions of the items Middle East* and North Africa* quarter-yearly in
2010 in the Guardian and the Telegraph
112 A. Partington
As a next step, two concordance-corpora were prepared with 500 char-

acters of co-text using the search items Middle East and North Africa, the
first from the first nine months of the year, MENAearly, the second of the
last three months, MENAlate. The two concordance-corpora were con-
trasted using the WordSmith Keyword tool, designed to find the key
(relatively more frequent) items in one corpus when compared to another,
to investigate what the papers were saying about the region in the last
quarter of 2010. Figure 4.7 shows there was no increase in mentions of
these regions, but perhaps the nature of the discourse about the regions
changed over the year, and maybe there was increased talk about political
dissent.
The items cables, cable, leaked and wikileaks were all prominent in the
MENAlate keyword list and these were then looked up manually in the
MENAlate concordance corpus and, when appropriate, the newspaper
articles in which they appeared were read. They related, of course, to
wikileaked cables discussing Middle Eastern affairs, mainly from US
sources (embassy—generally referring to US embassies—was also a
MENAlate keyword). The lion’s share of the occurrences are found in the
Guardian data which was one of the main conduits of the leaked docu-
ments; the item cables occurs 1089 times (105 per million words, pmw)
and wikileaks occurs 969 times (93 pmw) in the last three months of the
year. In the Telegraph in the same period they occur 282 times (21 pmw)
and 438 times (34 pmw) respectively (the Telegraph MENAlate data set
is slightly larger at 12.85 million words compared to the Guardian’s 10.4
million words).
The wikileaked topic which receives most attention in these newspa-
pers is the virulent animosity and fear reportedly expressed by several
Arab leaders towards the Iranian regime:
(7) The US embassy cables: Iran: Gulf neighbours: Spare us your evil: Arab
states scorn ‘Persian meddling’
King Abdullah of Saudi Arabia urged Iran’s foreign minister to ‘spare us
your evil’ in a meeting that reflected profound Arab hostility to the Islamic
Republic—a recurrent theme of high-level private conversations in the
Middle East in recent times. (Guardian, 29/11/2010)
One might speculate that this topic is deemed especially newsworthy

given its novelty to many Guardian readers who are perhaps used to view-
ing the Middle East as a homogeneous bloc and to being informed much
more often of Arab-Israeli hostility and of Iranian-Israeli animosities, but
very rarely of Arab-Iranian ill-feeling. This is another significant absence,
this time regarding the newspaper’s geopolitical representations.
Other topics covered include: US worries about the movements of
arms into and around the Middle East, from eastern Europe, from North
Korea, to and from Iran and Syria; concerns over the funding of al-Qaeda
groups and the Taliban by private individuals from Saudi Arabia and the
Gulf States and the poor capability of security forces in Yemen and Algeria
to respond to al-Qaeda attacks.
What is striking, especially in the context of the current study and the
hypothesis outlined above of whether the newspapers were conscious of
more dissent, is the dearth of discussion of malgovernance or of discon-
tent among Arab populations towards the region’s authoritarian regimes.
This could be either because the topic is simply little discussed in the
leaked cables themselves or because the newspapers are more interested in
the other issues listed above. There are just two articles which report the
matter, one on Tunisia:
(8) Deeply unflattering reports from the US embassy in Tunis, released by

WikiLeaks, make no bones about the state of the small North African
country, widely considered one of the most repressive in a repressive region.
(Guardian, 08/12/2010)
The second is a discussion of a leaked report by the US Ambassador to

Cairo which refers to President Mubarak’s ‘quasi-dictatorial regime’, and
affirms that the upcoming presidential elections ‘will not be free and
fair’(Guardian, 10/12/2010).
Finally, there is an article outlining leaked cables whose contents report
aspects of Colonel Gaddafi’s personality (‘mercurial and eccentric’), his
private life (‘not lavish in any way compared with the ostentation of the
Gulf oil state families’) and his reputation among other African leaders
where ‘he is regarded with fear and mistrust’ (Guardian, 08/12/2010).
114 A. Partington
There is, however, another highly significant absence; the article contains
no information concerning Gaddafi’s popularity or otherwise inside
Libya, that is, among the Libyan people.
4.3.3.3 The Individual MENA Countries
In this investigation to detect the presence or absence of any increase in

interest in the MENA region, the next step taken was an examination of
the concordances of the mentions in the last three months of 2010 of
some of the individual countries where uprisings were later to take place.
The search items were thus Tunisia*, Egypt*, Liby* and Syria*.
Tunisia
Of note is the fact that in October and November there are no reports in
either newspaper concerning politics in Tunisia. In December, the
Guardian carries two stories, one on the Wikileaks cable mentioned ear-
lier (example 8) and another after protests in Tunis headlined ‘Crackdown
threat in Tunisia after graduate protests: Two die in demonstrations
against unemployment President warns rioters but promises more jobs’
(30/12/2010). The Telegraph carries no political news on Tunisia in
December.
Egypt and Mubarak
The Telegraph shows very little interest in political events in Egypt over
these three months. Although the search item Egypt* produced 311
results (24.2 pmw), these were almost exclusively about tourism, culture
and history. The item Mubarak produced just eight relevant results, six in
November and two in December. There were parliamentary elections at
the end of November (27th) and so I concordanced politic* / election* /
poll* within a ten-item span of Egypt*, but this produced only four results,
from two articles, on November 29th and December 1st. Both emphasise
the ‘violence and fraud’ in which the elections were conducted. The sec-
ond reports the ‘sharp rebuke’ issued by the US to the Egyptian govern-
ment while the first underlines how the events are ‘leading the US to fear
for the country’s future as a key strategically in the Middle East’. All of
which leaves one wondering why there is an absence of any follow-up of
this story by the newspaper in the rest of the year.
The Guardian’s coverage is decidedly more assiduous. It has three
pieces previewing the situation and the inevitable outcome—a landslide
victory for the ruling party (the National Democratic Party, NDP)—two
by the newspaper’s Cairo correspondent and the third by Amira Nowaira,
Professor of English Literature at Alexandria University and occasional
contributor to the Guardian. She begins by listing the paradoxes of
Egyptian politics, chaotic and exciting but entirely predictable in terms
of electoral result, a country of young people ruled by extremely old men,
a ‘temporary’ state of emergency lasting 30 years, a ‘banned’ party—the
Muslim Brotherhood—openly campaigning on the streets and on the
airwaves. However, she also sees the present elections as potentially repre-
senting a real turning point, given how ‘out of touch’ the ruling party has
become:
(9) there are indications that the winds of change have started to blow …
while the outcome is assured, I feel sure things will never be the same again.
(Guardian, 28/11/2010)
She realises quite prophetically, as the ruling party does not, the sub-
versive power of the new communications media:
(10) [The NDP] doesn’t realise that the day might come when it could be
tweeted out of power. Nor is it able to understand that it won’t be able to
station the country’s security forces on the information superhighway as it
does on Cairo’s ring-roads. (Guardian, 28/11/2010)
Libya and Gaddafi
Libya features frequently in the UK news in 2010, due mainly to drawn-

out negotiations for the release and return to Libya on the grounds of
116 A. Partington
Table 4.2 How the two newspapers referred to the Libyan administration and
its leader throughout 2010
Guardian Telegraph Total
Gaddafi:

Libyan leader 28 13 41

dictator 1 1 2
Administration:
government 32 13 45
regime 1 1 2
ill-health of Mr al-Megrahi, imprisoned in Scotland for allegedly having

planted the aeroplane bomb which caused the Lockerbie massacre in
1988, and also due to often fraught commercial relations between the
UK and Libya.
Neither newspaper refers to any internal dissidence or dissatisfaction
with the Libyan administration or to any politicised factions within the
country. The administration is normally referred to as the Libyan govern-
ment, with 32 mentions in the Guardian and 13 in the Telegraph; it is
described as a regime just once by each newspaper. Gaddafi is normally
called the Libyan leader, 33 times in the Guardian and 47 times in the
Telegraph, often accompanied with Colonel. He is called dictator just once
by the Guardian and four times by the Telegraph (Table 4.2).
In other words, in 2010, there is an almost complete absence of the
highly negative language we saw used in the Guardian’s careful retrospec-
tive selection of its own reporting (Manhire 2012), mentioned earlier
(Sect. 4.3.3.1), where regime was the favoured term and Gaddafi himself
was the object of much name-calling.
Syria and Assad
Concordances of Syria*, after removal of instances referring to ancient

Syria, Syrian recipes and dances, ‘Syrian-born businessman’, and the like,
produced 81 occurrences related to politics from the Guardian and 30
from the Telegraph in the final three months of 2010. These referred almost
exclusively to Syria’s foreign relations, more precisely, its 20-year occupa-
tion of Lebanon, its patronage of Hezbollah, the possible involvement in
the murder of the Lebanese Prime Minister, its enmity towards Israel and
its close ties with Iran. There is a solitary article, in the Guardian, about the
arrest of a young female blogger on charges of spying, which hints that
Syria has a politically repressive administration: ‘The New York-based
watchdog Human Rights Watch raised the profile of Mallouhi’s case last
month when it described her detention as “typical of the cruel, arbitrary
behaviour of Syria’s security services”’ (Guardian, 05/10/2010). This is also
the only article in either newspaper to describe the Syrian government as
‘the Assad regime’ (my italics), though only on the second mention; the first
mention uses the usual formula ‘President Bashar-al-Assad’s government’.
This article on Syrian political repression is balanced by two mentions,
one in each of the newspapers, of Syria’s religious tolerance: ‘[t]here is a
[…] portrait of stability and freedom in Syria, where Christians comprise
up to 10% of the general population’ (Guardian, 24/12/2010), and ‘[a]s
a member of the minority Allawi strain of Shia Islam, Bashar al-Assad,
the Syrian president, has recognised the need to protect other vulnerable
faiths’ (Telegraph, 12/11/2010).
As was the case with Libya and Gaddafi, then, we find an almost com-
plete absence of the highly negative representations which the Guardian
decided to publish in Manhire (2012) as representative of its coverage of
Syria and Assad in 2011.
At the time of writing, Assad is the only one of the leaders of these
countries still in power. A concordance of Assad in a corpus of all the
Guardian articles published in 2013 (part of the SiBol 13 newspaper
corpus)5 revealed, curiously, that he is only once referred to as dictator in
the newspaper’s own voice (rather than in attributions to other voices)
but is called President on 178 occasions, yet another significant absence.
The newspaper would appear to have abandoned the vituperative lan-
guage of 2011 and reverted to the more respectful naming of the Syrian
leader which was apparent in 2010. The Telegraph is slightly less coy and
uses the term dictator for Assad in its own voice on 11 occasions. In con-
trast, the Guardian in 2013 is still happy to refer to Mubarak and Gaddafi
as dictator, which it does on 24 and 10 occasions, respectively. This redis-
covered respect for the still-powerful Assad but not for the ousted and
dead dictators may be the Guardian’s way of dealing with the ‘dictator
dilemma’ (see the next section).
118 A. Partington
4.4 Conclusion
4.4.1 Thoughts on the Lacunae in the Press
By employing a number of corpus-assisted discourse techniques we have

unearthed strong evidence of a complete absence of awareness or pre-
science that dramatic events were about to unfold in the Arab MENA in
2011. No questions were asked in the White House press room on the
area and the two highly reputable UK newspapers exhibit no particular
heightening of interest in the region until the occupation of Tahrir
Square.
We might speculate on why the Western press was caught so unawares.
It is highly likely that the attention of foreign correspondents in the
Middle East and elsewhere is oriented to the local urban elites; especially
if they do not speak the local language, they find themselves able only to
converse with English-speakers, who are usually to be found among the
educated urban middle classes. They may well also have little inclination
to spend time outside the comforts of cities like Jerusalem and Cairo and
among the inhabitants of poorer rural people, whose views—and dissat-
isfactions—may well therefore be underrepresented. It was interesting
that it was an Egyptian contributor, Amira Nowaira, who was the one
voice to warn of potential dramatic changes to come. However, having
local correspondents embedded in the socio-cultural milieu brings a
whole fresh set of problems. They may well be particularly politically
partisan and are certainly vulnerable to government or local political
interest-group pressure.
Second, we found an almost complete absence of criticism in the 2010
reporting of the internal repression practised by some of the main actors,
Mubarak, Gaddafi and Assad and their administrations, all of which
would be described in a variety of dramatically negative ways in 2011,
but only once popular protest had actually broken out. There is reference
in 2011 to ‘Gaddafi’s notorious Tripoli prisons’ (31/10/2011), but they
were not so notorious as to be mentioned by either the Guardian or the
Telegraph in 2010. The systematic murder, torture and mutilation of pris-
oners under the Syrian regime were also highlighted in the course of
2011 (Guardian, 20/09/2011) and yet, as with Libya, what happened in

Syria’s prisons was well known to anyone who cared to investigate well
before 2011, in times when the administration was being described by
the newspapers in far more respectful terms (see, for instance, Fisk, 2005,
p. 212). We might speculate on the effect of what we might term the
foreign correspondent’s ‘dictator dilemma’. While he or she can write
what they please about democratic governments, if he or she wishes to
have access to a repressive country, it may not be wise to be too critical of
its regime. For resident and local-born reporters, of course, the dangers
can be even greater.
The alert reader will have noted that I called all the above ‘speculation’
and not ‘interpretation’ or ‘explanation’. In the Introduction the question
was posed of what counts—not only in CADS but in all forms of dis-
course analysis, perhaps in all forms of text analysis from history to liter-
ary criticism—as ‘explanation’. In the best cases, interpretations are
informed inferences from the available data, and it is argued here that
corpus-assistance can constitute a valid way of gathering and organising
extra data. Here I define ‘explanation’ as the construal of those inferences
within the wider context—in linguistics—of a theory of language or—in
sociolinguistics and (critical) discourse analysis—a theory of language and
society. However, the precise nature of the inferences drawn will depend
not only on the personal primings of the person doing the inferencing,
including their knowledge of the world, but also their prejudices (by defi-
nition pre-formed) and, what is more, by the very theory of language (and
society) they have already adopted. The process of explanation is not just
subjective but even runs the risk of being circular, and explanations should
therefore be offered along with the readiness to engage in dialogue.
4.4.2 Corpus-assisted Discourse Studies and Absence
Corpus linguistics has occasionally been criticised for being blind to

absence, that is, it may have much to say about what is present in the
corpus being examined but it cannot enlighten us about what is absent,
what is not found therein (see the discussion in Baker, 2005, p. 35). But
such criticisms derive from a narrow simplistic view of corpus linguistics
120 A. Partington
as simple statistics gathering. Instead, the capacity of corpus techniques to

allow the researcher to compare and contrast large data collections actu-
ally adds new dimensions to the investigation of absence, as Taylor (2012)
remarks, ‘as corpus linguists and discourse analysts, we actually have a
rather impressive arsenal for investigating absence.’
Moreover, as we have seen during the course of this research, absences
come in different types. An important distinction was made between
‘known absence’—or rather suspected absence—and ‘unknown absence’.
A suspected absence can be searched for as, for instance, when we checked
the 2010 newspaper data to see whether various MENA leaders were ever
referred to as dictator. An ‘unknown absence’ is an absence revealed ser-
endipitously during the course of research as when stepping from one
concordance to another took us in various stages from Mubarak to the
discovery of a White House distinction between countries where both
sides were blamed for violence and those where only one side was blamed.
We can then list a number of ways in which corpus techniques, par-
ticularly various forms of comparison among corpora, can shed light on
absences in the analysis of a historical political event. They can:
Verify—or otherwise—suspected absences (as in the example of dictator

above).
Reveal previously unsuspected ‘unknown absences’.
Locate absence, that is, tell us from which part of the data a particular
message is missing, for instance, speculation about the forcible removal
of Mubarak was absent from the briefings but present in the CNN
news reports. If a corpus of spoken interaction is tagged for type of
speaker, for example, questioners and responders, it should be possible
to discover what sorts of messages are present or absent in the contri-
butions of each speaker type.
Quantify absence, for example, we saw there was no increase in references
to MENA countries in MENAlate, the corpus subsection containing
articles from the last quarter of 2010, compared to the rest of the year.
Track presences—absence, that is, the appearance and disappearance of
messages over time (see Sect. 4.1.2 on modern diachronic corpus-
assisted discourse studies). For instance, in the White House briefings

we tracked the rapid disappearance of Libyan government and its
replacement by Qaddafi regime.
As for the interpretation and consequent explanation (Sect. 4.4.1) of

absence, as with all interpretation, so much will depend on the skill,
knowledge of the world, and indeed the worldview and prejudices of the
researcher. However, the marriage of comparative statistical techniques
with researcher inferencing from data would seem to offer a more valid
basis for interpretative statements on absence than introspection and
conjecture alone, however educated, on what messages might be absent
from or should have been present in a particular set of texts. Indeed, it is
hard to see how, without corpus techniques, any confident statements on
many of the absences in the large data sets, as employed in this research,
could be made. Before debating ‘what is implied, inferred, insinuated or
latently hinted at’ (Baker et al., 2008, p. 296), one needs to ascertain
what actually was and was not said or written, and corpus tools and tech-
niques seem to be helpful and at times even vital in doing so.
Notes
1. Compiled by Giulia Riccio (2009).
2. Compiled by Anna Marchi, Lancaster and Bologna University. There are
arguments both for and against removing reports which are repeated.
3. Compiled by Charlotte Taylor, University of Sussex.
4. The remaining three refer not to Arab Spring protests but to the Israelis
and Palestinians.
5. SiBol 13 is part of the SiBol (Siena-Bologna) suite of newspaper corpora.
SiBol 13 contains the entire published production in 2013 of 13 English-
language newspapers: Guardian, Mail, Mirror, Telegraph, The Times from
the UK, the New York Times and Washington Times from the USA, The
Times of India, The South China Morning Post (Hong Kong), Daily News
Egypt, and Gulf News (UAE).
122 A. Partington
References
Baker, P. (2005). Public discourses of Gay Men. London: Routledge.
Baker, P. (2009). The BE06 Corpus of British English and recent language
change. International Journal of Corpus Linguistics., 14, 312–337.
Baker, P., Gabrielatos, C., Khosravinik, M., Krzyzanowski, M., McEnery, T., &
Wodak, R. (2008). A useful methodological synergy? Combining critical dis-
course analysis and corpus linguistics to examine discourses of refugees and
asylum seekers in the UK press. Discourse & Society, 19, 273–305.
Blommaert, J. (2005). Discourse. A critical introduction. Cambridge: Cambridge
University Press.
Davies, M. (2012). Expanding horizons in historical linguistics with the
400-million word Corpus of Historical American English. Corpora, 7,
121–157.
Duguid, A. (2007). Men at work: How those at Number 10 construct their
working identity. In G. Garzone & S. Sarangi (Eds.), Discourse, Ideology and
Specialized Communication (pp. 453–484). Bern: Peter Lang.
Duguid, A., & Partington, A. (in press). You don’t know what you’re missing.
Or do you? Using corpus linguistics to investigate absence/s. In A. Marchi &
C. Taylor (Eds.), Corpus approaches to discourse: A critical review. London:
Routledge.
Fisk, R. (2005). The great war for civilisation. London: HarperCollins.
Halliday, M. (1994). An introduction to functional Grammar. London: Edward
Arnold.
Hoey, M. (2005). Lexical priming: A new theory of words and language. London:
Routledge.
Manhire, T. (2012). The Arab spring. London: Guardian Books.
Marchi, A. (2010). ‘The moral in the story’: A diachronic investigation of lexi-
calised morality in the UK press. Corpora, 5, 161–190.
Marchi, A., & Taylor, C. (2009). Establishing the EU: The representation of
Europe in the press in 1993 and 2005. In A. Jucker, M. Hundt, & D. Schreier
(Eds.), Corpora: Pragmatics and discourse (pp. 201–224). Amsterdam: Rodopi.
Partington, A. (2003). The linguistics of political argument. London: Routledge.
Partington, A. (Ed.). (2010). Modern diachronic corpus-assisted discourse studies:
Corpora, 5: 2 (special issue).
Partington, A., Duguid, A., & Taylor, C. (2013). Patterns and meanings in dis-
course. Amsterdam: John Benjamins.
Riccio, G. (2009). White House press briefings as a message to the world. In

J. Morley & P. Bayley (Eds.), Corpus-assisted discourse studies on the Iraq con-
flict (pp. 108–140). London: Routledge.
Scott, M. (2008). WordSmith tools version 5. Liverpool: Lexical Analysis Software.
Taylor, C. (2010). Science in the news: A diachronic perspective. Corpora, 5,
221–250.
Taylor, C. (2012). And there it isn’t: (How) can we access the absent using
CADS? Talk given at CADSConf 2012, September 13, University of Bologna.
Taylor, C. (2013). What is absent in UK press discourses of migration? Talk
given at the International Language in the Media Conference, September,
23–30, Queen Mary, University of London.
5
Cross-media Studies as a Method
to Uncover Patterns of Silence
and Linguistic Discrimination of Sexual
Minorities in Ugandan Print Media
Cecilia Strand
5.1 Introduction
The media can be seen as a discursive battleground, where various social
actors present, advocate and propagate their position, often with the
unspoken intention of domination. Silverstone (2007, p. 27), influenced
by Hannah Arendt, argues that the media constitute a ‘space of appear-
ance’ where media also ‘provide the frameworks (or frameworlds) for the
appearance of the other and define the moral space within which the
other appears to us’. The media, or the Mediapolis—to use Silverstone’s
term to refer to all, new, old, print and broadcast media—however, not
only bring the world to us, they constitute and construct it at the same
time. This ‘doubling’ is why the media hold such power over us
(Silverstone, 2007).
At their best, the media constitute an open and inclusive space, reflec-
tive of society’s multitude of voices and their constant process of
C. Strand (*)
Uppsala University, Uppsala, Sweden

126 C. Strand
re-negotiating of preferred frames, i.e., interpretive schemes to under-

stand the world, its inhabitants and challenges. Silverstone (2007), how-
ever, highlights that some actors traditionally dominate this space,
making the Mediapolis a highly unequal space that is saturated with
power relations. Media discourses are thus often reflective of dominant
groups’ opinions and values, rather than those of suppressed minorities.
Dominant media discourses may not determine, but can surely influence
public opinion on various topics and subsequently social policy
(McCombs, Shaw, & Weaver, 2014). Consequently, there is a struggle for
visibility in media spaces, since it is through mediated narratives and
images that the world and its multitude of others become known to us.
Social actors thus come into existence through their textual appearance
and the frames in which they appear which influence our understanding
of not only that particular actor, but their ‘reality’.
Analysis of media texts should thus allow the study of such discursive
battle lines and provide a sense of the dominant actors’ and their adver-
saries’ propositions in favour of either the status quo or change. Engaging
analytically with discursive battles, or the actors’ struggle for attention
thus primarily revolves around constructing an analytical toolbox that
captures linguistic representations of actors and their positions, or at the
very least, traces of these.
This study subscribes to this notion of media as an important space of
appearance, where texts, i.e. both narratives and images, are reflective of
both current dominant frameworlds and their contestation. Furthermore,
the power to influence media appearance and in particular the opportu-
nity to self-represent are intimately connected with influence over dis-
course. An extension of this line of thinking suggests that absence from
discourse production sites could entail annihilation of a discursive exis-
tence or relegate silenced voices to the status of the other, which is
defined by more powerful actors. So what happens when certain social
actors are systematically absent from sites of media appearance and left
without an opportunity to influence the frames by which they are to be
understood? And how can we analytically engage with something that
may not be linguistically present and argue the existence of the non-
existent voice, as well as normatively discuss the implications of patterns
of absence?
Cross-media Studies as a Method to Uncover Patterns… 127
This chapter attempts to engage with the methodological dilemma of

studying absent voices—in our case, sexual minorities voices in Uganda—
by arguing that absence can become visible by indicating the possibility of
presence, and through analytically engaging with reflections of the absent
actors. Reflection means that the silenced actor’s input into the narratives
entailed in discursive battle—in this specific case, a human-rights-for-all
narrative—is present. This study utilizes a cross-media design to demon-
strate the possibility of presence of sexual minorities, as well the reflec-
tions left behind after their input has been filtered by Ugandan media’s
exclusion practices. A cross-media research design will contrast the main
LGBTQI (Lesbian, Gay, Bi-, Trans-, Queer and Intersexuals) organiza-
tion’s Twitter account with two leading Ugandan daily newspapers. In
this way, sexual minorities’ high level of activism in the former and their
invisibility for the most part in the latter are captured. The cross-media
design thus provides a tool to argue the possible existence of systematic
silencing practices in an important site of appearance, the Ugandan
mainstream media. Finally, the chapter discusses the potential implica-
tions of Ugandan sexual minorities’ systematic absence, or their occa-
sional appearance, which is often dominated by other claims makers’
prejudicial representation of them. It is suggested here that these medi-
ated representations further reinforce already existing homophobic dis-
courses which delegitimize sexual minorities’ claims to human rights, as
well as naturalize and legitimize existing discriminatory politics. Finally,
it is argued that the Ugandan media appear to continue a post-colonial
media tradition, where national media exclusively represent elite interests
(Skjerdal, 2012; Wasserman, 2010).
The chapter is organized as follows: since exclusion and subsequent
absence from discourse are always context-dependent, an understanding
of context is pivotal, so the chapter starts with an introduction of sexual
minority rights in Uganda, followed by an introduction to the Ugandan
mediascape’s historically discriminatory coverage and practices of silenc-
ing sexual minorities. The methodology section briefly presents the case
organization, Sexual Minority Uganda Network (SMUG), and the overall
research design. The results section will present the outcome of contrast-
ing the two platforms and discuss how a cross-media study can contrib-
ute to the identification of silencing practices and discursive absences.
128 C. Strand
A cross-media research design can effectively address the methodological

challenge of identifying existing patterns of absences. It is through the
contrast of different linguistic representations of the same subject matter,
in this case, a highly contentious and controversial social topic, that the
unsaid emerges, whereby many ‘unsaids’ form a pattern of silence that
ultimately shapes our understanding of a subject matter. The chapter
ends with a discussion of the potential implications of LGBTQI’s absence
in Ugandan print media discourse.
5.1.1 U
gandan State-sponsored Homophobia
and Politics of Non-belonging
In a comprehensive review of the constitutional protections of sexual

minorities, their civil, political, and socio-economic rights, as well as the
presence and efficacy of gay rights organizations; Dicklitch, Yost, and
Dougan (2012) classified Uganda as an ‘active persecutor’ of sexual
minorities. Several researchers have attempted to make sense of this sad
state of affairs by exploring the intricate historical roots (Englander, 2011;
Epprecht, 2010; Human Rights Watch, 2008; Tamale, 2013), and the
contemporary triggers of post-colonial state-sponsored homophobia
(Beyrer, 2014; Boyd, 2013; Kaoma, 2013; Oliver, 2013; Wahab, 2016;
Weiss & Bosia, 2013). Although it is beyond the scope of this chapter to
explore these factors in any detail, it is nevertheless important to acknowl-
edge that today’s discrimination is a direct product of historical factors, in
particular, colonialism, cultural practices and traditions, religion and reli-
gious re-colonialization, international financial flows, as well as Ugandan
post-colonial politics necessitating the management of globalization
angst and the assertion of domestic self-determination.
Although state-sponsored homophobia has been a permanent fixture
of post-colonial politics (Tamale, 2009), a revival of the construction of
sexual minorities as un-godly and un-African, an unwanted other, paved
the way for the latest efforts to strengthen already harsh laws against sex-
ual minorities (Kaoma, 2013; Ssebaggala, 2011). The 2009 Anti-
Homosexuality Bill, best known to outsiders for its archaic provisions of
the death penalty and life imprisonment, was marketed to the Ugandan
public as a way to protect Ugandan culture and traditional family values
against international sexual rights activists seeking to impose their values

of sexual promiscuity on the people of Uganda (Hollander, 2009; ‘The
Anti-Homosexuality Bill’, 2009). Indeed, the Bill was widely framed as a
tool to protect Ugandan sovereignty against Western imperialism and to
enforce the country’s right to self-determination in domestic matters
(BBC, 2011; Nyanzi & Karamagi, 2015).
The intense public debate on the Bill in 2009 and 2010 brought into
the open that rights which are indisputable for other Ugandan citizens—
most notably rights to privacy, health and indeed life—were openly ques-
tioned in relation to sexual minorities in parliament and in the mainstream
media (BBC, 2011; Strand, 2011, 2012, 2013). The public debate on the
Bill also exposed the Ugandan State and faith-based organizations within
and outside of Uganda as active propagators of furthering legal discrimi-
nation as a means of addressing the supposed problem of sexual minori-
ties in Uganda. The dominant framing of sexual minorities as un- African,
un-Christian, as a threat to Ugandan culture, and subsequently as not
really Ugandan (Boyd, 2013), rationalized political and socio-cultural
discrimination for the Ugandan citizenry. The law was promoted as a
means to an end, i.e. to handle and preferably remove unwanted and
unworthy members of Ugandan society, and ultimately to save Uganda
from unwarranted influences from the West. Interestingly, dominant
actors failed to highlight the irony of the Bill in that political homopho-
bia is a direct product of previous colonial masters (Semugoma, Nemande,
& Baral, 2012).
Despite strong public opinion in favour of the Bill (Boyd, 2013),
domestic resistance (Nyanzi & Karamagi, 2015) and international pres-
sure (Semugoma et al., 2012) managed to keep the Bill dormant until
late 2013, when it was finally passed. Celebrations among the Bill’s pro-
ponents were, however, to be short-lived as the Anti-Homosexuality Act
was successfully challenged in 2014 and overturned by the Constitutional
Court (Nyanzi & Karamagi, 2015). This legal victory has, however, not
been translated into less discrimination on the ground, and the human
rights situation of sexual minorities continues to be precarious. Human
rights abuse in the shape of extra-judicial violence, social discrimination,
denial of due legal process in connection with abuse, denial of employment
and housing continue unabated (Sexual Minority Uganda, 2016). As will
130 C. Strand
be elaborated in greater detail in subsequent sections, parts of the

Ugandan media have quite actively contributed to the community’s vul-
nerability by discriminatory coverage, as well as by the outing of alleged
LGBTQIs by publicizing their pictures, home and work addresses. The
Sexual Minorities Ugandan Network (SMUG) concludes in its 2016
report that despite the defeat of the Anti-Homosexuality Act in 2014, the
strongly held heteronormative ideals that brought the Bill into existence
in the first place, and society’s extra-judicial social policing of those ideals
remain.
5.1.2 Homophobia as a Hegemonic Media Discourse
The Ugandan media sphere is diverse and has as many as 48 publications

in circulation, eight TV stations and more than 200 radio stations
(UNESCO, 2012). The print media are dominated by four dailies; the
state-owned English-language paper, the New Vision, and its Luganda-
language sister paper, Bukedde; as well as the privately owned English-
language papers, the Daily Monitor and The Observer. Parallel to the main
newspapers, there are several tabloids and magazines such as The Red
Pepper, Rolling Stone and The Independent. The broadcasting entities cover
a range of different languages, but are controlled by privileged groups
such as politicians and religious leaders who have ties to the ruling elite
(UNESCO, 2012). The number of media products thus does not auto-
matically result in diversity for listeners and viewers (UNESCO, 2012).
Furthermore, there are indications that journalistic freedom in Uganda is
under attack with a government crackdown against media organizations
perceived to be too critical of the government (Freedom House, 2016;
Meyen, Fiedler, & Schamberger, 2016).
Decades of state-sponsored homophobia have had an impact on many
social spheres, and discriminatory attitudes towards sexual minorities are
not confined to the legal sphere, but affect other sectors of the Ugandan
society. Historically, the Ugandan media coverage of sexual minorities
ranged from blatantly discriminatory to active silencing. Previous studies
have found that sexual minority organizations and their representatives
are rarely awarded space to promote human rights for all, to self-define or
to provide a first-person narrative on their concerns and their commu-

nity’s plights (Strand, 2011, 2012). A case in point is that when local
human rights defenders highlighted concrete concerns about the 2009
Bill, these were only fully acknowledged after being endorsed by interna-
tional partners, notably mostly bilateral and multi-lateral donors and
then primarily in the privately owned press (Strand, 2011). Domestic
human rights defenders thus depended on outside support to become
part of the public discourse about the Bill, a situation which indicates
varying degrees of active silencing practices.
The exclusion of LGBT voices should, however, not be understood as
the complete absence of sexual minorities from Ugandan media dis-
course. Sexual minorities are present and appear to the reader, but mainly
through media gatekeepers’ choice of frames and other self-appointed
interpreters such as Ugandan political and religious actors. The existence
of sexual minorities in Uganda is thus acknowledged, but their visibility
is shaped by the interpretative lens of others, with the result that the
group is repeatedly discursively misrepresented, vilified and dehuman-
ized. The lack of rhetorical influence on representations has in the past
resulted in coverage that includes negative stereotypical representations
and unfavourable narratives of sexual minorities (Strand, 2012). In short,
Ugandan sexual minorities’ exclusion from discourse production sites
results in limited rhetorical ownership and limited influence on the inter-
pretive schemes by which the minority group becomes known outside
the LGBTQI community. Furthermore, (Adamczyk, 2017) argues that
in contexts where the LGBTQ community is excluded from the discourse
production processes and defined by other claim-makers, such as politi-
cians and religious organizations, mediated discourse consistently fails to
reflect the community and its concerns.
State-sponsored discrimination of sexual minorities is not restricted
only to the legal field. Repeated incidents of sanctions against media out-
lets for providing sexual minorities with a platform for human rights
advocacy indicate the existence of more or less clandestine coercion which
seeks to influence journalistic practices and to advance a culture of jour-
nalistic and editorial self-censorship in relation to sexual minorities
(Borlase, 2011; British Broadcasting Corporation, 2004; Human Rights
Watch, 2010; Johnson & Cameron, 2007). Sanctions in the past have
132 C. Strand
ranged from suspension of journalists to fines and to forcing media orga-

nizations to issue a public apology for providing a platform for public
indecency. In particular, journalists based outside the capital Kampala
face multiple hurdles in the form of local government sanctions and
undue pressure from media owners as well as negative audience reaction
(Borlase, 2011). Some media outlets, most notably the state-run the New
Vision, are confirmed to have developed editorial policies that restrict the
coverage of sexual minorities and thus have institutionalized silencing
practices (Friedrich-Ebert-Stiftung, 2012).
Given such obstacles to self-representation, sexual minorities are
actively exploring alternative spaces of the Ugandan Mediapolis. With
the introduction and increasing affordability of social media, previously
excluded and marginalized groups now have possession of a self-controlled
public space where they are in charge of their own narratives. Social
media provide an alternative site for appearing to others outside of the
LGBTQI community, as well as a platform to resist the dominant dis-
courses of sexual minorities as unAfrican, ungodly and as a manifestation
of subversive Western imperialism.
5.1.3 S
ocial Media as an Alternative Site
of Appearance
Social media offer new possibilities to resist the homophobia that perme-
ates political rhetoric and mediated discourses. It has been argued that
Twitter, as a few-to-many-platform combined with its less demanding
format in terms of demands of editorial capacity, gives previously unrep-
resented groups new opportunities to communicate. Techno-optimistic
interpretations of Twitter’s ease of use argue that the platform could open
up for participation of a broader range of actors beyond the established
media-savvy elite and thus expand the number of voices in public debates
(Jungherr, 2015). Twitter could thus support mediated reflections on
politics and organize action by mobilizing like-minded people (Gerbaudo,
2012).
Given the existing patterns of silence around sexual minorities in the
Ugandan mainstream media, social media could potentially provide the
account holder with an alternative space to challenge dominant dis-

courses which surround same-sex loving individuals in Uganda and to
support uncensored human right advocacy. To date, there is only one
study of Ugandan LGBTQI bloggers which concludes that accessibility
to the Internet provide self-identifying LGBTQIs a safe space where they
are free to express resistance to dominant heteronormative discourses as
well as to actively challenge hegemonic discourses on sexual minorities in
general (Valois, 2015). Blogs are used as a space to argue for the recogni-
tion of LGBTQIs as part of the Ugandan citizenry, and thus challenge a
discourse of queer identity as incompatible with citizenship in an African
nation, and with being African. Blogs were also used as platforms to claim
‘sameness’ with the broader Ugandan public and to demand equal rights
in ‘accordance with a neoliberal politics of normalization’ (Valois, 2015,
p. 145) where sexual minorities are regarded as no different from other
self-assigned identity labels. Although there is limited research on the
scope and impact of the LGBTQI community usage of social media as a
platform of resistance to homophobic discourses, the aforementioned
study indicates that an exploration and even colonialization of the space
have begun.
5.2 Methodology
This study proposes that absent voices can be identified by establishing
the possibility of their presence, as well as by engaging with the shadow
reflections of the absent actors. To credibly argue the possibility of pres-
ence of the absent voices and to explore patterns of silence, a research
design needs to be able to argue that the absent voice, is not absent
because of its own inaction or its choice to be silent, but because it has
been excluded as a result of gatekeepers’ more or less conscious silencing
practices. A cross-media analysis can address the methodological chal-
lenges connected to arguing the existence of the absent by contrasting
different communication platforms. Exploring absence from discourse
through the reflection of an absent voice entails capturing the absent
through the dominant discourse producers’ representation of the absent
voices or rejection of their proposition or demands. A politician’s public
134 C. Strand
rejection of human rights protection for sexual minorities reflects their

existence and demands for change, even if the group never appears
beyond presupposition. Even though this study primarily focuses on
establishing the possibility of the presence of LGBTQIs in the Ugandan
mediascape, and less on reflection of absent actors, engaging with what
can be labelled as discursive discrepancies can be rewarding. Discrepancies
between the absent voices’ attempts at discourse input and their more or
less recognizable reflections in spaces into which they are not invited are
indicative of more or less institutionalized and conscious silencing
practices.
In order to assess the possibility of discourse input, this study relies on
a cross-media design, where sexual minority community voices are
explored through the community’s umbrella organization SMUG’s
Twitter account and contrasted with two elite discourse producers repre-
sented by two daily newspapers, the government-owned New Vision and
the largest privately owned the Daily Monitor. These three sites of dis-
course production are compared and contrasted quantitatively and quali-
tatively in an attempt to explore patterns of silence surrounding LGBTQIs
in mainstream media and to deepen our understanding of possible impli-
cations of silencing.
5.2.1 SMUG and SMUG on Twitter
The Sexual Minorities Uganda (SMUG) network was established in 2004

as the main network for LGBTQI organizations, hosting eleven member
organizations. Its mission is to re-claim full inclusion of non-
heteronormative sexual orientations as part of the Ugandan citizenry
with equal rights as enshrined in the Ugandan Constitution (SMUG
website 2016).
SMUG started its Twitter account in 2012, a year before it opened up
its Facebook account in 2013. Twitter was selected based on the fact that
although SMUG is active on both Facebook and Twitter, the frequency
of posting via Twitter was higher at the time of the study (November
2015 and February 2016). Furthermore, Twitter is often positioned as
the established elites’ medium, such as political elites, and media workers
and journalists, which would suggest a potential for inter-agenda setting.

Finally, Twitter was the second most used social media platform in
Uganda after Facebook at the time of the study (Global Stats, 2016).
The data collection period was chosen so as to cover the Ugandan elec-
tion of 2016. The choice to follow the official campaign period was based
on the assumption that elections are opportune moments for mediated
debates on social concerns and future policy responses to societal chal-
lenges. Elections thus steer public deliberation towards assessing what
kind of society is desirable. By choosing the campaign (9 November 2015
until 16 February 2016) as data collection period, it becomes possible to
explore how SMUG used this period of societal deliberation to resist
hegemonic discourses of homophobia which deny them equal human
rights.
It is important to bear in mind that SMUG is a network advocacy
organization. Its social media platforms thus serve as a platform for all
network members. Subsequently, all tweets produced by SMUG and re-
tweets by member organizations were included in the analysis. The
Twitter platform includes five technological features for tweets: direct
messages, a form of public email to specified users; re-tweets, the reposting
of a tweet generated elsewhere but regarded as relevant for the network;
hyperlinks to other web material; hashtags to ensure that the tweet is more
easily found by non-followers interested in the topic (Small, 2011); and,
finally, user mentions, when the sender is talking about someone. Due to
the focus of this study, tweets were analyzed with regard to the message-
content of tweets and re-tweets disseminated from the SMUG account,
as opposed to looking into networking aspects of tweets.
5.2.2 Quantitative Analysis of Tweets
The SMUG account disseminated 418 tweets during the data collection
period which were analyzed quantitatively in a first step. The quantitative
analysis identified three peaks of high activity of tweeting which were
arranged on a chronological axis and confirmed the impression that
activity had distinct peaks. The three peaks constituted 39% of the total
tweets SMUG generated during the time period (Fig. 5.1). These events
136 C. Strand
Fig. 5.1 SMUG tweets distribution over time Nov 2015 to Feb 2016
were the first ever TV-broadcast live presidential debate (21 tweets), the
David Kato Memorial (54 tweets), and the inaugural SMUG gala (86
tweets). The frequent and sometimes even sport-reporting-like live narra-
tion of events as they were taking place, indicate their importance to the
SMUG network and its members. Even though intensified social media
activity may not equate the importance of an event, a closer and qualita-
tive inspection of the three peaks suggests that in these instances, inten-
sity was a sign of the event’s significance and was indicative of an active
attempt to gain attention.
5.2.3 Message-level Narrative Analysis of Tweets
As tweets in some instances were posted only minutes apart, the Twitter
feed had the resemblance of live narration of an event, intending to bring
the event to physically non-present followers. The subsequent analysis
was inspired by narrative analysis, which ‘permits a holistic approach to
discourse that preserve context and particularity’ (Smith, 2000, p. 327).
As part of a constructivist tradition, it allows for reflexivity, interpretation
and representations to take centre-stage in the analysis of stories or
accounts of experiences, as a mode of accessing how individuals or groups,
i.e. storytellers, interpret and attempt to make sense of events in their

lives (Riessman, 2005; Smith, 2000). Furthermore, narrative analysis not
only uses narratives as data to capture the narrator’s ‘reality’, but also
regards language as a means to construct meaning (Silverstone 2007).
According to Denzin, language and speech do not so much reflect experi-
ence, but rather ‘create experiences and in that process of creation con-
stantly transform and define that which is being described’ (cited in
Savin-Baden & Van Niekerk, 2007, p. 461).
According to Smith (2000), narratives consist of perspective and context
as well as frame. The narrative’s ‘perspective refers to the fact that a narra-
tive contains a point of view toward what happened, telling us what is
significant’ (Smith, 2000, p. 328). The context comprises the narrative’s
external influences, such as historical period, socio-political and cultural
context, as well as the immediate social setting, who the narrative is
directed towards and for what purpose. A frame is understood as an inter-
pretative lens based on prior expectations, upon which new texts are
interpreted. No narrative can be understood in isolation from its larger
and immediate surrounding textual and social context (Smith, 2000).
Most humans are innately gifted storytellers, and tell stories purposely.
Consequently, narratives have multiple functions, including to ‘raise con-
sciousness’ (Smith, 2000, p. 329), to create a shared group identity, and
provide a foundation for collective action. Narrative analysis has many
similarities with content analysis in that it systematically approaches
texts, allowing for varying degrees of inter-textuality to influence the
analysis and interpretation (Smith, 2000). Finally, Savin-Baden and Van
Niekerk (2007) argue that it is hugely important to recognize the central-
ity of the researcher in the interpretative process, especially when the
analysis of texts is not a co-creative interpretative process. This study
relied heavily on the author’s prior knowledge of the context to recognize
minority frames’ oppositional or reinforcing character vis-à-vis existing
hegemonic discourses. The minimalistic Twitter format also relies on the
receivers’ prior and contextual knowledge to de-code both abbreviations
and the compressed narrative in each unit, i.e. the tweet.
The following qualitative analysis is inspired by narrative analysis and
consists of two stages. A first stage entails re-constructing the narratives
around the three events by removing repeated tweets and by arranging
138 C. Strand
tweets thematically if parallel themes were pursued more or less simulta-

neously. In particular, when partner organizations tweeted parallel to
SMUG but appeared with a time delay during peak events in the SMUG
time line, their re-tweeted messages did not always appear chronologi-
cally in the SMUG main Twitter feed. By connecting the discombobu-
lated 140-character long mini-narratives, the study tried to re-construct
the larger narrative. The aim was not only to gauge sexual minorities’
level of activity in this space, but also to explore SMUG’s discursive input
in relation to the three significant events. As events were clearly demar-
cated, this analysis would facilitate an exploration and detecting the
reflections of the aforementioned discursive input in the mainstream
media spaces, should sexual minorities have failed to claim full visibility.
The second stage analysis also engaged in situating their discourse input
within the overall Ugandan context of discrimination, exclusion and
invisibility. Twitter, it would turn out, is a site of active resistance against
established patterns of discrimination of sexual minorities.
5.2.4 C
apturing News Media Discourses on Sexual
Minorities
Mainstream discourses are captured through two main English-language

Ugandan dailies; the government-owned and state-run newspaper, the
New Vision (NV) and the largest privately owned newspaper, the Daily
Monitor (DM). Both papers have online versions of their daily edition,
and it is these that were included in the sample. Both newspapers claim
that they seek to inform, educate and entertain the Ugandan public.
According to ABC (2011), the NV is the largest Ugandan newspaper with
a registered daily circulation of 32,218 copies, compared to 24,230 for the
DM, making the DM the third largest paper. Although daily newspapers
are primarily an urban and elite medium, it has been estimated that each
newspaper copy sold in Uganda is read by 10 people (Open Society
Initiative for East Africa, 2010). The NV, albeit self-proclaimed as inde-
pendent, is commonly known to side with the government in times of
elections and on key policy areas (Freedom House, 2014). Furthermore,
Freedom House’s Freedom of the Press Reports (2014, 2015, 2016) rarely
reports instances of intimidation of New Vision group’s media workers.
In order to explore the inclusion of sexual minority voices during the

official period of campaigning, all journalistic texts as well as submissions
by citizens between 9 November 2015 and 16 February 2016 which con-
tain one or several of the following words were included in the sample:
‘homosexual’, ‘homosexuality’, ‘homo’, ‘gay’, ‘gays’, ‘gayism’, ‘lesbian’,
‘LGBT’ and ‘kuchu/s/’ (a local term used in Uganda). The search gener-
ated as expected a small number of items for NV (seven items) and a
significantly larger sample for DM (28 items). A sampling procedure
which relies on direct references to sexual minorities in manifest language
may fail to capture all the instances where sexual minorities are implicitly
referred to. However, broadening the search and inclusion criteria could
undermine the validity of the sample.
5.2.5 Discursive Discrimination and Absent Voices
Boréus’s (2006) typology of discursive discrimination was used to explore

the implications of silencing practices in Ugandan newspapers. Boréus
(2006, p. 406) defines discursive discrimination as ‘discrimination car-
ried out through the use of language’. Unfavourable linguistic treatment
of a particular group or groups can take on multiple forms in texts, but
often includes four main types of discursive discrimination: (1) exclusion
from discourse, which manifests itself through denying the actor a voice,
where textual invisibility and voicelessness imply and textually signal that
the invisibles are inconsequential and generally unimportant; (2) nega-
tive other-presentation, i.e. negative descriptions and associations
ascribed to the absent actor; (3) discriminatory objectification, i.e., when
a group is visible only, or almost entirely, as the recipients of actions by
the other dominant actors, such as social policy interventions, and as not
as having agency or own needs and interests; and (4) advocacy to transfer
linguistic discrimination outside the textual discourse, i.e. advocating to
bring the unfavourable treatment of group members into action outside
the original text (Boréus, 2006).
Although this study did not use the typology for analyzing individual
media texts, it constituted an important intellectual point of departure to
understand the implications of exclusion. It is through the patterns of
140 C. Strand
Fig. 5.2 Exclusion and invisibility as enablers of other types of discursive

discrimination
exclusion in the Ugandan context that other types of discursive discrimi-

nation become possible (Fig. 5.2). It is thus argued here that the media
scripts of negative othering and discriminatory objectification, as well as
advocacy for unfavourable treatment of sexual minorities, are greatly
facilitated by excluding sexual minorities as actors and denying their
agency. Studying the discursive invisibility of sexual minorities is thus key
to understanding their marginalization within and outside of the media.
5.3 Results and Analysis

The result section will first of all present the study’s individual components
and then bring the separate analyses into dialogue. The section ends with a
discussion of the merits of cross-media studies as a methodology to advance
our understanding of the Ugandan media’s silencing practices and, its con-
sequence; the discursive absences of Ugandan sexual minorities.
5.3.1 U
gandan Media—Patterns of Silence Enable
the Reproduction of Homophobic Discourses
The analysis of the NV and the DM confirms earlier research in the field,
(Bompani & Brown, 2015; Strand, 2011, 2012, 2013). Historically,
Ugandan media awards minimal space to sexual minority voices, and
when they do appear, they are primarily defined through the lens of
Fig. 5.3 Daily Monitor silencing of local sexual minorities during the 2016 elec-
tion period
religious and political elites. This description is still largely accurate.

Ugandan sexual minorities only self-represent once in the DM in a pur-
chased piece from the international news agency AFP (Fig. 5.3). Ugandan
journalists did thus not at any point actively engage with domestic sexual
minorities even when the community’s activities were deemed newswor-
thy to international news agencies. This would suggest that both media
houses pursue some degree of silencing practices. As this study has not
interviewed journalists or editors, it is not possible to ascertain to what
extent journalists themselves are consciously denying space, or are gov-
erned by editorial guidelines, or simply reflective of a societal culture that
seeks to deny the existence of LGBTQIs.
As described earlier, the Ugandan media do not practise complete
obliteration of the subject, and the silencing practices thus do not result
in a complete absence of sexual minorities from discourse. There are how-
ever some noticeable differences between the two newspapers in terms of
how sexual minorities become visible to the readers. In the state-owned
NV, domestic sexual minorities are almost entirely absent (Fig. 5.3). NV
contains seven items in total and none award domestic sexual minority
142 C. Strand
individuals or groups neutral visibility. Instead, six out of the seven are
purchased material from AFP and relate to Western contexts, the US
entertainment industry and the Oscars to be specific. The items thus
place sexual minorities in a distant context and thereby reinforce the
hegemonic discourse of LGBTQI as being a Western phenomenon as
opposed to part of an African context. Only one item, a political com-
mentary, makes a brief reference to the existence of domestic sexual
minorities in relationship to one of the Presidential candidates’ failure to
conclusively answer a question whether or not he ‘supported gays’ during
the live TV-broadcast presidential debate. The material thus contains a
trace of sexual minorities being part of the first-ever TV-broadcast presi-
dential debate, but fails to include the fact that accusations of homosexu-
ality was featured prominently in the debate. The lack of any references
in texts produced by in-house staff, in particular in relation to the
Presidential debate where the community was a heated debate topic,
strongly indicates that the previous editorial policy on silencing domestic
sexual minority voices and LGBTQIs as a subject matter is most likely
actively pursued (Figs. 5.3 and 5.4).
Fig. 5.4 New Vision silencing of local sexual minorities during the 2016 election
period
5.3.2 D
iscursive Discrimination in the Privately
Owned Press
Sexual minorities are more frequently mentioned in the DM. With 28

items containing a reference to sexual minorities, one could be made to
believe that the paper is more open to inclusion of sexual minorities.
However, the DM sample contains only one purchased AFP piece, i.e.
not produced by local journalists, Gay Ugandans seek private Pope session
(DM, 25 November 2015), which consists of an interview with SMUG
Director Frank Mugisha and his attempts to arrange a meeting with the
Pope with the intention to convince him to speak out on LGBTQI
human rights during his forthcoming visit to Uganda.
The remainder of the articles follow a familiar pattern where LGBTQIs
only appear through the lens of others. DM does, however, include a
larger number of items than the NV, albeit mostly international news
agency items, that contains no manifest discursive discrimination. But 15
items replay familiar discriminatory frames, where negative othering, dis-
criminatory objectification and advocacy for unfavourable treatment of
sexual minorities are clearly present, and where local elite stakeholders
have been invited to do the defining. Most of these 15 items diligently
report on various Ugandan politicians who use sexual minorities as part
of their political programme and campaigning, vowing to ‘cleanse
Uganda’ from homosexuality by offering rehabilitation or, if need be,
through the removal of sexual minorities.
(1) Presidential candidate Benon Biraaro has promised to fight homosexu-

ally at a regional level after failed efforts by Uganda as a country. Maj Gen
Biraaro said if elected president, he will take a regional approach with the
East African Community (EAC) partner states against the practice since a
lone fight by Uganda will lead to international isolation and embargo.
(DM, 7 January 2016)
(2) Presidential aspirant Abed Bwanika has urged Ugandans to use the
power of the ballot to wipe out homosexuality in the country …
‘Homosexuals are bankrolling these elections as a base to spread their devil-
ish vices and Uganda should resist this so as not to be another Sodom and
Gommorah facing destruction and wrath of God and Pastor Abed Bwanika
is here to lead the crusade.’ (DM, 11 November 2015)
144 C. Strand
The discursive practices of negative othering and discriminatory objec-

tification were also frequent in relation to religious discourse producer.
(3) The Archbishop of Church of Uganda Stanley Ntagali has asked

Members of Parliament to shun the money sent by Western countries to
promote homosexuality in the country. He said over the weekend that
some of the 9th parliament legislators are still haunted by the dirty money
they received from such countries and people, urging Ugandans not to re-
elect them. (DM, 21 December 2015)
In conclusion, the analysis of the two elite discourse producers indi-

cates patterns of silence which systematically fail to provide editorially
controlled space to LGBTQI individuals or their organizations for self-
representation, even when they are directly addressed by others as during
the Presidential debate on national TV. However, there are differences in
the forms of discursive discrimination generated by the exclusion of
LGBTQI voices. The privately owned press, DM, provides visibility, i.e.
allowing LGBTQIs to be present linguistically in contexts both inside
and outside Uganda, and includes some neutral-toned representations in
particular outside the Ugandan context. Furthermore, DM appears to
take considerable cues from the outside when taking editorial decisions as
to whether or not to cover sexual minorities as a phenomenon. For
example, domestic sexual minorities featured, albeit as an accusation, in
the first TV-broadcast Presidential debate, and SMUG’s active lobbying
to meet the Pope during his visit. But even where sexual minorities are a
key actor and news source, they are only present through others, often
with the result of discursive discrimination in the shape of negative other-
ing, discriminatory objectification and the proposal of non-linguistic
discrimination.
Although this analysis merely affirms what other studies have shown,
systematic silencing practices in parts of the Ugandan media can only be
credibly argued if absence can be attributed to and pinned to editorial
and journalistic decisions. Without establishing that the LGBTQI com-
munity has actively engaged in discourse production at the same time,
absence could be deflected as a consequence of their inactiveness as social
agents in the Ugandan mediascape.
5.3.3 Qualitative Messages-level Analysis
As briefly alluded to in previous section, sexual minorities were actively

engaging with hegemonic discriminatory discourses during the period
under study. Although SMUG was active throughout the entire time
period, the three peak events were chosen, as they were believed to con-
stitute important points of activism, and to be at least potentially news-
worthy. Each sub-section contains a selection of tweets that both illustrates
the event’s narrative and poignantly highlights that the community was
far from silent or passively accepting the surrounding society’s and the
elite’s rhetorical ownership of Ugandan LGBTQIs. Indeed, an explora-
tion of SMUG’s communication through Twitter shows attempts to
actively resist existing homophobic discourses in the Ugandan context.
5.3.3.1 Th
e First-ever Live TV-broadcast Presidential Debate,
15 January 2016
SMUG’s director is the first to highlight the event’s importance by re-

tweeting another account’s tweet:
(4) ‘7:35pm #UGDebate16 yet to kick-off’ quickly followed by his own

prompt ‘Over 27m Ugandans waiting, start already’.
These pre-event tweets are meant to sensitize and alert SMUG account
followers to the importance of the upcoming debate. SMUG’s subse-
quent tweets on the debate consist of recounts of selected quotes from the
on-going debate combined with their own commentary;
(5) Presidential Candidate Mabiriizi asks @AmamaMbabazi whether he

supports homosexuality. #UGDebate16.
The tweet is followed by the accused’s emphasis of his heterosexual

status
(6) I’ve heard these stories (about homosexuality) making rounds, I’m not
gay. I’m married to a woman. @AmamaMbabazi #UGDebate16
146 C. Strand
as well as his response challenging his accusers,
(7) I’ve heard rumors that I’m gay and I put Minister Amongin to the test
to prove that I’m gay—Amama Mbabazi #UGDebate16
Another candidate’s positioning on the topic generates further tweets.
(8) While I’m the President of Uganda, the country will never be a homo-
sexual country—Abed Bwanika #UGDebate16
(9) There’s no proof that there are people who are born gay—Abed Bwanika
#UGDebate16
(10) will ensure that the gays are brought back to normal through rehabili-
tation—Abed Bwanika #UGDebate16
The SMUG Director sums up one of the candidates’ position:
(11) Uganda Presidential candidate Abed Bwanika, homosexuals are not

normal, he will fight them & rehabilitate them. #UGDebate16
The presidential candidates’ exchange of ardent denials of homosexual-

ity and rejection of homosexuality as a natural part of human sexuality
prompted one SMUG network members to voice concerns with the can-
didates’ nonsensical exchange, which SMUG retweets:
(12) thats why u are not President 3 times and you still talk about homo-
sexuality yet we have more pressing issues.
Subsequent tweets and re-retweets reiterate further accusations that

were raised between the various candidates and their attempts to distance
themselves and reinforce their heteronormative credentials by reminding
the audience that the candidate is a married man with children and even
grandchildren.
The analysis of SMUG’s Twitter activity shows somewhat surprisingly
that the platform in this particular instance is not used to resist dominant
homophobic discourses. Instead, SMUG’s self-controlled platform merely
reiterates the debate theme of homosexuality as a negative subject posi-

tion that requires immediate distancing. Indeed, the majority of the
tweets simply re-broadcast the candidates’ homophobic sentiments.
Homosexuality in the Ugandan socio-political context has been discur-
sively connected to un-Africanness, ungodliness and lack of Christian
morals, as well as believed to reflect Western imperialist loyalties. The
accusation of homosexuality in a political debate can thus function as a
powerful tool to discredit the political opponents. The lack of commen-
tary and resistance despite the absence of editorial filters might indicate a
degree of internalized homophobia in a context where explicit homopho-
bia has become the norm in political communication. As hegemonic dis-
courses, they have become taken-for-granted frames of orientation, even
by those negatively affected by them.
5.3.3.2 David Kato 5-Year Memorial, 26 January 2016
The 5th memorial for the murdered human right activist David Kato is
the second most tweeted event. Just like the presidential debate, it begins
with a notice of the forthcoming event, and a re-announcement when it
begins in real time.
(13) Tomorrow @SMUG2004 will hold a memorial service in honor of

David Kato, also Premier our first Documentary ‘And Still We Rise’.
Followed by
(14) 5th Memorial for David Kato begins
Subsequent tweets and re-tweets document the opening of the event

both verbally and visually. Tweets are tagged #WecelebrateYourlegacy to
increase ease of being identified as part of the memorial, as well as signal
community with the use of ‘We’ are coming together in celebration.
The main part of the Twitter feed is a recounting of the memorial ser-
vice led by the Ugandan sexual minority community’s religious leader,
Bishop Christopher Ssenyonjo, who has paid a high personal and profes-
sional price for standing up for his belief that his God embraces all human
148 C. Strand
beings. The service and Twitter feed incorporate several Bible quotes and
religious references, before the memorial service becomes political in the
sense that it directly rejects the dominant frames to understand homo-
sexuality in the Ugandan context.
(15) People are ignorant about the purpose of human sexuality—Bishop

Ssenyonjo
(16) People are who they are, accept them—Bishop Ssenyonjo
(17) God loves you as you are, some have tried to change but it’s difficult—
Bishop Ssenyonga
(18) If you are heterosexual and they tell you to change to homosexual you
can’t, why do you want others to change?—Bp Ssenyonjo
(19) Respect each other in spite of our differences—Bishop Christopher

Ssenyonjo
(20) Anything anti-love is anti-God, Bishop Senyonj
(21) Bishop Ssenyonjo; many people are using Religion to condem

Homosexuality instead of preaching love. God is love
The memorial service closes with a song which, according to its accom-
panying tweet, is intended to instill strength in the memorial attendees.
The memorial ends with a premiere viewing of a 70-minute-long docu-
mentary, And Still We Rise, which documents the community’s resistance
and successful overturn of the controversial Anti-Homosexuality Act.
Besides the news-like coverage of the actual event, SMUG also use the
opportunity to actively urge attendees and followers to engage in discourse
production through various channels. One of the last tweets of the event
urges the community to raise their voices and claim rhetorical ownership.
(22) It’s important that we tell our stories—SMUG PD
Unlike the live commentary on the TV-broadcast presidential debate,

most of the tweets from the event are oppositional to dominant discourses
on sexual minorities. Tweets generated during the memorial service led by

Bishop Christopher Ssenyonjo engaged directly with the dominant fram-
ing of sexual minorities as rejected by God. With a point of departure in
the Bible, the Bishop’s statement that same-sex loving individuals are part
of God’s large and diverse herd, and that God accepts you, regardless of
who you choose to love, directly rejects a dominant frame in the Ugandan
context that God’s love is conditioned on heterosexuality. This message of
divine acceptance activates a powerful counter-narrative on Ugandan
sexual minorities, where non-heterosexual status is preached as a disquali-
fier for divine grace as well as symbolic membership in Ugandan society.
Tweets 17–21 thus engage directly with unnamed religious claim-makers
who castigate same-sex loving individuals as sinners with no place in the
Christian community unless they alter their ways and seek redemption
for past sins. Furthermore, through the counter-frame of equality before
God, LGBTs are implicitly framed as the same as heterosexual individu-
als. Being the same as heterosexual Ugandans should entitle them to enjoy
the same human right as enshrined in the Ugandan Constitution.
5.3.3.3 The SMUG Gala, 17 December 2015
The 2015 SMUG Inaugural Gala, hashtagged #SMUGGala2015 is by

far the most important event during the analyzed time period. The gala
was launched with the characteristic pre-tweets and a notification that a
member organization is going to live-stream the event. The actual event
is kicked off with a SMUG welcoming note, and, most importantly, a call
for partnership.
(23) SMUG ED @frankmugisha calls for partnership with Uganda

Government and partners to end stigma and discrimination
(24) Tonight I call upon the govanment to continue to work with us to

achieve success in fighting Hiv@dr frank mugisha @SMUG2004
#SMUGGala2015.
A key feature of the Twitter feed is to highlight the presence of a num-

ber of high-level guests and their public endorsement of LGBTQs as
partners.
150 C. Strand
(25) Dr Uchenna from UN Human Rights Commission says introduction

of #AHA led to an increase in LGBTI persecution.
(26) All human beings are equal and should be treated equal—uchena
emolonye UNOHR rep #SMUGGala2015
Another key guest is former Vice-President of Uganda and United

Nations Special Envoy for HIV/AIDS in Africa, Dr. Wandera Specioza
Kazibwe.
(29) I represent the ministry of health at this gathering today. Dr Specioza

Wandera Kazibwe #SMUGGala2015
(30) What you are doing as sexual minorities in the fight against HIV/AIDs
is commendable, own it. Dr Specioza Wandera Kazibwe #SMUGGala2015
(31) If u countinue to stand in yo truth as who you are, the world will
honor you.-VP Specioza wandera kazibwe @SMUG2004 #SMUGGala2015
(32) As sexual minorities the work you are doing in fighting HIV/AIDS
is phenomenal.-VP dr. Specioza wandera kazibwe @SMUG2004
#SMUGGala2015
(33) People still don’t know about sex and sexuality, they are shy to talk
about it. Dr Specioza Wandera Kazibwe #SMUGGala2015
(34) Thank you for taking on the threats and turning them into
opportunities. Dr Specioza Wandera Kazibwe #SMUGGala2010
#HonoringHealthChampions
(35) God bless you in your love lives. Dr Specioza Wandera Kazibwe
#SMUGGala2015 #HonoringOurHealthChampions
Highlighting the presence of distinguished guests who represent not

only the international community through the UN Human Rights
Commission, but also the Ugandan government and both of whom come
from outside the LGBT community, provides the community with
much-needed legitimacy. The high-level guests’ public endorsement and
even praise of SMUG’s work validate its mission and position SMUG as
a valuable partner. This is a powerful narrative for a minority group that
has endured public condemnation and persecution from the President,
members of parliament, church leaders and at times was even afraid to
congregate in public after the introduction of the controversial Anti-
Homosexuality Bill in 2009. Through the voices of established actors
with discursive power to influence dominant frames, SMUG uses its
Twitter platform as a megaphone for challenging dominant discourses.
Since the purpose of the inaugural SMUG gala is to celebrate individu-
als who in various capacities have contributed specifically to the network’s
mission and who have championed in the area of health rights, a central
feature of the Twitter narrative is the acknowledgement of the health cham-
pions who tirelessly ensure sexual minorities’ access to health services.
(27) Today we are here 2 honor persons like dr sylvia tamale for having é
guts 2 fight 4 policies that benefit people @SMUG2004
(28) Tonight we honor heroes in the work of access to health for LGBTI
persons in Uganda-Dennis wamala @SMUG2004 #SMUGGala2015
The Twitter feed also documents other aspects of the gala, that has little
to do with the reason behind the gathering. Recurrent tweets feature inte-
rior pictures of a glamorous venue with smartly dressed individuals listen-
ing to speeches of prominent guests, supporting their community,
sandwiched with entertainment features. The Twitter feed reiterates the
script of a classic gala, with welcome addresses and distinguished s peakers,
presentation of awards followed by thank-you speeches and entertainment
in the shape of live music performances. The gala and its online represen-
tation could be seen as a bold attempt to claim public space and to consti-
tute the community visually and textually to an outside audience. The
claiming of space in an explicit way, both offline and online, signals a new
confidence and a wish to challenge years of self-imposed silence as a secu-
rity measure in the aftermath of the Anti-Homosexuality Bill.
152 C. Strand
5.3.4 C
ross-media Studies—A Tool to Capture
the Invisible?
Studying the exclusion of voices and the subsequent absence of discursive

elements on key discourse production sites such as national media, is
methodologically challenging. To credibly claim the existence of what is
not manifest in texts requires a research design that allows us to capture
the possible presence of missing elements in the discourse, and/or traces
of the absent. To address this methodological problem, this study used a
cross-media research design, where two different public platforms print
media and social media platform, Twitter, entered an involuntary dia-
logue in an attempt to trace the way in which the more powerful of the
two discourse producers—established newspapers—actively engage in
more or less intentional silencing practices.
By analyzing SMUG’s Twitter feed it became apparent that the com-
munity was actively engaging in constituting itself within and outside of
social media. The Twitter feed includes three bold claims of public space
with the Memorial Service of David Kato and the premiere of the docu-
mentary And Still We Raise and the SMUG Gala, as the most prominent
attempt at claiming space and providing discourse input.
The analysis of SMUG’s tweets shows that Twitter provides a vibrant
place for resistant and oppositional narratives, attempting to redefine and
influence how the community is textually constituted. The analysis of
SMUG’s tweets thus firmly establishes that absence in key discourse pro-
duction sites, i.e. the DM and the NV, cannot be attributed to lack of
possible input, should the journalists have sought it. Furthermore, being
a small country with a small movie industry and few premieres or galas,
these events could be argued to be newsworthy, according to basic stan-
dards of news criteria. But neither of the newspapers chooses to cover
these events.
Through this, research design-imposed dialogue between dominant
discourse producers, i.e. the DM and NV and SMUG’s social media plat-
form Twitter, it is possible to credibly claim that the state-owned newspa-
per NV pursues a stringent silencing practice, with a nearly complete
absence of linguistic representations of sexual minorities as a result.
Sexual minorities, as far as they are concerned, simply do not exist in the
Ugandan contexts. Despite sexual minorities being prominently featured
in the TV-broadcast presidential debate, the NV coverage indicates that
sexual minorities primarily exist as a feature in the American entertain-
ment industry.
The privately owned newspaper DM features different patterns of
silence, where sexual minorities are similarly denied an opportunity to
self-represent, even if they are the subject of the news story; but may
appear as reflections through established discourse producers’ framing,
such as politicians and religious officials. Even if the DM sample included
internationally generated neutral coverage, domestic sexual minorities,
for the most part, are visible through the discriminatory lens of others. In
conclusion, both newspapers appear, to various degrees, to pursue active
silencing practices. As stated earlier, this study cannot determine whether
these patterns of silence are recognized and actively debated inside respec-
tive media houses, or simply exist as automatic unquestionable filters.
But as LGBTQI’s discursive absence historically enable an endless cycle
of reproduction of hegemonic discourses on Ugandan LGBTQIs silenc-
ing practices should never be mistaken for being innocent and without
consequences.
5.4 Discussion
In a context where elite discourse producers’ silencing practices may facil-
itate non-linguistic discrimination that entails extra-judicial violence,
social discrimination, denial of due legal process in connection with
abuse and denial of employment and housing (Sexual Minority Uganda,
2016), the price for exclusion is arguably high for the invisible.
Twitter allowed SMUG to bypass some of the silencing practices and
gave this often misrepresented community an opportunity to self-
represent in a manner that, historically, Ugandan media would never pro-
vide space for. However, this study also demonstrates that contemporary
media are as reluctant as ever to provide space for this particular minority,
even when they are the subject and source of a story, such as in the
154 C. Strand
Presidential debate and the SMUG’s director’s attempts to meet the Pope
during his visit to Uganda to persuade him to speak out for sexual
minorities.
Although the study concludes that Twitter narratives were a stark con-
trast to the Ugandan newspapers’ patterns of silence, a pertinent question
is, however, if this new opportunity to self-represent makes a difference,
especially given Twitter’s questionable inter-agenda setting function.
There were, for example, no manifest linguistic traces of SMUG’s oppo-
sitional narratives in the print media, which arguably has a larger uptake.
It is very likely that in a short-term perspective, it may not change any-
thing. But the fact that Twitter provided a persecuted community, oper-
ating in a hostile environment, with new opportunities to deliberately
constitute themselves as a part of Ugandan society and to resist homo-
phobic discourses produced and re-produced by the state, religious insti-
tutions and media representatives, can prove to be important in a
medium- and long-term perspective. On a more profound level, the
social media platform gives the community a space to negotiate and
launch counter-narratives, where Ugandan LGBTs are presented as resil-
ient heroes and champions, active members of Ugandan society, award
recipients, as glamorous and enjoying the support of some elite members
of Ugandan society. In a long-term perspective, this may matter greatly as
it could strengthen the community’s sense of agency and thus support
their long journey to claim other spaces, with the aim of expanding their
current social environment’s limited understanding and acceptance of the
great diversity of human sexuality. Before social media platforms, there
were few, if any, alternative spaces that would allow a diverse set of voices
pushing the public’s understanding of the pluralism in human sexual
identities, or allow coordination of activism. Twitter alone will not change
anything, but it may be an important platform for creating a sense of
agency and to function as a useful backbone structure for activism.
Unquestionably, this slightly techno-optimistic interpretation needs to
be further explored. Longitudinal research in particular is needed to sys-
tematically assess the real impact of social media as tool for chipping
away at entrenched silences, and eventually create a large enough crack so
as to allow previously absent voices space on traditional discourse-
producing platforms such as national media.
References
Adamczyk, A. (2017). Cross-national public opinion about homosexuality: Examining
attitudes across the globe. Oakland, CA: University of California Press.
BBC (Writer). (2011). Is homosexuality un-African? BBC World Debate, RSA.
Beyrer, C. (2014). Pushback: The current wave of anti-homosexuality laws and
impacts on health. PLoS Med, 11(6), e1001658. https://doi.org/10.1371/
journal.pmed.1001658
Bompani, B., & Brown, S. T. (2015). A “religious revolution”? Print media,
sexuality, and religious discourse in Uganda. Journal of Eastern African Studies,
9(1), 110–126.
Borlase, R. (2011). Global journalism, local realities: Ugandan journalists’ views on
reporting homosexuality (MSc in Media, Communication and Development),
London School of Economics and Political Science, MEDIA@LSE.
Boréus, K. (2006). Discursive discrimination. European Journal of Social Theory,
9(3), 405.
Boyd, L. (2013). The problem with freedom: Homosexuality and human rights
in Uganda. Anthropological Quarterly, 86(3), 697–724.
British Broadcasting Corporation. (2004). Fine for Ugandan radio gay show.
Retrieved from http://news.bbc.co.uk/2/hi/africa/3712266.stm
Dicklitch, S., Yost, B., & Dougan, B. M. (2012, May). Building a Barometer of
Gay Rights (BGR): A case study of Uganda and the persecution of homo-
sexuals. Human Rights Quarterly, 34(2), 448–471. Published by Johns
Hopkins University Press. https://doi.org/10.1353/hrq.2012.0033
Englander, D. (2011). Protecting the Human Rights of LGBT people in Uganda
in the wake of Uganda’s ‘Anti Homosexuality Bill, 2009’. Emory International
Law Review, 25(3), 1263–1316.
Epprecht, M. (2010). The making of ‘African Sexuality’: Early sources, current
debates. History Compass, 8(8), 768–779.
Freedom House. (2014). Annual report Uganda 2014. Retrieved from https://
freedomhouse.org/report/freedom-world/2014/uganda
Freedom House. (2015). Annual report Uganda 2015. Retrieved from https://
freedomhouse.org/report/freedom-world/2015/uganda
Freedom House. (2016). Freedom on the Net Uganda Report. Retrieved from
https://freedomhouse.org/report/freedom-net/2016/uganda
Friedrich-Ebert-Stiftung. (2012). African Media Barometer—The first home
grown analysis of the media landscape in Africa Uganda 2012. Retrieved from
http://library.fes.de/pdf-files/bueros/africa-media/09427.pdf
156 C. Strand
Gerbaudo, P. (2012). Tweets and the streets: Social media and contemporary activ-
ism. Pluto Press.
Hollander, M. (2009). Gay rights in Uganda: Seeking to overturn Uganda’s anti-
sodomy laws. Virginia Journal of International Law, 50, 219.
Human Rights Watch. (2008). This alien legacy: The origins of “sodomy” laws in
British colonialism. Retrieved from Printed in the United States of America.
Human Rights Watch. (2010). A media minefield increased threats to freedom of
expression in Uganda. Retrieved from http://www.hrw.org/en/reports/2010/
05/02/media-minefield-0
Johnson, C. A., & Cameron, E. (2007). Off the map: How HIV/AIDS program-
ming is failing same-sex practicing people in Africa. International Gay and
Lesbian Human Rights Commission New York.
Jungherr, A. (2015). Analyzing political communication with digital trace data.
Cham, Switzerland: Springer.
Kaoma, K. (2013, October 25). The marriage of convenience: The US Christian
Right, African Christianity, and postcolonial politics of sexual identity. In
M. L. Weiss & M. J. Bosia (Eds.), Global homophobia: States, movements, and
the politics of oppression (1st ed., pp. 75–102). University of Illinois Press.
McCombs, M. E., Shaw, D. L., & Weaver, D. H. (2014). New directions in agenda-
setting theory and research. Mass Communication and Society, 17(6), 781–802.
Meyen, M., Fiedler, A., & Schamberger, K. (2016). ‘It is a crime to be abusive
towards the president’: A case study on media freedom and journalists’ auton-
omy in Museveni’s Uganda. African Journalism Studies, 37(3), 1–18.
Nyanzi, S., & Karamagi, A. (2015). The social-political dynamics of the anti-
homosexuality legislation in Uganda. Agenda, 29(1), 24–38.
Oliver, M. (2013). Transnational sex politics, conservative Christianity, and
antigay activism in Uganda. Studies in Social Justice, 7(1), 83.
Riessman, C. K. (2005). Narrative analysis. In N. Kelly, C. Horrocks, K. Milnes,
B. Roberts, & D. Robinson (Eds.), Narrative, memory & everyday life
(pp. 1–7). Huddersfield: University of Huddersfield.
Savin-Baden, M., & Van Niekerk, L. (2007). Narrative inquiry: Theory and
practice. Journal of Geography in Higher Education, 31(3), 459–472.
Semugoma, P., Nemande, S., & Baral, S. D. (2012). The irony of homophobia
in Africa. Lancet, 380(9839), 312–314. https://doi.org/10.1016/
S0140-6736(12)60901-5
Sexual Minority Uganda. (2016). “And That’s How I Survived Being Killed”
Testimonies of human rights abuses from Uganda’s sexual and gender minorities.
Retrieved from Kampala, Uganda.
Silverstone, R. (2007). Morality and the media. On the rise of the Mediapolis.
Cambridge: Polity Press.
Skjerdal, T. S. (2012). The three alternative journalisms of Africa. International

Communication Gazette, 74(7), 636–654.
Small, T. A. (2011). What the Hashtag? A content analysis of Canadian politics
on Twitter. Information, Communication & Society, 14(6), 872–895.
Smith, C. P. (2000). Content analysis and narrative analysis. In H. T. Reis &
C. M. Judd (Eds.), Handbook of research methods in social and personality
psychology (pp. 313–335). New York: Cambridge University Press.
Ssebaggala, R. (2011). Straight talk on the Gay question in Uganda. Transition,
106(1), B-44–B-57.
Strand, C. (2011). Kill Bill! Ugandan human rights organizations’ attempts to
influence the media’s coverage of the Anti-Homosexuality Bill. Culture,
Health & Sexuality, 13(8), 917–931.
Strand, C. (2012). Homophobia as a barrier to comprehensive media coverage of
the Ugandan anti-homosexual bill. Journal of Homosexuality, 59(4), 564–579.
Strand, C. (2013). The rise and fall of a contentious social policy option–narra-
tives around the Ugandan Anti-Homosexuality Bill in the domestic press.
Journal of African Media Studies, 5(3), 275–294.
Tamale, S. (2009). A human rights impact assessment of the anti-homosexualy
Bill. Public Dialogue 18 November 2009, at Makarere University in Uganda’s
Anti-homosexuality Bill-the Great Divide. Retrieved from http://pamba-
zuka.org/en/category/features/61423
Tamale, S. (2013). Confronting the politics of nonconforming sexualities in
Africa. African Studies Review, 5(6), 31–45.
The Anti-Homosexuality Bill. (2009). No. 47, Ugandan Gazette, Volume CII
Cong. Rec.
UNESCO. (2012). Assessing the media landscape in Uganda using the media.
Retrieved from http://www.unesco-ci.org/ipdcprojects/content/assessing-
media-landscape-uganda-using-media-development-indicators
Valois, C. (2015). Virtual access: The Ugandan ‘anti-gay’ movement, Lesbian,
Gay, bisexual and transgender blogging and the public sphere. Journal of
Eastern African Studies, 9(1), 145–162.
Wahab, A. (2016). “Homosexuality/homophobia is Un-African”?: Un-mapping
transnational discourses in the context of Uganda’s anti-homosexuality bill/
act. Journal of Homosexuality, 63(5), 685–718.
Wasserman, H. (2010). Popular media, democracy and development in Africa.
Routledge.
Weiss, M. L., & Bosia, M. J. (2013). Political homophobia in comparative per-
spective. In M. L. Weiss & M. J. Bosia (Eds.), Global homophobia (pp. 1–29).
Champaign, IL: University of Illinois Press.
6
Critically Illuminating Relevant
Absences in Public Sphere Arguments
via Digital Mining of Their Weblinks:
A Software-based Pedagogy
Kieran O’Halloran
6.1 Introduction
6.1.1 Orientation
In this chapter, I highlight and illustrate a software-based method I

have devised to assist with the detection of straw man arguments, par-
ticularly where the reader is unfamiliar with the standpoint attacked in
the argument and where the data describing the standpoint are exten-
sive. It is intended for use by undergraduates on modules where critical
thinking is a key pedagogical focus. The strategy helps establish efficient
critical purchase on ‘public sphere arguments’ on the world-wide web
with which readers are unfamiliar. Public sphere arguments are intended
for wide consumption in the public domain, appearing in mass plat-
forms such as newspapers—print or online—or other major online
forms of distribution such as blogs. They are part of our cultural
K. O’Halloran (*)
King’s College London, London, UK

160 K. O’Halloran
connective tissue, having the power to shape debate and agendas. With
the extraordinary advance of the world-wide web, written public sphere
arguments are more pervasive than ever. Learning how to evaluate effec-
tively such arguments, then, is an important skill, vital to meaningful
participation in online debate.
The critical analysis strategy of this chapter fastens on weblinks
inserted into an argument. Weblinking is a common practice for direct-
ing the reader to the text(s) being criticised in the argument. Weblink
content may be long, however, and thus may not have been checked
carefully by the argument’s author. What if the weblink contains mate-
rial which conflicts with the argument’s framing of the standpoint it is
criticising? If this framing is rendered inaccurate due to relevant absences,
then the argument’s credibility is in doubt. That is to say, it has been
identified as a ‘straw man’ argument and thus can be rejected. A key
pedagogical advantage of focusing on weblink content is a student is
able to attain a critical grip on a public sphere argument with an unfa-
miliar topic in a targeted, and thus non-open-ended manner, while
simultaneously increasing their knowledge of a new domain of debate.
Significantly, this strategy relies on easy-to-use software to extract recur-
rent information from the weblink content which is absent from the
argument. As I show, not only does the software-based data extraction
markedly decrease human labour, but it also has a key methodological
advantage: it enhances rigour in significantly reducing arbitrariness in
analysis of the argument, selection bias and human error.
6.1.2 Organisation
In the next section, I discuss ‘straw man’ arguments and how a digital text
analysis tool, Sketchengine1 (Kilgarriff et al., 2014), is useful for helping
to ascertain this fallacious type of argument where a reader is not familiar
with the standpoint being attacked in the argument, and where this
standpoint is long. In Sect. 6.3, I highlight how the cohesion of an argu-
ment, how it hangs together through its vocabulary and grammar, is
important to how it frames the standpoint it attacks. I also spotlight a key
point of this chapter: an argument may only appear cohesive because of
Critically Illuminating Relevant Absences in Public Sphere… 161
what it excludes. Revealing relevant absences from an argument can

impact on its cohesive structure and, in turn, the credibility of how it
frames the standpoint it attacks.
Section 6.4 includes the data I examine to illustrate the critical
approach to public sphere arguments I put forward. The data is an argu-
ment in a UK national newspaper which criticised the UK Parliament’s
decision, in September 2014, to launch air strikes against the ultra-violent
Islamist group, ISIS, in Iraq. The argument contains a weblink to the
parliamentary debate. This debate is a long text consisting of over 22,000
words.
In Sect. 6.5, I model how Sketchengine can be used conveniently to
extract the most recurrent words and expressions in this long debate
which are absent from the argument. On the basis of this quantitative
information, Sect. 6.6 performs a deconstructive analysis of the argu-
ment, revealing it to be a straw man. I accomplish this by highlighting
how the argument’s cohesion fractures because it omits relevant material
which was identified in the parliamentary debate via these recurrent
words and expressions.
6.2 Straw Man Arguments

6.2.1 Definition
I have mentioned ‘straw man’ arguments, but not yet given a definition:
the technique used when an arguer ignores their opponent’s real position
on an issue and sets up a weaker version of that position by misrepresenta-
tion, exaggeration, distortion or simplification.
(Bowell and Kemp, 2015, p. 252)
The straw man argument is thus fallacious because of what is absent from
it. Specifically, it is a dialectical fallacy. While ‘dialectic’ has meant differ-
ent things since the time of Plato, it is commonly used now to refer to the
exchange of opinions between debaters and the rules of engagement
which facilitate meaningful debate or, as Wenzel (1990, p. 14) puts it, ‘a
162 K. O’Halloran
procedure for regulating discussions among people’. So, a straw man

argument flouts a dialectical obligation that the arguer should represent
accurately the standpoint that they are attacking. This does not just apply
to spoken debate. A written arguer may also flout dialectical standards by
producing a straw man.
6.2.2 T
ypes of Straw Man and Reasonable Dialectical
Obligations
Some scholars go beyond a general definition of a straw man argument,

such as Bowell and Kemp’s, by discriminating sub-types. Talisse and
Aikin (2006) argue for two forms of straw man: (1) misrepresentation and
(2) weak. The first form involves a speaker or writer advancing an argu-
ment which misrepresents, in part, the standpoint they are attacking. The
second is not a misrepresentation. Instead, it involves the antagonist
selecting the weakest version of a protagonist’s argument, or non-central
aspects of the standpoint, because they are more readily criticisable.
Aikin and Casey (2011) build on the stance of Talisse and Aikin (2006)
by arguing for a further sub-type of straw man argument—the hollow
man argument. While the misrepresentation and weak man bear some
resemblance to the standpoint which is attacked in the argument, the
hollow man is a complete fabrication. The proponent of the standpoint
which is being attacked simply did not advance anything remotely simi-
lar to that standpoint.
While the straw man argument contravenes a dialectical obligation,
authors can hardly be obligated to engage with every single element of a
standpoint, especially where space is an issue. It would be unreasonable
to expect this, say, in a short newspaper opinion piece. What is dialecti-
cally reasonable to expect is that: (1) the arguer gives an accurate account
of the main element(s) of the standpoint of the other party, commensu-
rate with the space available; and (2) if they go on only to focus on a
particular element of the standpoint, perhaps a peripheral one, then they
are explicit that they are doing so and provide reasons for their specific
focus.
6.2.3 U
sing Sketchengine Software to Assist
in Ascertaining Straw Man Arguments
In this chapter, I show how a digital tool can be used to help conveniently
ascertain if an argument which criticises an unfamiliar standpoint is a
straw man. The tool, Sketchengine, can automatically find the most
recurrent words and expressions in one text which are absent from
another text. This is very useful for my purposes. If one of the texts is an
argument attacking a standpoint and the other text describes that stand-
point, use of this software function can assist with the following: rigor-
ously and efficiently ascertaining any significant absences from the
argument when it describes that standpoint. This is particularly useful
and convenient where the standpoint data is long. I will use Sketchengine
for these ends in Sect. 6.5.
Following on from the quantitative analysis, it is important that we
also inspect the standpoint data qualitatively to make sure that absences
from the argument are relevant absences. Many of the most recurrent
words and expressions in the standpoint data are likely to be indices of
the main elements of that standpoint. A methodological advantage of
having qualitative engagement with the standpoint data directed by the
contrastive quantitative results is that this qualitative engagement is
(1) targeted, and thus efficient; (2) not arbitrary. In turn, arbitrariness is
significantly reduced in judgement of where the argument misrepresents
the standpoint, facilitating in principle a scrupulous evaluation of the
argument’s dialectical status.
6.3 A
scertaining How the Argument Frames
the Standpoint it Attacks
6.3.1 An Argument’s Cohesion and Coherence
To facilitate evaluation of whether or not an argument is a straw man, in

the first place the analyst needs to know accurately how the argument has
164 K. O’Halloran
framed the standpoint. In my use of ‘framing’, I echo Robert Entman’s

well-known definition:
Framing essentially involves selection and salience. To frame is to select

some aspects of a perceived reality and make them more salient in a com-
municating text, in such a way as to promote a particular problem defini-
tion, causal interpretation, moral evaluation, and/or treatment
recommendation for the item described. Typically frames diagnose, evalu-
ate, and prescribe […]
(Entman, 1993, p. 52)
Since framing involves selection, by the same token it may involve signifi-
cant exclusion. The latter might involve absences which are deliberate
and/or inadvertent.
An argument’s framing can be seen in the recurrent vocabulary that the
arguer uses to describe the topic and standpoint they attack. Identifying
how an argument has framed its standpoint through recurrent use of the
same lexis and recurrent use of different but semantically related lexis, in
effect, reveals dominant lexical and semantic cohesion in the text of the
argument. By lexical/semantic cohesion, I mean how a text is tied together
by its vocabulary and meaning. For example, in:
Mary had a little lamb. Its fleece was white as snow.
cohesion is created across the sentences through the semantically related

lexis ‘lamb’ and ‘fleece’. Cohesion can also be grammatical. For instance,
the pronoun ‘its’ links cohesively to ‘lamb’.
Cohesion in a text is hardly trivial. Indeed, as the linguists Ronald
Carter and Walter Nash say, ‘The first requirement of any composition is
that it should “hang together”’ (Carter & Nash, 1990, p. 189). Just like
any effective text, the text of an argument needs to be well formed:
Cohesion distinguishes well-formed texts, focusing on an integrated topic,

with well-signalled internal transitions […]. It is founded on a very simple
principle: each sentence after the first is linked to the content of one or
more preceding sentences by at least one tie.
(Fowler, 1996, p. 83)
Cohesion is crucial, then, to the effectiveness of an argument and thus to

its persuasiveness.
6.3.2 A
n Argument May Appear Cohesive/Coherent
Because of What it Excludes
In linguistics, cohesion is often discussed in relation to another concept—

coherence. This is the experience in reading or listening that the meaning
of a text is unified. Coherence is a mental property. In contrast, cohesion
is a property of the text.2 Cohesion is usually necessary to ensure our
experience of reading a text is coherent, certainly where the text is not
short. But, other factors are required for coherence such as relevant back-
ground knowledge.
Like any text, a public sphere argument needs to be cohesive to be
effective. But, what if we find out that the argument has excluded key
aspects of the standpoint it attacks? Highlighting relevant absences in the
argument may disrupt its cohesion. If an argument’s cohesion suffers, if
its sentences no longer stick together on the page, then there may well be
repercussions for the sense we can make of it. If our reading comprehen-
sion suffers as a result of loss of cohesion, this must mean that the argu-
ment also lacks coherence. In turn, the credibility of how it has framed
the standpoint it attacks may suffer. The critical reading strategy of this
chapter thus rests on the following idea: an argument may appear cohesive
on the page and coherent in our reading because of what it excludes.
I move on now to the data I use to model this critical approach to
public sphere arguments.
6.4 Data
6.4.1 The Argument
The argument was written by the environmentalist and left-wing activist,

George Monbiot. The background for the argument is a British parliamen-
tary debate in the House of Commons on 26 September 2014. Members
166 K. O’Halloran
of Parliament debated whether or not to allow the UK Royal Air Force, as

part of a US military coalition, to conduct air strikes in Iraq against the
Islamic State of Iraq and Syria (ISIS), the ultra-violent Salafist militia.3 In
his argument, Monbiot criticises the content of the UK Parliament’s debate
and the ultimate decision to permit air strikes. Monbiot’s argument is 1067
words in length and was published in The Guardian on 30 September
2014. It can be found in Sect. 6.4.2 with my annotations. But, if the reader
would prefer to look at an unannotated version first, I have footnoted the
URL.4 At the bottom of Monbiot’s argument in The Guardian is the infor-
mation that a fully referenced version can be found at George Monbiot’s
personal webpage, www.monbiot.com. Here Monbiot provides a number
of weblinks in his argument. In Sects. 6.5 and 6.6, I will show how digital
mining of one of the argument’s weblinks leads to the deconstruction of the
argument’s framing of the standpoint it criticises, thus revealing Monbiot’s
argument to be a straw man.
A few words about why I chose this argument. I admire George
Monbiot5 and consume his journalism regularly. Reading the journalism
of a thinker one habitually agrees with, however, potentially creates a
confirmation bias or, in more contemporary parlance, reinforces the
reader’s ‘filter bubble’ (Pariser, 2011). In modelling this approach to criti-
cal argument analysis, I took the opportunity potentially to ‘burst’ this
bubble.
6.4.2 U
sing Software to Help Highlight the Dominant
Framings of the Argument
Before I can begin to explore whether or not Monbiot’s framing of the

standpoint he criticises fissures because of relevant absences, I need to iden-
tify this framing, or in linguistic terms, dominant aspects of its cohesive
structure. To help identify these as accurately as possible, I used Sketchengine.
It is advisable to use a software tool to find recurrent words and expressions
in an argument since this significantly reduces human error (and labour) in
identifying its framing of the standpoint it criticises. In turn, this enhances
the rigour of any subsequent evaluation of straw man status.
I loaded up Monbiot’s argument to Sketchengine6 and generated a
lemma frequency list for lexical words.7 Many words have different word
forms, e.g. ‘go’, ‘goes’, ‘going’, ‘gone’, ‘went’ are all part of the same word
family. The general word which encompasses different word forms—
morphologically equivalent to the simplest word form—is referred to by
linguists as the lemma. So go is the lemma for the above word forms.8
Generating lemmas from a large body of data is useful. Aggregating dif-
ferent morphologically related lexical word forms under one label pro-
vides a convenient bird’s-eye view on lexical content. For my specific
purposes here, this helps establish a keener sense of lexical repetition, and
thus lexical cohesion, across an argument. To help achieve the most effec-
tive ‘bird’s-eye’ view on the lexical content of Monbiot’s argument, I
treated all data as lower-case.
Table 6.1 shows frequencies for lexical lemmas which include word
forms repeated at least four times. That is to say, these could be words
contributing to dominant patterns of lexical cohesive structure and thus
repeated framing of the standpoint Monbiot attacks. I focus on lexical
words because I am interested in the argument’s conceptual framing. I
used a ‘stopword list’ to filter grammatical words,9 allowing me to
Table 6.1 Frequencies for lemmas in Monbiot’s argument with a threshold of

four using a stopword list
Rank Freq. Lemma Word Forms
1 10 isis isis (x10)
2 8 saudi saudi (x8)
3 8 bomb bomb (x2), bombs (x4), bombed (x2)
4 7 government government (x3), governments (x4)
5 7 people people (x7)
6 5 british british (x5)
7 5 death death (x5)
8 5 moral moral (x5)
9 5 murder murder (x2), murders (x1), murdered (x2)
10 5 shia shia (x5)
11 4 al qaida al qaida (x4)
12 4 humanitarian humanitarian (x4)
13 4 militia militia (x1), militias (x3)
14 4 muslim muslim (x4)
15 4 network networks (x3), network (x1)
16 4 obama obama (x4)
17 4 stop stop (x4)
18 4 war war (x4)
All data treated as lower case.
168 K. O’Halloran
appreciate frequent lexical content more clearly. Stopwords are words

that the software user wants automatically filtered in their word list gen-
eration. There are stopword lists freely available on the web.10 As the
reader can see from what is highlighted in Table 6.1, relatively frequent
lexical words in Monbiot’s argument include ‘moral’ (x5), ‘humanitarian’
(x4) and ‘bomb’ (x8).
My choice of threshold of four words in Table 6.1 reflects my need to
ascertain the dominant lexis across the argument without being over-
whelmed by too much data. At the same time, the threshold is an arbi-
trary one. Given this, the analyst should go on to inspect whether or not
there is other lexis in the argument, perhaps occurring once or twice,
which may contribute to dominant cohesive chains. This can easily be
done in Microsoft Word by looking for the stem of a word. For example,
now I know that bomb (x8) is relatively frequent, looking for ‘bomb’ also
generates the lexically related ‘bombing’ (x1), ‘bombings’ (x1) and ‘bom-
bardment’ (x1). There are then 11 words identified with the stem ‘bomb’.
If we treat these as a single lemma, then it becomes the most frequent in
the argument.
Lemma generation is certainly useful in facilitating systematic tracing
of an argument’s cohesion, and thus its framing of the position it attacks,
as well as helping to reduce errors in doing so. All the same, software can
only recognise lexis and not meaning. Given this limitation, the analyst
should eschew too much subservience to quantitative data when making
judgements about meaning, being free instead to group words semanti-
cally into various cohesive chains. For example, ‘moral’ and ‘humanitar-
ian’ are part of the same semantic field; together they provide semantic
cohesion in Monbiot’s argument. If we treat ‘moral’ and ‘humanitarian’
as a single item, it consists of 9 words—becoming then the third most
common ‘lemma’ in Monbiot’s argument. The most frequent lemma,
isis, denotes a key social actor in the text rather than how a topic or social
actor is framed. As is clear from Fig. 6.1, this means that the most domi-
nant framings of the argument are via the lemma bomb (broadly con-
ceived) and ‘moral’/‘humanitarian’. Figure 6.1 highlights these dominant
framings in Monbiot’s argument with three types of annotation:
• BOLD: ‘moral’ (x5); ‘humanitarian’ (x4);

• UNDERLINED: repeated two-word collocations containing ‘moral’

and ‘humanitarian’, i.e., ‘moral imperative’ (x3) and ‘humanitarian
arguments’ (x2);
• HIGHLIGHTED: ‘bomb’ (x2), ‘bombs’ (x4), ‘bombed’ (x2), ‘bomb-
ing’ (x1), ‘bombings’ (x1); ‘bombardment’ (x1).
Another reason for annotating these cohesive chains is that they relate
specifically to how the argument is framed in its headline—an important
foregrounding and thus rhetorical device in the argument. Lastly, I have
alphabetized the paragraphs of Monbiot’s argument in Fig. 6.1.
[A]
Why stop at Isis when we could bomb the whole Muslim world?
Humanitarian arguments, if consistently applied, could be used to flatten the
entire Middle East
[B]
Let’s bomb the Muslim world – all of it – to save the lives of its people. Surely
this is the only consistent moral course? Why stop at Islamic State (Isis), when
the Syrian government has murdered and tortured so many? This, after all, was
last year’s moral imperative. What’s changed?
[C]
How about blasting the Shia militias in Iraq? One of them selected 40 people
from the streets of Baghdad in June and murdered them for being Sunnis.
Another massacred 68 people at a mosque in August. They now talk openly of
‘cleansing’ and ‘erasure’ once Isis has been defeated. As a senior Shia politician
warns, ‘we are in the process of creating Shia al-Qaida radical groups equal in
their radicalisation to the Sunni Qaida’.
Fig. 6.1 Major lexical cohesive chains in Monbiot’s argument

170 K. O’Halloran
[D]
What humanitarian principle instructs you to stop there? In Gaza this year,
2,100 Palestinians were massacred: including people taking shelter in schools
and hospitals. Surely these atrocities demand an air war against Israel? And
what’s the moral basis for refusing to liquidate Iran? Mohsen Amir-Aslani was
hanged there last week for making ‘innovations in the religion’ (suggesting that
the story of Jonah in the Qur’an was symbolic rather than literal). Surely that
should inspire humanitarian action from above? Pakistan is crying out for
friendly bombs: an elderly British man, Mohammed Asghar, who suffers from
paranoid schizophrenia, is, like other blasphemers, awaiting execution there after
claiming to be a holy prophet. One of his prison guards has already shot him in
the back.
[E]
Is there not an urgent duty to blow up Saudi Arabia? It has beheaded 59 people
so far this year, for offences that include adultery, sorcery and witchcraft. It has
long presented a far greater threat to the west than Isis now poses. In 2009
Hillary Clinton warned in a secret memo that ‘Saudi Arabia remains a critical
financial support base for al-Qaida, the Taliban … and other terrorist groups’. In
July, the former head of MI6, Sir Richard Dearlove, revealed that Prince Bandar
bin Sultan, until recently thehead of Saudi intelligence, told him:‘The time is
not far off in the Middle East, Richard, when it will be literally ‘God help the
Shia’. More than a billion Sunnis have simply had enough of them.’ Saudi
support for extreme Sunni militias in Syria during Bandar’s tenure is widely
blamed for the rapid rise of Isis. Why take out the subsidiary and spare the
headquarters?
Fig. 6.1 Continued

[F]
<weblink>The humanitarian arguments aired in parliament last
week<weblink>, if consistently applied, could be used to flatten the entire
Middle East and west Asia. By this means you could end all human suffering,
liberating the people of these regions from the vale of tears in which they live.
[G]
Perhaps this is the plan: Barack Obama has now bombed seven largely Muslim
countries, in each case citing a moral imperative. The result, as you can see in
Libya, Iraq, Pakistan, Afghanistan, Yemen, Somalia and Syria, has been the
eradication of jihadi groups, of conflict, chaos, murder, oppression and torture.
Evil has been driven from the face of the Earth by the destroying angels of the
west.
[H]
Now we have a new target, and a new reason to dispense mercy from the sky,
with similar prospects of success. Yes, the agenda and practices of Isis are
disgusting. It murders and tortures, terrorises and threatens.<weblink> As
Obama says it is a ‘network of death’. <weblink> But it’s one of many networks
of death. Worse still, a western crusade appears to be exactly what Isis wants.
[I]
Already Obama’s bombings have brought Isis and Jabhat al-Nusra, a rival
militia affiliated to al-Qaida, together. More than 6,000 fighters have joined Isis
since the bombardment began. They dangled the heads of their victims in front
of the cameras as bait for war planes. And our governments were stupid enough
to take it.
Fig. 6.1 Continued

172 K. O’Halloran
[J]
And if the bombing succeeds? If – and it’s a big if – it manages to tilt the
balance against Isis, what then? Then we’ll start hearing once more about Shia
death squads and the moral imperative to destroy them too – and any civilians
who happen to get in the way. The targets change; the policy doesn’t. Never
mind the question, the answer is bombs. In the name of peace and the
preservation of life, our governments wage perpetual war.
[K]
While the bombs fall, our states befriend and defend other networks of death.
The US government still refuses – despite Obama’s promise – to release the 28
redacted pages from the joint congressional inquiry into 9/11, which document
Saudi Arabian complicity in the US attack. In the UK, in 2004 the Serious Fraud
Office began investigating allegations of massive bribes paid by the British
weapons company BAE to Saudi ministers and middlemen. Just as crucial
evidence was about to be released, Tony Blair intervened to stop the
investigation. The biggest alleged beneficiary was Prince Bandar. The SFO was
investigating a claim that, with the approval of the British government, he
received £1bn in secret payments from BAE.
Fig. 6.1 Continued
To sum up: by annotating the argument in this way, it becomes very

clear that Monbiot repeatedly frames:
• the parliamentary debate on ISIS in moral/humanitarian terms;

• the military action ensuing from the decision to engage ISIS in Iraq in
terms of aerial bombing. The implication is that responsibility for the
decision to bomb rests with the UK Parliament/US and western gov-
ernments more generally.
[L]
And still it is said to go on. Last week’s Private Eye, drawing on a dossier of
recordings and emails, alleges that a British company has paid £300m in bribes
to facilitate weapons sales to the Saudi national guard. When a whistleblower in
the company reported these payments to the British Ministry of Defence, instead
of taking action it alerted his bosses. He had to flee the country to avoid being
thrown into a Saudi jail.
[M]
There are no good solutions that military intervention by the UK or the US can
engineer. There are political solutions in which our governments could play a
minor role: supporting the development of effective states that don’t rely on
murder and militias, building civic institutions that don’t depend on terror,
helping to create safe passage and aid for people at risk. Oh, and ceasing to
protect, sponsor and arm selected networks of death. Whenever our armed forces
have bombed or invaded Muslim nations, they have made life worse for those
who live there. The regions in which our governments have intervened most are
those that suffer most from terrorism and war. That is neither coincidental nor
surprising.
[N]
Yet our politicians affect to learn nothing. Insisting that more killing will
magically resolve deep-rooted conflicts, they scatter bombs like fairy dust.
© The Guardian 2014
Fig. 6.1 Continued
The reader will have noticed that I have annotated portions of Fig. 6.1
with ‘<weblink>’. This is because, in paragraphs [F] and [H], Monbiot
includes a weblink to the official textual record of the British parliamen-
tary debate he criticises.11 In Sect. 6.5, I digitally mine the most recurrent
lexical words and expressions in this debate which are absent from
174 K. O’Halloran
Monbiot’s argument. Then, in Sect. 6.6, I show how these absences lead
to problems for the credibility of how Monbiot’s argument frames the
debate.
6.5 A
scertaining Absences from Monbiot’s
Argument that Are Recurrently Present
in the Parliamentary Debate
6.5.1 Rationale
There are 21 weblinks in Monbiot’s argument on www.Monbiot.com.12

Why do I choose the particular weblink in paragraph [F] for data extrac-
tion? First, the text around the weblink—‘The humanitarian arguments
aired in parliament last week’—echoes ‘Humanitarian arguments’ in the
title of Monbiot’s text. In other words, the text neighbouring the weblink
echoes a significant part of how Monbiot’s argument is framed. A second
reason is the size of the weblink’s content. The text of the parliamentary
debate consists of 22,433 words in total. Given its length, there may be
relevant material in this debate which Monbiot is unaware of and, in
turn, does not know is in conflict with his argument’s framing. A third
rationale is Monbiot’s repeated usage. The weblink appears twice—again
in paragraph [H].13
6.5.2 D
igitally Mining Presences in the Weblink
which Are Absent from the Argument
I now turn to the procedure for ascertaining words or expressions which

are the most frequent in the weblink but absent from the argument.
Users load up two text files to Sketchengine and then activate the ‘Word
List’ function. This function usefully reveals words which are absent
from one text, but relatively frequent in another. Moreover, I used the
same stopword list from 4.2 in order to remove the ‘noise’ of grammati-
cal words as well as other common words (e.g., temporal adverbs like
‘often’). Table 6.2, generated using Sketchengine’s ‘word list’ function,
shows lexical words repeated five or more times in the parliamentary

debate which are absent from Monbiot’s argument. I used five words as
the threshold since this was the default setting in the software; I have also
edited out words which reflect parliamentary interpersonal discourse
(e.g. ‘right honourable lady’).
A similar procedure can be conducted for two or more word expres-
sions. Indeed, since meaning is habitually conveyed in collocation, this is
both useful and necessary. Table 6.3 shows all two-word expressions in
the parliamentary debate which occur at least five times but are absent
from Monbiot’s argument.
6.5.3 A
Non-arbitrary, Targeted Qualitative
Engagement Directed by Quantitative Results
Just because Tables 6.2 and 6.3 show the most frequent words and two-
word expressions which are absent from Monbiot’s argument, it does not
mean they are relevant absences. In order to ascertain this, I need to
examine qualitatively how these words are used in the debate. This can be
done, simply enough, using the ‘Find’ function in Word. Alternatively,
all immediate co-texts of these words can be grouped together using the
‘concordance’ function of Sketchengine (see Sect. 6.6.3). Since I am only
Table 6.2 The most frequent lexical words in the parliamentary debate which are
absent from Monbiot’s argument
1. iraqi 53 14. kurds 15 27. genocide 8
2. ground 38 15. president 13 28. equipment 7
3. international 32 16. prime 13 29. regime 6
4. legal 28 17. minister 13 30. turkey 6
5. britain 25 18. muslims 12 31. violence 6
6. strikes 25 19. council 11 32. hostage 5
7. security 23 20. kurdish 11 33. legitimate 5
8. coalition 22 21. citizens 11 34. legitimacy 5
9. arab 22 22. democratic 10 35. peshmerga 5
10. strategy 21 23. islam 10 36. savagery 5
11. army 21 24. extremists 9 37. tribes 5
12. assad 19 25. resolution 9 38. christians 5
13. troops 16 26. barbarism 8 39. yazidi(s) 5
176 K. O’Halloran
pursuing up to 36 words, and up to 20 two-word expressions, with sev-

eral of the 36 words making up two-word expressions (e.g., ‘Iraqi army’),
this makes for a more efficient read of data in excess of 22,000 words than
a reading which is not targeted in this way. Moreover, since the most
recurrent lexical words in the debate which are not in Monbiot’s argu-
ment direct my qualitative engagement with the parliamentary debate,
this engagement is not then arbitrary.
It is important to highlight that sometimes there are lexical differences
between Monbiot’s argument and the parliamentary debate, but this does
necessarily translate into semantic differences. For example, ‘air strikes’ (x
21) is used in the parliamentary debate. In contrast, Monbiot prefers the
denotationally equivalent ‘bombs’, presumably because its connotation
carries greater rhetorical punch. The analyst thus needs to take care to
avoid automatically assuming a lexical absence from the argument trans-
lates into a semantic absence vis-à-vis the standpoint data.
I move on to discussion of words and collocations which are recurrent in
the parliamentary debate but relevant absences from Monbiot’s argument.
6.5.4 Relevant Absences from Monbiot’s Argument
6.5.4.1 The Legality of Air Strikes
In the parliamentary debate, there are 53 instances of ‘Iraqi’ (Table 6.2), of

which 23 occur in ‘Iraqi government’ (Table 6.3). This phrase is frequently
Table 6.3 The most frequent two-word expressions in the parliamentary debate
which are absent from Monbiot’s argument
1. military action 26 11. combat troops 7
2. iraqi government 23 12. air power 7
3. air strikes 21 13. united states 6
4. security council 10 14. president obama 6
5. iraqi army 10 15. security forces 5
6. legal basis 9 16. legal base 5
7. united nations 8 17. international community 5
8. international coalition 7 18. democratic state 5
9. national interest 7 19. british people 5
10. ground forces 7 20. arab countries 5
repeated in relation to the legal status of the air strikes. They were deemed
to be legal because the ‘Iraqi government’ invited air strikes by the UK/US
on ISIS targets in Iraq.
(1) The Prime Minister: The Iraqi Government have requested our help
and given their clear consent for UK military action, […] there is no ques-
tion but that we have the legal basis for action […] [In quotation (1) from
the parliamentary debate above and quotations (2), (3) and (4) from the
same below, I indicate in bold recurrent words and two-word expressions
which are absent from Monbiot’s argument.]
If a democratic (x10) government solicits the military assistance of

another democratic government, then subsequent military intervention
is legitimate within international law. The relatively frequent use of ‘legal’
(x28), ‘legal basis’ (x9) and ‘legal base’ (x5), as well as ‘legitimate’ (x5) and
‘legitimacy’ (x5), also reflected the importance to all parties in the debate
of the legality of air strikes (as compared with the contested legal status of
the US-Coalition invasion of Iraq in 2003).
6.5.4.2 Providing Air Strikes to Assist on the Ground Forces
Other important absences from Monbiot’s argument are ‘Kurds’ (x15)

and ‘Kurdish’ (x11). One of the reasons these words are relatively fre-
quent is the Kurdish President also requested British air support (‘MP’
refers to ‘Member of Parliament’):
(2) Nadhim Zahawi MP: The Kurdish President is on record as saying that
the Kurds do not want British servicemen and women on the ground
fighting the fight for them. What they need is better equipment, training
and the air support.
Echoing the above, the ‘Peshmerga’ (x5), the Kurdish military force,
wanted to remain in control of ‘ground’ (x38) offensives from Kurdish
territory; they were requesting more and improved ‘equipment’ (x7) to
facilitate this. They were not requesting Western ground forces. Iraqi
178 K. O’Halloran
security forces in non-Kurdish territory had exactly the same appeal. In a

nutshell, requested air strikes by the Royal Air Force were intended to
ease military engagement on the ground with ISIS by Peshmerga and
Iraqi security forces. Bombing of ISIS targets was, then, only part of a
requested and co-ordinated military strategy between UK/Western air
support and indigenous troops on the ground.
6.5.4.3 Demographic Complexity
Kurds are not the only ethnic/religious groups affected by the incursion
of ISIS into Iraq. Christians (x5) and Yazidis (x5) were targeted by ISIS,
as were Shias and other Sunnis. This is very clear from the motion that
the House of Commons is debating:
(3) That this House condemns the barbaric acts of ISIL against the peoples
of Iraq including the Sunni, Shia, Kurds, Christians and Yazidi and the
humanitarian crisis this is causing; recognises the […] request from the
Government of Iraq for military support from the international commu-
nity and the specific request to the UK Government for such support;
further recognises the threat ISIL poses to wider international security
and the UK directly through its sponsorship of terrorist attacks and its
murder of a British hostage […]
That mention of Kurds, Christians and Yazidis is absent from Monbiot’s

argument would suggest that he does not deal with the complexity of
social actors affected by the growth of ISIS.
6.5.4.4 UK/International Security
A key reason for supporting Iraqi security forces and the Kurdish
Peshmerga with air strikes was certainly humanitarian—to help prevent
the slaughter of Yazidis and Christians. But, there was another significant
reason—national and international security—which Monbiot does not
mention. National security is reflected in use of ‘Britain’ (x25) such as in:
(4) The Prime Minister: […] [ISIS] is a threat to Iraq and a threat to
Britain […]
as well as in ‘national interest’ (x7). ‘International security’ is reflected by

the relatively frequent use of ‘international’ (x32). Indeed, ‘international
security’ features in the debate motion extract in Sect. 6.5.4.3.
It is time now to see how these recurrent features of the parliamen-
tary debate, which are all relevant absences from Monbiot’s argument,
affect the coherence of the latter. Any effective text is dialogical in
anticipating and responding to the question posing of the reader.
Monbiot’s argument is, in fact, markedly dialogical because of the
number of explicit questions it contains (13 in total). In Sect. 6.6, I
model how a user could respond in a critical dialogue with Monbiot’s
argument by drawing attention to relevant absences from the stand-
point he attacks. In doing so, I show how Monbiot’s framing of the
standpoint he criticises cracks.
6.6 C
ritical Deconstructive Dialogue
with Monbiot’s Argument
6.6.1 A
bsence I: Legality of the Airstrikes and (Inter)
national Security
Monbiot’s framing of the British Parliament’s endorsement of aerial

bombing as ‘moral’/‘humanitarian’ omits its legal status. In other words,
it ignores a key reason why the Members of Parliament decided to endorse
aerial bombing. Another reason for air strikes that Monbiot omits is
(inter)national security. These absences affect the stability of the cohesive
chain of ‘moral’/‘humanitarian’ and thus the coherence of the argument
as shown in Fig. 6.2:
6.6.2 A
bsence II: Action with Iraqi/Kurdish Forces Is
Requested and Co-ordinated
Repeated instances of ‘bomb’ in Monbiot’s argument frame under-

standing of the proposed military strategy to deal with ISIS in Iraq.
However, absent from his argument is the information that air strikes
are not only requested, but are just one aspect of the military strategy.
180 K. O’Halloran
[A]
entire Middle East
[B]
Let’s bomb the Muslim world – all of it – to save the lives of its people. Surely
this is the only consistent moral course? Why stop at Islamic State (Isis), when
the Syrian government has murdered and tortured so many? This, after all, was
last year’s moral imperative. What’s changed?
[…]
‘What’s changed?’ Answer: the US

coalion was invited by a
democrac naon, Iraq, to
provide air support.
This invitaon guaranteed a legal

basis for this military assistance –
i.e., the basis for intervenon was
not just humanitarian / moral. You
have omied to menon this legal
basis .
Fig. 6.2 How the cohesive structure of ‘moral’/‘humanitarian’ and thus coher-
ence of the argument are disturbed by relevant absences
That is to say, they are part of a co-ordinated action for assisting on-
the-ground forces of the Iraqi military and the Kurdish Peshmerga. As
Fig. 6.3 shows, this relevant absence destabilises the cohesive struc-
ture, and thus framing, of Monbiot’s argument in relation to ‘bomb’.
Loss of coherence ensues:
[D]
What humanitarian principle instructs you to stop there? In Gaza this year,
2,100 Palestinians were massacred: including people taking shelter in schools

and hospitals. Surely these atrocities demand an air war against Israel? And
what’s the moral basis for refusing to liquidate Iran? Mohsen Amir-Aslani was
hanged there last week for making ‘innovations in the religion’ (suggesting that
the story of Jonah in the Qur’an was symbolic rather than literal). Surely that
should inspire humanitarian action from above?
[…]
[F]
<weblink>The humanitarian arguments aired in parliament last
week<weblink>, if consistently applied, could be used to flatten the entire
Middle East and west Asia. By this means you could end all human suffering,
liberating the people of these regions from the vale of tears in which they live.
You have omied the

perceived need to deal with
ISIS in Iraq because of
(inter)naonal security.
In other words, a key reason for providing air

support was (inter)naonal security interest,
and not only humanitarian/ moral concerns.
Fig. 6.2 Continued

182 K. O’Halloran
[G]
countries, in each case citing a moral imperative. The result, as you can see in
Libya, Iraq, Pakistan, Afghanistan, Yemen, Somalia and Syria, has been the
eradication of jihadi groups, of conflict, chaos, murder, oppression and torture.
[…]
[J]
death squads and the moral imperative to destroy them too –
[…]
Fig. 6.2 Continued
6.6.3 A
bsence III: Occluding Social Actors Affected
by Air Strikes Through General Categories
Omission in Monbiot’s argument does not just work through stark

absence. It also works through use of related general categories to occlude
more specific ones (whether deliberately or inadvertently). Another rela-
tively frequent category in Monbiot’s argument is ‘Muslim’ (x4).
Figure 6.4 shows all these instances in the parliamentary debate stacked
neatly on top of one another in a ‘concordance’ generated with
Sketchengine. As Fig. 6.4 shows, ‘Muslim’ collocates with categories of
general geography and polity—‘world’, ‘countries’ and ‘nations’.
Use of these general expressions with ‘Muslim’ occludes the ethnic
complication in Iraq (Kurds, Arabs, etc.) as well as the religious compli-
cation (not just Sunni and Shia Muslims but Christians and Yazidis).
[A]
entire Middle East
[B]
Let’s bomb the Muslim world – all of it – to save the lives of its people. […]
Pakistan is crying out for friendly bombs: an elderly British man, Mohammed
Asghar, who suffers from paranoid schizophrenia, is, like other blasphemers,
awaiting execution there after claiming to be a holy prophet.
[…]
[G]
countries, in each case citing a moral imperative.
[…]
[I]
Already Obama’s bombings have brought Isis and Jabhat al-Nusra, a rival
militia affiliated to al-Qaida, together. More than 6,000 fighters have joined Isis
since the bombardment began.
[…]
Fig. 6.3 Showing how the cohesive structure of ‘bomb’ and thus the coherence
of the argument are disturbed by relevant absences
Christians, Kurds and Yazidis who were being attacked by ISIS were
direct beneficiaries of the air strikes. While, in paragraph [H], Monbiot
acknowledges the barbarity of ISIS:
(5) [ISIS] murders and tortures, terrorises and threatens

184 K. O’Halloran
[J]
death squads and the moral imperative to destroy them too – and any civilians
who happen to get in the way. The targets change; the policy doesn’t. Never
mind the question, the answer is bombs.
But the policy was not just bombing. You have

omied to menon that air-strikes were part of an
invited and co-ordinated acon. Air-strikes were
intended to support on the ground military acon
by Iraqi security forces and Kurdish Peshmerga.
[K]
While the bombs fall, our states befriend and defend other networks of death.
[…]
[M]
[…]
Whenever our armed forces have bombed or invaded Muslim nations, they have
made life worse for those who live there. The regions in which our governments
have intervened most are those that suffer most from terrorism and war. That is
neither coincidental nor surprising.
Fig. 6.3 Continued

Use of ‘invaded’ above is irrelevant since US

coalion forces were not being commied in Iraq,
at the me. Importantly, this is because they were
not wanted by the Iraqi government / Peshmerga.
[N]
Yet our politicians affect to learn nothing. Insisting that more killing will
magically resolve deep-rooted conflicts, they scatter bombs like fairy dust.
You leave the false impression that air strikes

against ISIS were decided solely by the US coalion
when they were requested by the Iraqi prime-
minister and Kurdish president.
Fig. 6.3 Continued
1 Isis when we could bomb the whole Muslim world? Humanitarian arguments,
2 entire Middle East Let’s bomb the Muslim world – all of it – to save the
3 Obama has now bombed seven largely Muslim countries, in each case citing a
4 forces have bombed or invaded Muslim nations, they have made life worse
Fig. 6.4 Concordance for ‘Muslim’ in Monbiot’s argument
he does not include the grammatical objects of the verbs ‘murders’,

‘tortures’, ‘terrorises’ and ‘threatens’. These objects include ‘Kurds’,
‘Yazidis’, ‘Christians’. Thus, specific beneficiaries of the air strikes are
absent from the argument as well as occluded in the phrases with a
generalised reference fronted by ‘Muslim’. Indeed, the use of these
phrases leaves the impression that air strikes will be akin to the ‘carpet
186 K. O’Halloran
bombing’ of Iraq when they are intended to target ISIS positions only
(assuming the UK Parliament is to be believed).
6.7 Conclusion
I have shown how Monbiot’s argument is a straw man by drawing atten-
tion to relevant absences from the standpoint he criticises. His argument
is thus dialectically fallacious. More specifically, I would interpret it as a
misrepresentation straw man (Sect. 6.2.2). It is not a hollow man since it
accurately reports that Parliament voted for air strikes. All the same, it
ignores their legal basis, that air strikes have been requested by the Iraqi
government and the Kurds, and that the air strikes are part of a co-
ordinated action with Iraqi and Kurdish ground forces. Once relevant
absences are accounted for, the cohesive structure of the argument is
destabilised which, in turn, creates problems for the coherence of its
framing. In effect, I have demonstrated how a weblink in an argument is
not an outside supplement to the argument, but potentially a lurking
‘destabilising agent’. Moreover, I have highlighted that relevant absences
from an argument may occur through stark omission or through use of
general categories which obscure more relevant specific ones.
6.8 Reflection
6.8.1 U
sing Software to Help Detect Relevant
Absences from Arguments: Methodological
Advantages
This chapter has outlined a software-based method for helping to detect

relevant absences from public sphere arguments. In doing so, it echoes
the important case made by Alan Partington for employing software to
help rigorously disclose relevant absences from language use (Partington,
2014). As I have shown, employing text analysis software for evaluating
the straw man status of an argument affords a number of methodological
advantages. First, this software is very useful for efficient extraction from
big data, revealing the most recurrent words and expressions in a long
standpoint which are absent from an argument which attacks that stand-
point. Second, reading of weblink content which is directed by relatively
frequent words and expressions absent from the argument not only
diminishes the labour of reading potentially thousands of words, but aug-
ments rigour in significantly reducing arbitrariness in its qualitative study.
In turn, third, the chances that evaluation of the argument’s straw man
status is done in an arbitrary fashion is significantly decreased, making in
principle for a scrupulous assessment. See also O’Halloran (2017).
6.8.2 Pedagogical Advantages
Once an argument has been judged to be dialectically fallacious or sound,

this does not mean, naturally, that the student now needs to adopt the
standpoint criticised in the argument. On the contrary, now that they
understand both standpoints, they are in a position to decide where they
stand in the debate. The approach of this chapter can thus be used as a
pedagogy which not only engenders intellectual satisfaction and empow-
erment from highlighting straw man arguments, but also extends critical
awareness of domains of debate. Felicitously, students have a secure
(quantitative) foundation from which to develop their knowledge, and
progress to making an informed decision about where they stand. They
might, perhaps, end up rejecting both sides in favour of a fresh position.
Moreover, given that the student’s knowledge of a domain of debate has
expanded, they are better placed to assess the logical structure of related
arguments that they subsequently encounter.
Another pedagogical advantage is as follows. Since the student only
focuses on weblinks attached to the argument, data exploration is
bounded. This assists engagement with an argument attacking an unfa-
miliar standpoint in a manner which is targeted and thus not open-
endedly research-heavy. Furthermore, since students would be encouraged
to engage with public sphere arguments on topics with which they are
(largely) unfamiliar, rhetorical sensitivity is sharpened. For example, they
are in a better position to see where categories in an argument might be
insufficiently specific and differentiated in how they are used to describe
the standpoint it opposes.
188 K. O’Halloran
Echoing what I wrote in Sect. 6.4.1, the pedagogy can be used by stu-
dents to facilitate assessment of whether or not their favourite opinion
piece writers have constructed straw man arguments and, thus, in turn to
avoid confirmation bias. And, what if a student does not find any relevant
absences from their chosen argument? This is obviously a good thing! The
argument has passed a test of quality. Alternatively, because the software is
not labour-intensive to use, the student could just move on until they find
an argument which does misleadingly frame a standpoint it criticises.
6.8.3 C
ommunicating Argument Deconstruction
Via Social Media
Lastly, if the student has produced a scrupulous revelation of the dialectical

defectiveness of a current argument, they may wish to communicate this in
the public sphere. This could be done, for instance, on Twitter copying in
the username of the argument’s author if they have a Twitter presence, as
well as the website hosting the argument. In their tweet, the student could
link to their assignment on their blog. With this communication as an
optional end goal, the student may be further motivated to produce careful
digital extraction of weblink content together with rigorous use of this
information to probe the stability of the argument’s framing.
Acknowledgements I am grateful to The Guardian for permission to reproduce

‘Why stop at Isis when we could bomb the whole Muslim world?’ published on
30 September 2014.
Notes
1. A free 30-day trial is available for ‘Sketchengine’. See https://www.
sketchengine.co.uk [accessed March 2017].
2. On the cohesion/coherence distinction, see de Beaugrande and Dressler
(1981); Widdowson (2007, pp. 49–51).
3. The parliamentary debate also uses the alternative acronym, ‘ISIL’
(‘Islamic State of Iraq and the Levant’). Salafism is an austere branch of
Sunni Islam.
4. http://www.theguardian.com/commentisfree/2014/sep/30/isis-bomb-
muslim-world-air-strikes-saudi-arabia [accessed March 2017].
5. An account of Monbiot’s activism: http://www.monbiot.com/about/
[accessed March 2017].
6. ‘Plain text’ is the default format for digital text analysis, i.e., not format-
ted text.
7. Lexical words carry the main information content of a text and belong
to four classes: nouns (‘dictionary’); lexical verbs (‘walk’); adjectives
(‘hot’); adverbs (‘beautifully’).
8. Lemmas are conventionally represented in small capitals.
9. Grammatical words are non-content-based words such as determiners
(‘the’), conjunctions (‘if ’) prepositions ‘(in’), pronouns (‘he’) and auxil-
iary verbs (‘is’ in ‘he is keeping well’).
10. I used the following stopword list: https://cup.sketchengine.co.uk/stop-
words/english/ [accessed March 2017].
11. http://www.publications.parliament.uk/pa/cm201415/cmhansrd/
cm140926/debtext/140926-0001.htm#1409266000001 [accessed March
2017].
12. http://www.monbiot.com/2014/09/30/bomb-everyone/ [accessed March
2017].
13. I deleted irrelevant metadata, e.g. the time and date of the debate.
References
Aikin, S., & Casey, J. (2011). Straw men, weak men, and hollow men.
Argumentation, 25(1), 87–105.
Bowell, T., & Kemp, G. (2015). Critical thinking (4th ed.). Abingdon, UK:
Routledge.
Carter, R., & Nash, W. (1990). Seeing through language. Oxford: Blackwell.
de Beaugrande, R., & Dressler, W. (1981). Introduction to text linguistics.
London: Longman.
Entman, R. M. (1993). Framing: Toward clarification of a fractured paradigm.
Journal of Communication, 43(4), 51–58.
Fowler, R. (1996). Linguistic criticism (2nd ed.). Oxford: Oxford University
Press.
Kilgarriff, A., Baisa, V., Bušta, J., Jakubíček, M., Kovář, V., Michelfeit, J., et al.
(2014). The sketch engine: Ten years on. Lexicography, 1(1), 7–36.
190 K. O’Halloran
O’Halloran, K. A. (2017). Posthumanism and deconstructing arguments: Corpora

and digitally-driven critical analysis. London: Routledge.
Pariser, E. (2011). The Filter Bubble: What the Internet is hiding from You.
London: Penguin.
ing absences. International Journal of Corpus Linguistics, 19(1), 118–146.
Talisse, R., & Aikin, S. (2006). Two forms of the Straw man. Argumentation,
20(3), 345–352.
Wenzel, J. (1990). Three Perspectives on Argument. In R. Trapp & J. Schutz
(Eds.), Perspectives on argumentation: Essays in Honour of Wayne Brockriede
(pp. 9–16). Prospect Heights, IL: Waveland Press.
Widdowson, H. G. (2007). Discourse analysis. Oxford: Oxford University Press.
7
Silence and Absence in Chinese
Smog Discourses
Jiayi Wang and Dániel Z. Kádár
7.1 Introduction
Smog has become a living reality for a growing number of Chinese people
in China. While it has drawn increasing public attention in China and
around the world, interestingly, very little is known about the key features
of Chinese media discourses on this issue, or public reactions to these dis-
courses, especially outside the field of Chinese journalism. By focusing on
silence and absence in Chinese smog discourses, this study aims to fill this
gap. It is partly inspired by ecolinguistics which “analyses language to
reveal the stories we live by” because they influence how people think, talk
Kádár’s research has been sponsored by the Hungarian Academy of Sciences’s MTA Lendület
Research Grant (LP/2017-5).
J. Wang (*)
University of Central Lancashire, Preston, UK
D.Z. Kádár
Research Institute for Linguistics, Hungarian Academy of Sciences,
Budapest, Hungary

192 J. Wang and D.Z. Kádár
and act, and contributes to the “search for new stories” (Stibbe, 2015,
p. 183). We used the approach of corpus-assisted discourse analysis to
examine a corpus of 415 Chinese news articles with explicit mention of the
word “smog”. The findings suggest that the numbers of Chinese media
reports dramatically increase in periods when the Chinese government and
public media address this issue as part of political gatherings. However, in
such periods the issue is approached through the lens of notions that reflect
propaganda rather than genuine environmental considerations. The
silences surrounding smog coverage reveal that smog is usually back-
grounded in news articles, and in the minority of reports where smog is
foregrounded, the media tend to frame it in a way that silences the causes.
The smog discourses also tend to hide individual agency by, for example,
vague characterisation of social actors, i.e. no individuals and organisations
are named as in charge of resolving smog-related issues, but when the dis-
courses touch upon agency, it is almost always limited to a small group of
actors, i.e. the government, reinforcing the perception that only the gov-
ernment’s agency matters while promoting the authority of government.
China’s air quality has noticeably worsened in recent years. While
“smog” has become a buzzword in the Chinese media, interestingly, in
discourse and interaction studies, little is known about Chinese news
discourses on smog, and the Chinese public’s reactions to these discourses
(cf. Jia, 2014; Wang, 2014). We believe that examining this issue is
important not only from an academic but also from a public point of
view, considering the urgent need to mobilise and widen public partici-
pation in Chinese environmental protection—a need which has been
expressed by the Chinese government.
To illustrate the importance of ‘smog’ in China, let us overview the gen-
esis of the word wumai/雾霾/“smog” in Chinese language. In historical
texts, wu/雾and mai/霾 were used separately to refer to “fog” (‘Wu’雾,
1716) and “dust in the air” (‘Mai’霾, 1716) respectively. The English word
“smog” was coined in 1905, and it was translated into Chinese as yanwu/
烟雾. Although yanwu’s literal meaning is close to its English counterpart,
it was not a newly coined word in Chinese and has already had other com-
monly used meanings. Therefore, yanwu as a word for ‘smog’ was perceived
as a ‘foreign’ term to refer to air pollution in Victorian England specifically,
rather than as a word to describe a Chinese urban phenomenon. Perhaps
because of this, in around 2010, the Chinese word wumai came into exis-
Silence and Absence in Chinese Smog Discourses 193
tence as a combination of two old words/characters wu/雾/“fog” and mai/

霾/“dust in the air” to describe smog in general (Jiang, 2011), including
the smog that has become a reality of daily life in China by that time. Since
its occurrence, wumai has become conventionalised and widely used in
contemporary Chinese (Jiang, 2015; Shan, 2014), by becoming one of the
“top 10 most frequently used words of the year” in 2013 (Han, 2013).
The main reason for this dramatic difference between the old yanwu and
the modern wumai is that until 2010, the public had only been aware
that cities like Beijing are polluted, but few had realised the actual health
hazards caused by PM2.5 (the key indicator of smog), which is a particu-
late matter of 2.5 micrometers or less in size, capable of being embedded
deep in one’s lungs (PM25.com, 2014). Following the publication of a
report on PM2.5 data in China, which generated unprecedented public
attention, not only has the Chinese media picked up wumai, but also the
Chinese government has started to address wumai-related health hazards
in the forms of government work reports and press conferences. This pub-
lic attention represents a huge attitudinal change in Chinese politics.
After Deng Xiaoping’s Reforms, the Chinese leadership promoted the
industrialisation of China as an absolute national priority, and environ-
mental protection and public health issues caused by the industrialisa-
tion of the country were far less prioritised in governmental discourses. This
priority has been changing, to a certain degree, only recently.
In this study, we focus particularly on silence and absence in Chinese smog
discourses. We regard textual silence as “the omission of some piece of infor-
mation that is pertinent to the topic at hand” (Huckin, 2002, p. 348). As
Huxley (as cited in Huckin, 2002, p. 347) has put it, the “greatest triumphs
of propaganda have been accomplished, not by doing something, but by
refraining from doing. Great is truth, but greater still, from a practical stand-
point, is silence about truth” and by “simply not mentioning certain subjects
[…] propagandists have influenced opinion much more effectively than they
could have done by the most eloquent denunciations, the most compelling
of logical rebuttals”. The power of silence to affect communication is strong,
but how to identify and analyse silence in discourse has proven difficult. It has
no overt linguistic form, so the “discourse analyst has fewer formal cues to
work with and must compensate with more attention to sociopolitical, cul-
tural and rhetorical factors” (Huckin, 2002, p. 353). For example, using
qualitative discourse analysis to identify textual silence by noting all men-
tioned subtopics and giving them graded marks according to their promi-
nence, Huckin (2002) shows silence operates in the corpus of 163 newspaper
articles and editorials on homelessness published in the US; in a similar fash-
ion, Sweeney (2012) points out that in U.S. media there is a recurrent sense
of silence in the coverage of intimate partner homicide. In an earlier study of
strategic silence, or reverse agenda setting, in Depression-era news coverage in
Seattle, Haarsager (1991) has addressed the phenomenon of silence from a
qualitative angle. Interestingly, none of these previous studies have applied
corpus tools—which is a methodological gap that we aim to address in this
paper. As Taylor (2012) noted insightfully, there is a variety of approaches to
investigate absence in texts, and we believe that combining corpus with dis-
course analysis can offer valuable insights into this phenomenon.
Interestingly, there are only very few academic studies on Chinese
smog discourses, and issues that surround such discourses, (e.g. Jia, 2014;
Peng, 2013; Wang, 2014; Zhou, 2015), and all of these studies have been
published in Chinese language. Also, interestingly to us, academics
involved in the study of this topic tend to define their work as research on
journalism, rather than being contributions to linguistics. Such previous
research has covered various noteworthy areas. For example, Peng (2013)
explores how the notion of smog has been incorporated into Chinese
weather forecasts; Jia (2014) explores the ways in which smog related risk
tends to be covered in the state newspaper People’s Daily; and Zhou (2015)
and Wang (2014) examine the representation/coverage of smog online, on
the platform People.cn. These studies have touched on the issue that news
reporters excessively rely on quotations from official sources, and as such
fail to report on smog in an independent way, in a similar fashion to what
Tolan (2007) has observed on reporting on environmental protection in
Chinese media. We aim to contribute to this emerging research on
Chinese through the lens of discourse, by focusing on the phenomena of
silence and absence in Chinese smog discourses.
7.2 Data and Methodology

In order to study the key features of Chinese media coverage of wumai,
we collected news articles that explicitly mention the word wumai, by
studying three major newspapers, namely the People’s Daily, China
Table 7.1 The smog corpus

Number of articles
Newspaper Nature mentioning “smog”
People’s Daily 人民日报 Number one state 135
newspaper
China Environmental 中国环境报 Number one 211
News environmental
newspaper
Beijing Times 京华时报 Major Beijing local 69
newspaper
Total 415
Environmental News, and the Beijing Times, in the period spanning 1

January 2016 to 30 June 2016 (our rationale for selecting this time
span is that it covers a ‘politically active’ period when the government
and its news outlets bring up the smog issue as well as periods without
salient political activities). Our data collection has resulted in a corpus
of 415 newspaper articles containing 440,266 Chinese word tokens1
(Table 7.1).
We have adopted the approach of corpus-assisted discourse analysis
(see Partington, 2008, for an overview), combining the quantitative
approach of statistical data analysis with the qualitative approach of dis-
course analysis that examines particular stretches of texts of interest in
detail. We have analysed this corpus by using (1) the qualitative research
software NVivo for qualitative content analysis; and (2) the corpus
analysis software AntConc (Anthony, 2016). Specifically, we investigated:
(1) the distribution of frequencies of wumai across texts and newspapers
via frequency lists; (2) the contextual use of wumai through concor-
dances; (3) the verbs and nouns with which wumai collocates; and (4) the
unusually frequent words in the corpus studied via keyword lists. In order
to achieve task (4), we have compared our corpus with a reference corpus:
the Lancaster Corpus of Mandarin Chinese (LCMC), a 1,000,000-word
balanced2 corpus of written Mandarin Chinese created by McEnery and
Xiao (2004). It is pertinent to note here that keywords in our research
have been words with unusually high frequency in the specialised corpus
in comparison with the LCMC reference corpus. We have employed the
corpus software of AntConc Version 3.4.4 (Anthony, 2016) to produce
the keywords (see Table 7.4).
7.3 C
hinese Smog Discourses:
A Corpus-assisted Analysis
In what follows, let us illustrate the ways in which wumai occurs in the
sources studied. Table 7.2 shows the month-by-month breakdown of the
number of articles in respective newspapers.
Overall, Beijing Times had the smallest number of reports (69), the state
newspaper People’s Daily had 135, and China Environmental News, some-
what unsurprisingly, the largest (211), more than three times that of Beijing
Times. Drawing on the corpus-assisted discourse analysis (Partington,
2008, 2010; Taylor, 2010) of 415 newspaper articles containing 1016
explicit mentions of the word wumai, we have identified four major aspects
of silence and absence in the corpus. First, the peaks of media reports are
more correlated with political periods than with peaks of smog, revealing a
lack of media attention in non-politically salient periods. Second, smog
tends to be trivialised as background information. Third, the silences sur-
rounding smog coverage reveal that the causes, or certain factors, tend to be
excluded. Fourth, the representation of social actors is characterised by the
dominance of a small group of actors and the absence of others.3
7.3.1 G
reatest Media Attention in the Political
Season: More Severe Smog Attacks,
More Coverage?
It might be assumed that more severe smog attacks can lead to more cov-
erage in the media. Indeed, reporting was at a high level in January across
Table 7.2 Newspaper breakdown (1 January 2016–30 June 2016)

Newspaper
Month People’s Daily China Environmental News Beijing Times
January 39 63 24
February 22 31 7
March 41 68 19
April 13 26 6
May 14 23 8
June 6 0 5
Total 135 211 69
Number of news articles containing the word 'smog'

80
70
60
50
40
30
20
10
0
January February March April May June
People's Daily China Environment News Beijing Times
Fig. 7.1 Reporting trend in the first half of 2016
the board when air pollution was the most serious (The Economist, 2017)
in the selected time span. However, if we place the numbers of the reports
along the timeline, an interesting reporting trend can be observed, as
shown in Fig. 7.1.
As can be seen from Fig. 7.1, the overall reporting trend was similar
across the three major newspapers. Interestingly, the numbers of reports
peaked in March, the political season in China, when China’s air pollu-
tion, measured by the average PM2.5 concentrations, was the least seri-
ous (The Economist, 2017) in the year 2016. A closer examination of the
content of the reports revealed that the vast majority of reports in March,
and even in early April, cover the nation’s top two annual political meet-
ings. They are colloquially known as the ‘Two Sessions’ in China: the
meetings of the National Committee of the Chinese People’s Political
Consultative Conference (CPPCC), China’s top political advisory body,
and the National People’s Congress (NPC). The NPC and CPPCC meet-
ings—which start in early March every year and last for around two
weeks—represent an important bellwether for Chinese government pol-
icy (Campbell, 2016). It is during these political sessions that the central
government delivers a report that reviews the work of the government
over the past year and announces its priorities and policies for the
pcoming year, and when over 3000 deputies such as top provincial
u
administrators who attend are given an opportunity to submit proposals
for further (selected) deliberation. Extract 7.1 is a typical example of
news articles published during this period.
Extract 7.1 A News Article with One Mention of wumai

Actions to promote new ideas to realise the building of a moderately pros-
perous society in an all-round way (Voices [from the Two Sessions] 2016)
11 March 2016 People’s Daily
Deputy Luo Zhijun (Party Secretary of Jiangsu Province, Chairman of the
Standing Committee of the local People's Congress)
[…]
In order to achieve the transition from “developing at the expense of the
environment” to “the environment optimising the development”, [we]
need to strive to turn our province into a demonstration province for eco-
logical development, seize the key points, organise a tough fight and fight
a battle of annihilation, constantly make efforts to reduce smog , increase
blue-sky days, protect the clear water, control water pollution, protect non-
contaminated soil, control soil contamination, guard against touching the
red line by leaving leeway, focus on development while increasing the
green coverage, address the fundamental problems while changing the
[development] modes, cope with the difficult tasks bit by bit and step by
step, and achieve steady progress in the long run, so as to make Jiangsu’s
sky bluer, mountains greener, water clearer, air fresher and enable the peo-
ple to enjoy the fruits of new Jiangsu’s development in a good environ-
ment. (Italics added. Authors’ translation)
理念引领行动决胜全面小康(声音2016)
2016年03月11日人民日报
罗志军代表(江苏省委书记、省人大常委会主任)
……
要实现由“环境换取增长”向“环境优化增长”转型,努力建设全国生态文明建设
先行示范省,要抓住关键点,组织攻坚战,打好歼灭战,坚持不懈降䴮䵮、增蓝天,保
清水、治污水,护净土、治脏土,守红线、善留白,重建设、广增绿,抓根本、转方
式,一个一个骨头地啃,一步一个脚印地干,久久为功、稳步推进,让江苏大地天更
蓝、山更绿、水更清、空气更清新,让人民群众在青山绿水中享受新江苏建设成
果。
This typical example of news articles during this period shows that, on
the one hand, smog tends to be backgrounded (see the italics in the
English translation and the boxes in the Chinese and English texts),
which will be discussed in detail the section below, but on the other
hand, the topic itself, no matter how politically driven it is, at least has
begun to emerge more frequently in the news coverage in general. Our
data reveal that the number of newspaper reports on wumai surged when
these top political events took place, even though the smog attacks were
the least serious, and the concern of the general public was given a space
to be voiced through government channels during this politically, rather
than environmentally, salient period. By contrast, when more severe
smog attacks occurred (The Economist, 2017), the number of articles was
noticeably lower before the political season and also dropped dramati-
cally in subsequent months. The decreased media coverage could not
always be attributed to the reduction of smog attacks.
Significantly, this illustrates that the news coverage of wumai can be
driven by the political season in China and the severity of the issue may
not be the most important predictor of Chinese media attention, high-
lighting the relative silence, or lack of attention, in these major newspa-
pers at other times of the year.
7.3.2 Trivialising Smog as Background Information
In this corpus of 415 news articles with explicit mention of the word
wumai, the topic smog itself was frequently backgrounded as trivial con-
textual information. The word wumai tended to be mentioned in pass-
ing. Overall, more than 67 per cent of the newspaper articles in the
corpus, i.e. 279 articles, mentioned the word wumai only once at all (as
shown by the typical example in Extract 7.1), nearly 13 per cent (53
articles) twice, nearly 7 per cent (29 articles) three times, 3 per cent (13
articles) four times, nearly 2.5 per cent (10 articles) five times, and nearly
7.5 per cent (31 articles) six times or more. The vast majority of these
articles, especially the media’s scattered coverage outside the political sea-
son, used the word wumai sporadically to represent smog as trivial
background information which seems to be more or less taken for granted
by the reader. For example, it was brought up briefly in structures, such
as: (1) wumai tianqi pinfa/雾霾(天气)频发/“smog (weather) frequently
occur(red)” or wumai tianqi duofa/雾霾(天气)多发/“smog (weather)
frequently occur(red)” or pinfa de wumai tianqi/频发的雾霾(天
气)/“frequently occurring smog (weather)” (43 occurrences); (2) wumai

longzhao/雾霾笼罩/“smog engulfing” or longzhao zai wumai zhong/笼罩
在雾霾中/“engulfed in smog” (11 occurrences); or (3) (ru…) wumai
wenti (yanzhong)/(如) […] 雾霾问题(严重)/“(such as […]) smog prob-
lems (are serious)” (26 occurrences).
Furthermore, wumai tended to be listed as one of various social prob-
lems. The examination of the texts studied has revealed that wumai usu-
ally appears in compounds outside of the context of environmental
pollution, i.e. the main theme of the articles containing the word wumai
is usually not about the area of environmental protection. It is illustrative
to overview the common noun collocations of wumai as well. It tends to
co-occur with shuiwuran/水污染/“water pollution” (94 occurrences) on
the left, and quanqiu/qihou/daqi biannuan/全球/气候/大气变暖/
“global/climate/atmosphere warming” (27 occurrences) on the right (the
collocation settings were within a 50-word collocational window size, i.e.
showing 50 words in a concordance line). This shows that smog tends to
fall into the same category as water pollution and global warming in the
discursive constructions whereas water pollution and global warming
have a much longer history of communication than smog in China (Chen
& Chen, 2006; Jia, 2007, 2014). A further inspection of the concordance
lines indicates that it tends to be part of a general expression of the gov-
ernment’s commitment to control pollution, and in the case of zhongquan
zhili daqiwumai he shuiwuran/重拳治理大气雾霾和水污染/ “control
smog and water pollution with a heavy fist”, or part of the claim that
smog is partially related to global warming, e.g. wumai de chengyin he
[…] quanqiubiannuan yeyou yiding guanxi/雾霾的成因和 […] 全球变
暖也有一定关系/“the causes of smog also have something to do with
[…] global warming”. The nominal collocations imply that Chinese
newspapers tend to represent smog on a rather generalised level.
7.3.3 T
he Construction of Causes: What Are
the Causes of Smog?
The third major aspect of silence and absence in Chinese smog discourses
that emerged from our analysis is the construction of causes. The causes
of smog, which is an important subtopic itself, were excluded completely

in 294 news articles, nearly 71 per cent of the total.
In the 121 articles mentioning the causes of smog, there were 11 men-
tions stating that the causes are rather complex, 3 mentions claimed that
the causes are not clear yet. Four indicated that controversies still exist,
and 1 mention said the causes have basically been understood. These
generalised statements about the causes seemed to present conflicting
narratives, but it is noteworthy that the last article which contradicts the
other statements was a collation of important points that the NPC depu-
ties made at a political meeting during the Two Sessions, and the state-
ment that the causes are basically clear was made by a professor at Beijing
Normal University, while the other statements were all made by govern-
ment officials. Nine mentions called for action, e.g. tell the children what
the causes of pollution are to raise awareness.
In order to capture the prominence of subtopics,4 we noted all the
subtopics mentioned in relation to the causes of smog and gave promi-
nent subtopics extra points. For example, if a subtopic was mentioned in
the headline or the first paragraph, it received an extra point. It is worth
pointing out here that the categories of subtopics can be ‘leaky’ in the
sense that there were crossovers and hybrids, so if the subtopic of causes
of industrial coal burning was mentioned once, we noted it under both
coal burning (subcategory: industrial coal burning) and industrial pro-
duction (Table 7.3).
This approach to analysing textual silence yielded a weighted inventory
of nine subtopics concerning the causes of smog, with some typically
receiving more attention than others. It shows that the current public
discourse in China characterises smog as mainly caused by coal burning,
car emissions, and industrial production, whereas construction sites and
petrol burning seem to have a smaller impact and the impact of the rest
of the causes is minimal. If one compares the numbers on this list, one
can immediately see the imbalances among some of the causes. For exam-
ple, industrial production occurs to be given relatively less attention than coal
burning and car emissions.
The above-mentioned imbalance was even more obvious in the subcat-
egories of the first subtopic of coal burning, industrial usage was given
relatively little attention whereas considerable attention was given to
Table 7.3 Subtopics of causes mentioned in the corpus of 415 news articles pub-
lished in major Chinese newspapers, 1 January–31 June 2016 (440,266 Chinese
words)
Causes Weighted totals
Coal burning 567 (= 340 general + 215
household + 12 industrial)
General coal burning 340
Household coal burning 215
Industrial coal burning 12
Motor vehicle emissions 219
Industrial production (e.g. factories, steel 106
manufacturing, energy industry, etc.)
Petrol burning 59
Construction sites (e.g. dust, dump, etc.) 25
(Not caused by) nuclear 4
Straw burning 2
Nature and topography (e.g. wind and valley) 2
El Niño 2
individual household coal burning and coal burning in general, without

specifying its purpose. While scholarly literature has examined the effects
of energy use for industrial production (e.g. coal and petrol burning in
cement and steel industries) on smog and identified it as a top contribu-
tor (Ma & Zhang, 2014; Zhang et al., 2013), by excluding this impor-
tant factor in the construction of causes, the writer presented a selective
picture, which affected the communication of smog. This leads us to the
next finding concerning the representation of agent and agency.
7.3.4 T
he Representation of Social Actors: Whose
Agency Matters?
Smog problems in China and silences surrounding smog discourses have

led the general public to ask, “What can be done?” in addition to “What
are the causes?” For example, the Chinese people are crying out on non-
traditional media for more information on action that they personally
can take to help address the problems that have affected them immensely,
e.g. “What else can we do apart from wearing masks?” [除了戴口罩,还
能做些什么] (e.g. Xiaolan, 2015), “complaining” and “driving less”
(e.g. Qingnianzhisheng internet user_53813, 2015). The fourth major

aspect of silence and absence in Chinese smog discourses pertains to the
representation of agency.
Agency refers to an actor’s ability to make meaningful choices and
changes (Alsop, Heinsohn, & Somma, 2004). We identified this problem
as we undertook a keyword analysis that enabled us to identify items
“which are likely to be of linguistic interest in terms of the text’s aboutness
and structuring, and which can be expected to repay further study, e.g.
through concordancing to investigate collocation” (Scott, 2009, p. 80).
The keywords tool settings were: (1) negative keywords (words that are
unusually infrequent) to be excluded; (2) keyword generation method:
log likelihood (abbreviated as LL, the most commonly used method); and
(3) threshold value: top 100 (Qian & Tian, 2014). We used the Lancaster
Corpus of Mandarin Chinese (LCMC) as a reference corpus. The key-
word list in Table 7.4 represents the top keywords of our smog corpus.
This keyword list comprises the words that are the most unusually highly
frequent in our corpus of news articles with explicit mention of the word
wumai compared to the reference corpus. Notably, most of these words
are centred on environment and pollution, which is logical if one consid-
ers the fact that we compiled the corpus by extracting news articles that
contain the word wumai. However, two keywords in the list stand out,
namely, zhengfu/政府/“government” (ranking number 6 in the list) and
tuijin/推进/“push forward” (number 7), as they may not seem to
Table 7.4 Key lexical words

No. Frequency Keyness (LL) Keyword
1 1924 1.856 huanjing/环境/“environment”
2 1385 1.336 huanbao/环保/“environmental protection”
3 1354 1.306 wuran/污染/“pollution”
4 1247 1.203 shengtai/生态/“ecology”
5 1081 1.043 zhili/治理/“govern (or address/control)”a
6 999 0.964 zhengfu/政府/“government”
7 914 0.882 tuijin/推进/“push forward”
8 899 0.867 wumai/雾霾/“smog”
a
Please note, while the literal translation of the verb 治理 is to “govern or
address/control” which seems to indicate agency and responsibility, in the
Chinese contexts of smog control, the actor of this action is only limited to the
government. This point will be further discussed later in this section.
obviously fall into the same semantic fields. Why are zhengfu/政
府/“government”, tuijin/推进/“push forward”, and even the verb zhili/
治理/“govern (or address/control)” which seems to be confined to col-
locate mostly with a very limited group of actors, so prominent in the
smog discourses?
Clearly, the corpus-assisted analysis helped us to see some general pat-
terns, which served as an entry point into the issue of agency. We then
went back to examine the texts more closely. The findings suggest that the
news discourses tend to hide individual agency by: (1) use of Chinese
zero-subject sentences; (2) the use of active verbs carrying a passive mean-
ing; and (3) the vague characterisation of social actors. Yet, when agency
is touched upon, it is almost limited to a small group of actors, i.e. the
government.
First, the Chinese zero-subject sentence structure facilitates agent dele-
tion. It is worth pointing out that lack of overt subjects in sentences that
are not imperatives is commonplace in Chinese grammar (Zhao & Ng,
2007). Zero pronominal and nominal forms are also called zero anaphors
which are noun phrases that may be understood from the context (Yeh &
Chen, 2001), as Extract 7.2 illustrates.
Extract 7.2
Φ raise the concentrated use of coal and the technological standards of coal
burning. Φ promote efficient clean coal technologies. Φ improve the oil
quality. Φ make huge efforts to control smog . (Authors’ translation)
Φ提高煤利用集中度和燃煤技术标准,Φ推广煤炭清洁高效利用技术,Φ提升燃油品
质,Φ下大力气治理大气䴮䵮。
Note, the symbol Φ shows where the subject is omitted in the text.
This is a typical example of Chinese-specific zero-subject sentences con-
cerning the actions to address smog widely used in the corpus. It is an
excerpt taken from a news article on the report delivered by the NPC
Financial and Economic Committee, regarding its review of the govern-
ment’s implementation of its 2015 plan of national socio-economic
development and its draft plan for 2016. The main body of the article
consists of lengthy recommendations put forward by the Committee and
these recommendations are characterised by sentences such as are found
in Extract 7.2. As can be seen, all the subjects are missing. The actor for
the last action, i.e. to make huge efforts to control smog, may be relatively
easily inferred from the pattern of the use of the verb zhili/治理/“govern
(or address/control)” in the corpus. That is, the verb zhili almost always
co-occurs with the government as the actor and the last action Φ下大力
气治理大气雾霾/ “make huge efforts to control smog” in itself is an oft-
cited instruction from the Chinese Premier Li Keqiang. However, the
subjects for the first three actions are not this clear and seem to refer to
different actors. For example, the actor to promote clean coal technology
can be the country as a whole, the relevant governmental departments,
such as the National Energy Administration or local governments, or a
combination of them all, but with overt exclusion, it is not clear whose
responsibility it is and who will take action. Similarly, while the actor to
improve the oil quality can be the energy industry and/or oil manufactur-
ing companies, with subject omission, the actor is not named and the
responsibilities seem to be rather diffused.
Second, in addition to the use of the Chinese zero-subject sentence
structure to delete the agent, the smog discourses also tend to hide indi-
vidual agency by the use of active verbs carrying a passive meaning. The
verbal collocates of the word wumai in our corpus may illustrate this
phenomenon. Wumai seldom collocates with a verb that refers to the
general public’s possible action in relation to smog, but when it does, the
verb usually carries a passive meaning, e.g. duo/躲/“escape” (two occur-
rences), yuanli/远离/“stay away” (two occurrences), and baoyuan wuyong/
wu yi/抱怨无用/无益/“it is no use/no good complaining” (two occur-
rences). This pattern seems to imply that in the media constructions,
individual citizens are usually discouraged to take active actions against
smog, and it is implicated that their individual agency is not deserving of
mention (e.g. Xiaolan, 2015).
Third, the vague characterisation of social actors is also widespread in
the exclusion of individual agency in the discourses. This is achieved
mainly by the use of the Chinese pronoun women/我们/“we” or the
omission of this pronoun. The Chinese “we” is employed much more

broadly in the Chinese language than the royal “we”, or the editorial “we”
in English used by editors and writers to avoid the personal “I”. It is one
of the most frequently used Chinese words, which expresses a wide range
of meanings, and which can be applied so flexibly that in certain contexts,
even though it is absent, the reader can still fill it in easily (Baoerji, 2011).
Extract 7.3
To practise the new concept of development: Make changes together for a
green lifestyle
[We] should advocate public awareness and take ecology as [our] own
business. Only by doing so, can [we] integrate “beautiful genes” into the
blood of development.
To live a virtuous life is always linked to external constraints. In fact, what
we are going to eradicate is not barbecues [food stalls], but rather the envi-
ronmental pollution caused by barbecue [stalls]. Draw the bottom line of
rule of law for environmental protection, so that green is no longer empty
talk; rather, it will be a behavioural option that can truly influence the peo-
ple. (Authors’ translation)
践行新发展理念:为绿色生活方式一起改变
应该多倡导公共意识,把生态当作自己的事。如此,才能在发展的血液中融入“
美丽基因”。
过有德行的生活,往往离不开外在的约束。其实, ᡁԜ 要取缔的不是烧烤,而是
烧烤带来的环境污染。划出环保的法治底线,绿色就不会停留在口头上,而是真正
能影响人们的行为选择。
Taken from an article calling for public action, Extract 7.3 is a typical
example of the use of the Chinese “we”. Throughout the text, the actor is
usually the plural pronoun “we”, which is the distinctive subject in
Chinese calls for action (Baoerji, 2011). In the first paragraph of Extract
7.3, although “we” is omitted three times, drawing on the linguistic rep-
ertoire (Baoerji, 2011), the Chinese reader can effortlessly fill it in. In the
second paragraph, this subject is mentioned explicitly once. In all the
cases, the plural “we”, omitted or not, vaguely characterises the agent,
allowing the social actors—individual citizens and/or organisations,

responsible for the actions, ranging from taking environmental protec-
tion as one’s own business to eradicating outdoor pollution caused by
barbecue food stalls—to be excluded.
However, when agency is neither backgrounded nor excluded, or sup-
pressed in van Leeuwen’s (1996) terms, not all social actors are included
and those included are predominantly governments at all levels. This
partly explains why zhengfu/政府/“government” is one of the top key-
words in the corpus. Also, a closer examination of these news articles
shows that the word tuijin/推进/“push forward” is a very frequent phrase
to indicate the government’s action, and its use is not limited to the spe-
cific context of smog control. This is consistent with the findings of Qian
and Tian’s (2014) study of China’s government work reports (e.g. to push
forward the SOE reforms, p. 84). Therefore, even though the verb tuijin/
推进/“push forward” does not explicitly co-occur with the actor govern-
ment in some cases in our corpus, it still can be inferred from the context
that the implied actor is the government. The prominence of the govern-
ment in the representation of social actors in the smog discourses is best
illustrated in the verbal phrases zhili wumai (daqiwuran/kongqiwuran)/治
理雾霾 (大气污染/空气污染)/“govern or address/control smog (or air
pollution)”. The actor of this action, either explicitly mentioned or
inferred from the context (e.g. in a zero-subject sentence), is still pre-
dominantly the government, e.g. government at all levels or “our coun-
try” implying the Chinese government. A few exceptions include the
introduction of third-party experts, e.g. disanfang zhili/第三方治
理/“control by a third party”, and the occasional inclusion of other rele-
vant social actors when touching on the importance of synergy with the
public and businesses.
In short, the smog discourses tend to hide individual agency by use of
Chinese zero-subject sentence structure, by use of verbs carrying a passive
meaning, or by the vague characterisation of social actors. When the dis-
courses do talk about agency, only a very limited group of actors, i.e. the
government, are present or foregrounded; this reinforces the perception
that only the government’s agency matters in the fight against smog, and
this indirectly promotes the authority of the government.
7.4 Discussion and Conclusion

With respect to the reporting trend of Chinese smog coverage, the dra-
matic increase of media attention, which is measured by the number of
reports, during the political season is a new finding not previously
described. Clearly, the peaks of smog reports correlate more with political
periods than with peaks of smog. A closer examination of the texts
has revealed that smog is usually backgrounded in news articles, and in
the minority of reports where it is foregrounded, media outlets tend to
frame it in a way that silences the causes. The smog discourses also tend
to hide individual agency by, for example, the vague characterisation of
social actors, i.e. no individuals and organisations are named as in charge
of resolving smog-related issues, and if agency is brought up, almost only
the actor government is present or foregrounded, reinforcing the percep-
tion that only the government’s agency matters in the fight against smog.
The results of our corpus-assisted discourse analysis substantiate the
observations of previous studies (Jia, 2014; Peng, 2013; Wang, 2014;
Zhou, 2015) that Chinese media discourse relies heavily on official dis-
course, offering insights into this new, emerging area of research.
News about smog does not happen by itself. As Hansen argues, “News
about the environment, environmental disaster, and environmental issues
or problems does not happen by itself, but is rather ‘produced’,
‘manufactured’ or ‘constructed’” (Hansen, 2010, p. 72). The construc-
tion of smog problems and the proposed solutions in news discourse are
discursive in nature. Analysing how smog has been constructed in dis-
course, especially its lack and avoidance, is vital to raise awareness on
progress and deficiencies. Although the term “smog” has become a famil-
iar term and a salient topic in Chinese society, smog communication
remains much understudied (Jia, 2014) and still faces a lot of challenges.
Realisation of environmental deterioration does not automatically trans-
late into action. New discourses, even new stories to live by (Stibbe,
2015) in ecolinguistic terms, have to be created and publicised to mobil-
ise people to change their thinking and practices, considering alterna-
tives. What are the new discourses? How can change happen? These issues
could benefit from further research. Ultimately, we hope to find ways to
translate the Chinese people’s realisation of significant deterioration of air
quality into enhanced awareness and action.
Notes
1. Note that a Chinese word usually comprises two or more Chinese charac-
ters and there is no natural delimiter between Chinese words as in English,
so we first used a Japanese and Chinese segmenter, SegmentAnt Version
1.1.1, to segment the texts and then checked the segmentation manually.
2. A balanced corpus means that the corpus has been carefully designed so
that the sampled genres of texts are “proportional to the relative frequency
of occurrence of those genres in the language’s textual universe as a whole”
(Leech, 2007, p. 136).
3. We actually tried out different approaches to see which combination best
suited our study. For example, following Schröter’s (2013) meta-linguistic
approach, we collected a corpus of Chinese metadiscourse from the inter-
net, not from the media, mainly because the Chinese media seldom talked
about silence and absence in the media coverage of smog, but within our
specific context, these metadiscussions tended to be very superficial and
not as illuminating as those media texts and protocols of parliamentary
inquiries in Schröter’s (2013) study of political silence and concealment.
Therefore, we decided to abandon this approach and moved on to
others.
4. Note that our approach here coincides with that of Sweeney (2012), who
follows Huckin’s (2002) approach to analysing textual silence.
References
Alsop, R., Heinsohn, N., & Somma, A. (2004). Measuring empowerment: An
analytic framework. In R. Alsop (Ed.), Power, rights, and poverty: Concepts
and connections (pp. 120–146). Washington, DC: The World Bank.
Anthony, L. (2016). AntConc (Version 3.4.4) [Computer Software]. Tokyo,
Japan: Waseda University.
Baoerji, Y. (2011). ‘我们’是谁? [Who are ‘we’?]. Fangcao: Jingdian yuedu, 5,
64–65.
Campbell, C. (2016, March 3). Five things to know about China’s ‘Two
Sessions’.Time.Retrievedfromhttp://time.com/4245878/china-beijing-chinese-
communist-party-two-sessions-cppcc-npc/
Chen, L., & Chen, J. (2006). 松花江水污染事件中信息流障碍分析 [An
analysis of the barriers of communication in the information flow regarding
the water pollution incident of the Songhua River]. Xinwenjie, 6, 19–22.
Haarsager, S. (1991). Choosing silence: A case of reverse agenda setting in

depression era news coverage. Journal of Mass Media Ethics, 6(1), 35–46.
Han, J. (2013, December 13). 2013关键词 [Keywords of 2013]. Morning Post.
Retrieved from http://www.morningpost.com.cn/
Hansen, A. (2010). Environment, media and communication. London: Routledge.
& Society, 13(3), 347–372.
Jia, G. (2014). 中国雾霾议题传播的公共性辨析 [An analysis of the public
communication of China’s smog problems]. In B. Zheng (Ed.), Collection of
papers on journalism and communication. Beijing: Jingjiribao Press.
Jia, H. (2007). 全球变暖, 科学传播与公众参与 [Global warming, scientific
communication and public participation]. Kepu Yanjiu, 3, 39–45.
Jiang, D. (2011). 关于北京的PM2.5, 灰霾或雾霾 [On Beijing’s PM2.5 and
huimai or wumai]. [Blog post]. Retrieved from http://blog.sciencenet.cn/
blog-609047-506041.html
Jiang, X. (2015). 浅析新词’雾霾’的产生 [A preliminary analysis of the birth of
the new word ‘wumai’]. Qingnian wenxuejia, 9, 144.
Leech, G. (2007). New resources, or just better old ones? The Holy Grail of
representativeness. In M. Hundt, N. Nesselhauf, & C. Biewer (Eds.), Corpus
linguistics and the web (pp. 133–149). Leiden: Brill.
Ma, L., & Zhang, X. (2014). 中国雾霾污染的空间效应及经济, 能源结构
影响 [The steric effects of China’s smog pollution and effects of economic
and energy structure]. China Industrial Economics, 4, 19–31.
Mai (霾). (1716). In Kangxi dictionary (digitised version). Retrieved from http://
ctext.org/kangxi-zidian/zhs
McEnery, A., & Xiao, Z. (2004). The Lancaster Corpus of Mandarin Chinese:
A corpus for monolingual and contrastive language study. Religion, 17, 3–4.
Partington, A. (2008). The armchair and the machine: Corpus-assisted discourse
research. In C. Taylor Torsello, K. Ackerley, & E. Castello (Eds.), Corpora for
university language teachers (pp. 189–213). Bern: Peter Lang.
Partington, A. (2010). Modern Diachronic Corpus-Assisted Discourse Studies
(MD-CADS) on UK newspapers: An overview of the project. Corpora, 5(2),
83–108.
Peng, G. (2013). 从雾霾报道看气象新闻的拓展 [Extension of meteorologi-
cal news from the smog reports]. Media Observation, 13, 15–16.
PM25.com. (2014). PM2.5是如何走进中国人视线的 [How PM2.5 became
known to the Chinese people]. Retrieved from http://www.pm25.com/
news/216.html
Qian, Y., & Tian, H. (2014). A decade of change in China: A corpus-based

discourse analysis of ten government work reports. In Q. Cao, H. Tian, &
P. Chilton (Eds.), Discourse, politics and media in contemporary China
(pp. 77–96). Amsterdam: John Benjamins.
Qingnianzhisheng internet user_53813. (2015). 北京城又被雾霾刷灰了,
除了吐槽,我们还能做些什么? [The city of Beijing has been blanketed
by smog again. Apart from complaining, what else can we do?] [Online
discussion post]. Retrieved from http://sns.qnzs.youth.cn/index/show/

id/352514/s/54679
Schröter, M. (2013). Silence and concealment in political discourse. Amsterdam:
John Benjamins.
Scott, M. (2009). In search of a bad reference corpus. In D. Archer (Ed.), What’s
in a word-list? Investigating word frequency and keyword extraction (pp. 79–92).
Surrey, UK: Ashgate.
Shan, M. (2014). On Chinese neologisms and their English translation.
Pingdingshan daxue xuebao, 29(6), 75–80.
Stibbe, A. (2015). Ecolinguistics: Language, ecology and the stories we live by.
London: Routledge.
Sweeney, M. (2012). Listening rhetorically to textual silence: Intimate partner
homicide media coverage. International Journal of Listening, 26(3), 146–158.
Taylor, C. (2010). Science in the news: A diachronic perspective. Corpora, 5(2),
221–250.
Taylor, C. (2012). And there it isn’t: (How) can we access the absent using CADS
[PowerPoint slides]. Retrieved from https://sussex.academia.edu/
The Economist. (2017, January 4). Why China’s air pollution is on the rise
again. Retrieved from http://www.economist.com
Tolan, S. (2007). Coverage of climate change in Chinese media. UNESCO
Human Development Report, 8. Retrieved from http://hdr.undp.org/en/con-
tent/coverage-climate-change-chinese-media
van Leeuwen, T. (1996). The representation of social actors. In C. R. Caldas-
Coulthard & M. Coulthard (Eds.), Texts and practices (pp. 32–70). London:
Routledge.
Wang, Q. (2014). 人民网风险报道优劣分析——以雾霾报道为例 [An anal-
ysis of the advantages and disadvantages of People.cn risk reports: A case of
reports on smog] [online research article]. People.cn Research Institute. Retrieved
from http://media.people.com.cn/n/2015/0312/c221847-26680032.html
Wu (雾). (1716). In Kangxi dictionary (digitised version). Retrieved from http://
ctext.org/kangxi-zidian/zhs
Xiaolan. (2015, December 10). 雾霾来临除了戴口罩,还能做些什么? [Smog

coming, apart from wearing masks, what else we can do?] [Blog post].
Retrieved from http://blog.sina.com.cn/s/blog_a844802e0102wba2.html
Yeh, C. L., & Chen, Y. J. (2001). An empirical study of zero anaphora resolution
in Chinese based on centering model. In ROCLING.
Zhang, X., Sun, J., Wang, Y., Li, W., Zhang, Q., Wang, W., et al. (2013). 我国
雾-霾成因及其治理的思考 [Factors contributing to smog in China].
Kexue tongbao, 58(13), 1178–1187.
Zhao, S., & Ng, H. T. (2007). Identification and resolution of Chinese zero
pronouns: A machine learning approach. In EMNLP-CoNLL (Vol. 2007,
pp. 541–550).
Zhou, J. (2015). 人民网雾霾报道研究 (2012–2014年) [A study of smog reports
by People.cn (2012–2014)]. Unpublished master’s thesis, East China Normal
University, Shanghai, China.
Part II
Exploring Means that Produce
Silence and Absence
8
Theoretical and Methodological
Challenges in Identifying Meaningful
Absences in Discourse
Patricia von Münchow
Discourse can be characterized not only by what is said, but also by

what remains unsaid. However, identifying what can be called ‘mean-
ingful absences’ presents numerous theoretical and methodological
challenges for discourse analysts. First of all, they have to ask them-
selves why certain elements are absent. A series of answers to this theo-
retical question is possible, highlighting the fact that we need to
distinguish different kinds of absences in discourse. Researchers thus
have to decide which kinds of absences are relevant for them, within the
approach they practise.
The second series of questions is a methodological one: How do we
notice absences? In other words, how can discursive ‘presence’ help to
detect ‘absence’? What kind of distinctions can be made between differ-
ent kinds of relevant absences and by which means? Instead of absence,
on the one hand, and presence, on the other, can we conceive of degrees
of absence and presence?
P. von Münchow (*)

Université Paris Descartes, Paris, France

216 P. von Münchow
And, finally, how can these (different kinds of ) absences be interpreted

in terms of social representations? What is the relationship between ‘the
unsaid’ and social consensus? Is the unsaid always less consensual than
what is present? Does absence always correspond to suppression, as
Achino-Loeb (2006, pp. 13–14) thinks?
In this chapter, I will try to answer these questions, drawing on differ-
ent types of discourse studies, namely, Critical Discourse Analysis and
French Discourse Analysis, as well as on the results of my own research
on ‘the unsaid’, including excerpts from studies on data sets in French,
German and English from parenting guidebooks, intercultural parenting
books and history textbooks.
In line with what is announced in Chap. 1 in this volume, when
‘absence’—and its synonym ‘the unsaid’—are not umbrella terms, they
refer to discourse in a collective, ‘determined’, Foucauldian sense, whereas
‘silence’—when it is not part of a quote—is linked to individual
intentionality.
8.1 T
he Theory of the Unsaid: Different
Types of Absences in Different Types
of Discourse Studies
8.1.1 Critical Discourse Analysis
Within Critical Discourse Analysis (CDA), van Dijk’s socio-cognitive

conceptualization of context provides a convincing framework to under-
stand how absence ‘works’. He locates what remains unsaid in the ‘mental
models’ speakers and recipients more or less share:
[Di]scourse meaning is strategically construed on the basis of socially

shared generic knowledge and its application in the construction of situa-
tion models of events talked or written about.
The information in these models is much more detailed than the infor-
mation actually expressed in text and talk or sentence structures. […] As a
general rule, old or given knowledge in discourse tends to be implicit, pre-
supposed, reduced (as in pronouns) and unstressed (as in initial sentence
topics). And even much of the new knowledge conveyed by discourse need
Theoretical and Methodological Challenges in Identifying... 217
not be expressed but may be inferred from generic knowledge or old men-
tal models in the construction of a current situation model that defines the
subjective interpretation of recipients.
Thus, indeed, the metaphor of the iceberg is quite apt to describe the
rather limited role of ‘visible’ or ‘hearable’ grammatical expression of
knowledge in discourse: most of the old and even the new knowledge
involved in text and talk remains invisible and implicit. (2009, pp. 4–6)
This explains convincingly how it is possible to communicate informa-

tion without making it explicit—the speaker can count on the recipient’s
mental models to provide what remains unsaid—and also how communi-
cation can be disturbed by insufficient overlap between the speaker’s and
the recipient’s mental models. Furthermore, it indicates that the unsaid is
virtually limitless. And finally, as van Dijk (2014, p. 316) puts it, it gives us
‘an elegant definition of implicit information: all information in the situa-
tion model (and hence intended or understood) that is not expressed in the
discourse’. In effect, van Dijk limits the situation model to what is ‘intended
or understood’ because ‘implications are not just any proposition implied
by another proposition—which would include a vast number of possible
implications a speaker would not ever entertain and a recipient would not
and could not derive because they are irrelevant’ (2014, pp. 281–282).
In a similar fashion and also within a CDA framework, Schröter’s
(2013) conception of what she calls ‘meaningful silence’ rests on three
pillars, which are intention, expectation and relevance:
‘Meaningful, communicative silence depends on an intention (to remain

silent or to conceal) as much as on (a disappointed) expectation (of speech),
which is moderated by relevance; for there would normally not be any
effort to conceal nor an expectation of speech when the issue at hand is
completely irrelevant for all involved’ (Schröter, 2013, p. 7).
For her, ‘silence’ refers to ‘the absence of verbal communication as well as

concealment’ (Schröter, 2013, p. 19). This definition goes beyond the scope
of what van Dijk considers ‘implicit information’, though, since ‘[c]onceal-
ers must intend concealment, but also discourage recipients from figuring
out their intentions. […] In the case of concealment, this detection crucially
goes beyond the intention of the communicator’ (Schröter, 2013, p. 27).
218 P. von Münchow
Schröter does mention other types of absence than the ones she chooses
to investigate, though, and even describes some of them with great precision.
One of them is what Huckin (2002) calls ‘discreet silences’, that is silence
produced in a situation or concerning a topic which is ‘socially, legally or
culturally sensitive’ (Huckin, 2002, p. 351), for instance, at a funeral.
Schröter (2013, p. 32) differentiates between this type of (intentional) silence
and ‘culturally agreed silences, e.g. taboos that often require no specific
intention or effort to be circumnavigated’, such as not speaking in a non-
communicative situation, for example, waiting in line, when ‘there is no
intention to be silent, and no expectation to speak’ (Schröter, 2013, p. 32)
Another type of absence—which Schröter calls ‘silencing’—plays an
important role in discourse hegemony. ‘A major function of silencing is to
[…] identify[…] categories of persons and ideas about which speech and
texts will be unacceptable’, a process ‘complemented by the circulation of
acceptable speech and texts that express some things at the expense of
others’ (Thiesmeyer, 2003, p. 9; see also Schröter, 2013, p. 4). This con-
tribution to the ‘order of discourse’ in a Foucauldian sense need not be
intentional or conscious, as silencing—just like succeeding in participat-
ing in dominant discourse—is a gradual process within a struggle for
power.1 Schröter opens a separate category for ‘conspiracies of silence
which involve a collective effort of avoidance which renders certain topics
taboo’ (2013, p. 5). Unlike silencing, which entails one group wielding
power over another group, but which is generally not intentional when an
individual author engages in it, conspiracies of silence ‘presuppose mutual
denial’ (Zerubavel, 2006, p. 9). Denial ‘involves refusing to acknowledge
the presence of things that actually beg for attention, thereby reminding
us that conspiracies of silence revolve not around those largely unnotice-
able matters we simply overlook but, on the contrary, around those highly
conspicuous matters we deliberately try to avoid’ (Zerubavel, 2006, p. 9).
8.1.2 French Discourse Analysis
Almost from the very beginning, the investigation of ‘the unsaid’, or ‘the
preconstructed’ (‘le préconstruit’) (Pêcheux, 1975, p. 193) has played an
important role in French Discourse Analysis.2 From a theoretical point of
view, Orlandi (1994/1996)—probably the most prominent representative
of the Brazilian branch of the discipline—provides a quite comprehensive

account of what she calls ‘silence’, well beyond the range of ‘the precon-
structed’. She distinguishes ‘silencing’ (‘silenciement’) from ‘the silence of
the already said’ (Orlandi, 1994/1996, p. 73).3 According to her,
we say “x” in order to not [let] say “y” […]. That is how we “silence” the
meanings we want to avoid, meanings which could set up the significant
work of “another” discursive formation, “another” area of meaning. Silence
thus draws up the boundaries of discursive formations and therefore deter-
mines the boundaries of what can be said. (Orlandi, 1994/1996, p. 62)4
This is of course a matter of power. As Achino-Loeb (2006, pp. 13–14,

quoted by Schröter, 2013, p. 46) says:
In fact, it is at this primary level of selective assigning of salience that ideo-

logical manipulations are at their most elusive. […] The road to overt ideo-
logical domination rests on a bedrock of silence running through different
layers of suppression that bend perspective at the service of contextual
cohesiveness […].
However, unlike the approach the above-mentioned CDA scholars

adopt, Orlandi’s is clearly non-logico-pragmatic. For her, what is worth
explaining is not so much why communication is possible in spite of
silences, but why it is impossible without them. Whereas van Dijk’s and
Schröter’s concerns are with issues such as intention, comprehension,
expectation and relevance in order to detect what is intended to be under-
stood even if it remains unsaid or what stays unsaid although it is expected
to be communicated, Orlandi focuses on ‘constitutive silence’. She is indeed
convinced that, ‘in order to say you have to not say’ (Orlandi, 1994/1996,
p. 23). Choosing one nomination entails excluding another. ‘Silence’ thus
offers the (human) subject ‘a possible space of singularity’, allowing her not
to be caught in a specific discursive formation but to move around among
different formations while maintaining her identity (Orlandi, 1994/1996,
p. 73) as well as keeping a certain distance from social consensus without
taking a risk. In the same perspective, ‘silence’ is also what makes it possible
for a text to be coherent in spite of the subject’s heterogeneous positions. In
summary, among what remains unsaid, Orlandi insists in particular on
220 P. von Münchow
what does not need to be said because it is ‘already said’ or ‘obvious’, as part
of the interlocutors’ shared knowledge, on the one hand, and on what can-
not be said because it would not be acceptable, would seem inconsistent
and/or would make a text incoherent, on the other. These are the two types
of—mostly unintentional—absences I am primarily interested in as well.
‘Absences’ is in fact an appropriate term because while what remains unsaid
need not or cannot be uttered in a certain context, it is not non-existent or
impossible to verbalize in general.
Following this logic, the speaker’s intentions cannot be a criterion for
the analysis of the unsaid, of course. Heinemann (1999) and Kurzon
(1998, p. 8) might both consider ‘unintentional silences’ linguistically
irrelevant or meaningless (see Schröter, 2013, p. 22), but what matters in
the (French) discourse-analytical perspective I adopt is discursive rather
than linguistic meaning. Hence the investigation focuses on the condi-
tions of language use rather than speaker intention: ‘In general discourse-
analytical approaches are based on the thought that what is said contains
more than what is meant’ and that ‘[e]ach enunciation depends on spe-
cific circumstances of which we are partly conscious (e.g. legal limita-
tions) and partly less conscious (e.g. socio-cultural and thematic shaping
[…]). These conditions show themselves in semiotic enunciations as spe-
cific orders, the discourse’ (Dreesen, 2015, p. 59). In other words, as will
be shown in Sect. 8.3, what I am trying to investigate is neither what is
meant without being said (i.e. implication) nor what is intentionally con-
cealed or not mentioned by the speaker, but what remains unsaid albeit
constituting the condition for what is said—be it information from ‘old
mental models’ (in van Dijk’s terminology) that is necessary for the new
information to be comprehensible, or elements that cannot be made
explicit because they are contradictory to what is asserted. In a way, the
approach is about collecting ‘given off’ rather than given information
(Goffman, 1956), but the ‘given off’ information should be seen as being
hidden by ideology rather than by the speaker. The rules of the patriar-
chal system, for example, are a necessary prerequisite to understand par-
enting guide books in France and in Germany without ever being made
explicit, as I showed in a previous study (von Münchow, 2011). They are
never reaffirmed as being legitimate not only because they are part of
shared knowledge in both societies, but also because authors explicitly
adopt a gender-equal approach, which would be contradicted by an

equally explicit traditional understanding of parenting.
As van Dijk points out, though, what is said is only the tip of the ice-
berg of all knowledge implied by discourse (2014, pp. 306–307). So how
can the (French) discourse-analytical approach manage the ocean of
absences beyond the words? Obviously, there must be a criterion of rele-
vance even if it is not the speaker’s intention. Van Dijk (2014, p. 281)
also mentions the implications ‘a recipient would […] and could […]
derive’ as being the ones that would be relevant. Similarly, Schröter
(2013, p. 7) suggests relying on the recipient’s expectations. This cannot
be the criterion adopted by a discourse-analytical approach either, though,
since in this perspective—and following Foucault (1969)—not only the
speaker’s, but also the recipient’s autonomy need to be questioned
(Dreesen, 2015, p. 60). If speaker and hearer are not necessarily con-
scious of the discursive meanings that come or do not come across
through the unsaid, it is the analyst who needs to be the judge of discur-
sive—and not linguistic or interactional—relevance. The challenge for
the analyst is thus not to find out what the speaker’s intentions or the
recipient’s communicative expectations are, but to create the conditions
in which she will be able to identify the ‘discursively relevant unsaid’. But
which analytical procedures can be imagined in order to attain this goal?
8.2 T
he Methodology of Analysing
the Unsaid: Discovering and Interpreting
‘Absence’ Through ‘Presence’
The problem with ‘silence’, as Schröter puts it, is its ‘perception—hence
[its] […] phenomenological existence’ (2013, p. 43). Indeed, the unsaid
can only be ‘observed indirectly by (discursive) historical, critical, decon-
structivist methods’ (Orlandi, 1994/1996, p. 42). Schröter herself (2013)
studies metalinguistic reference in public discourse to politicians’ silence
in order to bring about its forms and meanings. Non-intentional absences,
though, are rarely identified by recipients or another kind of audience
and thus remain to be discovered by the analyst.
222 P. von Münchow
8.2.1 Detecting the ‘Preconstructed’

and ‘Prediscourses’
Throughout its history and under the initial influence of Althusser,

French Discourse Analysis has been searching for proof of ‘obviousness’
(‘évidences’) or for the ‘already there’ (‘le déjà-là’), as Pêcheux put it.5 In
the 1970s, Henry and Pêcheux conceptualized ‘the preconstructed’ (‘le
préconstruit’), which Pêcheux defined as ‘what relates back to an anterior,
exterior and at any rate independent construction as opposed to what is
“constructed” by the utterance’ (1975, p. 193). Maldidier points out that
for the two authors, determinatives, relative clauses, etc. are ‘the traces of
anterior constructions, of already-there discursive elements whose enunci-
ator has been forgotten’ (Maldidier, 1993, p. 114; emphasis in the origi-
nal). The ‘presence of the “unsaid” running through what is said without
an identifiable limit’ (Maldidier, 1993)6 was to become a central focus in
French discourse analysis starting in the 1980s. In 1982, Sériot resumed
his work on linguistic markers of ‘the preconstructed’, followed by a series
of other studies on the manifestations of interdiscourse within texts.7
Among others, one can mention Authier-Revuz’ publications on enun-
ciative heterogeneity (e.g. 1995), a variety of studies following Bakhtin’s
thoughts on dialogism (e.g. Brès, 1998 for dialogical markers) and also
Philippe’s (2002) and Rabatel’s (e.g. Rabatel, 2008, pp. 578–579) work
on the ‘marks of the formal instrument of enuniative fading out’.
Within a socio-cognitive framework, Paveau puts into practice the
notion of ‘prediscourse’, which is related to but different from ‘the pre-
constructed’ (2006, pp. 126–128). She defines prediscourses as ‘a set of
prediscursive collective frames (knowledge, beliefs, practices) that give
instructions for the production and the interpretation of meaning in dis-
course’ (2006, p. 118). Prediscourses have an anterior collective existence
within a group (Paveau, 2006) and are thus intersubjective, but speakers
are not conscious of them as such since they are convinced that they
themselves produce meaning. Prediscourses are immaterial, tacit, but also
discursive since they leave (indirect) markers in (the materiality of ) dis-
course (Paveau, 2006, pp. 119, 126). Paveau labels these markers ‘calls for
prediscourses’ or ‘signals of prediscourses within discourse’ (2006, p. 127)
and assembles them in three categories: discursive lines (etymologism,

lexicologism, lexicographism, call for collective wisdom, call to the
fathers, proper name), shared worlds (encyclopedic deixis, generic inter-
rogation, epistemic modality and evidentiality), textual-cognitive orga-
nizers (typology, metaphor, antithesis).
8.2.2 B
eyond the ‘Preconstructed’: New Tools
for the Discovery of Absences
My own studies are conducted within the theoretical and methodological

framework of what I call Cross-Cultural Discourse Analysis (CCDA), at
the crossroads of French Discourse Analysis, text linguistics and cross-
cultural studies.8 CCDA’s long-term aim is to come to an understanding
of ‘discursive cultures’, which can be defined by means of social
representations that are circulating within communities concerning social
objects, on the one hand, and the discourse to be held about these objects,
on the other (that is, what must, can and cannot be said about them and
how it can and cannot be said).9 The methodology consists in inferring
hypotheses on social representations from traces of discursive operations
in data belonging to a specific discourse genre. In this perspective—and
according to what is announced above (Sect. 8.1.2.)—I am trying to go
beyond searching for ‘the preconstructed’ and ‘prediscourses’ in order to
reach the most widely shared and thus ‘obvious’ social representations
within a community.
On the one hand, it is markers that are not specialized in signalling
shared representations that will (indirectly) help to detect what remains
unsaid—and thus unmarked as such—because it is ‘obvious’ for a given
group. On the other hand, we can only discover what is ‘given off’ through
non-specialized markers by means of what is said about something else.
Furthermore, some representations are so widely shared that they are not
even ‘given off’, but stay entirely unmarked. It is thus important to shift
attention from what to observe to how to observe. In effect, as pointed out
above, the unsaid is unlimited. In addition, Bourdieu draws our attention
to the fact that what is ‘hidden par excellence’ is also hidden to science
since it hides ‘in the eyes of the scientist’ (2001, p. 168). So if we want to
224 P. von Münchow
avoid a circular approach that would only lead to finding representations

the researcher is actively looking for we have to establish an observation
protocol on the basis of what is present in the data set(s) under study.
Four analytic procedures in order to identify absences can be put for-
ward (and will be put to work in Sect. 8.3):
–– searching for actors and actions associated with the ones who are
mentioned in an utterance;
–– making premises in argumentations explicit;
–– looking for ‘instabilities’ within a data set;
–– comparing different data sets.
The first two procedures are quite current, especially the search for
associated actors and actions (see van Leeuwen, 2008, for instance). The
search for instabilities corresponds to Orlandi’s suggestion to look for
silence in ‘cracks, breaches and faults’ where ‘it shows itself surrepti-
tiously’ (1994/1996, p. 42). Comparison, finally, is the foundation of
CCDA as defined above. In effect, the analyst has to be—to a certain
extent—a member of a community in order to access absences that occur
within the community. In other words, she must know which social rep-
resentations could be found in order to be able to infer them when they
remain unsaid. But she also needs to take enough distance from the com-
munity to catch a glimpse of what is considered obvious by its members,
which comparison helps to achieve.
8.2.3 T
he Interpretation of the Unsaid: Mapping
Presences and Absences in Discourse
Other than helping to reveal particularly strong and widespread social

representations, the study of absences can also play a role in highlighting
competing representations and thus social change. Indeed, according to
Orlandi (1994/1996, p. 14), ‘the movement, the uncertain relationship
between change and permanence indistinctly meet each other in silence’.
Unlike Orlandi, I believe—based on the results of my work (see in par-
ticular von Münchow, 2016)—that distinctions can be made between
different kinds and degrees of absence, which may help to reach a better
understanding of the relationship(s) between what is said and what is

unsaid. Table 8.1 shows the marking that may be expected of different
types of social representations (as the speaker sees them) and the analyti-
cal procedures which may give access to these representations. Other
than offering a synthetic overview of the descriptive tools which are dis-
cussed in the previous section, it may thus be used for the interpretation
of discursive ‘presences’ and ‘absences’ in terms of social
representations.
Table 8.1 Types of social representations, linguistic marking and analytical

procedures
Types of representations
(status within the Analytical procedures to
community) Marking access representations
1 Obvious No markers Searching for associated
actors, argumentative
premises and
instabilities within a
data set, comparison
2 Dominant (Assertion), All types of analyses
indexation (syntax, semantics,
enunciation, sequential
types, argumentation,
presupposition,
‘preconstructed’, calls
for prediscourses,
category 1 procedures
[…])
3 On their way Declining Assertion = object Content analysis
to of an utterance;
becoming object of a
dominant metadiscursive
utterance
4 Emerging Widely Indexation All types (see category 2)
challenged
5 Still Having No markers Searching for associated
unutterable become actors, argumentative
→ hidden unutterable premises and
→ hidden instabilities within a
data set, comparison
6 Inexistent No markers Comparison
226 P. von Münchow
First of all, I agree with van Dijk’s idea that it is always a cognitive
representation of a social situation ‘and not the “objective” social situation
[itself ], that influences the cognitive process of discourse production and
comprehension’ (2009, p. 5; emphasis in the original). In this p erspective,
the word ‘representations’ in the left column is to be understood as the
cognitive representations a speaker has of social representations within
the community in which the discourse will be received.
The general idea expressed in Table 8.1 is that the most explicitly
asserted representations (category 3) are the ones that are not domi-
nant—in the community of recipients, according to the speaker’s
(cognitive) representations—but are either on their way to becoming
dominant or already declining. Dominant representations, for their part,
do not need to be explicitly asserted but will only be indexically marked
(category 2).10 As for emerging or widely challenged representations (cat-
egory 4), they are not without social cost for speakers if made explicit in
discourse; hence they are mostly only indexed, if at all present. Unutterable
representations (‘conspiracies of silence’, according to Zerubavel, 2006),
that is representations that have become or are still taboo within a com-
munity (category 5), cannot be marked at all in discourse unless speakers
are willing to run the risk of being ostracized, persecuted or even killed,
depending on the community and the taboo that is broken. Finally, ‘obvi-
ous’ representations (category 1), that is the most widely shared and the
least questioned of all, cannot be marked in discourse because speakers are
not aware of them. Neither can inexistent representations (category 6), of
course.
One might ask where this conceptualization leaves the individual
speaker. First of all, it is coherent with a French discourse-analytical per-
spective to see the individual as ‘always speaking “under the cover of ”
implicit discourses of reference, of imaginations [imaginaires] function-
ing as social norms on behalf of which utterances acquire their semantic
meaning’ (Charaudeau, 2004, p. 30). Indeed, members of a group ‘co-
share’ [‘co-partagent’] knowledge, beliefs and values—‘representations’ in
my terminology—which allow them to recognize themselves as belong-
ing to the group in question. But that does not mean they are committed
to those representations, it just means that they know they exist
(Charaudeau, 2004, p. 29). So if in general utterances correspond to rep-
resentations in the categories 2, 3 or 4 of Table 8.1, a speaker can also
break a conspiracy of silence (category 5) or be the first one to introduce

a representation to a community (category 6) using explicit markers (cat-
egory 3). It needs to be understood, though, that in this case, the speaker
risks social marginalization (or worse), considering that she thinks and,
more importantly, speaks out of the box. The risk in explicitly uttering a
representation from category 1 is of a different kind. The speaker might
be considered an incompetent communicator or even be psychologically
marginalized, given that he thinks and speaks too much within the box, in
other words he does not provide anything new. In general—and whether
the (human) subject has actual agency as such or finds her individuality
in a unique combination of determinations (Lahire, 2013)—the speak-
er’s individuality becomes visible in the discrepancy between the category
of the representation to be expressed and the markers chosen to do so.
The assertion of dominant representations (category 2) also deserves
some thought with regard to the interaction between the individual
speaker and a community or its institutions. As Table 8.1 shows, domi-
nant representations generally do not need to be made explicit. If they are
nevertheless asserted, someone—either the speaker or a coercive power to
which the speaker has to obey—probably fears that they are not as widely
shared as they should be or could cease being widely shared if they are not
constantly reasserted. This can be the case in learning situations, for
example, or in authoritarian regimes. In both cases, what students and
citizens learn from what is asserted by the powerful speaker are the repre-
sentations which the power-holders want to be dominant rather than
those which are dominant in the sense of ‘widely shared’. From a strictly
discursive point of view, unlike indexation, explicit assertion always leaves
a door open for disagreement, which may of course be closed by the situ-
ational, generic and/or political context. In this case, the recipients of the
assertive discourse learn not so much what they have to think as what
they have to say (and not to say). As the Romanian-born German writer
Herta Müller famously said about both a learning situation and a totali-
tarian regime: ‘I learnt in school how to say what I didn’t think.’11
It is the more standard cases of correspondence between certain repre-
sentations and a variety of markers that I will illustrate in the following
section, though. As announced above, I will be focussing in particular on
the interplay of different categories of representations and of what is said
and what is not said as well as on the detection of ‘obviousness.’
228 P. von Münchow
8.3 Illustration: How ‘Presence’

and ‘Absence’ Interact in Discourse
In this section, the analysis of a few excerpts of data sets I examined in
three previous case-studies (see von Münchow, 2011, 2013a, 2013b,
2015) will illustrate some of the aforementioned analytical procedures
designed to reveal silences. First, I show how the search for actors and
actions associated with the ones who are mentioned helps to uncover
absences, then I focus on making premises in argumentations explicit.
Instabilities within data sets are then exploited, and finally I highlight the
heuristic value of comparison of different data sets in bringing about
what remains unsaid in one of them. The different types of silences are
systematically interpreted by means of the typical relationships between
social representations and marking accounted for in Table 8.1.
8.3.1 A
ssociated Actors and Actions, Argumentative
Premises
The following passage is extracted from what I called an intercultural

parenting book (von Münchow, 2015), written by an American life-style
journalist who observes Parisian child-rearing habits. It deals with the
matter of day care for young children12:
(1) Americans remain consumed by the question of how even normal day
care affects a child’s fragile psyche. There are headlines on whether day care
causes learning delays, makes kids more aggressive, or leaves them inse-
curely attached to their mothers. I know American moms who quit their jobs
rather than subject their kids to day care. […]
French mothers are convinced that the crèche is good for their kids. […]
French mothers do worry about pedophiles, but not at the crèche. […] ‘If
she’s going to be tête-à-tête with someone, I want it to be me,’ the mother
of an eighteen-month-old […] tells me. [Druckerman, P. (2012). Bringing
up bébé: One American mother discovers the wisdom of French parenting.
New York: The Penguin Press, pp. 104–105]
Mothers are not only the foregrounded parental actor here, but the
only one. As for fathers—who could be considered associated actors
when it comes to childcare—they are ‘backgrounded’ in the larger cotext
and outright ‘suppressed’ in the excerpt (van Leeuwen, 2008, p. 29).13
Stay-at-home mothers thus seem to be the only alternative to day care or
nannies. That is not explicitly said but it is a representation the reader
needs to activate in order to understand the passage (category 1). The
same is true for the representation that mothers are the ones who decide
which childcare arrangement to choose. In fact, the decision-making
authority of the mother seems to go much further:
(2) When the father of one of Bean’s school friends comes to fetch his daugh-
ter at our house one Sunday afternoon after a playdate, he hears Bean shouting
caca boudin as she runs down the hall. The father, a banker, looks at me warily.
I’m sure he mentions the incident to his wife. His daughter hasn’t been back
to our house since. [Druckerman, 2012, p. 167; emphasis in the original]
In this case, the analyst has to look for argumentative premises in order
to bring about the representation of the mother as the one deciding with
whom her child may or may not associate. Indeed the utterance is only
comprehensible if one links the incident being mentioned to the wife to
the daughter not coming back to play by the information that the
mother—unlike the father—has the authority to choose her daughter’s
playmates (again, a category 1 representation).
The following excerpt from a case-study on French and German
‘monocultural’ parenting books (von Münchow, 2011, 2013b) restores
the father’s decision-making power, in a way … It requires an identifica-
tion of backgrounded actions rather than actors:
(3) L’idéal est que père et mère partagent les moments du change. Mais si
le père ne souhaite pas le faire, respectez son choix. [Rufo, M., & Schilte,
C. (2003/2004). Élever Bébé. De la naissance à six ans. 2nd upd. edn. Paris:
Hachette, p. 122]
[Ideally, the father and the mother share diaper changing. But if the
father does not wish to do it, you should respect his choice.]
230 P. von Münchow
When it comes to changing diapers, the father is shown as the only

agent making choices, which the mother is expected to respect. Her own
actions—and their modality—are not mentioned but may be inferred
from the father’s choice making. Indeed, any reader will probably under-
stand that the mother has the duty to change diapers. Why is this repre-
sentation of the mother as the mandatory social childcare actor not made
explicit? Two answers to this question seem possible: Because, according
to the authors of the parenting guide, it is obvious for members of their
community that the mother is the principal caregiver (category 1 in
Table 8.1) or, on the contrary, because seeing childcare as the mother’s
exclusive duty has become unutterable (category 5).
The father is again the backgrounded actor in the next excerpt from
the above-mentioned intercultural parenting book:
(4) About the only thing wrong with Dietlind is that she can’t cook. Her
family subsists almost entirely on food from Picard, the French frozen-food
chain. [Druckerman, 2012, p. 101]
As for the argumentative premise that is necessary to understand the

utterance, it consists in the mother (Dietlind) being considered the only
possible cook in the family. Again, the representation that cooking is the
mother’s exclusive responsibility is not made explicit either because it is
so obvious that is does not need to be mentioned (category 1), or because
it has become so unutterable that it cannot be mentioned (category 5).
8.3.2 Instabilities Within a Data Set
Unlike argumentative premises, which generally remain silent and often

consist in ‘obvious’ representations, arguments or conclusions tend to be
explicit or even ‘overasserted’ (Maingueneau, 2012a), as excerpt (5) from
a French parenting book shows:
(5) C’est ensemble que les parents vont partager les joies et les soucis, mais aussi
les tâches matérielles. C’est donc au père et à la mère que je m’adresse. [Pernoud,
L. (1956/2004). J’élève mon enfant. upd. edn. Paris: Horay, p. 18]
[Together the parents share the joys and the sorrows, but also the tasks.
That’s why I address the father and the mother.]
The addressee is constructed here in a metadiscursive emphasized

assertion, that is, in an extremely explicit fashion. Moreover, the assertion
appears in the introduction of the book, which is a highly strategic loca-
tion. As Table 8.1 shows, this does not mean that the addressee of the
parenting book being both parents corresponds to a dominant represen-
tation, but rather to a representation on its way to becoming dominant
(category 3). This kind of addressee still needs to be justified, hence the
preceding argument. The emphasis (‘mise en relief ’) of the argument
(‘C’est ensemble que’) again makes the sharing of joys and sorrows, as
well as the tasks so explicit that it could not possibly reflect a dominant
representation, but only one that the author believes to be in the process
of becoming dominant.
Interestingly, the explicit construction of the addressee in the intro-
duction of this parenting book (as well as in four of the five other French
and German parenting books I analysed in von Münchow, 2011 and
2013b) does not coincide with the positioning of the addressee in the rest
of the book. This intratextual instability can be highlighted by means of
the following excerpt:
(6) Vous allez voir d’ailleurs. Bébé a faim, il tète ; si vous le caressez, il
s’arrête de téter ; si vous continuez à le caresser, il prolonge la pause tant il
est heureux de ce signe de reconnaissance qui lui est plus cher que la satis-
faction de la faim. Ou bien: son père parle doucement au bébé, « areu…
gligli… », tendrement, et, s’il sait attendre, le nouveau-né, délicatement,
presque imperceptiblement, réagit, clignote d’un œil, soulève légèrement
un coin de lèvre. [Pernoud, 1956/2004, pp. 178–179]
[You will see by the way. The baby is hungry, he suckles; if you caress
him, he stops suckling; if you continue fondling him he extends the
pause because he is so happy about this sign of recognition, which is
dearer to him than satisfying his hunger. Or: his father speaks softly to
the baby, ‘googoo… googoo…,’ tenderly, and if he is patient, the new-
born reacts delicately, almost imperceptibly, blinks, slightly lifts a cor-
ner of his lip.]
232 P. von Münchow
The deictic ‘vous’ can only refer to the mother here since the author
refers to the father in the third person. This instability reveals that several,
at least partly incompatible representations are circulating in the com-
munity in which the parenting book was written and to which it is
addressed at the same time. It is indeed likely that the author displays her
determination to address her guidebook to both parents in order to con-
form to contemporary social representations of parenthood (category 3),
but cannot persist because she is also (and probably less consciously)
influenced by a less recent set of representations, which remains in force,
but is widely challenged (category 4). It is also possible that the author
consciously endorses the representation of the mother as the primary or
exclusive addressee of parenting books (category 4), but claims to support
the more recent, ‘en vogue’, on its way to becoming dominant representa-
tion (category 3). In this case, what she deliberately tries to hide is only
imperfectly hidden, but in order for the text to be coherent, it cannot be
explicit and thus only ‘appears’ via indexation. It needs to be highlighted
that searching for intratextual instabilities is of paramount importance
here. Since there are no limits to what remains unsaid or what is not
explicit, the search for instabilities is necessary in order to detect signifi-
cant absences. Indeed, conflicting representations within a text are always
significant for discourse analysis since they are a powerful means to detect
social change as projected and constructed in discourse.
8.3.3 Comparing Different Data Sets
Another such analytical procedure is the comparison of different data

sets, which even enables the researcher to go beyond the detection of
‘quasi-silence’ (indexation) in order to reach ‘complete silence’, that is
category-1 representations. These representations are in effect dominant
enough to be absolutely unquestioned and thus obvious for members of
a community. The heuristic value of comparison will be illustrated here
by means of an example from a study about French and German history
textbooks (von Münchow, 2013a). History textbooks are eminently suit-
able data for the discovery of the unsaid. Indeed, they contain, as
Christophe and Schwedes (2015, pp. 10–11) point out, a number of
‘blind spots’ (‘blinde Flecken’) based on ‘societal consensus’
(‘gesellschaftliche[r] Konsens’). Of course, these blind spots tell us a great

deal about ‘the cultural world’ (‘die kulturelle Welt’) in which the corre-
sponding representations stay silent because they are so obvious. In order
to ‘escape the maelstrom of normalization’ (‘sich dem Sog der
Normalisierung entziehen’), it is worth identifying these blind spots
through comparison (Friedl, Christophe, & Schwedes, 2015, p. 152).
The following excerpt from the chapter on World War I in a German his-
tory textbook contains such a blind spot:
(7) Der Kriegsausbruch im August 1914 beendete eine längere

Friedensphase in Europa. Aber: Kein lähmendes Entsetzen erfasste die
Menschen, sondern es breitete sich eine große Kriegsbegeisterung aus, vor
allem in bürgerlichen Kreisen. Hier war die Bereitschaft junger Männer,
sich als Kriegsfreiwillige zu melden, besonders groß. Niedergedrückt war
zunächst die Stimmung in der großstädtischen Arbeiterschaft; das änderte
sich allerdings auch hier, als Mitte August die ersten Siegmeldungen ein-
trafen. [Lendzian, H.-J., & Mattes, W. (Eds.). (2005). Zeiten und Menschen
3. Paderborn: Schöningh, p. 194]
[The outbreak of the war in August 1914 put an end to a rather long
peace phase in Europe. However, no paralyzing horror spread, but great
enthusiasm for the war did, especially in bourgeois circles. There the will-
ingness of young men to volunteer for military service was particularly
important. The mood within the working class in the big cities was down-
beat at first but that changed there, too, when the first news of victory
arrived in the middle of August.]
Indeed, until the element ‘Niedergedrückt war zunächst die Stimmung

in der großstädtischen Arbeiterschaft’ (‘The mood within the working
class in the big cities was downbeat at first’), we do not know to which
geographical entity the text refers.14 One could think the authors are talk-
ing about the whole continent since ‘peace in Europe’ is mentioned, but
the next utterance helps to construe a more precise reference: ‘das änderte
sich allerdings auch hier, als Mitte August die ersten Siegmeldungen ein-
trafen’ (‘but that changed there too when the first news of victory arrived
in the middle of August’). It is now clear that the spatial referent is
Germany since Germany is the country that received news of victory in
August 1914. The fact that this remains unsaid presupposes and con-
structs Germany as an ‘obvious’ community, which is neither backgrounded
234 P. von Münchow
nor suppressed: it is unnecessary to specify that the authors are referring

to Germany because unless otherwise indicated, they are ‘obviously’ refer-
ring to Germany. In other words, we are dealing with a category-1 repre-
sentation here. Because of this obviousness, such informal and non-explicit
learning through a hidden curriculum probably leaves a more indelible
mark on students than explicit information that could be questioned. As
Baker (2006, p. 19) points out, ‘hegemonic discourse can be at its most
powerful when it does not have to be invoked, because it is just taken for
granted.’
But how does the analyst come to notice this significant one among all
other absences? It is the comparison to the French data in my study on
history textbooks that, by contrast, drew my attention to the ‘absent ref-
erent’. Indeed, one of the results of the study was the fact that in their
narrative of World War I, French authors tend to refer to Europe in gen-
eral or to European countries in the plural, as the following excerpts from
one of the textbooks show:
(8) L’Europe connaît un déclin économique. [Abrami, J. et al. (2012).

Histoire Géographie 3e. Paris: Belin, p. 38]
[Europe is in economic decline.]
(9) Les États organisent l’économie de guerre: ils reconvertissent les
industries en usines d’armement […]. [Abrami, 2012, p. 34]
[The states organize the war economy: They convert industries into
armament plants.]
The focus on what happened in all of Europe instead of what any one
country brought upon another is probably driven by the reluctance to
blame any of France’s now allies (and especially Germany) for the out-
break of the war or for specific actions in its course (see von Münchow,
2013a). The result is that by means of a series of linguistic markers French
authors seem to construct a European perspective, which, by contrast,
not only brings about the German perspective of German textbooks, but
also the significance of the absence in or through which it is conveyed. In
other words, it is only once the discourse analyst’s attention is drawn to
the spatial referent in French textbooks that she identifies its absence
from the explicit surface of German textbooks.
8.4 Conclusion
This study has shown that certain analytical procedures focussing on how
to observe rather than what to observe make it possible to detect absences
in discourse and thus to establish hypotheses on representations so widely
shared that speakers and recipients are probably not conscious of them. It
has also proven that absence is not always the result of suppression and
that the unsaid is often more consensual than the explicit contents of an
utterance.
The analytical procedures also give us access to conflicting representa-
tions within data sets. But we can go beyond this observation and make
hypotheses on the chronology of appearance (and disappearance) of the
different representations if we pay close attention to the kind of markers
that are in use. Indeed, contrary to what Orlandi says, ‘the movement,
the uncertain relationship between change and permanence’ do not
‘indistinctly meet each other in silence’. In fact, Table 8.1, constructed to
this end, helps to distinguish dominant from emerging and vanishing
representations by their characteristic marking. Ultimately, this proce-
dure constitutes a detection device for social change.
But what is detected among the countless absences in discourse
depends on the analyst, of course. In excerpt (4), for instance, I pointed
out that the father was backgrounded, which is not an abstract objective
interpretation, but an analysis carried out at a certain time in a certain
place. One could indeed just as well bring up the absence from cooking
duties of the children, the grandparents or any other member of the
extended family or even the larger community. It is because the father as
a regular cook seems possible in a certain context that his absence is
noticeable. Indeed, phenomena, like the unsaid in this instance, are
always meaningful or meaningless for someone, not ‘intransitively’ or in a
vacuum. As Chap. 1 in this volume states, it takes a certain ‘perceptive
framework’ to notice the unsaid. The challenge for the discourse analyst
is not to use the perceptive framework of an empirical or a model inter-
locutor or reader (as she imagines it). It is to develop her own framework
in order to detect the absence not of what she would expect to be present in
discourse, but of what she knows could be present, but is absent for reasons
that tell us quite a lot about the society at stake.
236 P. von Münchow
Notes
1. Schröter also refers to Herdina (1996, p. 30) on the subject of
silencing.
2. On the history of French Discourse Analysis, see, for example, Maldidier
(1993) and Maingueneau (2012b). In order to compare with Critical
Discourse Analysis, see Blommaert (2005, pp. 21–38).
3. All translations in the chapter are mine.
4. ‘Discursive formation’ is used here in the sense given to it by Haroche,
Henry, and Pêcheux (1971, p. 102), that is what determines ‘what can
and must be said […] from a given position in a given set of
circumstances’.
5. The next two paragraphs draw on a more extensive development on this
subject in von Münchow (2016).
6. Maldidier refers to a 1982 handwritten note from Pêcheux here.
7. See Pordeus Ribeiro (2015, pp. 163–164) for a recent summary of these
studies.
8. For a detailed presentation of CCDA, see von Münchow (2004/2009,
2010, 2015).
9. My definition of ‘social representations’ differs from the ones that social
psychologists like Guimelli (1999, p. 63) put forth in that I see them as
covering the beliefs, the knowledge and the opinions that members of a
group know about and are able to use (in whatever way) rather than pro-
duce and/or share.
10. Gumperz’s (1996) conception of ‘indexicality’ is defined as follows by
Kramsch (2004, p. 248): ‘indexicals’ […] indirectly refer to, or “index,”
the personal, social, cultural, and ideological subject position of the
speaker and require interpretation on the part of the participants’.
11. Der Spiegel 1/2011, p. 138.
12. Unless otherwise specified, the italics in all excerpts are mine and point
out elements that are particularly important for the analysis.
13. For van Leeuwen, ‘backgrounding’ entails that ‘the excluded social actors
may not be mentioned in relation to a given action, but they are men-
tioned elsewhere in the text, and we can infer with reasonable (though
never total) certainty who they are’ (2008, p. 29). ‘Suppression’ is a more
radical exclusion from the text.
14. The excerpt marks the beginning of a new section. The reader thus can-
not rely on previous information for the construal of reference.
References
Achino-Loeb, M.-L. (2006). Introduction: Silence as the currency of power. In
M.-L. Achino-Loeb (Ed.), Silence. The currency of power (pp. 1–19). New York
and Oxford: Berghahn.
Authier-Revuz, J. (1995). Ces mots qui ne vont pas de soi. Boucles réflexives et non-
coïncidences du dire (Vol. 2). Paris: Larousse.
Baker, P. (2006). Using corpora in discourse analysis. London: Continuum.
Blommaert, J. (2005). Discourse: A critical introduction. Cambridge: Cambridge
University Press.
Bourdieu, P. (2001). Science de la science et réflexivité. Paris: Raisons d’agir
éditions.
Brès, J. (1998). Entendre des voix: de quelques marqueurs dialogiques en fran-
çais. In J. Brès et al. (Eds.), L’autre en discours (pp. 191–212). Montpellier:
Université Paul Valéry – Montpellier 3, Praxling & Université de Rouen,
Dyalang.
Charaudeau, P. (2004). Tiers où es-tu ? In P. Charaudeau & R. Montes (Eds.),
La voix cachée du tiers. Des non-dits du discours (pp. 19–42). Paris: L’Harmattan.
Christophe, B., & Schwedes, K. (2015). Einleitung. In B. Christophe &
K. Schwedes (Eds.), Schulbuch und Erster Weltkrieg. Kulturwissenschaftliche
Analysen und geschichtsdidaktische Überlegungen (pp. 7–13). Göttingen: V&R
unipress.
van Dijk, T. A. (2009). Society and discourse. How social contexts influence text and
talk. Cambridge: Cambridge University Press.
van Dijk, T. A. (2014). Discourse and knowledge: A sociocognitive approach.
Dreesen, P. (2015). Sprache – Wissen – Kontingenz. Die Kontrastive
Diskurslinguistik in der Schulbuchforschung am Beispiel deutscher und pol-
nischer Geschichtsschulbücher. In J. Kiesendahl & C. Ott (Eds.), Linguistik
und Schulbuchforschung. Gegenstände – Methoden – Perspektiven (pp. 53–83).
Göttingen: V&R unipress.
Foucault, M. (1969). L’archéologie du savoir. Paris: Gallimard.
Friedl, S., Christophe, B., & Schwedes, K. (2015). Den Ersten Weltkrieg unter-
richten –Didaktische Prinzipien und Herausforderungen im Spiegel von
Schulbüchern. In B. Christophe & K. Schwedes (Eds.), Schulbuch und Erster
Weltkrieg. Kulturwissenschaftliche Analysen und geschichtsdidaktische
Überlegungen (pp. 93–154). Göttingen: V&R unipress.
238 P. von Münchow
Goffman, E. (1956). The presentation of self in everyday life. Edinburgh: University

of Edinburgh Social Sciences Research Centre.
Guimelli, C. (1999). La Pensée sociale. Paris: PUF.
Gumperz, J. J. (1996). The linguistic and cultural relativity of inference. In J. J.
Gumperz & S. C. Levinson (Eds.), Rethinking linguistic relativity
(pp. 374–406). Cambridge: Cambridge University Press.
Haroche, C., Henry, P., & Pêcheux, M. (1971). La sémantique et la coupure
saussurienne. Langages, 24, 93–106.
Heinemann, W. (1999). Das Schweigen als linguistisches Phänomen. In
H. Eggert & J. Golec (Eds.), ‘…wortlos der Sprache mächtig.’ Schweigen und
Sprechen in der Literatur und in sprachlicher Kommunikation (pp. 301–314).
Stuttgart and Weimar: Metzler.
Herdina, P. (1996). The manufacture of silence (or how to stop people doing
things with words). In G. Grabher & U. Jessner (Eds.), Semantics of silences
in linguistics and literature (pp. 29–44). Heidelberg: Winter.
& Society, 13, 347–372.
Kramsch, C. (2004). Language, thought and culture. In A. Davies & C. Elder
(Eds.), The handbook of applied linguistics (pp. 235–261). Oxford: Blackwell.
Kurzon, D. (1998). Discourse of silence. Amsterdam and Philadelphia: John
Benjamins.
Lahire, B. (2013). Dans les plis singuliers du social. Individus, institutions, sociali-
sations. Paris: La Découverte.
van Leeuwen, T. (2008). Discourse and practice. New tools for critical discourse
analysis. Oxford: Oxford University Press.
Maingueneau, D. (2012a). Les phrases sans texte. Paris: Armand Colin.
Maingueneau, D. (2012b). Que cherchent les analystes du discours?
Argumentation et Analyse du Discours, 9. https://doi.org/10.4000/aad.1354
Maldidier, D. (1993). L’inquiétude du discours. Un trajet dans l’histoire de
l’analyse du discours: le travail de Michel Pêcheux. Semen, 8, 107–119.
von Münchow, P. (2004/2009). Les journaux télévisés français et allemands. Plaisir
de voir ou devoir de s’informer. Paris: Presses Sorbonne Nouvelle.
von Münchow, P. (2010). Langue, discours, culture: quelle articulation? Signes,
discours et sociétés, 4. Retrieved from http://www.revue-signes.info/docu-
ment.php?id=1439 (part I); http://www.revue-signes.info/document.
php?id=1452 (part II).
von Münchow, P. (2011). Lorsque l’enfant paraît. Le discours des guides parentaux
en France et en Allemagne. Toulouse: Presses Universitaires du Mirail.
von Münchow, P. (2013a). Le discours sur les conflits entre nations dans les
manuels d’histoire français et allemands. Représentations de la Première
Guerre Mondiale. In A.-C. Gonnot, N. Rentel, & S. Schwerter (Eds.),
Dialogue(s) entre langues et cultures (pp. 85–106). Frankfurt: Peter Lang.
von Münchow, P. (2013b). A discourse-analytical perspective on representations
of fatherhood and motherhood in French, German and American parental
guide books. In V. Sibirtseva (Ed.), Multi/interkulturalism? Opyt Avstrii, Rosii,
Evropy (pp. 222–239). Nijni Novgorod: Dekom.
von Münchow, P. (2015). ‘Believe me when I say that this is not an attack on
American parents’: The intercultural in intercultural parenting books.
Multilingua: Journal of Cross-Cultural and Interlanguage Communication, 34,
797–820. https://doi.org/10.1515/multi-2014-0058
von Münchow, P. (2016). Quand le non-dit n’est pas l’implicite : comment
rendre visibles les silences dans le discours? Signes, discours et sociétés, 17.
Retrieved from http://ec2-52-11-162-110.us-west-2.compute.amazonaws.
com/lodel/revue-signes/index.php?id=108
Orlandi, E. (1994/1996). Les formes du silence. Dans le mouvement du sens. Paris:
Editions des Cendres.
Paveau, M.-A. (2006). Les prédiscours. Sens, mémoire cognition. Paris: Presses
Sorbonne Nouvelle.
Pêcheux, M. (1975). Les Vérités de la Palice. Linguistique, sémantique, philosophie.
Paris: Maspero. Partly reproduced in Pêcheux, M. [texts from] (1990).
L’inquiétude du discours. texts selected and presented by D. Maldidier
(pp. 175–244). Paris: Éditions des Cendres.
Philippe, G. (2002). L’appareil formel de l’effacement énonciatif et la pragma-
tique des textes sans locuteur. In R. Amossy (Ed.), Pragmatique et analyse des
textes (pp. 17–34). Tel-Aviv: Presses Universitaires de Tel-Aviv.
Pordeus Ribeiro, M. (2015). ‘Droite’ et ‘gauche’ dans les discours d’un événement
électoral. Une étude sémantique et contrastive des presses brésilienne et française.
Les élections présidentielles de 2002 au Brésil et de 2007 en France. Doctoral
thesis, Université Paris 3-Sorbonne nouvelle.
Rabatel, A. (2008). Homo narrans. Pour une analyse énonciative et interactionnelle
du récit. Tome II. Dialogisme et polyphonie dans le récit. Limoges:
Lambert-Lucas.
Schröter, M. (2013). Silence and concealment in political disourse. Amsterdam
and Philadelphia: John Benjamins.
Sériot, P. (1982). Langue de bois et discours de vent. Essais sur le discours sovié-
tique, 2, 5–39.
240 P. von Münchow
Thiesmeyer, L. (2003). Introduction: Silencing in discourse. In L. Thiesmayer

(Ed.), Discourse and silencing. Representation and the language of displacement
(pp. 1–33). Amsterdam and Philadelphia: John Benjamins.
Zerubavel, E. (2006). The elephant in the room: Silence and denial in everyday life.
9
What’s Not in a Frame? Analysis
of Media Representations
of the Environmental Refugee
Nina Venkataraman
9.1 Introduction
The original impetus for this chapter came with the recognition that very
little is known about environmental refugees. Despite the increased focus
on climate change in the recent past, very little is understood about vic-
tims caught in a debate about climate change and displacement. One of
the reasons why this issue has failed to gain traction in elite newspapers is
the selective framing of the issue. In addition, this issue is infrequently
covered in the newspapers, with only an average of two or three articles
per year featuring them.
To understand how environmental reporting frames an issue that may
lead to selective absences of other factors that remain ideologically impor-
tant to the discussion, two conceptual tools were studied together: frames
and patterns of absences. This chapter echoes this volume’s editor’s sug-
gestion that the ‘choices as to what is considered to merit perception and
communication are usually not made with the deliberate intention to
N. Venkataraman (*)
National University of Singapore, Singapore, Singapore

242 N. Venkataraman
exclude others but absences result from the process of choosing …’

(Chap. 1 in this volume). It promotes the idea that frames need to be
examined alongside patterns of absences as they provide insights into
how elite newspapers (or any newspaper, institution, or social group)
might mute discussions or even marginalise the importance of an issue.
Thus far, media studies of framing have suggested absences as only a by-
product of the framing strategy itself. They have therefore not delved into
the linguistic devices (used consciously or otherwise) that show absences
of alternative perspectives or voices created by the salient frames. In
studying patterns of absences, this study aims to contribute to framing
studies by showing that both conceptual tools should be applied together
if we are to show that an issue is framed in a certain manner that is (dis)
advantageous to certain groups. Without application of this approach,
we are unwittingly participating in selectively making absent the evidence
for the projection of the frame. This is also an important step in examining
how key players choose to discuss the issue, how minimal coverage equals
less engagement at the policy level, and how counter-arguments about
the issue are muted or made irrelevant.
The emphasis in this study is to establish the nature and scope of
absences by distinguishing different manifestations of absences both
within and outside of single clauses and to examine the repeated use of
certain words to mask the reality. This study will examine three catego-
ries of absences: traces, masks and voids (Stibbe, 2015). Traces and
masks1 are the resultant effects of selective choices being foregrounded
against information that may not be so salient and discursive. The void,
on the other hand, while falling under the umbrella term of absence,
does not have a possible presence against which an absence can be iden-
tified in the newspaper article itself as there is no textual trace, yet,
access to a cross-media comparison and access to academic literature on
the same issue allows one to understand these voids are arguably rele-
vant to the discussion.
Once can then conclude that one of the most important tools through
which this value-based research can validate claims made about differen-
tial projections of hegemony and power is precisely through examining
frames and patterns of absences together.
What’s Not in a Frame? Analysis of Media Representations… 243
9.1.1 Environmental Refugees
In preparing for a warmer world, movement of people in response to

climate change can be an adaptive strategy that includes escaping danger
and shows increasing resilience (Morrissey, 2009; Tacoli, 2009). People
who are forced to move across borders due to natural disasters2 resulting
from climate change are widely identified as ‘environmental refugees’.3
Biermann and Boas (2010, p. 67) define these victims of climate change
as ‘people who have to leave their habitats (and home territory) immedi-
ately or in the near future because of sudden or gradual alterations in
their natural environment related to at least one of three impacts of cli-
mate change: sea-level rise, extreme weather events, and drought and
water scarcity.’ What is problematic about this definition is that the
cause-and-effect relationship between environmental change and migra-
tion remains little understood (Black, Kniveton, & Schmidt-Verkerk,
2013), is poorly theorised and lacks detailed empirical evidence (Stal,
2009). Yet there are other researchers, such as Myers (2002) and
El-Hinnawi (1985), among others, who show how environmental change
contributes directly to migration by forcing people out of areas that are
becoming increasingly uninhabitable.
The numbers of people permanently displaced primarily because of
climate change-related phenomena and environmental deterioration
were estimated in 2010 to be 50 million (UNFCCC, 2007) and are pro-
jected to be between 200 million and one billion by 2050 (Christian Aid,
2007; Stern, 2007; Brown, 2008). While these predictions are frequently
cited, they are also being increasingly challenged because of the method-
ological difficulties involved in separating environmental drivers from
other triggers of migration (Black et al., 2013) and of confidently mea-
suring future environmentally induced population movements. Part of
the problem arises from the difficulty involved in determining whether
people who move are doing so voluntarily or are being forced, whether
their relocation is permanent or temporary, and whether the movement
is internal to the country in question or takes place across international
borders. Additionally, these figures are simply estimates of the numbers
of people at risk and not necessarily of those who are likely to move
244 N. Venkataraman
(Tacoli, 2009). In fact, many poor people at risk cannot afford to move
(Matthew, Barnett, McDonald, & O’Brien, 2010).
Moreover, these discussions are complicated by international legal
regimes that rest on the consent of many countries with heterogeneous
stances on climate-induced population flows and often conflicting com-
mitments to reducing the number of refugees entering through their
borders.
The descriptive terms ‘climate refugee’ and ‘environmental refugee’
have been questioned in academic circles (McAdam, 2012) for the rea-
sons mentioned above. Yet these terms are used interchangeably by jour-
nalists as well as those geographers and environmentalists (El-Hinnawi,
1985; Myers, 2002) who wish to draw attention to the plight of these
victims of climate change.
These terms are also seen as erroneous in legal circles as climate change
is not classified as a source of persecution, a dimension that is crucial to
the definition of a refugee in the 1951 Refugee Convention. At present,
there is no explicit mention of climate change being a contributory fac-
tor, thus, there appears to be no coherent legal or policy response to these
victims of climate change.
9.2 Operational Definitions

The operational definitions of these key terms used in this chapter are
presented below.
9.2.1 Analytical Tool 1: Frames
The conceptual tool of frames is used with the assumption that frames are
an inherent part of cognition (Goffman, 1974) employed to c ontextualise
and organise issues, events, and occurrences. They are externalised as
linguistic tools that can be used by actors seeking to influence the way
people perceive or discuss an issue or problem.
The basic definition of a frame includes the notion that there is, ‘an
emphasis on the salience of different aspects of a topic’ (de Vreese, 2005,
p. 27). By emphasising some elements of a topic over others, a frame
provides a way to understand an event or issue by stimulating certain
constructs and values. Frames in the news can be examined and identified
by ‘the presence or absence of certain keywords, stock phrases, stereo-
typed images, sources of information, and sentences that provide the-
matically reinforcing clusters of facts or judgments’ (Entman, 1993,
p. 52).
This research focuses on the issue-frames approach (Gamson, 1992;
Nelson & Oxley, 1999; Entman, 2004) by understanding how specific
frames relevant to the topic in question are constructed. Issue-frames pro-
vide a high level of specificity and details relevant to the issue in question
(Hartman & Weber, 2009). The frames that emerge may or may not be
like the news-frames commonly seen in news media.4
9.2.2 Analytical Tool 2: Patterns of Absences
Frames rest on two dimensions: the selection of and access to sources by

claim-makers, and, the presentation and evaluation of the arguments,
evidence and actors involved in the news report. This implies that some
sources, arguments, and judgments will not be selected (Huckin, 2002).
These aspects of an issue which are not mentioned or are presented in
non-explicit manner, tend to alter the character of the issue by making
unmentioned aspects of the issue appear less important than those that
are mentioned (Entman, 2004, p. 54). If these instances of absences are
repeated over time, then a pattern of absence is observed. Therefore, to
examine patterns of absences, it is essential to study repeated instances of
the absences to show how people, their claims and issues are suppressed,
backgrounded, and excluded from texts.
It is important to acknowledge that a discursive representation of an
issue cannot contain all possible information about an event. There must
be some ‘discursive simplification’ (Jessop, 2000, p. 324). Therefore, the
246 N. Venkataraman
limits to a news-maker’s knowledge, his or her biases, the patterns of

ideology and behaviour dictated by the society of which the news-maker
is a member, and the features of the text and discourse context all act as
selective ‘filters’ (Fairclough, 2003, p. 42), allowing some information to
make it into the construction of reality and barring other information
from doing so.
To examine patterns of absences, van Dijk’s (2009) notion of context
helps distinguish between what is important and relevant to the discus-
sion and what is not. Context is defined as ‘the structure of all proper-
ties of the social situation that are systematically relevant for the
production, comprehension, or functions of discourse and its struc-
tures’ (van Dijk, 2009, p. 130). Context includes the professional or
social domain, genre, purpose, location, date, time, circumstances, par-
ticipants’ role, that help construct a news story. To determine what is
and is not important in the context can be only be done by comparing
what ‘information may be left implicit’ and what should be ‘explicitly
stated’ (2009, p. 30). A text analysis can therefore reveal through con-
text the distinction between what could have been said and yet was not.
Thus, the fact that news reporters have access to voices that include
environmentalists, politicians or even the victims themselves for the
purpose of writing a news article forms relevant context for this issue.
The genre of news reporting allows only limited space in the newspaper
for discussion on the issue, which, in turn influences the selective
emphasis on authoritative voices rather than people on the ground.
Already the context shapes the discourse in an unfair fashion with more
coverage devoted to the authoritative voices rather than the victims’
voices. One acknowledges that these authoritative voices help project
trust and a sense of surety in terms of quality resources used by the
newspaper, but, if only less than 0.6 per cent of space on environmental
refugees over a 30-year period includes the victims’ own voices, then we
see context itself as being a powerful tool to hegemonically construct
certain frames.
Once the context is detailed, to arrive at an accurate overview of how
an issue is constructed in news reporting, we need to examine what is
missing from the text and what is alluded to but not presented directly
along with what is framed. The specific absences examined were textual
and thematic absences.
Traces, much like the textual silences suggested by Huckin (2002),
leave a trace behind in the text that is retrievable either at the clausal,
sentential, or textual level. The basic premise here is that a trace allows for
an omission of some piece of information that is pertinent to the topic at
hand (Huckin, 2002, p. 348) and it is retrievable. Thus, a trace is ‘some-
thing [that] is erased but still present’ (Stibbe, 2015, p. 149).
Traces include the study of presuppositions and implicatures (as shown
in Fig. 9.1). Presuppositions have assumptions built into the text, and
absence is built into the clause itself, whereas implicatures are suggestions
connoted by the text, with each reader constructing the implicative
absence suggested by the implicatures differently. As suggested by the
Editors, presuppositions and implicatures need not be considered
absences as ‘it is hard to explicate the information all the time’ (Chap. 1
in this volume); the focus here is on the repetitive pattern of using
implicatures and presuppositions to suggest information, while directing
the attention of the reader to what is being said rather than what is left
absent. The focus in this chapter is on a pattern being established and not
on the occasional instances of their usage.
A mask consists of what is erased and ‘replaced by a distorted version
of itself ’ (Stibbe, 2015, p. 149). These include the repetitive use of sugges-
tive lexical items, metaphors, nominalisations, and transitivity patterns. In
these repetitive uses, what is absent is reframed differently. Nominalisations
express actions or processes as nouns, thus making processes and partici-
pants less explicit (Halliday & Martin, 1993). Passive voice use also allows
for the omission of agents. Here in agentless discourse, specific parties
responsible for unethical actions may be hidden. A less obvious risk is
that responsibility may be ‘diffused to a generic we, in other words, blam-
ing individuals inappropriately for systemic problems caused primarily
by institutions’ (Schleppegrell, 2001). Even resistive actions by people on
the ground can be muted with the passive voice.
Voids (Stibbe, 2015) may be explained as thematic absences that ignore
a topic, theme, or subject altogether (Kurzon, 2007). These absences,
unlike the previous two ‘epistemologically salient cases of absences’
248
Aim:
Paern of
Absences
N. Venkataraman
Linguisc device 2:
Linguisc Device 1:
Framing Textual and Themac
Absences
Category 3:
Reasoning Category 2:
Framing Devices Themac
Devices Textual absences
absences
Appeal to principles
Roots (causal
Metaphors Catchphrases (a set of moral TXT1: Traces TXT 2: Masks V1: Voids
analysis)
claims)
Consequences
Examples Descripons (a parcular type of T1:Implicatures T2:Presupposons M1:Metaphors M2:Metonyms
effect)
M3:Transivity
M4:Nominalisaons
paerns
Fig. 9.1 Overview of the conceptual tools

(Chap. 1 in this volume) can only be deemed important to the discussion

when one reads academic literature on the issue. For example, coral
bleaching is an important effect of the warming of sea waters that is
repeatedly discussed in the climate change literature (Jaleel, 2013). Yet
the connection between coral bleaching and how it affects the lives of
people living in Maldives is not something that is foregrounded in any
discussion of the selected news-reporting sources. Threats to the existence
of coral reefs come from coral and sand mining, solid waste disposal,
dredging for construction, warming sea waters and ocean acidification
(Jaleel, 2013). There is a complete void regarding these issues and how
they affect low-lying nations, their way of life and their livelihoods. To
claim that these absences are deliberate is questionable. The distinguish-
ing feature of these voids is that these missing topics constitute an impor-
tant part of the issue but are not mentioned at all by both news-reporting
sources. Yet these absences remain unjustified when we understand how
relevant it is to the discussion. While academic literature provides multi-
faceted perceptions on the problems confronting these victims and their
countries, newspapers need not necessarily subscribe to anything that is
seemingly low in terms of newsworthiness or due to the journalists’ own
value judgments or the political leanings of the institution in question.
Finally, a ‘pattern’ of absence is established if traces, masks or voids are
found in over 20 per cent of the data universe. In most cases, a systematic
analysis of texts will uncover patterns of absences. If these patterns frame
environmental refugees by presenting certain facets only, then one is enti-
tled to ask whether these representations are balanced and fair.
9.3 Methods and Design

The dataset for this research was acquired from news reports published in
The Times and The Guardian from 1985 to 2015. It was generated using
the LexisNexis database with specific keyword search terms ‘climate
refugee/migrant’ and/or ‘environmental refugee/migrant’.5 The list was
generated using the ‘major mentions’ search functions with start and end
dates of 1st January 1985 to 31st December 2015. The major functions
limit the search to include the terms ‘environmental refugees/migrant’,
250 N. Venkataraman
‘climate refugees/migrant’ in the headlines, lead paragraphs and indexing

tags. The dataset includes all types of articles including news articles, edi-
torials, opinion pieces and features. Articles were selected only if they had
more than 100 words per article, excluding the headline and summary
The final number of articles used for this research was 96 in number.
One advantage in using LexisNexis was the appearance of fewer fram-
ing mechanisms––no pull quotes and no logos. Photographs were pres-
ent, but they were not included for this study. The focus was entirely on
textual analysis.
A 30-year time period was chosen as it presented an overview of the
issue from the time when climate change and the resulting population
flows were first presented to the reading public when a few landmark
publications ‘raised the issue and provided alarming estimates of num-
bers of people foreseen to move’ (Piguet, Pécoud, & de Guchteneire,
2011, p. 4), to present-day newspaper representations that recycle old
arguments on the same issue. The choice of the rather long time period
ensures that the results would not be distorted by specific climatic events
and their resultant massive displacement like that of the Tsunami of 2009
or the 2010 Haiti earthquake.
9.3.1 Four Issue-frames
Each article collected was read closely. The central organising idea of each
text or frame was then established. This assignment of the central
organising idea is conditional and is determined by the presence of spe-
cific syntactic structures, patterns in the arrangement of words and
phrases, and the thematic and rhetorical structures. Once the frames were
identified, the news articles were categorized into the four issue-frames
(shown in Table 9.1). The close reading of the newspaper articles also
helped provide a comprehensive list of subtopics6 relevant to the issue.
Analysis was done in multiple iterations.
The framing devices (Carvalho, 2008) examined were metaphors,
exemplars (historical examples from which lessons are drawn and/or anal-
ysis is provided), catchphrases and depictions in the form of explana-
tions, descriptions and stereotypes. The reasoning devices examined were
Table 9.1 Overview of four issue-frames

Name of Framing devices
issue-frames Relevant exemplars used
Effects of [1] Spring comes earlier, winters are milder Description of
climate and summer hotter. Frogs are croaking climate change’s
change and flowers blooming unnecessarily. consequences
(other than (Nicholas-Lord, The Times, 2006, p. 2) detailed.
security risk)
Climate [2] The report said that security services Consequence
change as a would be challenged increasingly by the posited
security risk numbers of refugees and the
Government would need to consider
stronger border control. (Smith, The
Times, 2008, p. 7)
Migration as [3] Europe and the US face increased Countries
a problem pressure from people driven from North mentioned,
Africa and Latin America by causal link
deteriorating soil and water conditions. provided.
(Adam, The Guardian, 2005, p. 24)
Protecting our [4] He said there was a danger of a The metaphor
shores backlash in rich countries. ‘the climate in ‘doors will shut’
Europe, North America and Australia is explained.
not conducive to a relaxed debate about
increasing migration. There is a worry
doors will shut if we start that
discussion,’ he said. (Randerson, The
Guardian, 2009, para. 14)
roots (causal analysis), consequences and appeals to principles (a set of

moral claims). Decisions to name a specific issue-frame were based on the
appearance of a frame, and if multiple frames appeared, then a time-space
rule was used to decide which theme was given the most space and that
was coded as that article’s issue-frame. The articles were coded manually,
a frame for each article was simply coded as 1 for present and 0 for not
present and multiple stances of an issue-frame were not recorded for the
number of different instances of the same frame within an article.
As the articles were coded manually, the coding validity was verified by
enlisting the help of two colleagues unconnected to the project to indi-
vidually code a random sample of 20 per cent of the articles. The
Krippendorff”s α result was 0.72 for this three-way intercoder agreement.
Given that Krippendorff’s α is a rigorous and conservative index for inter-
252 N. Venkataraman
coder agreement and that a three-way test punishes incongruent coding

more than a two-way test, the coding and the resultant data are accept-
able at this coefficient level.
Table 9.1 gives an overview of the framing and reasoning devices.
There were two consequences that emerged with the classification of
the four issue-frames: one brought to attention the patterns of absences
(discussed further in Sect. 9.3.2); and, the other consequence was to
observe that the focus was on presenting the issue as a problem based on
criteria of newsworthiness rather than addressing intrinsic humanitarian
justice.
The conclusion one reaches after examining the issue-frames is that the
multifaceted problem is presented in a limited manner, mostly as observ-
ers from the outside, without access to the lived experiences of the victims
from the inside. Individual countries affected by the phenomenon are
actors charged with administrating any new regime designed to reduce
carbon emissions and change the way of life of their populations with
expert nations on the outside. Any help offered to the affected countries
was in the form of monetary aid and intellectual sharing of knowledge
and expertise.
9.3.2 Patterns of Absences
While issue-frames were categorised, simultaneously the context (van

Dijk, 2009) for this issue was defined too. For the context to be detailed,
the researcher first generated a list of subtopics relevant to this issue.7 The
subtopics were generated by also reading relevant academic literature and
other media reports on the issue. The comprehensive list of subtopics
helped determine how participants, actions, ideas, arguments, and judge-
ments are highlighted or alternately suppressed, backgrounded, and
excluded from texts by comparing their presence as against their absence.
Thematic patterns generated over the 30-year period were listed. The the-
matic patterns highlighted the issue-frames that readers were exposed to
time and again, along with the presence of some subtopics that were not
framed too often. Each subtopic was weighted and a theme was fore-
grounded if it found itself within the first few paragraphs, given the fact
that reporters use these paragraphs to pack in the punch, and, readers use
a top-down approach to reading the newspaper. It was also foregrounded

if it was discussed in over 60 per cent of the space in the article. In addi-
tion, constructing the context included information like the conventions
of news-reporting that were mentioned before.
Let us work with one brief example given below. Located within the
first paragraph of an article, the following information is foregrounded:
[5] Climate change will have devastating effects on millions of people,

animals and plants as fertile land turns to deserts and coasts flood, accord-
ing to a confidential United Nations report on global warming. (Brown,
The Guardian, 1990, para. 1)
By examining the text, we understand that it is weighted under the

subtopic of climate change as a problem. The other consequence men-
tioned in the article included migration as a problem. In detailing the
context in this fashion, one understands what is foregrounded is humans
and what is absent in the reportage is how humans who have no choice
must stay. Environmental migration is presented as result of stresses on
other systems like land, water and health.
In sum, it is vital to understand context if we are to understand what
is ideologically important, and the powerful stances projected to support
them. Yet the pattern of absences is revealed ‘through its systematic
absence, backgrounding, or distortion in text’ (Stibbe, 1995, p. 146).
When journalists omit details as well as certain stances by claim-makers,
it can be presumed that it is common-sense knowledge that is left out. Yet
an analysis of the data and of the context of the issue reveals patterns of
absences surrounding the issue-frames.
9.4 R
esults: Issue-frames and Patterns
of Absences
The issue-frames provide organised ways of understanding the world,
whereas the presuppositions, implications and suggestive vocabulary typ-
ify the experiences of the victims for the readers through selective framing
and absences of other aspects. This chapter only presents a brief overview
of the four issue-frames and a few interesting patterns of absences.
254 N. Venkataraman
9.4.1 F rames 1 and 2: Climate Change and Migration

as Problems
Climate change as a problem (in 56 per cent of the articles) and migra-
tion as a problem (72 per cent) are two issue-frames often repeated in
both newspapers. The ‘climate change as a problem’ frame is constructed
through arguments, examples, and judgments establishing that the main
driver of migration is climate change itself. Articles construct the reality
of climate change in terms of its effects on humans, water, soil and health.
Populations are shown to have lost resources and their social networks.
Climate change is blamed squarely for the deterioration of the quality of
life of these affected communities, which are poised to relocate or have
already done so.
[6] Everyone is aware of the environmental problems of global warming and

deforestation on one hand and the social problems of increasing poverty and
growing shanty towns on the other, Dr. Astrid Heiberg, the president of the
international federation, said: ‘But when these two factors collide, you have
a new scale of catastrophe.’ (Brown, The Guardian, 1999, para. 3)
People whose lives rely heavily on their natural capital (land and its
resources) are most vulnerable to changes in the ecosystem. While
descriptions of geographical locations and communities affected are pre-
sented from the perspectives of climatologists, environmentalists, and
politicians, all of them agree that climate change has adverse impacts, this
consequential dimension of an otherwise multi-faceted problem provides
only partial explanations for the phenomenon, making absent the effect
on non-human species. In addition, there is a void regarding the discus-
sion of what happens to people who do not have the ability to move
either further inland, into neighbouring urban sprawls or to different
countries.
‘Migration as a problem’ is an issue-frame that presents itself in two
ways. One as an argument against the relocation of these victims, the
other as characterisation of the effects of climate change as a problem.
What is framed is the recommendation that there is capacity for a social
system to respond to climate change by moderating or avoiding its con-
sequences which depends solely on promoting adaptive measures within

countries. As these communities’ resources are often meagre, collabora-
tion and international funding constitute the means of circumventing
migration as an option:
[7] As the movement of refugees would be unwelcome and could be even

resisted by force, governments must work to manage the problem them-
selves. (Ardill, The Guardian, 1989, p. 4)
One interesting point to be made about the ‘migration as a problem’

frame is the frequency with which environmental refugees are posited as
a security risk (Page & Redclift, 2002). These articles almost always (in
83.7 per cent of articles in this frame) discuss security as a means by
which the First World has been warned of the consequences of popula-
tions crossing international borders.
[8] The security council should join the general assembly in recognising
climate change as a threat to international peace and security. It is a threat
as great as nuclear proliferation or global terrorism. (Goldenberg, The
Guardian, 2011b, para. 16)
The consequence of outward migration is portrayed more often, rather

than detailing how ecological disasters are the starting point for these
stresses, followed by making a connection and explaining how these
stresses might result in outward migration. Thus, empathy is stirred for
host nations straining to meet the influx of these refugees whereas areas
of drought, water scarcity and coastal flooding are paired as characteristic
descriptors of developing nations. There is more emphasis on protecting
shores from the migration flows that already include victims of civil war
and unrest.
9.4.2 Pattern of Absence: People and Places
In projecting climate change and migration as issue-frames, both news-

papers described specific geographical regions affected by climate change.
Livable tracts of land had disappeared, forcing populations to change
256 N. Venkataraman
their way of living and sources of livelihood. In addition, what was made
absent is that using natural capital (land) is the only way of life for these
people, forcing them to become unskilled labour that moves to urban
spaces and suffers within their country; or, become unskilled labourers
which makes it harder to cross international borders.
Repeated presentation of environmental refugees as farmers (in 87 per
cent of the instances) also makes absent the fact that this problem affects
more than a small group of people who live off the land. If climate change
was framed as something that affects a few farmers, then there is no rea-
son for policy-makers to deliberate on ways in which to address this issue.
The following example limits the depiction of environmental refugees to
those people whose income is sustained by the land only.
[9] Half of all farmlands are so eroded that they are ‘unreclaimabale’ and in
some localities, much of the landscape has lost virtually all its soil cover.
(Myers, 1995, The Guardian, p. T24)
The repeated representation of a few farmers limits the universe of pos-

sible social meanings contained in each frame, thereby erasing how they
can be included in subsequent discussions at policy level. It also makes
absent the rest of society that may include people other than those mak-
ing a living through agriculture.
When a few references are made to whole societies being affected, then
reference is only made to low-lying coastal cities. Coastal areas are shown
to face the imminent dangers of climate change.
[10] As the Greenland and Antarctica ice sheets melt, rising sea levels will
threaten coastal cities worldwide … (Meachers, The Times, 2006, p. 21)
The repeated foregrounding of coastal areas and their destruction

makes absent other land-locked areas that suffer desertification, drought
and destruction. These areas include large areas of Africa, the Americas
and Asia and even Europe.
Interestingly, metaphors as masks present climate change as an invad-
ing force. Examples include: ‘Contingency plans are being prepared for
the evacuation of the population to other parts of the island’ (Guest, The
Guardian, 1989, para. 4.). ‘The threats to the atmosphere from industrial
gas emissions would be compounded’ (Mccarthy, The Times, 1990, para.

11); ‘… appealed for help in halting the mass migrations of their people,
who are fleeing encroaching deserts’ (Brown, The Guardian, 1998, para.
5); ‘… battle the Pacific to stop’ (Vidal, The Guardian, 2005, para. 1).
Contextualising climate change as an invading force frames the problem
as important; implying it remains a threat to the economy, people and
society at large. Yet what these masks make absent is how climate change
is a threat causing the extinction of all species, including other non-
human species and entire ecosystems.
In addition, closer examination of the use of masks shows they are used
quite often to set the critical gaze’s focus on the problems associated with
climate change, yet there seems to be no attempt to posit specific, defini-
tive, plausible solutions to help these victims of climate change. The dif-
ference is, these masks help augment the consequences of climate change
by evaluating its consequences on humans, at the very same time they
make absent the vision of a sustainable future that goes beyond just cut-
ting carbon emissions, but includes a visionary and pragmatic response
that includes all creatures.
To maintain the sustained interest of the readers, masks are used across
the articles, portraying anthropogenic climate change as a disastrous phe-
nomenon, yet very few specific details make causal links between the
beginnings of ecological crises and climate change as a security risk.
Finally, one notices a choice of words––the mandatory use of nouns
like ‘warfare’, ‘refugees’, ‘farmer’, ‘victims’, ‘poor’, along with verbs like
‘affected’, ‘abandon’, ‘accept’, which are repeated time and again; people
who read the news often tend to link a certain cluster of features with a
group of people. They help forge a recurring pattern in the newspapers of
the passive solicitation of the environmental refugees. This makes less
important the active resistance and agency by the victims themselves. (as
discussed in Sect. 9.5.1.)
9.4.3 Frame 3: Climate Change as a Security Risk
Representations of environmental refugees in these two newspapers reflect

the shift in climate change discourse from sustainable development for all
258 N. Venkataraman
humans to the issue being framed as one that will cause a security risk
(Collins et al., 2012). Environmental change becomes re-conceptualised
as a threat to human security (Barnett, 2001; Dalby, 2002). Climate
change as a security risk thus complements the previous frame of migra-
tion as a problem.
[11] The Pentagon and other military establishments have long recognised
climate change as a ‘threat multiplier’ with the potential to escalate existing
conflicts and create new disputes as food, water, and arable land become
significantly scarce. (Goldenberg, The Guardian, 2011a, para. 9)
Feelings of insecurity are stirred when news reports present ‘security

threat’ and ‘environmental refugees’ within the same article. The ready
availability of problem indicators such as scarce land, overpopulation,
drought, lack of resources, is presented to show an amplified and cumula-
tive effect of the problem, which then operates as a principal lever against
which arguments are made against migration across international
borders.
Appeals like the following were made:
[12] ‘If governments simply respond with traditional attempts to maintain

the status quo and control insecurity, they will ultimately fail,’ Chris Abbot
said. ‘The security consequences of climate change will not manifest them-
selves “over there;” there will be domestic concerns for both developed and
developing nations alike.’ (Smith, The Times, 2008, p. 7)
This frame manifested in 63.3 per cent of the articles on environmen-

tal refugees, thus making it an issue that requires reframing of victims as
potential trouble-makers as they migrate.
9.4.4 Pattern of Absence: The ‘Blame-game’
Both newspapers squarely posit that climate change’s main contributors

are the industrialised nations and/or the West. The issue of the culpability
of Western nations to take responsibility for the present state of affairs
regarding climate change although repeated over the years is done in the
form of lexical masks of generic reprimand to all industrialized nations.

The most significant means of creating focus on the problem and not
drawing attention to the culpability of specific countries is by the generic
use of words like ‘rich countries’, ‘the West’, and ‘us.’
In some examples, we know which countries are being referred to.
[13] The proposal is fiercely resisted by the G8+4 major polluters who seek
limited immigration by skilled workers, rather than taking on environmen-
tal refugees. (Adger, The Guardian, 2007, p. 28)
Yet the frequent generic reference to ‘industrialised nations’ or ‘rich

nations’ makes absent the efforts of some of these countries clubbed as
western or rich that have begun to make a change to their sustainable liv-
ing styles.
In fact, none of the articles in the elite newspapers frame the two-fold
challenge faced by developing and emerging economies. That is, it is crit-
ical for these nations to meet the needs of billions of people who still lack
access to basic energy and services while still participating in the global
transition to clean, low-carbon energy production. This is merely listed as
a subtopic and does not get foregrounded. In addition, these very coun-
tries survive on complex networks of producing goods and providing ser-
vices for the industrialised nations. Instead, the reader’s attention is drawn
to ‘the minutiae and statistics’ of climate change and its effects on a gen-
eral scale (Vidal, The Guardian, 2009, para. 3).
A pattern of implicatures masks the transition from pointing fingers at
specific polluters and contributors to anthropogenic climate change to
strategically arranging a call for collective-action behaviour. Initially,
there were directed implicatures about countries not willing to make the
transition of cutting carbon emissions or even taking climate change
seriously:
[14] Katrina made the Bush Administration take climate change seriously.
(Meachers, The Times, 2006, p. 21)
[15] It will probably need a climate-induced catastrophe on their own turf

for Americans to get with the Kyoto programme. (Brayfield, The Times,
2000, para. 15)
260 N. Venkataraman
The transition to recommending collective global change could have

accompanied the understanding that climate change itself is caused by
complex causal mechanisms that were extensive, it is indeed hard to point
fingers at specific countries to specify scales of one’s culpability in this
global phenomenon.
Instead, climate change offers an interdependent global system not
ontologically specific.
[16] Abdul Maal Abdul Muhith, Bangladesh’s finance Minister called on

Britain and other wealthy countries to accept millions of displaced people.
(Randerson, The Guardian, 2009, para. 1)
[17] It is our fault as much as anyone’s. (Vidal, The Guardian, 2011, p. 35)
[18] What is new is that this time the problem is of human making. (‘The
Guardian view on climate change and social disruption’, The Guardian,
2015, para. 5)
This transition makes absent the structural adjustments made by some

governments to some aspects of the system but nevertheless not complete
transformations. With calls for collective change what is made absent is
that the large polluters have not made a paradigm shift to cleaner ener-
gies, or even provided specific laws to help these victims of change.
The positioning of the pronoun usage of ‘we’ and ‘our’ has two binary
constructs in the newspapers: one of active resistance intended and the
other of active resistance in action. What is interesting though is the
journalist(s) hardly used the word ‘we’ or ‘our’ to suggest any sort of
agency or action on the part of developed countries. This is illustrated in
Table 9.2 which provides an analysis of how the ‘West’ or ‘industrialised
nations’ take responsibility, or are even reprimanded for the present state
of affairs. This provides us with an overview of a pattern of absence eras-
ing responsibility, yet, textually leaving a trace.
As seen from Table 9.2, The Guardian, compared to The Times, pre-
sented a fairer representation of intention and action by both the devel-
oping countries affected and the First World countries’ intent. Changes
are promoted from the outside of the developing nations through tech-
Table 9.2 Sample list of the use of the pronoun ‘we’ in The Times and The
Guardian
The Times The Guardian
[19] … It is tempting to observe that [20] The longer we all argue about
we just need to look around us to minutiae and statistics, the more
see that we are in the grip of rainforest disappears. (Vidal, 2009,
global warming. (Nicholson-Lord, para. 3)
2006, p. 2)
[21] We need rational, realistic [22] However, we should prepare now to
responses to climate-change, not define, accept and accommodate this
knee jerk reactions that create new breed of refugee. (Adam, 2005,
new problems like refugee influx p. 24)
and security. (Nicholson-Lord,
2006, para. 6)
[23] George Bush’s leading climate [24] We need an accounting system to
modeller, Jim Hansen, said a avoid an environmental Enron—new
month ago that we have ‘at most indicators to measure real progress.
ten years’ to make the drastic cuts (Vidal, 2001, p. 28)
in emissions that might head off
climatic catastrophe. (Meachers,
2006, p. 21)
[25] We should worry, but what [26] We know about energy efficiency,
about? (Brayfield, 2000, para. 4) renewable energy, and how to reduce
deforestation, but we seem reluctant
to apply them. (Simms, 2002, p. 4)
[27] ‘We say mass migration is likely [28] These people have a right to their
yet we are going to reduce land. It would also be a loss to the
immigration controls, and then world of a culture and a language. We
not aim to increase net are doing so much to save animals and
immigration,’ he said. (McCarthy, plants from extinction. How can we
1991, para. 4) tolerate the extinction of a nation?
(Brown, 1997, para. 1)
nology and scientific help extended, yet there were few and obvious
examples of self-reprimand.
9.4.5 Pattern of Absence: Migration as a Problem
The trace suggested by the use of presuppositional iteratives makes absent

the earlier and later occurrences suggested by specific words. By using
iteratives frequently, a holistic view of all amenable and non-amenable
262 N. Venkataraman
actions and reactions to the issue of climate change and its effects on
humans are not foregrounded. Iteratives were used every time climate
change’s disastrous effects were spelt out, yet with the use words ‘fre-
quently’, ‘again and again’, ‘redrawn’, and ‘repeatedly’, the readers are left
with suggestive trace of the inconvenience suffered by the victims. In the
following quote an unnamed victim projects understanding of what is
made absent when people are forced to relocate ‘again’.
[29] We lost everything. We are not happy, because we must move again.
Climate change is making thousands of people homeless. (Vidal, The
Guardian, 2013b, para. 4)
Presenting issues in the form of episodic8 frames along with iteratives

leaves only a trace of the fact that these catastrophic climate events have
occurred in the past and been a reason for people to be displaced more
than once. In some regions (for example, the Sub-Saharan region and
Bangladesh), these sporadic events result in internal displacement within
large geographical spaces and they are different when compared to cross-
ing international borders and do not evoke questions of refugee status,
yet these people endure immense suffering.
However, in island nations like the Maldives, Tuvalu, and Kiribati,
international relocation seems inevitable as habitable land is limited and
already overcrowded. In addition, in suggesting that disastrous climate
events are responsible for the movement of human populations neglects
the fact that other effects of climate change may be slow and gradual.
Instead, the repeated frames of climatic disasters and their human dis-
placement suggest that climate change takes place only sporadically and
does not constitute an immediate or daily threat. There were times sea-
level rise was mentioned as a cause of environmental refugees movement,
yet there is a void in terms of discussing global warming effects on coral
reefs. The warmed water not only bleaches the corals, but also makes
brittle the base of these low- lying island nations. Thus, the effect of the
destruction of corals is discussed but a void exists of the cause of the
destruction. This contributes to the absence of the one more important
cause of the sinking islands––the rising water temperatures.
In sum, the above analysis of frames and patterns of absences suggests

that representations of environmental refugees are neither fair nor bal-
anced. Instead, readers are presented with a minimalist representation of
the nature of the problem.
9.5 Discussion
By using the two conceptual tools together––frames and patterns of
absences––one understands that despite intermittent coverage on envi-
ronmental refugees over the years, they remain minimally defined or
described.
9.5.1 M
uted Voices of Resilience, Resounding Echoes
of Fear
In the coverage of environmental refugees the geopolitics of fear does

seem to affect the very nature in which climate change has been projected
in elite newspapers. Newspapers project climate change in episodic events
that are sudden, violent and dramatic. We have burgeoning literature on
how climate change projects urgency, foreboding doom and misgivings
(Chaturvedi & Doyle, 2015); even the few articles that echo the urgency
of the movement of environmental refugees are projected likewise. Yet
the pertinent point to note about the discussion is that with urgency and
fear being the key characteristic of these discourses, the urgency is of the
helping host nations coping with the strain of a new breed of refugees.
In addition, the resilient efforts by people on the ground are muted or
made less important. Frames of the victims as mostly passive people wait-
ing to be helped, ignore the dimension where they actively resist the idea
of a refugee status.
[30] We do not want to leave the Maldives, but we also do not want to be
climate refugees living in tents for decades. (Ramesh, The Guardian, 2008,
p. 1)
264 N. Venkataraman
The presence of this counter-discourse runs contrary to the projection

of victims as passive recipients of economic and technological help from
Western nations. Instead, they counter-frame dependency, ‘once a
general-purpose term for all social relations of subordination’ (Fraser,
1997, p. 142), as they resist being identified as ‘environmental refugees’.
In addition, there is potential for a counter-discourse engaging with the
dominant discourse. The focus on migration as an adaptive strategy
potentially acknowledges the environmental injustices suffered by these
people as this is one way of acknowledging defeat in the discursive battle
against climate change and its future effects (McNamara & Gibson,
2009). The notion of accepting refugees into other countries as they cross
international borders can be framed as something both positive and nega-
tive. If migration is an adaptive strategy, there will be no reason to per-
suade the richer nations to stop further damage to the affected
countries.
The constraints on adaptive capacity—the poverty, the weak gover-
nance, divided political parties, overpopulation—are framed to show
they add to climate stresses already suffered.
[31] To raise cash, his government will sell off state assets, reduce the cabi-
net, and turn the presidential palace into the country’s first university.
(Ramesh, The Guardian, 2008, p. 1)
Yet the attempts by leaders to build and strengthen diversified liveli-

hoods, to find opportunities to design and implement appropriate adap-
tive measures are featured minimally in these two newspapers. When
they are featured, one still gets the perspective of the desperate situation
not resilient efforts.
9.5.2 Lack of Polyphonic Voices and Agency
Elected representatives of countries have a special kind of authorisation

to speak for their people. The responsibility and accountability of the
affected regions towards activating change and doing something about it
are universalised into one voice—that of their leaders. This virtual absence
of polyphonic voices makes absent the different perspectives on the prob-

lem of climate change’s effects on the land and lives. In both newspapers,
it can be concluded that the power to define the problem and to postulate
solutions consists of a top-down approach that mainly lies in the hands
of a few experts, public authorities, and policy-makers, most of whom
stand on the outside of the problem. While expressions of the effects of
climate change come from the public as well as elite claim-makers, expres-
sions of possible solutions always came from the top. For these newspa-
pers, this is a way of constructing certainty as part of journalistic routines
amid uncertainty (Nelkin, 1995; Hornmoen, 2009). However, this strat-
egy also suppresses the voices of the victims of climate change and espe-
cially the changes they might envisage.
Another strategy used by both papers was the use of nominalisations to
focus on qualities and processes. They also play an important function
abstracting qualities and processes from things and time respectively.
‘Vulnerabilities’, ‘negotiations’, ‘adaptation’ focus on the actions rather
than actors. In most cases, nominalisation were used as an effective means
by which to hide the agency of the actions. Nominalisations perform an
important role in driving home the message that one needs to focus on
generalised actions suggested by the nominalised words rather than the
actors or the actions themselves.
9.5.3 A
re We Framing and Erasing the Same
Narrative?
Every news story is created in a social world that the individual journalist
operates in and the organisational world of the news agency. If we see the
news story as a product of a complex interaction between a sense of indi-
vidual and organisational definitions of news worthiness (Schlesinger,
1977), access to news sources and information, and the persuasive influ-
ence of similar stories in not creating a spiral of silence (Noeille-Neumann,
1974), then we have a better understanding of the product as we see it.
Yet there is one concept that can perhaps account for the absence of
certain dimensions of the complex issues. The news item itself reporting
climatologists, environmentalists, migration experts, politicians and the
266 N. Venkataraman
victims who enter the discourse are all positioned in the news reports as
though they are all entering the discourse at the same time. They enter
the discourse at different points in time, with cross-purposes and differ-
ent agendas. Each one of them has a divergent and competing view on
ecological and sustainable living or even disasters. The cross-purposes of
each point of view are absent when these experts are cited together in the
newspapers. One example is detailed below:
[32] Tony Oliver-Smith, a natural hazards expert at the University of

Florida, said: ‘Around the world vulnerability is on the increase, due to the
rapid development of megacities in coastal areas. Combine this trend with
rising sea-levels and the growing number and intensity of storms and it is a
recipe for a disaster, with enormous potential to create waves of
environment-driven migration.’ Dr. Bogardi called for an intergovernmen-
tal panel on environmental degradation to be established to assess the situ-
ation, feed advice to politicians and distinguish genuine environmental
refugees from economic migrants. ‘If you see a group of young men trying
to climb over a fence, it’s very difficult to say those are environmental refu-
gees, because, if you are an environmental refugee you bring your whole
family with you. The whole population would be on the move.’ (Adam,
The Guardian, 2005, p. 24)
In the above example, we can see how The Guardian attempts to proj-
ect a multidimensional perspective on the urgency of the matter, yet there
is actual absence of loss of resources as the cause of this displacement and
silencing of the effect of this displacement—that is it hard to distinguish
an environmental refugee from any other refugee. This lack of distinction
between the classes of refugees makes absent that these refugees do not
have a legal status yet. Thus, in selectively positioning points of views
together, we are creating a cacophony of voices that project voices of
doom and gloom, yet imaginative geographies of climate-induced dis-
placement and migrations are void.
Frames of the environmental refugees are presented in the same tone as
other issues of climate change. This narrows the characteristics of this
issue to a fragmentary, discrete presentation of episodes. In the presenta-
tion of the issues, the overwhelming message repeated time and again
were warnings of warmer worlds, the disastrous effects of migration and
loss of resources. Both the cause and effects of repeated issue-frames are
that they mute essential discussions of the narrative which include the
current protection gap that exists within the international legal system
(Williams, 1990) for these victims of climate change. Sporadically both
newspapers call for help to be rendered to these victims, but neither of
them spell out the gaps in the legal system to distinguish an environmen-
tal refugee from other refugees, or move the discourse for a shift in policy-
making about these victims.
Alternate imagined geographical spaces for the prospective environ-
mental refugees come from the victims themselves.
[33] The Maldives will begin to divert a portion of its billion-dollar annual
tourist revenue into buying a new homeland—as an insurance policy
against climate change that threatens to turn the 300,000 islanders into
environmental refugees, [….]‘We can do nothing to stop climate change
on our own and so we have to buy land elsewhere. It’s an insurance policy
for the worst possible outcome.’ (Ramesh, The Guardian, 2008, p. 1)
There is complete absence of discussion of the imagined geographical

spaces for these victims of climate change in positive terms from the
experts on the outside. There are discussions surrounding technological
help and even monetary help, but substantive help in terms of finding
alternative geographical space is not mentioned. Admittedly, it is hard to
debate alternative imagined geographies in real discursive terms when it
is not something other countries would like or even know how to do. It
is the victims themselves who drive discussions with no help from the
experts on the outside.
[34] Sri Lanka and India were targets because they had similar cultures,
cuisines and climates, he said. Australia was also being considered because
of the amount of unoccupied land available. (Ramesh, The Guardian,
2008, p. 1)
Yet that needs to be brought to the table for discussion as it seems to

be the only alternative for island-nations that are suffering sea-level rise
(individual migration aside).
268 N. Venkataraman
9.6 Conclusion
By understanding how environmental refugees are framed, we also under-
stand how humans are framed in climate change debates both as victims
and as perpetrators. This chapter shows how values and worldviews are
produced, reproduced, and transformed in media discourses while others
are excluded from them (Barnett, 2001; Fairclough, 1995). The value is
that serves it as a reminder that despite being complex and challenging to
report, environmental issues need greater care in framing and addressing
the assumptions that lie in the unsaid. To achieve balanced reporting of
environmental issues, journalists need to construct collective articula-
tions and expressions of meaning to show who and what acts as agents in
a causal narrative, what constitutes the event, and what are the possible
effects of the process on people and places.
Finally, despite best efforts to show how patterns of absence can be
observed against what is framed, one finds there were instances where the
nexus between the frames on environmental refugees and the absence is
not a straightforward analysis of a frame versus an absence. One frame
could include many absences of its subtopics, yet the frame is the starting
point for analysis to understand what is marginalised in terms of compet-
ing values, worldviews and arguments.
Next, the methodological problem of deciding whether each presup-
position and implication was relevant for the discussion on absences
rested on the assumption that the same kind of presupposition formed a
pattern that required discussion only when it was found in over 20 per
cent of the texts. By observing patterns, claims about the role of absence
could be made with a certain amount of confidence. Yet again, it meant
that were some presuppositions and implications that did not warrant
discussion in this chapter as they did not form a pattern, yet seemed
rather important for this issue; thereby making absent some interesting
observations.
The lexical choices of ‘risk’, ‘large numbers’, ‘tidal wave of humans’
accompanying texts that describe the situation play a large role in how
the report is received. Those presuppositions and implicatures that leave
suggestive traces in the text work to hint at a reading that a valid action
plan for these victims is present, while allowing for space for a range of
futures that may be different from that envisioned by the newsmakers.
Masks are effective ways in which the concern shown by people who
speak on behalf of these victims or even those who contest the very exis-
tence of these categories and how they are played up or played down by
the newspaper. These masks play a pivotal role in allowing attention to be
framed on the disastrous effects of climate change in the future but need
not necessarily compel or motivate action today. Finally, voids are not
glaring absences because of the lack of limelight on them. As all aspects
of a complicated issue cannot be covered by detailed explanations in
newspapers due to constraints of newsworthiness and space, what is made
available is the focus on the controversial construct rather than the
specifics of how to solve them or even of how they were created.
Most news items chase a story for its newsworthiness. In matters of
climate change, debate cannot be constructed based on a few constructs
of selective frames. However, news reportage of this issue must accom-
modate all issues, whether closely aligned or departing substantially from
stories about a small subset of actors and their actions. Ultimately, the
final impact of climate change will depend on a combination of human
and environmental factors that are interdependent. As Susan George
once said, ‘There are no ecological problems, only social and political
problems that underline and cause ecological damage’ (George, 1990,
p. 225). These two newspapers do not present climate change and its
effects on humans any differently.
Appendix: List of Subtopics

The following list of subtopics were listed with the following
information:
1 . Writer(s) and complete Reference:

2. Word count:
3. Unique observational angle if any:
4. Unique phrasing (e.g. adjectives, metaphors, descriptions, etc.) if any:
270 N. Venkataraman
5 . Comments on findings about framing:

6. Comments on pattern of absences:
7. Weighted Total:
Extrinsic Pro-migration Pro-adaptive Measure(s)

Effects of climate Negative effects of climate Negative description of
change on change on species, thus effects of climate change
species reason to move. The effects on species locally, thus
were listed in terms of: reason to adapt locally. The
Health effects were listed in terms
Species of:
Ecosystems Health
Humans Species
Ecosystems
Humans
Destruction and thus Destruction (which may be
movement away from slow and gradual) thus
epicentre (irreversible) change in a way of life
Migration as a Recommends action for Resistance against moving
solution movement
Climate change Violent conflict and wars Positive advantages of
as a security between states staying back
risk Competition for scarce Competition for scarce
resources in host countries resources in native country,
thus adaptation
recommended
Empathy for host countries Empathy for people moving
Protecting Negative portrayal of costs Positive attributes of saving
national: shores (international) funds
Influx, a strain on the system Benefits of helping from a
in terms of competition for distance (inclusive of funds)
jobs
Intrinsic Pro-migration Pro-adaptive Measure
Risk Negative description of Positive outcomes of non-risk
(Generalities) accepting behaviour locally
refugees/migrants
Uncertain future of these Better future adapting to
people(migrants/refugees) home ground changes
Spread of diseases, impact on Glocal-adaptive strategies to
socio-economic fabric of improve the socio-economic
the nation fabric of the nation
National interest Climate action in favour of Climate action not
individual nations recommended
Notes
1. Refer to Sect. 9.2.2 for a detailed discussion on distinguishing the terms
traces, masks and voids.
2. Not all ‘natural disasters’ can be blamed on climate change, e.g. the after-
math of volcanic eruptions, earthquakes, and tsunamis, which result from
massive geological shifts within the earth are not from human activity at
the surface. Exceptions exist like that of fracking that cause earthquakes.
That said, it is possible to argue that high numbers of casualties in earth-
quakes in Nepal or Haiti are due to inferior construction, which may be
related to poverty, clearly a dimension of human activity, though still
unrelated to climate change since poverty and shoddy construction in
these places have a long history in these places.
3. The terms ‘environmental refugee’ and ‘climate refugee’ are often used inter-
changeably. In this research, the term ‘climate refugee’ refers to a specific
subset of refugees who are victims of human-induced climate change
whereas ‘environmental refugee’ describes populations that move due to dra-
matic and violent climatic events that may or may not be humanly induced.
4. Relevant examples follow in Sect. 9.3.1.
5. The choice of using the search term ‘migrant’ is a conscious one, as there is
academic literature that suggests that the use of the term migrant is better
suited as it is a choice people make to move across international borders.
6. The subtopics are listed in Appendix.
7. Please refer to Appendix for a list of subtopics.
8. Not detailed in this chapter but an equally important phenomenon con-
tributing to masking climate change and its consequences on species is
presenting the issue in the form of episodic frames. The issues are pre-
sented in the form of concrete instances or specific events instead of the-
matic frames that focus broader social trends at a more abstract level and
discuss general outcomes (Iyengar, 1990).
References
Barnett, J. (2001). The meaning of environmental security: Ecological politics and
policy in the new security era. London: Zed Books.
Biermann, F., & Boas, I. (2010). Preparing for a warmer world: Toward a global
governance system to protect climate refugees. Global Environmental Politics,
10(1), 60–88.
272 N. Venkataraman
Black, R., Kniveton, D., & Schmidt-Verkerk, K. (2013). Migration and climate
change: Toward an integrated assessment of sensitivity. In T. Faist & J. Schade
(Eds.), Disentangling migration and climate change: Methodologies, political
discourses, and human rights (pp. 29–53). Berlin: Springer.
Brown, L. R. (2008). Troubling new flows of environmental refugees. Retrieved
June 14, 2016, from http://www.earth-policy.org/plan_b_updates/2004/
update33
Carvalho, A. (2008). Media(ted) discourse and society: Rethinking the frame-
work of critical discourse analysis. Journalism Studies, 9(2), 161–177.
Chaturvedi, S., & Doyle, T. (2015). Climate terror: A critical geopolitics of climate
change. Hampshire: Palgrave Macmillan.
Christian Aid. (2007). Human tide: The real migration crisis. London: May.
Collins, M., Chandler, R. R., Cox, P. M., Huthnance, J. M., Rougier, J., &
Stephenson, D. B. (2012). Quantifying future climate change. Nature
Climate Change, 2, 403–409.
Dalby, S. (2002). Environmental security. Minneapolis: University of Minnesota.
de Vreese, C. (2005). News framing: Theory and typology. Information Design
Journal + Document Design, 13(1), 51–62.
El-Hinnawi, E. (1985). Environmental refugees. Nairobi: United Nations High
Commission for Refugees.
Entman, R. M. (1993). Framing: Toward clarification of a fractured paradigm.
Entman, R. M. (2004). Projections of power: Framing news, public opinion, and
US foreign policy. Chicago, IL: University of Chicago Press.
Fairclough, N. (1995). Critical discourse analysis: The critical study of language.
New York: Longman.
Fairclough, N. (2003). Analysing discourse: Textual analysis for social research.
London and New York: Routledge.
Fraser, N. (1997). Justice interruptus: Critical reflections on the ‘postsocialist’ condi-
tion. New York: Routledge.
Gamson, W. A. (1992). Taking politics. New York: Cambridge University Press.
George, S. (1990). Ill fares the land. Harmondsworth: Penguin.
Goffman, E. (1974). Frame analysis: An essay on the organization of experience.
New York: Harper & Row.
Halliday, M. A. K., & Martin, J. R. (1993). Writing science: Literacy and discur-
sive power. London: Falmer Press.
Hartman, T. K., & Weber, C. R. (2009). Who said what? The effect of source
cues in issue- frames. Political Behaviour, 31(4), 537–558.
Hornmoen, H. (2009). ‘What researchers now can tell us’ – Representing scien-
tific uncertainty in journalism. Research and Knowledge in Communication,
3(4), 1–6.
& Society, 13(3), 347–372.
Jaleel, A. (2013). The status of the coral reefs and the management approaches:
The case of the Maldives. Ocean & Coastal Management, 82, 104–118.
https://doi.org/10.1016/j.ocecoaman.2013.05.009
Jessop, B. (2000). The crisis of the national spatio-temporal fix and the tangen-
tial ecological dominance of globalizing capitalism. International Journal of
Urban and Regional Research, 24(2), 323–360.
Kurzon, D. (2007). Toward a typology of silences. Journal of Pragmatics, 39(10),
1673–1688.
McAdam, J. (2012). Climate change, forced migration, and international law.
McNamara, K. E., & Gibson, C. (2009). We do not want to leave our land:’
Pacific ambassadors at the United Nations resist the category of ‘climate refu-
gees. Geoforum, 40, 475–483.
Morrissey, J. (2009). Environmental change and forced migration; A state of the art
review. Oxford: Refugee Studies Centre, University of Oxford.
Myers, N. (2002). Environmental refugees: A growing phenomenon of the 21st
century. Philosophical Transactions of the Royal Society B: Biological Sciences,
357(1420), 609–613.
Noeille-Neumann, E. (1974). The spiral of silence. A theory of public opinion.
Journal of Communication, 24, 43–51.
Page, E., & Redclift, M. R. (2002). Human security and the environment:
International comparisons. Northampton, MA: Edward Elgar.
Piguet, É., Pécoud, A., & de Guchteneire, P. (2011). Migration and climate
change. Paris: Cambridge University Press/UNESCO.
Schleppegrell, M. J. (2001). Linguistic features of the language of school.
Linguistics and Education, 12(4), 431–459.
Stal, M. (2009). Flooding and relocation in central Mozambique: Case study.
Report for the EACH-For Project. World Populations Prospects, the 2004
Revision. New York: United Nations.
Stibbe, A. (2015). Ecolinguistics: Language ecology and the stories we live by.
London: Routledge.
UNFCCC. (2007). Information on adaptation approaches, strategies, practices and
technologies at the regional, national and local levels in different sectors, as well as
on experiences, needs and concerns. Bali: UNFCCC.
274 N. Venkataraman
UNHCR. (2011). The 1951 convention relating to the status of refugees and its
1967 protocol. Geneva: Switzerland.
van Dijk, T. A. (1991). Media contents: The interdisciplinary study of news as
discourse. In K. B. Jensen & N. W. Jankowski (Eds.), A handbook of qualita-
tive methodologies for mass communication (pp. 108–120). New York:
Routledge.
van Dijk, T. A. (1998). Ideology: A multidisciplinary approach. London: Sage.
van Dijk, T. A. (2009). Society in discourse. How context controls text and talk.
Newspaper Sources
The Times
Anonymous. (1995, August 10). The world could soon have more than 200
million “environmental refugees.” The Times.
Anonymous. (2010, November 29). £173bn a year ‘will avert huge migration of
climate refugees’. The Times.
Anonymous. (2013a, September 28) Global lukewarming need not be cata-
strophic. The Times.
Anonymous. (2013b, October 18). Climate change refugee battle. The Times.
Aslet, C. (2007, January 6). Farmer Giles and the elephant grass. The Times.
Binyon, M. (1996, May 29). Red cross predicts ten-year doubling of refugee
total. The Times.
Brayfield, C. (2000, November 13). Too little, too late. The Times.
Buckoke, A. J. (2001, April 14). Clear thinking on climate change. The Times.
Cato, A. (2009, December 9). A regular musician’s guide to climate change. The
Times.
Kenneth, D. (2008, May 13). After the storm; Now the survivors face an epic
struggle to stay alive. The Times.
Lagan, B. (2014, January 18). Bats drop dead out of trees as Australia buckles in
the heat. The Times.
Lewis, S. (2007, December 3). Climate change means tropics are much bigger
than maps say. The Times.
McCarthy, M. (1990, May 26). Thatcher sets tough global warming target. The
Times.
McCarthy, M. (1991, September 21). Green demands the abolition of all immi-
gration curbs/SCT. The Times.
McConnell, T. (2009, October 22). 23 million are starving in the drought; But
rain will bring only more disaster. The Times.
McConnell, T. (2011, August 3). Millions could starve unless $1.4bn aid sent
now, warns UN. The Times.
Meachers, M. (2006, February 10). Ten years to prevent catastrophe. The Times.
Nicholson-Lord, D. (2006, December 15). How the world woke up to global
warming. The Times.
Nuttall, N. (1997, November 29). Islanders ready for Pacific evacuation. The
Times.
Nuttall, N. (1998, August 12). Earth blazes into the next millennium. The
Times.
Page, J. (2007a, August 6). Disease fear over millions driven from homes by
monsoon flooding. The Times.
Page, J. (2007b, February 2). Climate of fear in sinking country. The Times.
Page, J. (2008, November 3). Forest of doom: Man and tiger in deadly struggle
for survival amid floods and shrinking territory. The Times.
Page, J. (2009, December 5). The glacier is dying: Scientist climb mountains to
reach an unwelcome truth. The Times.
Ryder, R. S. (1991, May 2). Need for world fund on afflictions. The Times.
Simons, P. (2007, May 3). Cold climate’s devastating role in the Act of Union.
The Times.
Smith, L. (2008, January 24). Climate change will lead to warfare over food and
water. The Times.
Smith, L. (2009a, December 3). Climate change means tropics are much bigger
than the maps say. The Times.
Smith, L. (2009b, March 14). Global warming “can be controlled.” The Times.
Sylvester, R., & Thomson, A. (2011, March 26). Why Britain needs a new way
of responding to a disastrous world. The Times.
Thomas, C. (1990, November 20). Maldives fights to keep its head above rising
waters. The Times.
Valley, P. (1985, July 17). Famine errors that have killed thousands. The Times.
Valley, P. (1987, March 11). Refugees’ hungry but safe from war. The Times.
Webster, B. (2010, November 29). 173b a year will avert huge migration of
climate refugees. The Times.
Webster, B. (2013, September 25). Climate change forecasters “failed to see
human costs.” The Times.
276 N. Venkataraman
Wright, P. (1989a, February 15). Joint 2 million pound venture to save wildlife
and protect forest. The Times.
Wright, P. (1989b, June 6). Greenhouse effect will bring refugees. The Times.
The Guardian
Adam, D. (2005, October 12). 50m environmental refugees by end of decade,
UN warns. The Guardian.
Anonymous. (1990, November 30). Famine in the coldest climate. The
Guardian.
Anonymous. (1999, June 25). Natural disasters produce majority of refugees.
The Guardian.
Anonymous. (2006, October 31). Climate change: Stern warning. The Guardian.
Anonymous. (2008, June 18). Climate change fuels refugee crisis. The Guardian.
Anonymous. (2009a, May 18). We have taken every measure we can think of to
stop the desert moving closer and submerging our crops and villages. The
Guardian.
Anonymous. (2009b, May 19). The eco-refugee: Huang Cuikun: “To survive we
must control the desert”. The Guardian.
Anonymous. (2009c, May 27). Prince Charles: Delay on rainforest will have
catastrophic consequences. The Guardian.
Anonymous. (2009d, June 19). Climate change: What it means for US. The
Guardian.
Anonymous. (2013c, May 14). America’s climate refugees: Changing picture.
The Guardian.
Anonymous. (2013d, October 14). G2: The climate change survivor’s guide.
The Guardian.
Anonymous. (2013e, December 5). Can the Carteret atoll refugees of climate
change bank on chocolate? The Guardian.
Anonymous. (2015, March 8). Climate change and social disruption. The
Guardian.
Ardill, J. (1989, June 6). Climate changes “threaten refugees crisis.” The
Guardian.
Balch, O. (2015, February 25). The Pacific Islands: Tomorrow’s climate refugees
struggle to access water today. The Guardian.
Black, I. (2001, April 3). UK “most racist” in Europe on refugees: Commission
blames media and government. The Guardian.
Borger, J. (2008, June 17). Conflicts fueled by climate change causing new refu-
gee crisis, warns UN. The Guardian.
Brown, P. (1990, May 22). UN warns of global time bomb: Greenhouse effect
set to redraw world map in 60 years. The Guardian.
Brown, P. (1999, June 24). Environment more than warfare overtaken as refugee
cause. The Guardian.
Brown, P. (2000a, December 20). Sahara jumps Mediterranean into Europe:
Climate change threatens to create dust belt around globe. The Guardian.
Brown, P. (2000b, May 15). West’s pollutions blamed for world disasters. The
Guardian.
Brown, P. (2003, September 30). Refugee warning to global polluters. The
Guardian.
Brown, P. (2005, February 3). Climate conference hears degree of danger. The
Guardian.
Carroll, R. (2003, December 18). How rice farming and refugees brought
malaria to vulnerable corner of Burundi. The Guardian.
Goldenberg, S. (2011a, February 22). Ban Ki-moon takes climate change effects
to Hollywood. The Guardian.
Goldenberg, S. (2011b, July 20). UN Security Council to consider climate
change peacekeeping. The Guardian.
Goldenberg, S. (2013, May 13). America’s first climate refugees. The Guardian.
Guest, I. (1989, October 4). Flood threat to Pacific island. The Guardian.
Harvey, F. (2011, October 10). Climate change could trap hundreds of millions
in disaster areas. The Guardian.
Jowit, J. (2012, January 26). Flooding to cost UK billions as climate changes.
The Guardian.
Lawrence, F. (2011, July 5). Famine factors: Extreme weather and conflict lead
to crisis. The Guardian.
Leahy, S. (2013a, November 13). US fear climate talks will focus on compensa-
tion for extreme weather. The Guardian.
Leahy, S. (2013b, November 14). US fears focus on “blame and liability” in
climate talks. The Guardian.
Milman, O. (2013, July 30). Marshall Islands calls for leadership from Australia
on climate change. The Guardian.
Myers, N. (1995, June 27). The first wave of human flood. The Guardian.
Pendleton, A. (2007, June 21). China not to blame for climate change. The
Guardian.
Ramesh, R. (2008, November 10). Paradise almost lost: Maldives seek to buy a
new homeland: Maldives seeks new home as sea level rise. The Guardian.
278 N. Venkataraman
Randerson, J. (2009, December 4). UK should open borders to climate refu-

gees, says Bangladeshi minister. The Guardian.
Sample, I. (2006, October 4). Extreme droughts will spread; warn forecasters.
The Guardian.
Sherwood, H. (2014, June 20). More than 51m people forced to leave homes.
The Guardian.
Simms, A. (2002, August 27). Earth Summit 2002: Comment: Now is in the
time for rich to match poor’s generosity. The Guardian.
Smith, A. D. (2013, December 3). Chad struggles to prevent disease in midst of
humanitarian emergency. The Guardian.
Smith, D. (2009c, July 10). Volatile river levels bring devastating floods to
Zambia. The Guardian.
Smith, D. (2009d, June 6). Climate change: Zambia: Routine floods are coming
in too soon, too strong. The Guardian.
Tan, M. (2012, October 16). Forced migration in the 21st century: Urbanized
and unending. The Guardian.
Taylor, L. (2013, June 30). Kevin Rudd attempts to plug political weak spots
with change. The Guardian.
Vaughan, A. (2010, October 28). Future London: The capital after climate
change. The Guardian.
Vidal, J. (1996, February 16). Global warming disaster “on way:” Scientist
warns of risks from rising sea levels. The Guardian.
Vidal, J. (1997, September 27). South East Asian smog crisis: Poison fog blanket
threatens world climate. The Guardian.
Vidal, J. (2001, June 8). Causes for concern: Social and environmental cam-
paign groups at risk. The Guardian.
Vidal, J. (2005, November 25). New proof of global warming: Pacific Atlantis:
First climate change refugees. The Guardian.
Vidal, J. (2007, March 2). Build inland, UN climate report warns. The Guardian.
Vidal, J. (2009a, December 4). Migration is the only escape from rising sea
levels in Bangladesh. The Guardian.
Vidal, J. (2009b, May 28). Charles in rainforest plea to Nobel laureates. The
Guardian.
Vidal, J. (2011a, February 4). Climate change not expected to lead to mass
cross-border migration. The Guardian.
Vidal, J. (2011b, July 22). To pin the Somalia crisis on drought is wrong: This is
an entirely predictable, man-made calamity. The Guardian.
Vidal, J. (2013a, January 29). Sea change: The Bay of Bengal’s vanishing. The
Guardian.
Vidal, J. (2013b, January 30). Encroaching sea turns islanders into climate refu-
gees. The Guardian.
Williams, N. (1990, October 11). Surge tide threat to Thames barrier. The
Guardian.
10
A Discourse Analysis of Absence
in Nigerian News Media
Taiwo Oluwaseun Ehineni
10.1 Introduction
This chapter attempts to describe empirically the manifestation of absence
in the news media discourses context. Focusing specifically on Nigerian
news, strategies of achieving absence in the media and the linguistic fea-
tures that produce absence are discussed. The chapter argues that absence
can be reflected through the way certain topics or social actors are por-
trayed in the news media and this may include strategies such as topical-
ization, focus on detail, and omission. Furthermore, absence in the news
media may arise from the interplay of a range of devices. These strategies
work in synchrony to accomplish absence. Following this view, discus-
sion on the manifestation of absence in this chapter is organized, with
reference to language use in the news media (Sect. 10.1.1) and the issue
of absence (Sect. 10.1.2). Aside from providing some background to the
study, these sections show that the use of language in the media may be
ideologically driven and this process of re-presenting things in the media
T.O. Ehineni (*)

Indiana University, Bloomington, IN, USA

282 T.O. Ehineni
can provide insights into the issue of absence in the new media. In Sect.
10.2, I discuss the methodological framework for the chapter which
relates to how critical discourse analysis (CDA) may be used to explore
features of absence. Using CDA, I uncover the social orientations and
perceptions that underlie absence in the news. In other words, absence is
meaningfully informed by the social context. A brief overview of the cen-
tral news focus of the study is given in Sect. 10.3, which chronicles the
abduction of the schoolgirls in Northern Nigeria. This section creates a
background for the data that is analyzed in the study. The analysis of the
news data are given in Sect. 10.4, where I discuss how discourse strategies
of topicalization and omission are used to achieve absence, and also how
linguistic features such as rhetorical questions, ambiguous acronyms,
silent pronouns, and unclear numbers may contribute to absence in the
news media context. In essence, I approach absence in the news data in
this study through the analysis of representations of social actors and
events in the news reports. More specifically, I look at how events are
portrayed in terms of deliberate use of certain word choices and lexical
forms in a way that some other important information are not reflected.
Also, I explore how detail may be omitted as a way of foregrounding, to
background some other information in the news media context.
10.1.1 L anguage and Re-presentation in the News

Media
The media often portray events, ideas, topics or situations from certain
ideological or value perspectives. More often than not, the media strive to
‘re-present’ a reality (events) to a public audience. Significantly, these re-
presentations shape public opinion on events in a reported context as van
Dijk noted that ‘the media primarily have the potential to control to
some extent the minds of readers or viewers’ (van Dijk, 1996, p. 10).
Since language is a way of generating meaning characterized by the con-
stant making of choices, in both aspects of production and interpretation
(Verschueren, 2008), it is deliberately employed by journalists and news
experts to convey ideas to the public. In other words, journalists’
re-presentations of happenings and events are always made from a p articular
A Discourse Analysis of Absence in Nigerian News Media 283
ideological position (Richardson, 2007; Taiwo, 2007) and so, news is

often a nuanced not a neutral representation of facts. What is said and
how it is said in a way reflect the reporter’s orientation of the events.
Thus, news re-presentations are orientations of reality. In fact, as we shall
see in the news data, the same events are relayed in the newspapers differ-
ently. Hence, while the reality may be similar, the re-presentations vary.
As Reah (1998, p. 50) puts it: ‘Newspapers are not simply vehicles for
delivering information. They present the reader with aspects of the news,
and present it often in a way that intends to guide the ideological stance
of the reader.’ That is, the way language is used in texts can carry ideo-
logical meaning (Fairclough, 1995, p. 25). Similarly, van Dijk (1995)
reflects on the fact that lexical choices are not only evaluative judgements
but often project the ideological position of the user. The choice of
words, therefore, may evoke certain feelings in the audience which would
eventually influence their views. Furthermore, Matu and Lubbe (2007,
p. 402) claim that ‘linguistic choices play a fundamental role in the
propagation and perpetuation of implicit and dominant ideologies, and
that there are certain ideological differences that are conveyed either tac-
itly or overtly in newspaper reporting’. It is beyond the focus of this
chapter to elaborate on the complex concept of ideology, but a minimal
clarification is in order. Ideology may be conceived of as ‘any constella-
tion of fundamental or commonsensical, and often normative, beliefs
and ideas related to some aspect of “reality” (Verschueren, 1999, p. 238).
It basically relates to views, perspectives and opinions on issues. However,
it must be noted that these views may be both explicit and implicit in
newspapers. Thus, absence may be implicit or explicit in its manifesta-
tion in the new media.
10.1.2 The Concept of Absence in the News Media
First, absence is viewed as a general notion that includes all forms of iden-
tifiable and meaningful omissions in discourse. To discuss absence is an
attempt to be concerned with occurrences that carry meaning, the way
they are structured and how they are interpreted. Consequently, in the
news media context, absence deals both with situations when information
284 T.O. Ehineni
is omitted, and with cases when whole events are suppressed or when
certain interpretations of events are suppressed as long as they are all sig-
nificant to the discourse context. That is, it includes both the ‘unjustifi-
able omissions of significant facts’ as well as ‘under-emphasis of certain
aspects of an event’ which typify the media bias in news reporting
(McQuail, 1992, p. 192). Hence, absence in the news reports may not
necessarily be reporting mistakes or editing errors, they are meaningful,
significant instantiations in the news discourse. Though news reports
portray actual events or happenings, they may also be constructions or
framings of these events given by the journalist. Schudson explains:
To say that journalists construct the world is not to say they conjure the
world. Journalists normally work with materials that real people and real
events provide. But by selecting, highlighting, framing, shading, and shap-
ing in reportage, they create an impression that real people—readers and
viewers—then take to be real and to which they respond in their lives.
(2003, p. 2)
In Schudson’s view, journalists influence news reports by making

deliberate choices in how they convey information to the public. The
notion of ‘shading’ in Schudson’s term seems to approximate the idea of
absence where some detail may be deliberately made unclear to the reader.
Absence may manifest itself in a number of ways in the news media.
Some of these details could include not revealing the identity of victims
of a crime. In a 1990 survey, for example, 10 per cent of newspaper edi-
tors thought rape victims’ names should never be printed, 40 per cent
thought names should be printed only with victims’ permission, and 44
per cent believed names should only be printed in exceptional cases
(Winch, 1991). In some other cases, it is a policy observed by the media
to withhold the names of sexual assault victims (Thayer & Pasternack,
1994). It is also implicit in the laws and practices protecting juvenile
offenders from media exposure (Davis, 2000). It could be to mitigate the
effects of a horrible event, if described in detail, to the readers.
Accordingly, it has been observed that the presence or absence of news
facts reflects what is considered important or given prominence by the
journalists. This is also a way of encouraging people to think in certain
directions on social and public issues (Price & Tewksbury, 1997). In

many cases in the media, as this chapter will further show, absence of
detail may occur through an attempt to foreground or background
aspects of an event or occurrence in the society based on sociocultural,
psychological, political and other ideological reasons. Absence of details
may also be a strategy of self-protection or defence in reporting. It could
give the reporter some ‘reporting immunity’ such that he/she is not held
accountable by anyone involved in the event reported. For instance, eye-
witness’s names in crime events may be backgrounded in the Nigerian
media news to avoid any future attacks on them.
10.1.3 Framing and Absence in News
The concept of absence/silence in the news media may also relate to the
process of framing. Framing may be connected to the realization of absence
since while all facts of an event or incident are covered, they are not all pre-
sented to the reader. The news content that is relayed to the public through
the media depends on the journalist’s choices. Therefore, framing is an
attempt to highlight certain aspects of issues while obscuring and ignoring
others, which may in turn lead audiences to different conclusions (Entman,
1993). However, it has been noted that ‘framing differs from persuasion in
that it does not necessarily alter belief content but rather alters the perspec-
tive of what aspects of an issue are important’ (Druckman, 2001, p. 1044).
In terms of function, Druckman further explains that framing can be used
to shift topics, for instance, shifting the feasibility of a new housing devel-
opment from environmental concerns to economic concerns (2001,
p. 1044). This idea of framing in news media may involve the placement of
stories, the length of stories, the headlines, images, and overall tone (Kendall,
2005) or use of ‘condensing symbols’ such as metaphors and taglines along
with ‘reasoning devices’ such as moral claims (Gamson, 1996, p. 125).
Similarly, headlines may be creatively designed or framed to promote domi-
nant views, public interests and hegemonic orientations. This, for instance,
is exemplified by portraying reckless taxi drivers as constant catalysts of road
accidents, while safety officers are typified as protectors of the citizens
(MacRitichie and Seedat, 2008). In achieving absence or silence, framing
286 T.O. Ehineni
may be used to make some information ‘more noticeable, meaningful and

memorable’, making the readers more likely to pay attention to it (Entman,
1993, p. 53), but at the same time, backgrounding some other details from
the public. Furthermore, Huckin (2002) reinforces the idea that a topic
could be framed or knowledge about an issue constructed through certain
manipulative devices. This is often an attempt to draw attention to an issue
while allowing for the omission of other important facts.
10.2 Methodology
10.2.1 Data Description
Data for this study focused on news headlines and reports on the 200
abducted girls in Northern Nigeria in April 2014. I focused on this news
story because it was a major incident in Nigeria that drew much national
and international concern from 2014 to 2016 (see Sect. 10.3). Also, the
choice of headlines was motivated by the fact that news headlines give the
gist of news events. They are also designed to be captivating and memo-
rable so as to draw people’s attention to the news. Furthermore, in the
Nigerian context, people often read the headlines more than the entire
news stories, based on my personal observation while collecting data at
newspaper stations in Southwest Nigeria. People pass by these newspaper
stations to glance at the headlines to have a sense of what’s happening in
society. Moreover, headlines are often glimpsed on public transport, or
displayed on fliers, which makes their impact wider. Hence, based on the
importance of the headlines in news stories, I explore which strategies are
used in portraying the news about the girls’ abduction in the news head-
lines and reports, and how this relates to the issue of absence.
The news headlines and reports were collected from national newspa-
pers at newspaper stations in Southwest Nigeria (especially Ibadan and
Lagos) in 2016. These Nigerian national newspapers include Vanguard,
Punch, The Sun, and This Day, which are popular and widely read public
newspapers that express a plethora of political opinions, and also comment
extensively on news about the Chibok girls’ abduction. The headlines
were analyzed in terms of how certain information was made ‘absent’
within the reported news, and linguistic strategies that facilitate these
forms of absence are discussed. The collection of the news articles was also
determined in terms of whether they discussed the Chibok girls’ incident
or not. Thus, the main search keyword was ‘Chibok girls’. As discussed in
Sect. 10.2.2, I use the framework of Critical Discourse Analysis to discuss
the instantiations of silence/absence in the Nigerian news reports.
10.2.2 A
nalyzing News Media: Critical Discourse
Analysis (CDA)
My analysis of the news data is informed by the social constructionist

perspective of Critical Discourse Analysis (CDA) (Fairclough, 1992,
1995), where discourse is categorized as representations, reflections and
constructions of society by ascribing meanings to social identities, rela-
tions and practices. Language is therefore considered to be a construction
of the social world as well as other social practices (Phillips, 2006). CDA
is thus mainly used to explore how discourses constitute social identities
and social relations, and essentially how these relations are enacted, repro-
duced and challenged by discourse (Richardson, 2007). This implies that
discourse should not be reduced to language alone. Hence, social con-
structionist approaches require that discourse should be empirically ana-
lyzed within its social context (Jørgensen & Phillips, 2002). Critical
discourse analysts typically stress ‘patterns of domination whereby one
social group is dominated by another’ (Phillips, 2006, p. 288).
Furthermore, CDA does not focus on analyzing texts or processes of
production and interpretation, but on analyzing the relationship between
texts, processes and their social conditions, both the immediate condi-
tions of the situational context and the more remote conditions of insti-
tutional and social structures (Fairclough, 1989). Fairclough divides
discourse into three dimensions: texts, interactions and contexts. Applying
this conceptual framework to the news media discourse, texts would refer
to the news reports of events in the newspapers, interactions deal with the
process of production and consumption of the news while contexts
involve the situational and social background in which the events have
been reported.
288 T.O. Ehineni
The concept of absence or silence is a very important aspect in analyzing

media discourse since it also reflects social orientations and perceptions in
the news. In this view, Richardson (2007, p. 93) underscores the necessity
to ‘recognize that textual or journalistic meaning is communicated as
much by absence as by presence; as much by what is “missing” or excluded
as by what is remembered and present’. Richardson’s view therefore exam-
ines news as the product of the way in which events are constructed by
journalists rather than being inherent characteristics of the events being
reported. This may be achieved by the use of extracted topics and the pack-
age approach, in which the news focus is not on the main event, which is
probably already known to the audience, but rather on other non-central
elements, such as the consequences of that event and the reaction of the
participants in it. This may also relate to the strategies of topicalization
which I explore in this study.
However, the issue of absence in news media is ambiguous since it is
open to various interpretations. The absence of a thing may not always
imply very specifically what is unsaid. Thus, the context of situation plays
a crucial role in interpreting what is implied in absence, just as it has been
noted that discourses are influenced by context (van Dijk, 2008). This is
relevant, since, in media discourse, news as ‘texts’ in Fairclough’s (1989)
terms often relates to different contextual categories—social, political,
geographical, or economic events. These different aspects of contexts pro-
vide the grounds for our interpretation of absence.
10.3 Background on the Abduction

of the 200 Girls in Northern Nigeria
The abduction of the schoolgirls was a terrorist event carried out by the
Islamist group in northern Nigeria called Boko Haram. The Boko Haram
sect (which has been linked to Al-Qaeda) is motivated by the desire to set
up an Islamic caliphate in Nigeria and is particularly opposed to
western-style modern education, which they claim drives people away
from following Islamic teaching as a way of life. By 2014, tens of thou-
sands of people had been killed in attacks perpetrated by the group, and
the Nigerian federal government declared a state of emergency in May
2013 in Borno, Yobe and Adamawa states in its fight against the insur-
gency. This consequently led to the massacre of hundreds of Boko Haram
members, with the remainder retreating to mountainous areas from
where they began increasingly to target civilians. Several attempts by
government-organized campaigns failed to stabilize the country. Boko
Haram began to target schools in 2010, killing hundreds of students by
2014. Boko Haram’s attacks intensified in 2014. In February, the group
killed more than 100 Christian men in the villages of Doron Baga and
Izghe.
On the night of 14–15 April 2014, a group of militants attacked the
state Girls Secondary School in Chibok, Borno State, Nigeria. They
broke into the school, pretending to be guards and told the girls to get
out and come with them. During this process, a large number of students
were taken away in trucks, possibly into the Konduga area of the Sambisa
Forest where Boko Haram was known to have fortified camps. They were
estimated by news sources to be about 278 girls. It was later noted that
Nigerian military force confirmed that they had had advance notice of
the attack but that their over-extended forces were unable to mobilize
reinforcements. The incident generated lots of campaigns and activities
both locally and internationally. These were also covered by a number of
national newspapers to convey the details of the incident and reactive
campaigns to the public. Recently, there are new updates in the Nigerian
newspapers that the Chibok girls have been released. This study will focus
on the analysis of news about the supposed recent release.
10.4 Analysis
My analysis will focus on how absence is realized in news headlines and
reports on the abduction of the girls, which was one of the most reported
events in the Nigerian national media. I will focus on strategies and lin-
guistic features of absence in these news reports. I do not compare the
newspapers in terms of the degree of absence but only illustrate how
absence was achieved in the news reports from the newspapers. However,
over 90 per cent of the 50 news articles collected in 2016 reflect the strat-
egies and features of absence identified in this study.
290 T.O. Ehineni
10.4.1 Discourse Strategies of Achieving Absence
This study identifies topicalization and omission as major strategies of

achieving absence in the Nigerian news. These features were identified by
scrutinizing how the news reports were presented in the news articles. My
interaction with people while I was in Nigeria showed that there are more
details the news articles do not report. Some Nigerians who live or have
families in the region where the event took place informed me of other
facts of the incident that were missing in the news articles.
10.4.1.1 Topicalization
This is a process of highlighting details in news reporting. This is often

used to give prominence to certain aspects of the story where the reader’s
focus is shifted to what is emphasized only.
(1) Breaking News: 21 kidnapped Chibok girls released—Reports.

There are reports that twenty-one of out of the over 200 Chibok school
girls kidnapped by Islamist Boko Haram insurgents in April 2014 have
been released. A government official said the ‘insurgents group released the
21 Chibok girls to the Nigerian government’. Other sources revealed that
the Chibok girls were rescued in Banki area of Borno state where Boko
Haram militants had left them on Thursday by military helicopter.
(Vanguard, 13 October 2016)
The news report here focuses specifically on the idea that the girls have
been released and that the ones released were 21 girls. The word release is
more emphasized since Nigerians have been waiting for the release of the
girls after many campaigns and concerted steps have been taken by the
government on the issue. However, the news only focused on the 21 girls
who were released. By mentioning only 21 Chibok girls, the media seems
to imply that only these girls were kidnapped since no reference is made
to the remaining missing girls at all in the news. The specific focus on ‘21’
and ‘release’ only does not allow the reader to grasp the remaining details
about the abducted girls.
(2) 21 Chibok Girls Abducted by Boko Haram Found

No fewer than 21 girls of Government Secondary School, Chibok,
Borno State, who were abducted in the dead of the night two years ago by
the Boko Haram sect have reportedly been found. (This Day, 13 October
2016)
Similar to the previous news report, there is a specific focus on the 21

girls released. No information was given about the remaining girls who
are actually more in numbers—over 200 girls. The news report seems to
foreground the idea of release as more prominent information. Since
their release has been long awaited by Nigerians, this news report, just
like the former also, emphasizes the fact that they have been released as a
way of giving consolation to the public. Thus, providing details about the
fact that many more girls have not been released may make the public feel
sad, so focus is given to the release aspect only.
(3) Chibok Girls: From abduction to release of 21

Nigeria confirmed Thursday that 21 girls had been released following a
prisoner exchange deal struck with Boko Haram. (Punch, 13 October
2016)
(4) CHIBOK GIRLS REUNITED 21 kidnapped girls released by Boko

Haram in Nigeria reunited with their families (The Sun, 17 October
2016)
Looking at the information structure of the headlines, it is clear that

the main objective of the news report was to promote awareness that the
girls have moved from ‘abduction’ to ‘release’. The news report first intro-
duced the idea that these girls who have been kidnapped for over two
years have been released. No detail is given about how they were released
and especially, the remaining girls who may still be suffering in the
custody of the terrorist group. The news report and headline given in (4)
specifically emphasize the idea of family reunion as a result of the release
of the Chibok girls. It should be noted that capitalization is also used as
a major strategy of foregrounding information. The capitalized framing
292 T.O. Ehineni
CHIBOK GIRLS REUNITED draws the attention of the reader to the

idea of reunion and harmony that the girls now enjoy.
The capitalization is a way of diverting attention to a specific detail in
the news such that the media is silent on other more important facts. For
instance, what about the experiences of these girls in the Sambisa forest?
How did they survive these horrible times and more importantly, what
about others still there? What steps is the government taking to rescue the
remaining girls? All these gaps occur in the news reported here. However,
more focus is given to the fact that the rescued girls are reunited with
their families. It is a form of framing which draws attention only to the
event’s resolution, leaving out other details. It is a way of focusing the
attention of the public on something more specifically in such a way that
other facts which are expected in the events are not reported.
10.4.1.2 Omission
This may be seen as removal of detail that was initially present in a text
(Jaworski & Galasiński, 2000) or taking important information out of a
text (Huckin, 2002). This is a process where certain facts of an event are
omitted in the news.
(5) Boko Haram sets conditions to free 83 more Chibok girls

As negotiations for the expected next phase of the release of 83 Chibok
girls begins this week, there were indications, last night, that the leadership
of the Boko Haram sect might table two major conditions to be met by the
Federal Government to seal the deal. (Vanguard, 16 October 2016)
The news report does not provide details about the remaining girls—
there are still over 170 girls. The use of numbers is a strategy of omission
here where the news only gives information about a fraction of a large
number while omitting important facts about the other fraction. In
essence, omission is seen here in terms of how the news report leaves out
information about the remaining girls. The lack of this information may
trigger questions about the state of the remaining girls.
(6) Boko Haram’ll keep promise to release Chibok girls, as negotiations

continue—FG
A top military source, who confided in Vanguard, said: ‘It is true that the
sect demanded the release of some of their fighters and the Federal
Government is ready to meet some of their conditions to secure the girls’
freedom. There are high hopes that the girls will be released; the military is
ready to move in and bring out the girls when this is done.’ (Vanguard, 20
October 2016)
(7) FG, Boko Haram negotiating release of 83 more Chibok girls

Three days after the release of 21 girls abducted from Government
Secondary School in Chibok, Borno State by Boko Haram, the Federal
Government and the insurgents are already negotiating the release of 83
more. On April 14, 2014, 219 girls were abducted from their dormitory by
Boko Haram. (The Sun, 17 October 2016)
The news in (6) gives some hope and optimism to Nigerians in the
midst of despair and continuous worry over the abducted Chibok girls.
However, the identity of the source is not revealed by the newspaper:
‘A top military source, who confided in Vanguard, said’. While the
information about the source is not given, it is unclear if the promise
given can be fully trusted. Also, the headline indicates that the state-
ment is from the FG (Federal Government) but in the news content, it
is indicated that the news report comes from ‘a top military source’.
This is misleading since a top military source may not necessarily rep-
resent the FG. In (7), the news report omits information about the
other girls indicated in the 219 figure given at the end of the brief
story. It may suggest that these remaining girls have died or have
become wives for the terrorists. Also, the unreported detail about the
girls may be a way the reader is not allowed to think about them.
Hence, silencing aspects of an issue has the function of preventing such
questions from arising.
10.4.2 Linguistic Features that Produce Absence
Linguistic forms are used to promote absence in the Nigerian news media
reports. Some forms identified are incomplete headlines, silent pronouns,
rhetorical questions and ambiguous acronyms.
294 T.O. Ehineni
10.4.2.1 Incomplete Headlines
News headlines signal the topic of a news report and they are often the
most conspicuous part of any news story. Thus, incomplete headlines
would reflect some gaps in news details.
(8) Chibok girls: Talks with Boko Haram to continue if… Buhari (The
Sun, 18 October 2016)
The structure of the above headline reflects that some details are not
reported. This headline basically tries to focus on the fact that negotia-
tions are continuing with Boko Haram. However, the conditions for such
negotiation are concealed through deletion of the other segments. Van
Dijk (1991) suggested that a headline may bias the understanding pro-
cess and influence the interpretation of the news by the readers, because
it encapsulates what the news media agents view as the most important
aspect of an event. The headline necessarily implies an opinion or a spe-
cific perspective on the event. Therefore, it is not merely a summary of
the text of the news item, but an ideological representation of what the
news event is. Similarly, in the headline:
(9) Chibok girls alive; Many impregnated, carrying VVF, other dis-
eases... (Vanguard, 25 October 2016)
(10) Who is Afraid of BBOG? (This Day, 21 September 2016)
It can be seen that some details have not been offered indicated by the
use of ellipsis in (9). The use of ellipsis here seems to foreground the VVF,
Vesicovaginal fistula, as the major disease that the girls have contracted.
Since ellipsis means something is left ‘unsaid’ in a text, the reader is only
able to focus on what is said. Thus, VVF is foregrounded by the absence
of other diseases. The headline in (10) is a comment on a burning issue
where some detail are not provided. The news reveals that people may be
afraid of BBOG (BBOG again may not be readily clear). BBOG is a
slogan that means Bring Back Our Girls. However, the significant detail
about who may be afraid is absent in the news report.
10.4.2.2 Rhetorical Questions
In the news data examined, there are sometimes questions whose answers
are not provided in the news reports. This is used to present brief details
on an event and asks readers to discuss it. Here, the newspaper journalist
brings the reader to discuss issues while the journalist’s thoughts remain
‘invisible’.
(11) Should FG swap B’Haram detainees for Chibok girls? (Punch, 17

August 2016)
(12) Do you think the Chibok girls can still be rescued alive? (Punch, 6
February 2016)
(13) Are abducted Chibok girls still alive, pregnant or even dead?
(Vanguard, 17 April 2016)
Rhetorical questions in newspapers tend to elicit the views of the

reader rather than the view of the journalist. They provide an opportu-
nity for news media agents to hide their thoughts on a topical issue of
public interest. However, questions have some implicatures in the news
context. Even though it seems the news media is trying to question the
public on an issue instead of expressing specific views on it, these ques-
tions in fact show that the issue is significant. They foreground how
important and crucial the issue is to the news media. Thus, the fact that
the media is not directly offering opinions on something may reflect
their silence on it. But, the fact that the media brings up a question for
the audience to consider, on this issue, shows the media’s consideration
of it.
10.4.2.3 Silent Pronouns
Silent pronouns are those whose reference may not found in the struc-
ture. I refer to them as silent pronouns because they do not give addi-
tional information on what is being represented in the news context.
296 T.O. Ehineni
Thus, they are ‘silent’ on meaning by being vague. Some of these pro-
nouns may have no antecedent or more than one antecedent within the
news report.
(14) We’re suffering over failure to reconcile (The Sun, 10 June 2016)
It is unclear who is being referred to in the above headline. The pro-

noun ‘we’ is used to conceal the identity of the speaker. Additionally,
there may be other kinds of silent pronouns in the news headlines. For
instance:
(15) They said I did it—Adeleke (Punch, 10 July 2016)
The pronoun ‘They’ in Example (15) is used to confer responsibility on

other people in the media in such a way that the personal opinions of the
media on certain issues are not reflected. It is a way of speaking through
someone else who could be identified as ‘they’ or ‘them’. This affords the
opportunity to distance oneself from an event being reported such that
the journalist is not held responsible for what was said.
10.4.2.4 Ambiguous Acronyms
Acronyms are terms formed from a collection of words to designate con-

cepts and their meanings are often widely known by the public. However,
it is not uncommon to see acronyms that are unclear to the public. The
media may sometimes use these acronyms to refer to ideas that might
have been formed by the journalists or editors. Here, significant detail
may still be missing through their use.
(16) BBOG demands rescue of Chibok girls two years after (Punch, 14
April 2016)
(17) Making a case for Chibok girls and IDPs (Punch, 2 March 2016)
The meaning of BBOG and IDPs here is unclear, they are not a popular
acronymd like FG, (Federal Government) or EFCC (Economic and
Financial Crimes Commission) that Nigerians would easily understand.
However, this acronym plays a vital role in attracting attention to the stories
since the reader is interested in knowing what they mean. In essence, the
vague acronyms here have been used to conceal the information in the
headlines. While acronyms are very common in Nigerian newspapers
(Ehineni, 2014a, 2014b; Taiwo, 2007; Chiluwa, 2005), they can be used to
conceal information in the news media. For instance in Example (17), the
audience do not know what IDPs means. In a similar vein, the headline:
(18) Chibok girls alive; Many impregnated, carrying VVF, other diseases
….. (Vanguard, 25 October 2016)
This news headline uses an acronym which may not be clear to name
the diseases that the released girls now have. This might be a way of mak-
ing less explicit the nature of the girls’ diseases. While it is very common
to use acronyms as a way of concealing information in Nigerian hospital
discourse (Odebunmi, 2003), they may also occur in the news headlines
to veil the exact meaning of diseases.
10.4.2.5 Headless Numbers
Headless numbers are used as modifiers where the noun that they modify
is not supplied. In this case, the reader only knows the number of some-
thing but does not know what the number is talking about. This is a very
common feature of news headlines in Nigerian news media (Ehineni,
2014a, 2014b), which may also be observed in other news headlines.
However, it should be noted that while headlines may not always func-
tionally spell out everything, the information that is spelt out and what is
not spelt out may be important, since it reveals in some sense what the
news reporters or editors what the audience to focus on.
(19) Chibok Girls: Army declares three wanted (Punch, 27 October

2016)
298 T.O. Ehineni
(20) Army declares two others wanted over Boko Haram (Vanguard, 14
August 2016)
From the above headlines, the significant detail about who is wanted is
missing. The reader only knows the number of those wanted but not any
information about who these people are. Numbers are used to take the
place of nouns where the reader only knows partial information on a
news report. As initially stated, note that while it is a feature of Nigerian
news headlines (and news headlines in general) to be syntactically short,
projecting numbers as nominal heads (Ehineni, 2014a, 2015), what is
reflected in the headlines is also crucial. What is reflected in the headlines
may also be informed by journalistic choices as a way of foregrounding or
backgrounding news aspects to capture the audience’s attention.
Consequently, the reader is made to read the news story with a focus in
mind. For instance, in the headline (19), ‘declares three wanted’ creates a
major interest for reader who is reading the headline, and is not paying
attention to other aspects of the story like the fate of the Chibok girls,
which the story is based on. Thus, what is omitted may make something
more prominent, so that other details are backgrounded. In other words,
headlines may suggest a framework for the readers when reading a news
report, and what is missing could foreground some detail or background
others.
10.5 Conclusion
This chapter has examined forms of absence in recent Nigerian news
reports especially on the Chibok girls’ abduction saga, which generated a
lot of public concern both within and outside the country. First, I have
identified strategies or devices that were engaged in achieving silence and
absence in the news report. This could include topicalization and omis-
sion. Topicalization is a process of foregrounding information while
omission is a way of concealing information in the media. Second, I dis-
cussed some linguistic strategies such as incomplete headline structures,
ambiguous acronyms, rhetorical questions, silent pronouns and headless
numbers. Essentially, I argue that absence and silence, as exemplified in
the discussion of the news data, have implicatures of meaning. This may
include an attempt to conceal information from the public, shift the
reader’s attention, show the prominence of some facts over others. In
essence, absence in the news media deals with not only what is absent but
how this is presented—the strategies and the features of this process—
which may be motivated by journalistic intentions and media
ideologies.
References
Chiluwa, I. E. (2005). Discourse pragmatics of news headlines and lead stories in
Tell, The news and Newswatch, 1996–2002. Ph.D. thesis, Department of
English, University of Ibadan. xiii+212
Davis, C. K. (2000). How Iowa editors are using law expanding access to names
of juveniles. Newspaper Research Journal, 21(4), 38–48.
Druckman, J. N. (2001). On the limits of framing effects: Who can frame?
Journal of Politics, 63(4), 1041–1066.
Ehineni, T. O. (2014a). A syntactic analysis of lexical and functional heads in
Nigerian newspaper headlines. International Journal of Linguistics, 6(5),
9–21.
Ehineni, T. O. (2014b). Lexical, structural and rhetorical features of Nigerian
English print media. Journal of Arts and Humanities, 3(11), 26–32.
Ehineni, T. O. (2015). Syntactic government in Nigerian newspaper headlines.
Journal of Arts and Humanities, 3(11), 40–51.
Entman, R. (1993). Framing: Toward clarification of a fractured paradigm.
Fairclough, N. (1989). Language and power. London: Routledge.
Fairclough, N. (1992). Discourse and social change. Cambridge: Polity press.
Fairclough, N. (1995). Media discourse. London: Edward Arnold.
Gamson, W. A. (1996). Media discourse as a framing resource. In A. Crigler
(Ed.), The psychology of political communication (pp. 111–132). Ann Arbor:
University of Michigan Press.
& Society, 13, 347–372.
Jaworski, A., & Galasiński, D. (2000). Strategies of silence. Omission and ambi-
guity in The Black Book of Polish Censorship. Semiotica, 131(1-2), 185–200.
300 T.O. Ehineni
Jørgensen, M., & Phillips, L. (2002). Discourse analysis as theory and method.
London: Sage.
Kendall, D. E. (2005). Framing class: Media representations of wealth and poverty
in America. Lanham, MD: Rowman & Littlefield Publishers.
MacRitichie, V., & Seedat, M. (2008). Headlines and discourses in newspaper
reports on traffic accidents. South African Journal of Psychology, 38(2),
337–335.
Matu, P. M., & Lubbe, H. J. (2007). Investigating language and ideology: A
presentation of the ideological square and transitivity in the editorials of three
Kenyan newspapers. Journal of Language and Politics, 6(3), 401–418.
McQuail, D. (1992). Media performance: Mass communication and the public
interest. London: Sage.
Odebunmi, A. (2003). Pragmatic features of English usage in hospital interactions
amongst medical practitioners and patients in Southwestern Nigeria. Unpublished
Ph.D. thesis, Ile-Ife.
Phillips, L. (2006). Doing discourse analysis: A brief introduction to the field.
In N. Carpentier et al. (Eds.), Media technologies and democracy in an enlarged
Europe (pp. 285–294). Tartu: Tartu University Press.
Price, V., & Tewksbury, D. (1997). News values and public opinion: A theoreti-
cal account of media priming and framing. In G. Barett & F. Boster (Eds.),
Progress in communication sciences: Advances in persuasion. Greenwich, CT:
Ablex.
Reah, D. (1998). The language of newspapers. London: Routledge.
Richardson, J. E. (2007). Analyzing newspapers: An approach from critical dis-
course analysis. New York: Palgrave Macmillan.
Schudson, M. (2003). Sociology of news. New York: Norton.
Taiwo, R. (2007). Language, ideology and power relations in newspaper head-
lines. In R. Taiwo, A. Odebunmi, & A. Adetunji (Eds.), Perspectives on media
discourse. Munchen: LINOM Publishers.
Thayer, F., & Pasternack, S. (1994). Policies on identification of people in crime
stories. Newspaper Research Journal, 15(2), 56–64.
van Dijk, T. (1991). Racism and the press. London: Routledge.
van Dijk, T. (1995). Ideological discourse analysis. In E. Ventola & A. Solin
(Eds.), Special issue interdisciplinary approaches to discourse analysis. New
Courant (Vol. 4, pp. 135–161).
van Dijk, T. (1996). Power and the news media. In D. Paletz (Ed.), Political
communication in action. Cresskill, NJ: Hampton Press.
van Dijk, T. (2008). Discourse and power. New York: Palgrave Macmillan.
Verschueren, J. (1999). Understanding pragmatics. London: Edward Arnold.

Verschueren, J. (2008). Context and structure in a theory of pragmatics. Studies
of Pragmatics, 10, 13–23.
Winch, S. P. (1991). On naming rape victims: How editors stand on the issue.
Paper presented at the national convention of AEJMC, Minneapolis, MN.
Part III
Analysing Surface Indicators of
Silence and Absence
11
What the f#@$!: Policing
and Performing the Unmentionable
in the News
Crispin Thurlow and Jamie Moshin
This chapter examines the multimodal tactics used by newspaper journal-

ists and their editors to represent taboo language, as with the following
extract:
I’m referring to the C word … not the one you think, the one no one is
ever supposed to use … you’ll be able to grasp it (deliberate choice of
phrase) if I tell you the second and third syllables are ‘sucking’. (Birmingham
Post, Alabama, USA)
For us, this is a neat case study of the complex communicative and
cultural functions of silence, crudely and commonly understood as the
absence of words. Following Adam Jaworski’s (e.g. 1993, 1998) well-
known theorizing of silence, we are equally interested in the metapragmatic
and ideological dimensions of these absences as acts and/or performances
C. Thurlow (*)
Department of English, University of Bern, Bern, Switzerland
J. Moshin
University of Michigan, Ann Arbor, MI, USA

306 C. Thurlow and J. Moshin
of silencing. As Jaworski notes, silence is seldom, if ever, an all-or-nothing

phenomenon; rather, it is always gradable on a more-or-less scale. This is
something that becomes very apparent in this chapter where we see news-
makers pointedly and/or ostensibly omitting taboo words through all
sorts of otherwise quite obvious, visible devices. Furthermore, these
newsmakers partly respond to the smothering effects of societal pressures
and institutional directives concerning the use of ‘appropriate’ or ‘good’
language, but we also observe them doing so in ways that go well beyond
basic requirements. In keeping with Bourdieu’s (1991; see Hanks quoted
below) famous terms, newsmakers are actually regulating and ‘silencing’
both themselves and others.
Standardization and legitimation sanction certain ways of speaking,

rewarding some while silencing others. The effect is to intimidate and cen-
sor speech without any discrete acts of intimidation or censoring. … To
euphemize one’s speech, consciously or not, is to self-regulate … (Hanks,
2005, p. 76)
As a critical discourse study, our analysis is grounded in empirical evi-

dence: primarily, a dataset of nearly 300 newspaper articles, mostly from
the USA. (The cultural specificity of our data is certainly relevant but
does not, we think, detract from the broader communicative principles
or practices under discussion.) Our analysis starts with the top-down
muzzling effects of government regulation and newspaper style guides.
We then turn to a content analysis of our dataset, focusing on shit and
fuck, as well as two high-profile ‘talk scandals’ (Ekström & Johansson,
2008) about these same words.1 One of the immediate methodological
challenges in conducting this kind of investigation is finding instances of
absence; how, for example, does one search a database for the omission of
shit when people are precisely not actually saying shit? The issue is not
simply a practical one, but has deep theoretical and epistemological
implications for the study of discourse analysis more generally. Mike
Billig (1997) noted some time ago that a major failing of discourse s tudies
(broadly conceived) is its undue—exclusive, even—attention to what is
present in discourse rather than what is absent, what is said rather than
what is not said or left unsaid. With its grounding in social theory and
What the f#@$!: Policing and Performing the Unmentionable... 307
under the influence of cultural studies, Critical Discourse Studies is more

attentive to the unspoken, omitted or hidden; even so, it too readily pre-
sumes to know the relevance or intentionality of what is said. In our
chapter, we are concerned with identifying actual absences (i.e. moments
of censorship), examining how these absences are accomplished and/or
performed, and, following Melani Schröter’s (2013) example, attending
to the metadiscursive framing of absences—although we are equally
interested in both explicit and implicit framings.
Through the course of the chapter, we move between a descriptive
analysis of the omission tactics themselves, an interpretive analysis of
their institutional framing and communicative functions, and a criti-
cal analysis of the ideological ramifications of it all. We will demon-
strate the tension that emerges in this mediatizing of taboo language
between the newsmakers’ self-declared commitment to reporting the
truth and their own bourgeois desires for civil discourse, a tension
compounded in the USA by the constitutional protection of so-called
free speech. In effect, we witness a cultural-political tussle between
expression and quiescence. Newsmakers often, but only partly, resolve
this tension through a playful-cum-prurient performance of omission,
one rooted in dominant language ideologies (cf. Woolard & Schieffelin,
1994). Following Crispin’s earlier work (e.g. Thurlow, 2006, 2014,
2018), we see here a community of language workers at work and at
play, simultaneously crafting and policing language while also demon-
strating (or performing) their mastery and artistry as wordsmiths. Far
from reporting the facts, journalists are making up the news. Far from
being silent, they are using a lot of words and making an awful lot of
noise.
Against this brief introductory backdrop, we turn directly to our anal-
ysis, which we have organized into two parts. In Sect. 11.1, we consider
some of the top-down ‘silencing’ strategies at work in the reporting/rep-
resenting of taboo language, and the way newsmakers’ discursive, or
rather metadiscursive, practices are societally and institutionally
structured.
11.1 T
op-down Strategies: Institutional,
Regulatory ‘Silencing’
We start by going back in history and to a long, contorted story of gov-
ernmental intervention and regulation. This is more than mere scene-
setting. It is precisely the convolutedness of this story that makes our
point. We also want to draw attention to the actual words—the taboo
language—at the heart of this drawn-out process, and to the way the
print media take up the story and represent the words. One important
piece of information is that the Federal Communications Commission
(FCC) is the US government’s independent body authorized to oversee,
among other things, the broadcast media—including its standards of
language.
Whoever utters any obscene, indecent, or profane language … shall be

fined under this title or imprisoned not more than two years, or both.
FCC, Section 1464
Where the FCC had for many years consented to the use of ‘fleeting
expletives’, their general policy was suddenly reversed in 2003 when, in
pursuing action against pop star Bono’s use of fuck in that year’s Golden
Globe awards, the Commission declared that any use of fuck would be
deemed indecent. In news reports at the time, Bono was reported as actu-
ally having said: this is really, really fucking brilliant. So, not fuck after all.
Then, in March 2004, the FCC actually overturned its decision; accord-
ing to the North Carolina News and Observer, the FCC’s statement
included this wording: ‘in the context presented here, [the word] did not
describe sexual or excretory organs or activities’. (It was the newspaper’s
decision to censor fuck in quoting the FCC statement.) This kind of
unclear caprice on the part of the FCC works precisely to keep media
outlets nervous and self-censoring; they are, as one broadcaster (the
Public Broadcasting System, PBS) put it, ‘compelled to err on the side of
caution’. In spite of the exception made for the Bono case, the FCC pro-
vided little reasoned argument for its ruling on fuck, the assumption
being that fuck and shit are unacceptable because they refer to sexual or
excretory acts (see below). As such, the FCC’s policy hinged on a formal-
ist, referential conflation of ‘dirty’ words and ‘dirty’ meanings. This view
of language necessarily overlooks their expletive potential, their common
function as expressions of frustration, excitement or play (see Seizer,
2011). This way of thinking about verbal taboo also misrecognizes the
inherently situated, interactional and performative nature of both lan-
guage and taboo (Flemming & Lampert, 2011; Irvine, 2011). Regardless,
we note for now how the FCC rulings are quickly taken up by newspa-
pers and often reported in ways that avoid using the regulation-breaking
words themselves. The story continues.
In June 2007, a second-tier appeals court in the USA struck down (i.e.
reversed) a series of different indecency rulings by the FCC against four
major television stations for reporting ‘obscene language’. Once again,
most mainstream news reports about the court’s decision assiduously
avoided any mention of the actual words which had prompted the origi-
nal FCC decisions. The New York Times, for example, used the legalistic
terms ‘indecency’ and ‘fleeting expletives’ as well as euphemisms such as
‘blurted obscenities’ and ‘vulgarities’ or ‘a vulgarity’. (Referring also to a
completely separate incident, the same New York Times article discussed
an ‘angry obscene version of “get lost”’ used by Vice-President Dick
Cheney; this happens to be one of the ‘talk scandals’ we examine later.) It
was only after a considerable amount of internet searching, that we our-
selves managed to discover what the original infractions had been: a word
used by Cher during the Billboard Music Awards five years earlier (People
have been telling me I’m on the way out every year, right? So fuck ’em); two
words used by Nicole Richie at the following year’s Billboard Music
Awards (Have you ever tried to get cow shit out of a Prada purse? It’s not so
fucking simple); various words used in several episodes of the television
show NYPD Blue (bullshit, dick and dickhead); and, lastly, a live interview
on CBS’ The Early Show where a contestant from the television show
Survivor referred to another contestant as a bullshitter. One of the most
significant facts at this juncture, are the grounds for the federal appeals
court’s ruling against the FCC: the censoring of language contradicts the
First Amendment protection of free speech. At the time, various televi-
sion networks hailed the court of appeal’s ruling because, as the New York
Times put it, the FCC would be ‘chilling artistic expression’. In contrast,
the head of the FCC declared the ruling a ‘disappointment for American
families’.
This battle between the government and the broadcast media, together
with its coverage in the print media, provided us with an ideal pretext for
examining the mediatization of taboo, the representation of the unspeak-
able, and the policing of the unmentionable. It also brings us to the first
in a series of interpretive moves concerning the nature and politics of
silencing. We see here a perfect manifestation of the way all attempts at
linguistic proscription (i.e. both literal and metaphoric silencing; cf.
Jaworski, 1998) unavoidably end up producing the very discourses they
mean to stifle. We have, therefore, a veritable incitement to discourse in
accordance with Foucault’s (1978) repressive hypothesis. In seeking so
busily to censor the speaking of fuck, for example, they are obliged to do
nothing else but speak (or write) about fuck. (We are reminded of Paul
Baker’s 2004 paper about British parliamentary debates concerning age-
of-consent equality which saw the ‘Lords and Ladies’ of the upper house
debating some of the finer details of anal sex.) Ultimately, efforts to dis-
tract people from taboo acts merely draw attention them. It is a very
verbose kind of silencing.
Although the FCC’s government-backed policing of taboo language is
an important institutional and contextual factor, it is somewhat less
interesting to us. What really interests us is the kind of self-censorship
happening in the print media (e.g. in the New York Times above) which
actually falls completely outside the jurisdiction of the FCC. In fact,
there is no equivalent government agency in the USA for regulating the
print media. We cannot help but wonder how newspaper journalists and
their editors rationalize this type of censorship with their otherwise
mythologized claim to truthful, ‘factual’ reporting. How, also, do they
explain the avoidance of specific words while digressing in other factually
unrelated stories? This brings us to the more immediate institutional
rules and norms of the newsroom.
When it comes to the use and depiction of taboo language, the print
media is not a completely unregulated playing field, even without an
FFC-like body. Journalists and their editors are almost always under the
sway of self-imposed style guides. Most major news sources have in place
these top-down mechanisms for overseeing the use of language: grammar,
punctuation, spelling and, notably, ‘obscene’ or ‘indecent’ or ‘offensive’

language. These in-house guides vary from paper to paper, especially
when it comes to the norms and conventions for using/reporting exple-
tives, ‘foul language’, and so on. By the same token, each news organiza-
tion has its rules for speaking the supposedly unspeakable. Here are three
examples:
Obscenity, vulgarity and profanity: The Times writes unblushingly about

sexual behavior … opening its columns to any newsworthy detail, however
disturbing … provided the approach is dignified and the vocabulary clini-
cal rather than coarse. …. The Times virtually never prints obscene words,
and it maintains a steep threshold for vulgar ones. [New York Times, USA]
… do not use them unless they are part of direct quotations and there is a
compelling reason for them. … do not, however, change the offending
words to euphemisms. Do not, for example, change ‘damn it’ to ‘darn it.’
If one cannot drop a quotation containing an obscenity … replace letters
of an offensive word with a hyphen. When the subject matter of a story
may be considered offensive… flag the story at the top: ‘Editors: The con-
tents may be offensive to some readers. [Associated Press, USA]
We are more liberal than any other newspaper, using words such as cunt
and fuck that most of our competitors would not use. … remember the
reader, and respect demands that we should not casually use words that are
likely to offend. … there is almost never a case in which we need to use a
swearword outside direct quotes. … the stronger the swearword, the harder
we ought to think about using it. … never use asterisks, which are just a
copout … [The Guardian, UK]
These extracts point nicely to newspapers’ very different attitudes

and policies towards taboo language and how it is represented. Not
only do these style guides manifest a particular language ideology vis-à-
vis swear words and the like, but we also see the kind of meta-meta-
language by which papers use their regulation of linguistic resources for
styling themselves—the paper itself, that is—as a particular kind of
paper. Given its left-wing politics and progressive brand identity, the
Guardian newspaper, not surprisingly, takes a more unorthodox stance,
one which it performs twice over in the extract above: first, by insisting
journalists actually quote the taboo words without euphemism or
redaction; second, by defining themselves in opposition to their, by
implication, conservative competitors. What is particularly germane is
how the very absence or presence of taboo language is a key resource by
which these stances are enacted. The Guardian claims its distinction by
being willing to speak the words, positioning itself favourably vis-à-vis
a newspaper like the Telegraph. Accordingly, we learn that absence =
timid or prudish, while presence = bold/daring. Once again, however,
what is even more important is not simply what is or is not absent, but
how this absence is performed in practice—what typographic, lexical
and other discursive resources are used to create the absences and,
indeed, to draw attention to them. This brings us to the second major
part of our analysis.
11.2 Bottom-up Tactics: Performing ‘Absence’

As the initial step in a slightly larger investigation of mediatized taboo, we
started by looking at two commonly used (and commonly regarded)
swear words: shit and fuck. The choice of these two particular words was
also motivated partly by the two much-talked-about talk scandals at the
time which we had also decided to examine; both words had also been
such an issue with the FCC.2 Very briefly, our dataset was generated
through a LexisNexis search of major international, but mostly US
American, newspapers ranging from broadsheets (US: newspapers of
record) to tabloids, and including all sections of the papers. For shit and
fuck, we selected articles from the six months leading up the date when
the FCC ruling was struck down (see above); this produced 108 shit arti-
cles (i.e. shit, sh-t, s--t) and 59 fuck articles (i.e. fuck, f--k, f-ck, f*ck). For
the two case studies, we selected articles in the month following each
incident, producing 69 articles on George Bush’s shit and 57 articles on
Dick Cheney’s fuck. Although not a point of interest for this chapter, we
coded our dataset for different types of article: reports about the incident;
a report with a passing reference to the incident; an editorial about the
incident; and/or meta-commentary about this type of incident (e.g. a

discussion about the word ‘shit’ in general, or about its inclusion in ‘civil
discourse’). We mention this because it is important to acknowledge
there are different degrees of coverage in terms of the amount of copy (e.g.
a dedicated report or an extended opinion piece), the location of stories
(e.g. front-page news or back-page gossip), and the prominence of the
taboo words (e.g. in the headline or inside the article).
As we say, studying the unspoken or unspeakable has interesting meth-
odological implications for discourse analysts. We faced a similar episte-
mological challenge, but also one that was a lot more logistical. At issue
for us was, initially at least, not how to argue for the significance of omis-
sions but how to find them in the first place. So, for example, how to
determine from our search when a taboo word—say, shit—had been
avoided, deleted or replaced. How does one look for something that is
not there? Another related challenge arose when searching for commonly
used euphemisms like s**t or s--t which are interpreted by most databases
as ‘wild-card’ searches, and just as likely to return shot, shut and spit as
they are shit. Partly following Billig’s lead (see above), but also in the spirit
of ideological critique, the only solution was to pursue traces, or what we
might call ‘explicit absences’—those which call attention to themselves.
As it happens, this was not all that difficult because we found that jour-
nalists and their editors more often than not call attention to the absences
themselves.
11.2.1 Depicting Shit and Fuck
In spite of the style guides, what happens in practice feels a lot less regu-
lated and a lot more open to individual creativity, albeit still convention-
alizing in its own way. Bearing an uncanny resemblance to putative
text-messaging style, the reporting of familiar taboo language like swear
words finds journalists jumping through hoops. In organizing our data,
we following some of descriptive labels used by others (e.g. Allan &
Burridge, 2006) and present a range of linguistic, typographic and other
tactics used to represent shit and fuck (with one orthophemistic piss) and
to avoid saying them. Or to appear to avoid saying them. For each
c ategory, we include different examples selected from different articles.

By ‘institutional censorship’ we mean journalists evidently following style
guide rules; meanwhile ‘formulaic contextualization’ describes the way
meaning is established through familiar phrasing.
Institutional censorship: “This is (expletive) awesome!” | (expletive
deleted) | a major league (expletive) | he
used a curse word
Deictic references: it | that word | the obscenity
Typographic substitutions: s--- | s**t | f k | F#ck | SH!T | s—t |
sh*t | f---in’ | s…t ?
Formulaic contextualization: WHAT THE BLEEP? | I’m f------ seriously pissed
| F**k you! | cow s--t | eff you | shut the f..k
up | &^%$ it | s**t-hole | are you f king
kidding me?
Initialisms: WTF | they can F right O | FU
X-word formula: f-word | F word | word-bomb | f-bomb | the
s-word
Euphemisms/infantilisms: frikkin | friggin’ | No. 2s | fig | fug | fudge
Orthophemisms: ORDURE | an angry obscene version of ‘get
lost’ | “If Joe was on the side of the road on
fire, I wouldn’t [urinate] on him to put him
out.”
This array of typographic, orthographic and other visual tactics is evi-

dence of just how inventive journalists can be in representing the suppos-
edly unmentionable. Take, for example, just one of these tactics: the
so-called x-word formula. Cornog and Perper (1991, p. 326) highlight
this as an example of ‘linguistic oppression’; as they explain, ‘The
“X-word” paradigm revolves around speaking by not speaking: it is a way
to mark linguistically a specific word as belonging to a class of words––
like fuck itself––that only the reckless utter.’ We ourselves refer to this
tactic as a ‘formula’ because the same principle is applied all over the place
with a host of different taboo words. As Cornog and Perper further
explain, the parallel that gets rhetorically established by extending the
format to other terms such as the c-word (cunt), the n-word (nigger), the
l-word (lesbian) problematically renders them all equally, or at least simi-
larly, taboo.
What all of the different types of tactics and the different variations of
each tactic ultimately demonstrate is, as we indicated above, just how
productive proscription can be (Flemming & Lampert, 2011). These are
quintessential examples of ‘ostentatious taboo’ (Read, 1964), clearly
flouting Gricean maxims of quality and quantity in what Judith Irvine
(2011, p. 17) refers to as the ‘game of containment’ (we return to this
notion later). In short, far from avoiding the taboo language, journalists
are busy playing with witty ways of speaking it nonetheless and leaving
little or no doubt in the reader’s mind about what the word is. We have
absences being marked by essences and, sometimes, abundances. We also
have acts of silencing which are patently far from silent, but which repro-
duce the belief that some things should be silenced or best left unspoken.
In effect, therefore, we see taboo language and taboo in general being
reinscribed. Nowhere is this more apparent than in our two case studies.
11.2.2 Making (Up) ‘Talk Scandals’
To make sense of these two cases, a little background detail is necessary.

George Bush’s audio-recorded use of shit took place on 17 July 2006
when, as US president attending a Middle East summit, he was caught in
an off-record moment hailing Tony Blair (the UK Prime Minister) with
‘Yo, Blair’ and then characterizing the situation in the Middle East as this
shit. Vice-President Dick Cheney’s purported use of fuck (or fuck off or
fuck you or go fuck yourself) took place on 24 June 2004 during a photo
shoot, and was addressed to Democrat Senator Patrick Leahy. Unlike the
Bush incident, the Cheney incident was not recorded and so no-one
knew what was actually said. Needless to say, this potential absence of
information did not stop newsmakers.
We start with Bush’s shit and a descriptive analysis organized into three
broad categories together with multiple, typical examples. We use capital-
ization in the extracts to indicate headlines, and all quote marks in the
extracts are from the original text; in other words, the way journalists
themselves represented reported speech.
Circumlocutions: This is a particularly good term for capturing how one

word becomes many. In their effort to censor shit, journalists are often
compelled to make even more noise—literally, to use more words.
Once again, these tactics are quite ostentatious, drawing attention to
themselves both visually and cognitively (i.e. one has to work quite
hard at times to figure out what is going on). As with so much lan-
guage play (Crystal, 1998), these rebus-like games are often quite
witty; as such, they offer themselves as indexes of professional skill,
erudition and, we would argue, social class.
DUBYA’S REAL DOO-ZY—THE $*%! HITS THE FAN | the one that
rhymes with spit | excremental phrase | S-H-One-T | mild lavatorial rudery |
euphemism for offal
Meta-pragmatic/discursive framing: Everything happening in these news

stories is of course fully metalinguistic (Coupland & Jaworski, 2004);
we have journalists and editors inherently commenting on the nature
of language, whether through the use of typographic substitutions or
euphemisms, and through their act of censoring some language in the
first place. There are two other ways newsmakers often comment
explicitly on Bush’s shit. First, they furnish their own metapragmatic
assessments of intention, mood or force (as in the first set of examples).
Second, they make metadiscursive value-judgements about swearing
which points, in language-ideological terms, to the framing of Bush as
a particular kind of person and/or politician. So, not only more words,
but opinionated words.
‘frustrated’ | unguarded | impatient | irritable | a moment of frustration and

passion | at least Bush’s words were honest
raw language | plain-folk talk | refreshing blast of candor | earthy | unvarnished

| unpolished | straight-talking | regular guy | vulgar man | base, incurious cow-
boy | rough-edged
Verbal puritanism: The level of language-ideological judgement merely

rises with our next set of examples. We have chosen to frame these as
puritan for their overtly pious, moralizing tone. Suffice it to say, jour-
nalists here have moved well beyond reporting the facts and into the
terrain of unapologetic opinion and personal evaluation. We found
examples like this from across our dataset: broadsheets and tabloids,
editorial pieces and opinion pieces. We draw special attention to the
last example here which is a perfect act of ideological recursivity (Irvine
& Gal, 2000), whereby swearing is pejoratively linked to digital dis-
course and bad grammar, and, by implication, low-status youth styles
(see Thurlow, 2014). These kinds of extrapolations are at work also in
references to personal regard (‘respect’), social standing/status (‘man-
ners’) and competence (‘unprofessional’).
[speaking with his mouth full] | world leaders dispense with the normal con-
straints of syntax and grammar | swearing … causes people to respect them less
| swearing is still a sign of poor manners | unprofessional and unseemly | a kind
of texted-equivalent of the English language instead of bothering with properly
constructed sentences
Turning now to Cheney’s fuck, the key thing to remember is that this
was not recorded and so no-one knows exactly what was said at the time;
all that seems to have been understood is that he used some rendition of
fuck.
Managing absence of proof: In the context of uncertainty, journalists some-

times find ways of hedging their story, although it is seldom or not
necessarily made clear who/what their sources were—who alleged,
who reported, and so on. As we say, however, an incident which is
effectively grounded in gossip and hearsay is still newsworthy because
it involves taboo language used by a speaker and in a context framed
culturally and/or ideologically as inappropriate. In spite of having no
actual proof or direct evidence, newsmakers get to work producing
plenty of copy.
allegedly used the f-word | reported use of a vulgarism | either ‘f*** off’ or ‘go
f*** yourself ’ | which of the two preferred: the priceless two-worder—‘[verb]
you’—or the more expansive three-worder, a directive beginning with ‘go’
The same old s**t: Not surprisingly, we find many of the same ortho-
graphic and typographic tactics as with Bush’s shit and elsewhere in the
dataset. Remember, many of these decisions are obligated by institu-
tional guidelines. We see here a number of variations on the theme of
fuck, presumably all or mostly all fabricated. (They cannot all be cor-
rect.) What is striking is also the somewhat disingenuous use of quote
marks (see Waugh, 1995, on the peculiarities of reporting speech in
journalistic discourse). This is made all the more comical in our final
example where it seems highly unlikely Cheney would have said go
fuck himself. As Crispin notes elsewhere (Thurlow, 2006, p. 687), a
special product of academic research is that, in reading the same news-
paper several times and in reading dozens or hundreds of newspapers,
we are able quickly to spot patterned practices, cultural narratives and
topical (in)consistencies such as these here.
f--- yourself | ‘f..k off’ or ‘go f...k yourself ’ | either ‘f--k you’ or ‘f--k yourself ’ |
‘fuck yourself ’ | ‘Go F**k Yourself ’ | ‘(expletive) yourself ’ | ‘f… off’ or ‘go f…
yourself ’ | ‘Go f- yourself ’ | ‘go f--- himself ’
Prurient excess: In this last set of examples we want to highlight and reiter-
ate the tendency for journalists to produce excessive talk in the face of
apparent absence. Once again, we find the usual euphemisms and cir-
cumlocutions (first set of examples). But we also find journalists pro-
ducing an oxymoronically prurient performance of modesty: not
speaking the words themselves, but seizing on the chance to write
about any number of other incidents of high-profile/high-status swear-
ing, even from more than 200 years ago.
urging Mr Leahy to perform a sex act that’s anatomically impossible | Cheney’s

F-word. (Not Fox, the other one.) | … the vice president’s demand that Senator
Leahy commit an act of auto-eroticism
White House has not always been so forgiving of obscenity | Bush used the
F-word | Bush called someone a ‘major-league [expletive]’ | John Kerry waxed
profane | The Earl of Sandwich … uttered the f-word in the House of Lords in
1783.
We will shift in a moment to our more decidedly critical analyses, but

there remain a few interpretive observations about the bottom-up strate-
gies we have seen across our extensive dataset. In purely practical terms,
for example, we are left wondering how much orthographic detail can be
omitted before meaning is lost altogether or at what point so little is
omitted as to render the censorship completely ineffectual. To this end,
compare the completely obtuse ‘I’m not going to sit up here and b-------
you’ (Ottawa Citizen, Canada) with a very common form like ‘motherf-
-cker’ which leaves little to the imagination. The finer details of these
decisions and the circumstances under which they become conventional-
ized is also unclear (e.g. the principle of using one dash for every letter
omitted).3 Both within and across different papers, there is no consis-
tency of standard; why, for example, we see bastard, bitch and ass being so
variably censored. Likewise, we wonder what newsmakers really think is
to be gained from the following pretences of omitting fuck: ‘go (expletive)
yourself ’, ‘It don’t mean (expletive) right now.’, ‘Well, I think that’s
(expletive), and I hate that’, ‘Happy (expletive) New Year’, ‘You’ve got to
be (expletive) kidding me.’ Embedded in these formulaic phrases, the
technical omission of fuck, fucked-up or fucking certainly does not ensure
semantic omission.
We could not help but be amused by the irony of journalists having to
avoid taboo language even in feature articles about taboo language or
censorship, or in a review of a book about taboo language. Similarly,
there were a few moments of pure Schadenfreude in one news source
reporting on another news source’s accidental slips of the tongue/pen: in
2006, for example, the UK’s Guardian newspaper reported on the BBC’s
being reprimanded by the government regulator for using fuck in the
subtitles of a broadcast when the word had already been excised from the
soundtrack. We again have the Guardian styling or branding itself ‘cool’
in opposition to the BBC which, as a broadcaster, is held to tougher stan-
dards anyway. Speaking of style, our final observation about the overall
dataset is the consistency with which taboo language was used, in passing
or in some marked way, to depict certain types of people and not others.
Swearing is apparently only remarkable when used by presidents, prison-
ers or pop stars, and the like. These selective choices clearly expose lan-
guage ideologies. They also highlight ways in which silence is structured
by relations of power; not simply regarding the party-political power of

swearing presidents and vice-presidents, but the cultural-political power
of legitimating only some people’s ways of speaking. This brings us to the
three interrelated critical remarks in the way of a conclusion.
11.3 C
ritical Remarks: Containment Games
(or ‘Silent but Violent’)
It closed with a reference to the constituent as a certain anatomical part.
No, not that one. The one you sit on. (San Antonio Express-News, USA)
The moral life of language does not reside in the linguistic properties of
utterances alone, nor only in the moment of interaction. The words not
spoken, the discourse contexts, the interactional and societal histories, the
responses by interlocutors, the conventions of genre, the regimes of lan-
guage, truth, and knowledge that prevail in the interlocutors’ social
worlds—all these are relevant as well. (Irvine, 2011, p. 35; emphasis ours)
In the quote immediately above, Judith Irvine points to a fundamen-

tal feature of discourse: meaning lies not only or simply in what is pres-
ent, what is heard or seen, but also in what is absent—what is unheard,
what is not immediately visible, what lies beneath the surface or in the
past. Absence speaks, and it can speak volumes. By the same token, and
following Jaworski (1993) again, that which appears silent is often fully
articulate and resonant, just as apparent acts of silencing may, in both
effect and practice, be really quite loud and boisterous. In the case of
our news media data, this occurs quite literally and tangibly. We see this
in the data extract above. Undoubtedly stifled by in-house regulations
of the San Antonio Express-News, the journalist has opted for a coy cir-
cumlocution which is both wordy and, at least initially, obscure. What
is more, through their performance of not saying ass (presumably), they
pretty much say it anyway AND they also pretty much say dick (pre-
sumably) as well. The absentee is fully present, and the silence quite
effusive.
11.3.1 Containment Games (and Taboo)
Sigmund Freud famously described the inherent double-bind of taboo:

‘Anyone who has violated a taboo becomes taboo himself because he pos-
sesses the dangerous quality of tempting others to follow his example …’
(Freud, 1950 [1913], p. 32). He also goes on to note that people may be
ostracized for merely embodying the potential to violate a taboo. It is this
which enables social control to be exercised so powerfully through taboo
language, whether in the actual speaking of words, making reference to
the words, or by association with people who speak the words. In this
sense, we are all trapped, and one has to feel some sympathy for journal-
ists who are in the more invidious position of being professionally obli-
gated to dabble in the taboo-breaking of others. Or perhaps not. From
the start, one of our central problems with mediatized taboo has been the
contradiction (perhaps hypocrisy?) in the reporting of taboo language
given journalism’s ‘occupational mythology’ (Aldridge & Evetts, 2003).
What we actually witness throughout our dataset is a decidedly less deter-
mined attempt to legitimize the otherwise coveted ‘neutralistic stance’
(De Smedt, 2012: xx). We know from Allan Bell (1991) that the news is
always an act of storytelling; what we see across our data, however, is
more the kind of fabrication which Crispin (Thurlow, 2006) uncovered
in his study of news coverage of digital discourse. Journalists evidently
delight in their role as animators (i.e. releasing taboo language) while
carefully distancing themselves from authorship (e.g. through the use of
‘reported’ speech) (cf. Goffman, 1981). It is a careful balancing act or,
better still, containment game.
We take up the notion of containment primarily from Luke Fleming
and Michael Lempert’s (2011; also Irvine, 2011) introduction to their
special issue on ‘Verbal Taboo and the Moral Life of Language’, where
they push the study of taboo language away from its referential and deno-
tational properties (e.g. Allan & Burridge, 2006) towards its more perfor-
mative and indexical functions. In making this case, they pin-point four
key qualities of taboo language use, including the strategies of avoidance
and containment. Both are key for our analysis here. Very briefly, avoid-
ance and the performance of avoidance unavoidably reinscribe the belief
that the performative power of taboo language is somehow in the word

itself. And yet, just as Irvine observes in the quote above, the significance
(in both senses) of words is seldom immediately or obviously present; it
is always/also elsewhere, it is ‘absent’. Second, through their constant
efforts at containing taboo language, all speakers (not only journalists)
unavoidably call attention to the taboo itself. Furthermore, attempts to
reduce one’s responsibility for the taboo (e.g. reporting it as someone
else’s speech) or lowering one’s commitment to speaking it (e.g. a raised
eyebrow or wink) lead inevitably to the conventionalizing of both con-
tainment/avoidance tactics, and of the taboo itself.
So, through their various containment games, journalists are able to
perform their duties as upholders of public decency while also asserting
their professional status as clever, witty wordsmiths. Through the use of,
for example, euphemisms and circumlocutions, they can name and talk
about the taboo without explicitly speaking it. If they are to be released
from the risk of taboo-by-association, however, they must go further and
be as explicit as possible. It is this, we think that partly explains the
noticeable slide from ‘reporting’ to ‘opining’, from ostensible facts to
unapologetic statements of moral approbation. Fuelled either by profes-
sional obligation (or constraint) or through personal social or moral prin-
ciple, journalists overtly sanction others’ taboo violations while covertly
(or not) dancing around the taboos themselves. Either way, the taboos are
rehearsed and, effectively, validated.
11.3.2 T
he Classing of Silence (or the Silencing
of Class)
Newspapers are a particularly compelling site for discourse analysts for

a number of reasons, not least of which is that they produce a never-
ending source of manageable, conveniently accessible data. They are
also powerful, influential mechanisms of representation, and so obvi-
ous, worthy targets for critique. But what we also have is a community
of professional language workers who are not only prolific but also skil-
ful at using language (see Thurlow, 2018). What this also means is that
the stakes are unusually high for journalists when it comes to the
i mposition and policing of language standards or norms. In this regard,

we offer two final sets of examples from our newspaper articles about
the presidential shit.
barely comprehensible grunts you hear from teenage boys | frat-boy | ANIMAL
HOUSE SUMMIT | he can make even a global summit meeting seem like a
kegger | demeanour of a petulant adolescent
Homey G-8 | ‘not much dawg’ | Bush’s gangsta rap summary of the crisis in
Lebanon | they conversed not as statesmen but rather as semi-articulate home-
boys | his rap with George Bush | with all the diplomatic—and eating habits—
of a Cossack
These are, of course, more examples of the metadiscursive commentary

and verbal puritanism we picked up on earlier. What is different or spe-
cial about these examples, however, is their stereotypical referencing of
very particular groups or classes: specifically, young people and Black
people. In the first instance, we have comments that are reminiscent of
the pejorative texting reference in an example above, with Bush’s swearing
being framed as juvenile. As is typical of this kind of language ideology
(Woolard & Schieffelin, 1994), neither Bush nor young people come off
well from these flippant but clearly ageist jibes. And then we have the
arguably racist invocation of ‘Homey’, ‘gangsta’, ‘homeboys’, ‘rap’ and,
eventually, ‘Cossack’. Surely, this choice was prompted by Bush’s own use
of mock dialect in ‘Yo, Blair’, but the journalists deliberately and enthu-
siastically chose to run with it. In the last example, we also see a pointed
mapping of swearing onto social class (‘eating habits’) which we think is
also telling.
We have already considered how mediatized performances of/around
taboo language are the means for newspapers to brand themselves and for
journalists to style themselves professionally. We have also pointed (here
and earlier) to the strategic use/representation of taboo language for styl-
izing particular social groups. There is a third way we see taboo language
being deployed for stylistic purposes by newsmakers. Grammatically, sty-
listically or culturally ‘bad’ language is particularly bothersome for pro-
fessional, educated ‘elites’ (cf. Aldridge & Evetts, 2003; van Dijk, 1993)
who, as Stallybrass and White (1986, p. 191) note, are likely to be caught
up in some particular self-constitutive preoccupations: ‘The bourgeois

subject continuously defines and re-defines itself through what it marks
as “low”—as dirty, repulsive, noisy, contaminating. Yet the very act of
exclusion is constitutive of its identity.’ Try as hard as we might to dis-
tance ourselves from the objects of our distaste, we, the bourgeoisie, find
ourselves unavoidably defined by these very things. It is akin to what
Stuart Hall (2013, p. 353; following Freud and Homi Bhabha) calls
‘fetishism disavowal’: a desire or fascination is simultaneously indulged
and denied and two contradictory beliefs/values are at once embraced.
In these terms, it is perhaps not altogether surprising that our analysis
of newsmakers’ performance and strategic deployment of ‘absence’ bears
a close resemblance to the way Crispin and Adam Jaworski (Thurlow &
Jaworski, 2010) view silence functioning in the context of elite discourse.
Our mediatized taboo practices similarly entail two interlocking, meta-
level uses of absence or silence: first, their use as a multimodal resource
for representing taboo words; second, as a way to understand the cultural-
political consequences of these mediatized representations. Absence is a
key marker both of taboo and, like silence in Thurlow and Jaworski’s
study, of social status; all of which is patently refracted through the bour-
geois privileges and anxieties of newsmakers. So much of what is happen-
ing in our data may, therefore, be understood not only in terms of
institutional/professional practice, but also in terms of class. And this
bring us to our final and summing-up remark.
11.3.3 Turning to Violence
What we see, in this particular case, is an institution or profession which

touts its commitment to reporting ‘the facts’, but one in which language
is regularly censored and the words of others are either euphemized or
altogether repressed. It seems to us that this is a daily exercise of what
Pierre Bourdieu might consider symbolic violence and which he explains
thus:
Symbolic violence, to put it as tersely and simply as possible, is the violence

which is exercised upon a social agent with his or her complicity. … To say
it more rigorously: social agents are knowing agents who, even when they
are subjected to determinisms, contribute to producing the efficacy of that
which determines them insofar as they structure what determines them.
(Bourdieu & Wacquant, 1992, pp. 167–168)
Throughout our analysis, we have been concerned with newsmakers’

uses of language to regulate and control readers, as well as the people at
the centre of their reports and, indeed, themselves. The reporting of taboo
language—in practice, if not in principle—reveals an awful lot of self-
regulation while also policing the boundaries of language and broader
societal norms which may or may not have anything to do with language
per se. All of which is achieved consensually through often quite playful,
humorous manipulation. There is, of course, an inherent symbolic vio-
lence exercised in all language ideological practices or statements (cf.
Thurlow, 2010). In the current case, for example, we see young people
and Black people’s ways of speaking being negatively and reductively ste-
reotyped in newsmakers’ efforts to reprimand and shame high-profile
government figures for their ways of speaking. Whether or not one
chooses to see symbolic violence in these discursive acts, the main point
we have tried to make is this: what looks like absence seldom is. In the
face and performance of absence we invariably find, at the very least,
traces and essences. Sometimes, in fact, absence is substituted with abun-
dance. In this way, we have also sought to complicate and problematize
silence. To treat silence simply as an on-off or either-or phenomenon is to
misunderstand its complexity and to misrecognize its power. And so, we
think it fitting to finish by using academic conventions to demonstrate
just how easily one can speak without speaking. We end, therefore, not
with our own words (we will be perfectly silent), but with the words of
Adam Jaworski (1993, p. 24): ‘Silence has many faces.’
Acknowledgements This piece has been a very long time in the making or, at
least, the publishing. It started life back in 2007 right before Crispin became a
dad (two times over) and when the wheels fell off rather. The chapter had its
most significant showing in a panel Sociofuckinglinguistics: Mediatizing Taboo
organized by Crispin for the 21st Sociolinguistics Symposium, Berlin, August
2012. This was shortly before his father died very unexpectedly, followed by a
continental shift of the home front. In these years, Jamie finished his PhD, took
up his first of three academic appointments, got married and became a dad.
These are not excuses; they are simply the realities in which—and in spite of
which—academic life takes place. For his help and enthusiasm in pulling
together our dataset, we continue to be hugely grateful to Alex Bash, our under-
graduate research assistant at the University of Washington, and who, it seems,
has himself become a dad!
Notes
1. We use italics in the body of the chapter to mark out instances of taboo
language use.
2. The FCC-Supreme Court debate about fuck and shit continued rolling for
years; notoriously conservative judge Anton Scalia referring to ‘foul-
mouthed glitterati from Hollywood’ in a 2009 opinion and, in a 2012
opinion, stating that ‘the government is entitled to insist upon a certain
modicum of decency’.
3. Concerning these and other points of taboo language, the marvellous
Language Log can always be relied on for titbits and insights: http://lan-
guagelog.ldc.upenn.edu/
References
Aldridge, M., & Evetts, J. (2003). Rethinking the concept of professionalism:
The case of journalism. The British Journal of Sociology, 54, 547–564.
Allan, K., & Burridge, K. (2006). Forbidden words: Taboo and the censoring of
language. Cambridge; New York: Cambridge University Press.
Baker, P. (2004). ‘Unnatural acts’: Discourses of homosexuality within the
House of Lords debates on gay male law reform. Journal of Sociolinguistics, 8,
88–106.
Bell, A. (1991). The language of news media. Oxford: Blackwell.
Billig, M. (1997). The dialogic unconscious: Psychoanalysis, discursive psychol-
ogy and the nature of repression. British Journal of Social Psychology, 36,
139–159.
Bourdieu, P. (1991). Language and symbolic power (G. Raymond & M. Adamson,
Trans). Cambridge: Polity.
Bourdieu, P., & Wacquant, L. (1992). An invitation to reflexive sociology. Chicago:

University of Chicago Press.
Cornog, M., & Perper, T. (1991). More on the ‘X-Word’. American Speech, 66,
325–330.
Coupland, N., & Jaworski, A. (2004). Sociolinguistic perspectives on metalan-
guage: Reflexivity, evaluation and ideology. In A. Jaworski, N. Coupland, &
D. Galasinski (Eds.), Metalanguage: Social and ideological perspectives
(pp. 15–52). Berlin: Mouton de Gruyter.
Crystal, D. (1998). Language play. Chicago: University of Chicago Press.
De Smedt, E. (2012). Professionalism in political broadcast talk: The perfor-
mance of a distancing journalistic self in formatted pre-election debates.
Discourse, Context & Media, 1, 114–122.
Ekström, M., & Johansson, B. (2008). Talk scandals. Media Culture Society, 30,
61–79.
Flemming, L., & Lampert, M. (2011). Introduction: Beyond bad words.
Anthropological Quarterly, 84, 15–40.
Foucault, M. (1978). The history of sexuality. Volume I. New York: Random
House.
Freud, S. (1950[1913]). Totem and taboo: Some points of agreement between the
mental lives of savages and neurotics. New York: W. W. Norton.
Goffman, E. (1981). Forms of talk. Philadelphia: University of Pennsylvania
Press.
Hall, S. (2013). The spectacle of the ‘other’. In Representation (pp. 215–287).
London: Sage.
Hanks, W. F. (2005). Pierre Bourdieu and the practices of language. Annual
Review of Anthropology, 34, 67–83.
Irvine, J. T. (2011). Leaky registers and eight-hundred-pound gorillas.
Anthropological Quarterly, 84, 15–40.
Irvine, J. T., & Gal, S. (2000). Language ideology and linguistic differentiation.
In P. V. Kroskity (Ed.), Regimes of language: Ideologies, polities, and identities
(pp. 35–84). Santa Fe, NM: School of American Research Press.
Jaworski, A. (1993). The power of silence: Social and pragmatic perspectives.
Newbury Park, CA: Sage.
Jaworski, A. (1998). The silence of power and solidarity. Studia Anglica
Posnaniensia, 33, 141–152.
Read, A. W. (1964). A type of ostentatious taboo. Language, 40, 162–166.
Schröter, M. (2013). Silence and concealment in political discourse. Amsterdam:
John Benjamins.
Seizer, S. (2011). On the uses of obscenity in live stand-up comedy. Anthropological

Quarterly, 84, 209–234.
Stallybrass, P., & White, A. (1986). The politics and poetics of transgression.
London: Methuen.
Thurlow, C. (2006). From statistical panic to moral panic: The metadiscursive
construction and popular exaggeration of new media language in the print
media. Journal of Computer Mediated Communication, 11, 667–701.
Thurlow, C. (2010). Speaking of difference: Language, inequality and intercul-
turality. In R. Halualani & T. Nakayama (Eds.), Handbook of critical intercul-
tural communication. Oxford: Blackwell.
Thurlow, C. (2014). Disciplining youth: Language ideologies and new tech-
nologies. In A. Jaworski & N. Coupland (Eds.), The discourse reader (3rd ed.,
pp. 481–496). London: Routledge.
Thurlow, C. (2018). Critical discourse studies in/of applied contexts: Missed
opportunities, fraught possibilities. In J. Flowerdew & J. Richardson (Eds.),
Handbook of critical discourse analysis (pp. 328–341). London: Routledge.
Thurlow, C., & Jaworski, A. (2010). Silence is golden: Linguascaping, anti-
communication and social exclusion in luxury tourism representations. In
A. Jaworski & C. Thurlow (Eds.), Semiotic landscapes: Image, text, space
(pp. 187–218). London: Continuum.
Van Dijk, T. (1993). Elite discourse and racism. Newbury Park, CA: Sage.
Waugh, L. R. (1995). Reported speech in journalistic discourse: The relation of
function and text. Talk & Text, 15, 129–173.
Woolard, K., & Schieffelin, B. B. (1994). Language ideology. Annual Review of
Anthropology, 23, 55–82.
12
The Use of No Comment by Suspects
in Police Interviews
Joanna Garbutt
12.1 Introduction
This chapter focuses on the metadiscourse of silence, with silence defined
as the absence of talk for certain purposes (Alagözhi & Sahin, 2011) and
as such is a discursive act whereby ‘… silencing results from an act of
language where language is used in order to enable some kinds of expres-
sion and to disable others’ (Thiesmeyer, 2003, p. 11). By focusing on the
use of the phrase no comment in the police interview context, I analyse
how officers respond to these metadiscursive silences used by the suspect,
identifying how officers manage ongoing questioning strategies while
adhering to the needs of those listening to the interview subsequent to
the event.
The right to silence is an important part of the caution rights which are
provided to suspects at the beginning of the police interview in England
and Wales and found in other jurisdictions (e.g. the Miranda warnings in
the US). The suspect will often use the phrase no comment to invoke this
J. Garbutt (*)
London, UK

330 J. Garbutt
right. In this chapter, I will analyse examples from UK police interviews

where suspects have used this phrase, either consistently or inconsistently,
and investigate how officers manage ongoing interaction and the account
production process when faced with the suspect’s resistance in respond-
ing to questions.
While forcing people to speak imposes certain expressions or wording on
individuals (Thiesmeyer, 2003, p. 8), the police interview is different as the
suspect can remain silent despite the fact this is often a dispreferred response
(Heydon, 2011). Kurzon (1995) identifies how suspects or witnesses in
legal proceedings will either provide silences which are unintentional or
that are intentional which invoke the right to silence. Such unintentional
silences may result from the suspect’s nervousness (Kurzon, 1995) or poten-
tially from cross-cultural uses of silence within conversation more generally,
such as seen by Eades (2003) when analysing how Australian aborigines’
silences were often misinterpreted as resistance during police interviews.
In England and Wales, where the right of silence is outlined in the
preliminary parts of the interview, the phrase no comment is used by the
suspect to realise this right to silence, providing a clear legally recognised
response (Stokoe et al., 2016). Invoking this right places constraints and
limitations on the officer’s institutional objective of collaboratively creat-
ing an account of the suspect’s actions and motives for potential future
legal proceedings. This absence of information impacts on the officer’s
discursive strategies, and analysing these absences provides further infor-
mation regarding how officers control and manage interview discourse in
the light of suspect resistance. The analysis will also consider what officers
infer from the suspect’s use of no comment, whereby officers display cer-
tain information or opinions for those who listen to the interview record-
ing subsequently to the interview itself. The use of no comment and silence
during the interview is shown to have mixed interpretations as regards
the guilt or innocence of the suspect.
12.1.1 The Right to Silence in the Police Interview
Since the 1984 Police and Criminal Evidence Act (PACE) in England
and Wales, a number of important procedural elements have informed
The Use of No Comment by Suspects in Police Interviews 331
and impacted on police interview interaction. An important part of this

legislation is the provision and explanation of the caution at the begin-
ning of the interview itself. The caution wording in England and Wales
in its current incarnation is:
You do not have to say anything but it may harm your defence if you do
not mention when questioned something which you later rely on in court.
Anything you do say may be taken in evidence. (Criminal Justice and
Public Order Act, Part III, 1994)
The right to silence is clearly stated in the caution but so is the potential
penalty for invoking it. How these rights are explained to suspects differs
on an interview-by-interview basis, often as a result of the assumptions
and beliefs of the interviewing officer themselves. Rock (2007) analysed
how these rights are reformulated and introduced during interview inter-
action, exploring the officers’ methods of explaining these rights.
Sometimes, the officers would provide problematic reformulations, caus-
ing difficulties for the interviewee, because as Cotterill notes, officers often
find ‘the task of gauging linguistic and comprehension level problematic’
(2000, p. 10). It was not the vocabulary of the caution that was difficult
to comprehend but the difficulty lay in the sequence of information as it
was formulated, i.e. in how these phrases are ‘thrown’ together. As a result
of transformation, we see shifts in register, the emphasis of certain assump-
tions, comprehension, attention and knowledge (Rock, 2007).
While it is important that the right to silence is explained to the sus-
pect, it is also important that officers check when suspects do invoke this
right and accept the suspect’s silence regarding certain lines of question-
ing. However, in previous literature, officers were shown to have greater
control over the interview interaction (Shuy, 1998, inter alia) and would
therefore ignore pragmatic meaning when it did not adhere to their aims
for the interview. An example of this was found when suspects used
certain speech acts to invoke the right to silence (Ainsworth, 2008).
References were made, such as ‘I think I should have a lawyer present’
which were indirect and were often ignored. Ainsworth argued that,
while indirect speech acts are tolerated and understood as performative in
everyday conversation, in police interviews, these are not performative
332 J. Garbutt
speech acts and less institutionally recognised. Heydon (2011, p. 2311)

notes that the ability to invoke the right to silence during interview results
in ‘considerable pragmatic impediments’ with the potential for such
silence being misinterpreted. Citing Blimes (1988), she argues that
remaining silent or responding no comment after the accusation is stated
can lead the court to interpret this response as evidence of guilt. Similarly,
silences in court are potentially damaging to defendants, as Fridland
notes how attorneys will narrate witness testimony as the primary narra-
tive ‘by fragmenting and managing through the use of particular question
forms, the alleged victim’s attempts at narration and constructing alterna-
tive “non-occurring” narrative clauses to which the alleged victim has no
possible “good” response’ (2003, p. 121).
While courts may interpret silence as evidence of guilt, under the
terms of the caution, the officer has to accept the suspect’s silence or oth-
erwise violate those terms. However, Carter (2011) identified instances in
her data where officers would laugh in reaction to no comment responses.
Carter argued that in doing so, officers were using laughter to challenge
and discredit the suspect during the interview, using indirect means and
mitigating any challenge which resulted. Such indirectness enabled offi-
cers to breach these codes of conduct and aided ‘conversational asides’,
where the officer would provide an opinion or express doubt over the
suspect’s responses. Carter also highlighted how such laughter would
hide the officer’s annoyance at a suspect’s dispreferred response, inferring
that what the suspect had said was unlikely and problematic.
As Ainsworth (2008) and Heydon (2011) both argue, invoking the
right to silence during interviews is problematic and any attempts to do so
by suspects can be dismissed by the interviewing officer. Officers are seek-
ing to fulfil institutional objectives, thereby to move suspects from resis-
tance to participation. This objective can often be at odds with the suspect’s
best defence strategy, to remain silent (Oxburgh, Myklebust, & Grant,
2010). As Kurzon (1992) notes, silence can be a powerful d iscursive strat-
egy when used by the suspect, challenging what is held to be the ‘normal’
power base in these encounters. Further difficulties arise from the fact that
the caution itself warns against using silence even though such silence is
accepted as an appropriate response. It is also seen as somehow uncoopera-
tive, possibly indicating that the suspect has something to hide.
Not to cooperate with the authorities—the police, the court—seems to be

a characteristic of groups that live either on the boundary between law and
crime, or beyond that boundary—well into the area of crime… This lack
of cooperation—behaviour that goes against the accepted social code—is
suspect. (Kurzon, 2008, p. 177)
12.1.2 The Use of ‘No Comment’
No comment provides an unambiguous statement which invokes this

right to silence and is conventionally recognised within legal institutional
interaction. As such, it performs a speech act, that the suspect is provid-
ing this non-response in assuming these legal rights. Previous research
looking directly at the use of the phrase no comment in police interviews
is limited, though Stokoe, Edwards, and Edwards (2016) consider how
the no comment response can impact on interview interaction, specifi-
cally, how effective using no comment is when invoking the right to
silence. They drew on a large corpus of 125 interviews, identifying the
‘sequential placement, action orientation and design and uptake’ of the
use of no comment, particularly in regard to the officers’ responses. As
such, there were two broad areas they identified: those uses of no comment
which were spontaneous from the suspect; and those uses of no comment
which were informed and initiated by the solicitor present.
First, it was noted by the authors how officers would respond to no
comment with the ‘smiley’ voice inflection, thereby providing some ironic
response or in other ways highlighting no comment as a dispreferred
response. This type of response to no comment is similar to that noted by
Carter (2011) where officers laughed. Such responses by the officers were
often found at the beginning of the interview, where suspects would
begin using the no comment response to questions relating to the offence.
Such ironic receipts by the officer were seen as a result of the ‘interaction-
ally peculiar, almost absurd, practice of continuing with a series of ques-
tions to which it is understood that no substantial answer will be
forthcoming’ (Stokoe et al., 2016, p. 295).
In contrast, officers were also shown to note the significance of no com-
ment, acknowledging not just the response itself but also the ‘stance’
334 J. Garbutt
taken by the suspect during this process. No comment was important as

‘the action of not commenting’ as a ‘device for doing not-answering’
(Stokoe et al., 2016, p. 311). In responding to no comment, officers were
shown to mark receipt and further responded with no problem to indicate
a ‘normative expectation’ rather than attempting to further pursue a line
of enquiry. As such, suspects would often consistently use no comment to
respond to questions related directly to the offence and attempt to fore-
stall any specific lines of enquiry.
A comparison with previous research on ‘not-answering’ in media
interviews indicates that suspects are more restricted in how they respond
to such questions than public officials. When remaining silent for a given
topic, politicians would introduce other topics or subtopics, explicitly
noting how they ‘don’t want to rake this up’ or ‘don’t want to talk about
these issues’ (Alagözhi & Sahin, 2011, p. 3012). Rasiah (2010), in analys-
ing interviews with politicians in Australia, saw how there were numer-
ous methods to evade questions, remaining silent and leaving gaps or
absences, using many overt or cover practices such as justifying shifts to
other topics or repeating the terms of the question. Clayman (1993)
noted in using Conversational Analysis (CA), how types of reformulisa-
tions were often very specific and detailed. In his research, journalists
would either repeat questions or provide the opinion that the interviewee
is being evasive. Such responses are not possible for police officers unless
they risk contravening interview procedures.
While previous research identifies how officers’ reactions to no com-
ment can potentially contravene the rights specified in the caution (Carter,
2011) or how the way in which these rights are provided in the caution is
problematic (Stokoe et al., 2016), this chapter investigates the impact of
the phrase no comment on immediate subsequent interaction and what
this non-response and absence of information mean for how officers
proceed with account-setting talk, examining how officers will pursue
further information or topic switch onto other procedural matters. Such
choices seemed individual to the officer and not as a result of the suspect’s
consistent or inconsistent use of no comment. As such, the use of silence
by suspects in police interviews provides further insight into how partici-
pants respond when faced with a dispreferred response of silence.
12.2 Data and Method

The data was taken from a corpus of 22 police interviews which had been
collected by the author from one UK police constabulary as part of a
wider study concerned with police-suspect interaction. All the interviews
were with suspects and analysed for incidences of the use of no comment.
No comment was found to occur in seven interviews, either used inconsis-
tently by the suspect in response to certain questions during the interview
(four interviews) or used consistently in response to all questions relating
to the offence itself (three interviews). The decision to analyse no com-
ment rather than silences in general, where suspects did not speak, was
taken because it metalinguistically indicates intentional silences (Kurzon,
1995) but also because the phrase no comment was frequently seen to
occur within the data. The suspect not speaking at all in response to par-
ticular questions was very rare.
The interviews themselves were for relatively minor offences (such as
shoplifting or minor criminal damage). The initial recordings were made
by officers during their investigations, between 2005 and 2007, onto cas-
sette tapes. The interviews varied in length from 5 to 45 minutes though
it should be noted that the shorter interviews often contained the use of
no comment consistently. These tapes were then reviewed and transcribed
using a simplified version of Jefferson’s transcription system (1984). A
key to this transcription is given at the end of this chapter (see Appendix 2).
All personal information, such as names or locations, were removed dur-
ing the transcription process. The interviews were anonymised so no pri-
vate data was held and alternative names provided when discussed during
the interview.1 Any preliminary elements of the interview were omitted
from the transcripts where suspects provided such private information.
The tapes were listened to numerous times to ensure the accuracy of the
resulting transcription.
The information regarding the seven interviews which contained inci-
dences of the suspect using the no comment phrase is provided in Appendix
1. It must be noted that it is clarified in which interviews a solicitor was
present as this would often have a direct impact on whether suspects
336 J. Garbutt
would use the no comment response, according to Stokoe, Edwards, and

Edwards (2016).
The method of analysis uses tools from Conversation Analysis (CA).
An applied rather than pure CA approach is adopted to ensure a greater
appreciation of the process of the police interview is captured within the
analysis, for example, the effect of distal recipients or DRs on interaction.
This approach draws on ten Have’s model of CA (2007) and on Sacks,
Schegloff, and Jefferson (1974) and Schegloff, Jefferson, and Sacks
(1977). This includes:
1. Turn taking: how turn construction units (TCUs) and turn relevant
placements (TRPs) are organised.
2. Sequences: how adjacency pairs are organised and lead to sequence
expansion, where one interaction leads onto another.
3. Repairs: the organised strategies which speakers use to deal with
potential interaction breakdown.
4. Turn design: how turns will be viewed by their potential audience and
are designed by speakers accordingly.
By using the tools of CA, a line-by-line analysis of the interview

extracts provides an insight into how no comment is used by suspects and
how officers manage interactional strategies following such statements by
the suspects. Using CA tools in the analysis of silence contributes greater
insight into the turn-by-turn process by which silence, either as verbal-
ised or non-verbalised absences, impacts on interaction as participants
coordinate their responses accordingly. In the current study, a particular
focus on turn-taking and sequence organisation is developed through
analysing the verbalised absences, the no comment phrase as used by sus-
pects, and its impact on how officers direct the following question-and-
answer sequence. The officer is shown to attempt discursive strategies
which progress interview talk to accomplish procedural tasks, such as
raising challenges or providing the alleged victim’s statement. In a tradi-
tional CA approach, researchers often focus on pauses and non-verbalised
absences, the significance of which is highlighted when no uptake is
shown and the impact this has on turn-taking during the interactional
event (Schegloff, 1968, cited in Ten Have, 2007). However, turn design,
as noted as point 4 above, takes a lead from CA but develops a greater
ethnographic understanding of what occurs in these interviews that is a
result of the institutional constraints and expectations placed on the
interactional event. How an officer responds to the suspect’s use of no
comment is important, both in legal terms but also for how interaction in
the interview is managed and how these management strategies are per-
ceived by those listening to the tape after the event. Analysing the
instances of metadiscourse about silence helps to identify how partici-
pants react to and make sense of the silence, providing evaluations, as
previously noted by Stokoe, Edwards, and Edwards (2016), but also in
terms of observations participants make regarding the ongoing process of
account creation.
12.3 Analysis
In analysing the data, two broad patterns of response from officers were
noted when the suspect used the phrase no comment. These were:
1. Officers would attempt to elicit a further response from the suspect

regarding their original line of questioning beyond no comment.
2. Officers would accept the use of no comment but continue to question
the suspect on other relevant topics and/or continue with other inter-
view procedures.
These two different types of response are now considered separately in

this analysis, analysing how the officer either attempts to fill the absence
by obtaining a further response from the suspect or accept the absence in
certain parts of the account and move on to completing other procedural
requirements. I will also identify how the use of no comment by the sus-
pect varies from other methods of remaining silent on certain topics, for
example, such as those found in media interviews with politicians or
other public figures (Alagözhi & Sahin, 2011; Clayman, 1993; Rasiah,
2010). As such, this analysis will examine the impact of no comment on
338 J. Garbutt
subsequent interview interaction, where absence is tolerated as identified

in the right to silence and/or where the officer still attempts to obtain
further responses on behalf of future court proceedings.
12.3.1 Attempts to Fill the Absence in Account Detail
Despite suspects using the phrase no comment to certain questions, offi-

cers were shown to continue certain lines of questioning subsequent to
these non-responses. Officers have a professional obligation to continue
questioning after the suspect has invoked their right to silence and it
was found that such responses from the suspect are often dispreferred
(such as shown in Carter, 2011). This dispreference arises from the
impact the lack of a full response has on the account-setting process.
The use of no comment was shown to create a gap or absence in these
accounts which led to ambiguity. Such ambiguity may indicate guilt
(Heydon, 2011) but it does not provide an account for the officer to
challenge or a confession. The analysis of the extracts in this section
examines how officers attempt to fill these gaps or absences in account
detail.
Extract 1 provides an example of how an officer continues to ask ques-
tions following a no comment response regarding an injury the suspect has
sustained. Instead of accepting the no comment response or providing an
implicit challenge (such as laughter in Carter, 2011 or sarcasm, Stokoe
et al., 2016), the officer maintains the line of questioning in order to lead
on to identifying important evidential information. As such, the officer
attempts, and in this case succeeds, in obtaining further information
from the suspect, thereby filling the absence or gap in the account regard-
ing an injury sustained by the suspect.
The suspect has been arrested for burglary due to a description which
was provided by an eye witness as the suspect is believed to fit this physi-
cal description. However, the officer attempts to elicit further informa-
tion regarding an injury sustained by the suspect which would have
affected his ability to leave the scene quickly.
Extract 1 (Interview 5)
1 OF: Okay (.) erm you’re walking with a slight limp I saw [that]
2 SUS: [yeah]
3 OF: Today can you tell me about how that injury was caused?
4 SUS: No comment
5 OF: (coughs) (5.6) did you sustain that injury yesterday?
6 SUS: No comment
7 OF: When you were arrested yesterday were you carrying anything
else?
8 SUS: No comment
9 OF: Were you erm (.) walking with anything else? (.) do you
understand what I’m
10 getting at?
11 SUS: Yeah my crutch (laughs)
12 OF: Yeah so you were weren’t you?
13 SUS: Yeah
14 OF: Okay (.) so (.) where did you get that crutch from?
15 SUS: The hospital
The suspect has used no comment inconsistently, giving a full response

to some questions and no comment to others. Such responses have been
used by the suspect to indicate those questions which they do not feel are
relevant to the interview process and show resistance to certain lines of
enquiry. The silence and absence of a response are not just showing resis-
tance as relating to guilt, but are used by the suspect to object to or resist
lines of enquiry which they deem inappropriate to the investigative pro-
cess, and therefore provide some evaluative comment on the officer’s
questioning strategy (see Extract 4). The suspect is able to potentially
obtain some control over the interaction (as was also seen in Johnson,
2008, with regard to interview discourse more generally). As Kurzon
(1992) notes, using no comment provides a powerful discursive strategy,
testing the power which the officer assumes over interview interaction.
However, during this extract, the suspect moves from resistance to
cooperation. The initial question from the officer is provided with the no
comment response before a pause indicating what information is required
340 J. Garbutt
with the suspect’s initial response identified as vague, so a so-prefaced ques-

tion addresses this issue (line 12). As such, the officer adopts a confirmation-
seeking approach. Despite the previous no comment responses, the officer
enables a move into participation by directing turn transition, asking the
suspect to respond to questions within a fixed structure, either yes or no.
Although this process implies a challenge, it also shows how both inter-
locutors attempt to obtain control of the discussion. By asking for this
final confirmation in line 12, the officer provides further information for
DRs which is important after the initial no comment responses.
Such processes show the difficulty which arises when an absence or gap
in the account is not accepted but the right to silence must be, which
leads to the process of the officer obtaining a subsequent response as
shown in this extract and the others in this section. What is unusual in
this extract is the officer is attempting to obtain further information
regarding shared knowledge. This process is shown in lines 9–10 where
the officer implies what needs to be stated for DRs.
In the second extract, the officer’s continued questioning enables interac-
tion to fill a gap or absence in the account. In the previous extract, this elicita-
tion was shown to occur through a stepwise process as the information which
was elicited subsequent to no comment was evident to those in the room.
However, the officer in Extract 2 manages to obtain a confession by reformu-
lating the question slightly so the topic is revisited and the gap or absence is
highlighted. The confession arises because of the ambiguity and inconsis-
tency of the suspect’s account, particularly his words at the time of the arrest.
The suspect has been arrested for shoplifting and is explaining their
involvement and their friends’ involvement. The police officer who
arrested him had seen the group run away at the sight of the police. This
fact is discussed within the interview as an implication of their guilt. The
suspect is inconsistent in his use of no comment, choosing to respond to
some questions but not all.
1 OF: Did you see who did steal them?

2 SUS: Did I what?
3 OF: Did you see who stole them?
4 SUS: Yeah no comment though

5 (.)
6 OF: But it wasn’t you?
7 SUS: No comment
8 (2.4)
9 OF: So when you were arrested you said it wasn’t me it was Josh (.)
what did you
10 mean by that?
11 SUS: Josh stole the stolen goods I had nothing to do with it
12 OF: All of them?
13 SUS: Most of them (.) I might as well change my story then I admit
it I walked out
14 with a few stolen goods that’s it
The officer provides the initial question within line 1 which leads to a
sequence expansion. The suspect responds with mostly no comment till
later within the sequence. This no comment is slightly modified in line 2
as the suspect attempts to expand on the question previously given, clari-
fying the nature of the question for which the officer is seeking a response.
However, when the officer does clarify this question in line 3, the suspect
replies with a modified no comment response (line 4). Although the sus-
pect uses no comment, their initial response of yeah does constitute a full
reply to the officer’s question. The subsequent no comment though acts as
a barrier to further discussion and implies a resistance to elaborate on the
officer’s question in line 6, thereby attempting to control the discussion
(Kurzon, 1992).
This extract shows how the use of no comment provides some partial
resistance to the officer’s line of questioning. The further use of no com-
ment in line 7 creates a pause in the interview, thereby indicating a break
to the line of questioning so the officer introduces further information to
ascertain the suspect’s knowledge of who stole the goods, taken from the
significant statement, the words which were spoken by the suspect at the
time of the arrest. The officer uses this statement to further develop this
line of enquiry, addressing what was meant by the suspect, which further
contradicts the no comment responses provided in lines 4 and 7. Such
342 J. Garbutt
inconsistency with no comment causes issues with the resulting account as

the suspect follows these no comment responses with the fact that he saw
his friend stealing before providing a comment that the suspect had noth-
ing to do with it. The gaps in the account are subsequently filled despite
the earlier absence of information through the use of no comment. Such
inconsistencies are evident to the suspect himself which leads finally to
his confession in lines 13 and 14, where instead of allowing an absence or
gap in information so that the evidence remains ambiguous despite the
officer’s previous inferences, the suspect decides to change my story (line
13).2 This confession partly results from the officer’s persistence in a spe-
cific line of enquiry, managed through continuing to question the suspect
after the no comment response is provided but also due to the inconsisten-
cies becoming evident to the suspect himself who does not leave gaps or
absences of information.
This extract differs in challenging the suspect directly regarding the gap
in the account. This is a common process in police interviews but unusual
when the suspect has invoked their right to silence. Potentially, the offi-
cer’s continued questioning could contravene PACE guidelines, and
therefore make a subsequent confession problematic.
In the third extract, the officer continues to question the suspect fol-
lowing a no comment response but instead of using a stepwise transition
(Extract 1) or a direct challenge to the inconsistencies of the account
(Extract 2), the officer uses the terms of the caution to warn the suspect
about the potential outcome of remaining silent. This rephrasing of the
caution results in an implicit challenge in obtaining a different response
from no comment, expanding the first question into a longer sequence.
While the gap or absence in the account is not emphasised, the suspect
is unable to be completely unresponsive and is forced to interact
beyond the use of no comment, even if this is at a minimal level. The
officer goes further in attempting to obtain some response from the
suspect, moving the suspect from resistance to participation, making it
clear that the suspect is aware of the potential repercussions of remain-
ing silent.
1 OF: Okay so again erm we’ve asked relevant question you’ve not
answered any
2 questions and the court or whoever listens to this tape or reads
these
3 transcripts will be able to wonder why that is
4 (.)
5 OF: Do you understand that?
6 (.)
7 SUS: No comment
8 (.)
9 OF: Okay now it’s a question for me whether you understand it or
not you are
10 obviously entitled to say no comment but I would like to
know that you fully
11 understand the circumstances for what you you’ve answering
the questions for
12 the court will be allowed you =
13 SUS: = yes
14 OF: You understand?
15 SUS: Yes
16 OF: Okay (.) thank you
The question in line 5 is expanded, explained fully in lines 9–12, as the

suspect indicates some acceptance in line 13. While the officer does obtain
a response which is not no comment, the suspect’s resistance is still noted
despite the officer’s repeated challenges. The officer marks receipt in line
16 with okay though still marks this response as dispreferred. The confir-
mation provided only pertains to the suspect’s awareness of the potential
repercussions and no further information is provided. The officer pro-
vides a sense of closure with thank you (line 16). The officer’s final response
notes that though they may challenge, the suspect can remain silent if
they wish to, despite the potentially negative consequences. Therefore,
the gaps in the account remain due to the resistance of the suspect.
344 J. Garbutt
As in Extract 3, Extract 4 shows an example where an officer is unable

to elicit further information from a suspect despite subsequent attempts
following a no comment response. Unlike the previous extract, the suspect
highlights why they are not responding, providing a critique of the offi-
cer’s process of questioning. The gap or absence therefore is highlighted
by the suspect as not existing as the line of questioning is not relevant to
the interview itself.
The suspect has been arrested on suspicion of burglary following an
eye witness account which allegedly placed him at the scene. In a later
section of this interview, the officer questions the suspect regarding a
camera which was found at the suspect’s home, attempting to obtain
some information regarding its origin. However, the suspect resists this
line of questioning, using no comment to retain some control over interac-
tion and thereby, test the officer’s power (Kurzon, 1992).
1 OF: I’m just a little bit perplexed about why you’ve given such a
good account of
2 your actions and you’re adamant that you’re not involved in
this burglary but
3 yet when it comes to that camera you’re not saying anything
about it is there a
4 reason for that?
5 SUS: No comment
6 OF: Is it because it’s stolen?
7 SUS: No comment
8 OF: Or is it because you just feel it’s got no relevance to this
investigation?
9 SUS: No relevance to this conversation really
The officer introduces a challenge in lines 1–4 with the implication

that because the suspect has answered previous questions thoroughly, it
is ‘perplexing’ that they are not responding to the questions regarding
the camera found in their home. The challenge identifies the suspect’s
inconsistency in using the no comment response which infers that they
might have something to hide regarding the camera. Meaning is drawn

from the absence of information regarding this point (Heydon, 2011).
The caution rights indicate that the suspect does not have to respond,
only that the court will potentially infer guilt if further information is
given subsequently. However, the officer challenges the suspect’s invok-
ing of the right to silence at this point by identifying it as logically incon-
sistent with the account the suspect has previously provided.
The explanation as to why he maintains this silence is provided in line
9, that information regarding the camera has no relevance to the investi-
gation. Such challenges indicate control over what absences or gaps in the
information provided are permitted or not. The officer provides an opin-
ion which challenges the suspect’s lack of comment, and as such, the
officer designs his responses with DRs in mind. Such interactions show
how officers attempt to discourage a no comment response.
This section has shown how officers will pursue certain information
despite a no comment response from the suspect, providing insight into
how interviewing officers respond and the discursive strategies they use.
The extracts show that this occurs regardless of whether no comment is
used consistently or inconsistently by the suspect. The analysis has shown
the different methods of these attempts such as stepwise sequences
(Extract 1) and direct challenges (Extract 4). Section 12.3.2 will analyse
extracts where officers use different methods to progress within interview
interaction while tolerating these gaps in the account.
12.3.2 T
olerating the Presence of Absences
in Account Detail
While it is important to note how officers pursue lines of questioning

subsequent to a no comment response, it was noted in the data that though
officers continue to ask questions, there are often topic switches as a result
of a no comment response. This has been little discussed elsewhere as
research concentrates mostly on how officers may potentially contravene
PACE guidelines (for example, Carter, 2011). However, the data used in
this study shows how the officer will not always pursue a line of question-
ing to obtain a response from the suspect. It will also be shown how
officers often conduct such interviews as box-ticking exercises, providing
346 J. Garbutt
a list of information for the suspect but not expecting a response beyond
no comment.
Extract 5 is taken from the beginning of an interview in which the sus-
pect has used no comment consistently. The officer does not attempt to
reformulate questions (as noted in Extracts 1 and 2) but instead moves
onto the next topic. This therefore provides an example of box-ticking, a
process in the police interview highlighted by Gaines (2011), whereby
topics which need to be covered are put to the suspect for comment. No
comment, while a non-response, makes it clear that the question was asked
and the suspect decided to remain silent. The suspect has taken a particular
stance to the process of questioning or line of enquiry that the officer has
introduced, performing ‘the act of not answering’ (Stokoe et al., 2016).
The questions asked by the officer fill the absence of account detail from
the suspect such as shown in the questions regarding motivation and/or
intent (lines 13 and 18) which, while the suspect has not confessed, enables
the officer to identify possibilities why the suspect remains silent.
The suspect has been arrested on suspicion of assault. During the inter-
view, the officer shows a recording of CCTV footage of the alleged offence
taking place.
1 OF: Okay? and it is being tape recorded so it’s either read or the
tape played if it
2 should go to court okay? erm what I’m interviewing you about
today is about
3 an alleged assault in Hills Road can you tell me what hap-
pened today?
4 SUS: No comment
5 OF: No comment all right where were you going?
6 SUS: No comment
7 OF: Who with?
8 SUS: No comment
9 OF: Erm who’s the male you kicked?
10 SUS: No comment
11 OF: The male in the stripy jumper?

12 SUS: No comment
13 OF: Why did you kick him?
14 SUS: No comment
15 (.)
16 OF: Who was the male with the hat that you slapped?
17 SUS: No comment
18 OF: Okay why did you slap him?
19 SUS: No comment
20 OF: Had you been drinking?
21 SUS: No comment
Turn-taking is organised as a question-and-answer sequence, with the

officer and suspect providing first and second parts of the adjacency pairs.
Though there is a lack of engagement by the suspect, the officer still
attempts to expand the sequence, shown in lines 9–12 where the officer
elaborates on the male who is described. Such expansion indicates that
the officer wants to address the issue of who is being spoken about despite
the strong indication from the suspect that they do not wish to respond.
As a result, in this extract, and in most of the interview, the relevant ques-
tions which the officer provides are listed for DRs, and therefore it is clear
what information the suspect was not willing to provide during the inter-
view. In the third turn repeat of no comment in line 5, the officer indicates
some expectation of how the interview will unfold, and that such listing
of questions will provide the main structure of information, thereby pro-
viding an indication of what information the suspect has refused to give
(as highlighted by Stokoe et al., 2016). This sequence also indicates how
the officer is working on behalf of the court, seen particularly in lines 13
and 18 which assume the suspect’s guilt when no information has been
provided. As the suspect’s account has significant gaps, it identifies the
suspect as unwilling to provide this information. The officer fills in these
gaps through the questioning procedure they adopt with this absence of
comment highlighted for future legal proceedings.
In Extract 6, the officer is in a similar situation, that of a suspect using
no comment consistently to respond to all questions. However, instead of
348 J. Garbutt
the officer filling in the gaps or absences in the account through the ques-
tions, the officer states the alleged victim’s account, so this set of events
becomes the focus for interview talk. A similar process was discovered by
Fridland (2003) as occurring in court, where the victim’s testimony filled
in the gaps caused by the defendant’s silence.
The suspect has been arrested for assault, following an incident which
occurred at the alleged victim’s house. As with interview 13 in Extract 5,
the suspect uses no comment consistently to all questions asked in relation
to the alleged offence. What does vary is how the officer introduces case
evidence to the interview process and how the questioning is structured.
1 OF: Okay we have a statement to say you went round there to 43

South End today
2 to see someone called Becky (.) what is your relationship with
her?
3 SUS: No comment
4 OF: What did you go round there for?
5 SUS: No comment
6 OF: Okay (2.9) in the statement that we’ve got it said so that
you’ve got round
7 you’ve gone in for a cup of tea and then you’ve gone and got
in her bed (.)
8 Becky hasn’t been happy about this and has asked you to leave
9 SUS: No comment
10 OF: You haven’t got anything to say (3.2) okay also in the state-
ment it says that an
11 argument’s taken place between you guys and you’ve hit her
over the head
12 with a coat hanger (.) is this true?
13 SUS: No comment
The initial questions in lines 2 and 4 show the officer asking for infor-
mation integral to the process of creating an account: how the suspect
knew the alleged victim and why the suspect had gone to see the alleged
victim. However, following the lack of response regarding these two
points, the officer adopts a different questioning strategy to accommo-
date for the absence or gap in information which results when the suspect
remains silent. The officer changes their strategy, varying from that in
interview 13 in Extract 5, so that the alleged victim’s statement is pro-
vided to fill in the information required for what occurred. It is the expe-
riences of the victim which take the lead role within the interview
interaction rather than the suspect’s recollections of what occurred dur-
ing the incident in question.
In lines 6 and 10, the officer confirms receipt of the suspect’s no com-
ment responses and moves from closing explanation-seeking questions to
those which are confirmation-seeking. By confirming receipt with the
marker okay, tasks are shown to be complete before moving onto the next
piece of information (Gaines, 2011). However, there are variations to this
sequence such as seen in line 8 where the officer does not produce a direct
TCU in order to obtain a no comment response. In line 10, where the
officer states that you haven’t got anything to say, the use of no comment
does not necessarily mean that the suspect has nothing to say. However,
the officer reinforces the absence of information which the suspect is
refusing to provide. As the absence is filled with information from the
alleged victim’s statement, their account is provided as evidence in much
the same way as the DVD was used within interview 13 in Extract 5.
Though the officer is less direct in indicating that the silence notes guilt
(Heydon, 2011), the silence is still noted as undesirable and dispreferred
(line 10).
The two previous extracts showed how officers would not always pur-
sue certain lines of questioning when suspects provided the no comment
response consistently. The officer would either move through a list of
questions quickly or recount the alleged victim’s statement. However,
Extract 7 provides an example whereby it was not just in the suspect’s
consistent use of no comment that such switches are made.
Following the suspect’s no comment response (line 4), the officer marks
acceptance before moving on to the next part of the process (the summary
of crime-relevant information previously discussed). Using no problem in
350 J. Garbutt
such a response is seen within other research (Stokoe et al., 2016) where
it was suggested that the absence is tolerated. However, as with Extract 2,
it is the suspect who self-initiates filling in the absence in the narrative
without subsequent questioning by the officer.
The suspect was arrested for distributing a class C substance, cannabis.
The officer is attempting to obtain information regarding who the sus-
pect had been selling the cannabis to, following a discussion of the can-
nabis which had been found at the suspect’s home.
1 OF: Okay (.) erm where or how would you sell these on?
2 SUS: Where or how?
3 OF: Yeah like where would you sell them from?
4 SUS: No comment on that
5 OF: Okay erm (.) okay that’s no problem so you’ve pretty much
said to me that
6 these were located =
7 SUS: = they’re to whoever I can get rid of them to
8 OF: Yeah okay
Throughout interview 2 and Extract 7, the suspect responded to ques-

tions and confessed to the offence but decided to respond no comment to
the officer’s question in line 3 regarding who was buying the cannabis.
The question in line 3 has resulted from an expansion of the question
asked in line 1. Previously the suspect has remained silent regarding this
particular information but this is only partial resistance. As seen in
Stokoe, Edwards, and Edwards (2016), by using no comment, the suspect
takes a stance to this line of enquiry. This is acknowledged by the officer
as okay that’s no problem (line 5) before going on to pursue other account
detail. However, the suspect quickly provides an additional response to
no comment, filling this absence with the point of her not selling to any-
one specifically, rather than providing the practical details of how the
cannabis is sold. The no comment response enables the suspect to only
partially respond to the previous question (line 1) rather than the full
information which the officer has requested.
This example illustrates that importantly, when the suspect has used no
comment, the absence in the narrative is viewed as problematic by the
suspect even when inconsistencies and absences are not pursued by the
officer. In this extract, the officer already appears to be switching topic
but accepts the partial response given by the suspect (line 8) to the previ-
ous question.
In the final extract of this chapter, taken from the same interview as
that of Extract 7, another example of not pursuing a line of enquiry is
shown. However, what marks this as different is that the officer pref-
aces the question with an indication that a full answer is not expected.
Though it is implicit in all the extracts previously discussed in this
section that responses are not necessarily expected, this extract shows
this presumption of a lack of response is expected and tolerated. The
officer is even partially apologetic in marking the question itself as
inevitable.
1 OF: Anything else?

2 OF2: (hhhh) the inevitable question but where did you get the stuff
from?
3 SUS: No comment
4 (5.9)
5 OF: Okay so an overall round up
As mentioned with Extract 7, the suspect in interview 2 has been

arrested for the distribution of cannabis and has used no comment incon-
sistently during the interview, providing an otherwise full confession.
However, in the final stages of the interview after the first officer has
questioned the suspect regarding their account, the second officer is given
the floor to ask the inevitable question (line 2) regarding from where the
suspect obtained the cannabis. This question is asked after a sigh from the
352 J. Garbutt
officer marking a lack of expectation of a response regarding this informa-

tion from the suspect. The suspect does not give a partial response as seen
in Extract 7, despite being provided with the floor during a considerable
pause (line 4) and before the first officer regains the floor to provide a
brief summary.
Such absences or gaps in the account appear expected and indicate
other strategies found to cope with such absences beyond the chal-
lenges identified previously. The suspect is able to engage with their
right to silence with the officer marking receipt with okay in line 5
before proceeding into the summary section. This questioning strat-
egy goes beyond acknowledging the suspect’s stance noted in previous
literature (Stokoe et al., 2016) to highlight where certain lines of
enquiry are prefaced with an expectation of the suspect’s resistance. As
such, this lack of response and the gap within the evidential account
remain as they are with neither the officer nor suspect seeking to fill
them.
12.4 Discussion
This study has provided insight into how tools from CA, including turn
design, can be used to analyse meta discursive references to silence and
how these instances of talking about silence and absence impact on sub-
sequent interaction between participants. These methods enable an
examination of how people react and judge such absences, with the most
common meaning or inference drawn as guilt and potentially a lack of
remorse for guilty actions, creating a view of the suspect which is impos-
sible for them to correct (such as Heydon, 2011). In addition to this, it
was shown that even if suspects do provide further information subse-
quent to a no comment response, the lack of consistency in the suspect’s
responses can also be held against them and used as a challenge, such as
seen in interview 5.
By considering a turn-by-turn analysis, it is evident that the use of no
comment and therefore, a meta discursive reference to their wish to remain
silent, enables suspects to obtain some control over ongoing interaction.
What is also important is how these responses are received by the officer,
as certain responses could be seen as contravening these rights (Stokoe
et al., 2016) due to the fact that the negative repercussions of silence are
often asserted and reasserted during the interview process. Such silences
are shown to cause difficulties for the interviewing officer when forming
evidential accounts for DRs and this chapter has shown how officers
attempt to address these difficulties and ensure progression through insti-
tutional objectives.
The use of no comment showed how suspects would resist the officer’s
line of questioning, resulting in a gap or absence in the case detail dis-
cussed, and subsequently, a gap within the suspect’s evidential account.
How officers reacted to the no comment response varied, either by pur-
suing the same line of enquiry by reformulating questions and/or chal-
lenging the suspect or by switching topic. In other contexts, most
notably media interviews with public officials, questions could be
evaded and silence maintained implicitly (Clayman, 1993; Rasiah,
2010) but would also be verbalised such as ‘I don’t want to talk about
these issues’ (Alagözhi & Sahin, 2011). In comparison, the analysis of
police-suspect interviews shows how suspects need to verbalise such
silences in a standard no comment response with the risk of these
absences in the account being used against them. No comment responses
show how suspects and officers both coordinate to the absences or gaps
in the investigative account. The extracts analysed in this chapter indi-
cate that there is a mixed response to no comment from both the suspect
and the officer which varies between acceptance, partial acceptance and
non-acceptance which sometimes led to challenges. As such, the analy-
sis of no comment responses highlights what occurs when there are
absences in discourse within the police interview. Absences do not just
apply to the interview interaction as it is spoken but also to the longer-
term consequences, the recording of which clearly shows the use of no
comment and to what questions this response is provided. Therefore,
such absences are not only important for the current discussion in how
officers use certain conversational strategies to deal with such responses
or non-responses, but are also important for the legal proceedings fol-
lowing the interview itself.
354
Appendix 1
Table 12.1 Interview data
J. Garbutt
Interview Number 2 3 4 5 11 13 20
Length of Interview 18m29s 9m14s 4m48s 41m20s 25m20s 7m9s 4m35s
Offence Investigated Possession Theft Possession Burglary Theft Assault Assault
with intent of an
to supply offensive
Class C weapon,
substance affray and
possession
of a
bladed
article
Consistent/Inconsistent Inconsistent Inconsistent Inconsistent Inconsistent Consistent Consistent Consistent
use of No Comment
Present Participantsa OF1, OF2, OF, SUS, SOL OF, SUS, SOL OF, SUS, SOL OF1, OF2, OF, SUS, OF, SUS,
SUS SUS, SOL SOL SOL
Number of uses of no 3 5 2 37 76 18 11
comment by the suspect
a
OF refers to officer, so OF1 is the first interviewing officer, OF2 is the second interviewing officer. SUS refers to the
suspect. SOL refers to the solicitor.
Appendix 2: Transcription Key

(.) pauses of less than a second
(number) pauses of over a second, provided within seconds, e.g. (3.2) is
a pause of 3.2 seconds
[word] overlapping speech
- stuttering speech where a certain sound is repeated
? questioning intonation
↑ rising intonation (not necessarily a questioning intonation)
↓ falling intonation
(h) exhalation with the number of ‘h’s indicating length of breath
word at a louder volume
°word° at a lower volume
(laughs) laugh
<word> slow speech
>word< fast speech
= turns which run on with no pause between speakers
Notes
1. This condition was an important part of enabling this research. The inter-
views themselves were contributed by the police on cassette tape for tran-
scription on police office premises. I made contact with the British
Criminological Association, who confirmed that anonymised transcripts
do not constitute personal data, and as such can be used for research pur-
poses. The use of the data for research was approved by the university
ethics committee in consultation with the police constabulary regarding
the steps taken to ensure confidentiality. For further information regard-
ing anonymity in linguistic research, see Rock (2001).
2. Heydon (2005) similarly notes how officers would often provide the floor
to the suspect so that they would provide a confession due to the pressure
of needing to respond and fill the silence.
356 J. Garbutt
References
Ainsworth, J. (2008). ‘You have the right to remain silent… but only if you ask
for it just so’: The role of linguistic ideology in American police interrogation
law. The International Journal of Speech, Language and the Law, 15(1), 1–21.
Alagözhi, N., & Sahin, S. (2011). Silence as a multi-purpose speech act in
Turkish political discourse. Procedia—Social and Behavioral Sciences, 15(1),
3008–3013.
Carter, E. (2011). Analysing police interviews: Laughter, confessions and the tape.
London: Continuum.
Clayman, S. (1993). Reformulating the question: A device for answering/not
answering questions in news interviews and press conferences. Text, 13(2),
159–188.
Cotterill, J. (2000). Reading the rights: A cautionary tale of comprehension and
comprehensibility. Forensic Linguistics, 7(1), 1350–1771.
Eades, D. (2003). Participation of second language and second dialect speakers
in the legal system. Annual Review of Applied Linguistics, 23, 113–133.
Fridland, V. (2003). Quiet in the court: Attorneys’ silencing strategies during
courtroom cross-examination. In L. Thiesmeyer (Ed.), Discourse and silenc-
ing: Representing and the language of displacement (pp. 119–139). Philadelphia,
PA: John Benjamins.
Gaines, P. (2011). The multifunctionality of discourse operator okay: Evidence
from a police interview. Journal of Pragmatics, 43(14), 3291–3215.
Heydon, G. (2005). The language of police interviewing: A critical analysis.
Basingstoke, UK: Palgrave Macmillan.
Heydon, G. (2011). Silence: Civil right or social privilege? A discourse analytic
response to a legal problem. Journal of Pragmatics, 43(9), 2308–2316.
Jefferson, G. (1984). On stepwise transition from talk about a trouble to inap-
propriately next-positioned matters. In J. M. Atkinson & J. Heritage (Eds.),
Structures of social action: Studies in conversation analysis (pp. 191–222).
Johnson, A. (2008). ‘From where we’re sat…’ Negotiating narrative transforma-
tion through interaction in police interviews with suspects. Text & Talk,
28(3), 327–349.
Kurzon, D. (1992). When silence may mean power. Journal of Pragmatics, 18(1),
92–95.
Kurzon, D. (1995). Right of silence: A model of interpretation. Journal of
Pragmatics, 23(1), 55–69.
Kurzon, D. (2008). The silent witness. In J. Gibbons & M. T. Turell (Eds.),
Dimensions of forensic linguistics (pp. 161–178). Amsterdam; Philadelphia,
PA: John Benjamins.
Oxburgh, G. E., Myklebust, T., & Grant, T. (2010). The question of question
types in police interviews: A review of the literature from a psychological and
linguistic perspective. The International Journal of Speech, Language and the
Law, 17(1), 45–66.
Rasiah, P. (2010). A framework for the systematic analysis of evasion in parlia-
mentary discourse. Journal of Pragmatics, 42(3), 664–680.
Rock, F. (2001). Policy and practice in the anonymisation of linguistic data.
International Journal of Corpus Linguistics, 6(1), 1–26.
Rock, F. (2007). Communicating rights. Basingstoke, UK: Palgrave Macmillan.
Sacks, H., Schegloff, E. A., & Jefferson, G. (1974). A simplest systematics for
the organization of turn-taking for conversation. Language, 50(4), 696–735.
Schegloff, E. A., Jefferson, G., & Sacks, H. (1977). The preference for self-
correction in the organization of repair in conversation. Language, 53(2),
361–382.
Shuy, R. (1998). The language of confession, interrogation and deception. London:
Sage.
Stokoe, E., Edwards, D., & Edwards, H. (2016). “No comment” responses to
questions in police investigative interviews. In S. Ehrlich, D. Eades, &
J. Ainsworth (Eds.), Coercion and consent in the legal process: Linguistic and
discursive perspectives (pp. 289–318). Oxford: Oxford University Press.
Ten Have, P. (2007). Doing conversation analysis: A practical guide (2nd ed.).
London: Sage.
Thiesmeyer, L. (2003). Introduction. In L. Thiesmeyer (Ed.), Discourse and
silencing: Representing and the language of displacement (pp. 1–37).
Philadelphia, PA: John Benjamins.
13
Conspicuous by Presence: The Empty
Signifier ‘Interdisciplinarity’
and the Representation of Absence
Dorte Madsen
13.1 Introduction
Absence of precisely what and from where is the highly relevant question
asked by Partington (2014). The different types of absences identified and
discussed in the literature (Partington, 2014; Schröter & Storjohann,
2015; Taylor, 2013, 2014) are important for corpus-assisted studies and
corpus linguistics and can be valuable in locating and quantifying absence.
However, such approaches seem to depend on linguistic data to identify
absence and presence. But what about the cases where the linguistic data
represent an ontological absence, for instance, in the form of myth or social
imaginary? Discourses include imaginaries (Chiapello & Fairclough,
2002, p. 195) described as ‘representations of how things might or could
or should be’ and as ‘projections of possible states of affairs, “possible
worlds”’. In the same vein, Glasze (2007, p. 661) addresses the fascination
of myths, such as e. g., the ‘promised land’ or the ‘ideal society’ referring
to the ‘perception or intuition of a fullness that cannot be granted by the
D. Madsen (*)
Copenhagen Business School, Copenhagen, Denmark

360 D. Madsen
reality of the present’ (Laclau, 1990, p. 63). Myths and imaginaries are
conceived of as an ‘absent totality’ (Laclau, 1996a, 1996b, p. 42).
Representations of an absent totality may well be present in the lin-
guistic data, but what they represent is an ontological absence. Thus,
from a general social science perspective, what is lacking ontologically
should also be accounted for when theorizing absences; social imaginaries
‘provide a horizon for meaning and action that is structured around ten-
dentially empty and essentially ambiguous signifiers’ (Laclau, 1990a,
p. 65; Torfing, 1999, p. 115). We may have narratives that promise ‘a
fullness-to-come’ (Glynos, Howarth, Norval, & Speed, 2009). These
absences are meaningful in that ‘projections of “possible worlds”’ for
instance, in the form of discussing ‘the ideal society’ or ‘the unity of sci-
ence’ may constitute new spaces of representation, spaces in which it is
possible to discuss something as if it exists. Jørgensen and Phillips (2002)
explain:
We continuously produce society and act as if it exists as a totality, and we

verbalise it as a totality. With words like ‘the people’ or ‘the country’ we
seek to demarcate a totality by ascribing it an objective content. But the
totality remains an imaginary entity. (p. 39; italics in original)
In this quote, ‘the people’ and ‘the country’ are empty signifiers. Empty
signifiers signify an absence of this totality. Other examples of empty
signifiers often quoted in the literature are, for example, ‘liberation’, ‘rev-
olution’ (Torfing, 1999, p. 176), ‘full employment’, ‘equality’, ‘liberty’
and ‘security’, as mentioned by Dahlberg (2011, p. 43), and ‘integration’
(Zienkowski, 2017). Empty signifiers are so over-coded with meanings
that they mean everything and nothing at the same time (Torfing, 2005,
p. 301). As formulated by Zienkowski (2017):
They mean everything in the sense that our identities seem to depend on
their realisation. They propel us towards action and political mobilisation.
And they mean nothing in the sense that they signify ideals that can never
be fully realised. (p. 54)
This chapter uses the example of ‘interdisciplinarity’ as an empty signi-

fier based on the literature in the field of interdisciplinarity studies, a field
Conspicuous by Presence: The Empty Signifier… 361
that is seen as politicized because of the frequent hijacking of ‘interdisci-

plinarity’ for purposes that are not necessarily relevant to the scientific
field. Thus it is the politicized nature of this field that warrants the use of
a political theory of discourse. However, a signifier can also be fixed or
floating, and as it is only the empty signifier that represents absence, the
challenge is how to distinguish the empty signifier from a floating or
(partially) fixed signifier. The form of the signifier is identical in all three
cases, but their function is not. This chapter is an exploration of the chal-
lenges of theorizing absence through the identification of the empty
signifier.
In the context of this introduction it seems appropriate to briefly
sketch out the major analytical distinctions that this chapter is based on.
The chapter theorizes absence and its representation in the empty signi-
fier within poststructuralist discourse theory which generally subscribes
to a broad understanding of discourse as ‘a relational system of signifying
practices that … provides a contingent horizon for the construction of
any meaningful object’ (Torfing, 2005, p. 8). Discourse theory, according
to Laclau and Mouffe and the so-called ‘Essex School’ (Phelan &
Dahlberg, 2011), and in contrast to (Critical) Discourse Analysis, builds
on a social ontology which, in turn, implies that its point of departure is
signifying practices rather than linguistic analysis. Thus, this chapter does
not as such address the absence of language but focuses on absence as an
ontological category. Signifying practices involve different means of rep-
resentation, i.e. a signifier that is either (partially) fixed, floating or empty;
while the means of representation and their function are different, their
linguistic manifestation, i.e. their form is identical. And as it is only the
empty signifier that can represent absence, the challenge is how to distin-
guish the empty signifier from signifiers which are either floating or (par-
tially fixed). These analytical distinctions are illustrated in Fig. 13.1.
This chapter provides a number of contributions. First, in using Laclau
and Mouffe’s discourse theoretical framework, it contributes to discourse
studies’ line of enquiry into absence by offering a model that can serve as
a guide to analyze how absence relates to hegemony by way of empty
signifiers. Second, the chapter provides a methodological contribution in
exploring the challenges of identifying the empty signifier as distinct
from the floating and the fixed signifier. So far, linguistic aspects of Laclau
362 D. Madsen
Linguistic
manifestion
manife
f stion signifier
f
Sig
i nifier’s
Signifier’s
Analytical distinction
fform
fo r
rm
Possible
means of ffixed
(partially) fi xed floating
ffloating signifier
f empty
emp
m ty signifier
f
representation signifier
f
Signifier’s
Sig
i nifier’s
ffunction
fu nction
Social
ontology ABSENCE
Fig. 13.1 Analytical distinction between a signifier’s form, its function and
absence as an ontological category
and Mouffe’s conceptual framework remain under-researched. This chap-

ter differentiates, as indicated in Fig. 13.1, not only between the signifier
and the signified but also between the form and the function of the signi-
fier, in an attempt to isolate the linguistic manifestation of the sign in
preparation for connecting the discourse theoretical framework more
readily to the use of text corpora and corpus analysis tools. And, further,
suggesting a more explicit elaboration of the rather complex conceptual
framework of Laclau and Mouffe’s poststructuralist discourse theory, the
study also contributes to the ongoing discussion on the methodological
challenges in this framework (Zienkowski, 2012, 2017). Third, empiri-
cally, the study aims to trace differences or conflicts between concrete
practices and the logics of signification in academic discourse.
The chapter is organized as follows. In Sect. 13.2, the empirical field of
interdisciplinarity studies is briefly presented, highlighting how ‘interdisci-
plinarity’ as a signifier may represent an ‘absent fullness’, such as ‘innova-
tion and surprise’ or ‘the unity of science’. The discourse theoretical
framework of Laclau and Mouffe is introduced in Sect. 13.3, and the dif-
ferences and similarities between their social theory of discourse and other
approaches to discourse are addressed. Section 13.4 takes up the challenge
of investigating the logics of signification, identifying the possible signifieds
of the signifier ‘interdisciplinarity’, including absence, and the workings of
the logics of equivalence that imply a loss of meaning. The conceptual argu-
ment—how absence is represented by an empty signifier—is developed
throughout the analysis and summarized in Sect. 13.5 that also develops a
model of signification and the logic of hegemony (Fig. 13.2).
13.2 ‘Absent Fullness’ in the Empirical Field

of Interdisciplinarity Studies
All of them speak, or appear to be speaking, of one and the same thing; …
But in their totality and their variety they form neither a composite work
nor an exemplary text, but rather a strange contest, a confrontation, a
power relation, a battle among discourses and through discourses …
(Foucault, 1975, p. x)
This quote borrowed from Foucault’s Foreword to the Riviere dossier

may illustrate the impression a researcher typically gets when meeting the
vast literature on ‘interdisciplinarity’. The empirical field of interdiscipli-
narity studies is characterized by a lack of common definitions, ambigui-
ties and different understandings of ‘interdisciplinarity’. As noted by
Frickel, Albert, & Prainsack (2016), interdisciplinarity means different
things to different people, ‘it is generated through different practices in
different areas of research; and it is used in different ways by different
groups with different interests, goals, and expectations’ (p. 8), and they
relate the heterogeneity to the lack of a common definition of the con-
cept. Klein (1996, 2010) repeatedly refers to interdisciplinarity as a plu-
ralistic idea (Klein, 2010, p. 8), and Jacobs (2013) addresses the ‘plethora
of terminology’ used by interdisciplinarians and characterizes discussions
in this area as sometimes ‘plagued by an overabundance of terms.’ Frickel
et al. (2016) suggest the identification of ‘conceptual ambiguities that can
impede understanding and theory development’.
From the perspective of discourse theory, these ambiguities point us
in the direction of the operation of different discourses that compete to
fix the meaning of ‘interdisciplinarity’. Madsen (in review) identifies two
different macro discourses that struggle to invest ‘interdisciplinarity’
with meaning. One major challenge for investigating discourses on
364
Filling Filling Filling

D. Madsen
relative
fixation
(sedimen-
tation)
of meaning
Emptying Emptying Emptying
impossibility of
ultimate fixation
Absence
Remainder of particularity Tendentially empty

Particular Universal
Fig. 13.2 Signification and the logic of hegemony

‘interdisciplinarity’ is to delimit ‘the scholarly literature’ from academic

discussions more broadly conceived and thus to identify what may be
seen as the scientific field (Bourdieu, 1999) and theorized as an order of
discourse (Chouliaraki & Fairclough, 1999; Fairclough, 1992). The use
of ‘interdisciplinarity’ is widespread, and the following quote serves to
illustrate some of its most common connotations:
Interdisciplinarity is one of the great paradoxes of the research community

today. Given the knowing nods all around when the subject is raised,
apparently everyone embraces the concept. In most academic circles, sim-
ply speaking the word can warm a room, synonymous as it has become
with all things modern and creative and progressive about science. (Caruso
& Rhoten, 2001, p. 5)
Weingart (2000) is one of the few authors within the field who
addresses connotations and highlights how the object ‘interdisciplinarity’
has been persistently articulated to and thus has come to connote ‘inno-
vation’. Weingart states: ‘interdisciplinarity (or transdisciplinarity and
similar derivatives) is proclaimed, demanded, hailed, and written into
funding programs, but at the same time specialization in science goes on
unhampered, reflected in the continuous complaint about it.’ And he
goes on to analyze the polarized structure of interdisciplinarity vs. special-
ization or disciplinarity: ‘Disciplines carry the connotation of and are
valued (!) as being static, rigid, conservative, and averse to innovation.
Interdisciplinarity carries the connotation of and is valued as being
dynamic, flexible, liberal, and innovative’ (2000, p. 29). Weingart
concludes:
The idea of interdisciplinarity as the mode of innovation and progress has

taken the place of the promise of the unity of science … Interdisciplinarity
is not the promise of ultimate unity, but of innovation and surprise by way
of recombining of different parts of knowledge, no matter which.
(Weingart, 2000, p. 41)
Thus, the conclusion offered by Weingart—interdisciplinarity as the

promise of innovation and surprise—represents a ‘fullness’ that is
unachievable, as was the promise of ‘the unity of science’. As formulated
366 D. Madsen
by Laclau (1996a, 1996b, p. 53) ‘although the fullness and universality of

society is unachievable, its need does not disappear: it will always show
itself through the presence of its absence’. Therefore, the premise of this
chapter’s analysis of ‘interdisciplinarity’ is that the presence of ‘interdisci-
plinarity’ as a signifier, may represent an ‘absent fullness’, be it ‘innova-
tion and surprise’ or ‘the unity of science’. Given that the aim of this
chapter is to theorize absence, a choice is made to focus on one of the two
macro discourses identified in Madsen (in review), namely, the discourse
that Stone (2013) refers to as ‘the dominant, epistemologically oriented
approach’. This approach adheres to the following definition of interdis-
ciplinarity by the Committee on Facilitating Interdisciplinary Research,
National Academy of Sciences (subsequently referred to as the ‘NAS
definition’):
Interdisciplinary research (IDR) is a mode of research by teams or indi-

viduals that integrates information, data, techniques, tools, perspectives,
concepts, and/or theories from two or more disciplines or bodies of special-
ized knowledge to advance fundamental understanding or to solve prob-
lems whose solutions are beyond the scope of a single discipline or area of
research practice. (Committee on Facilitating Interdisciplinary Research,
National Academy of Sciences, (2004, p. 2)
This discourse is based on the premise of integration of knowledge, the

competing macro discourse is not, and for the purposes of this chapter,
the contestation of this ‘dominant discourse’ is only addressed where rel-
evant for the analysis of absence. It should be noted that this definition
serves as an example of an attempt at fixing the meaning of the signifier
‘interdisciplinarity’, which, however, because it is contested by a compet-
ing discourse, can only be a partial fixation. Bruun, Hukkinen,
Huutoniemi, and Klein (2005) developed a taxonomy for interdisciplin-
ary research explicitly distinguishing what they call ‘interdisciplinarity in
the specific sense’ from ‘interdisciplinarity in a generic sense’ to address
‘the double meaning of the concept of interdisciplinarity’ (p. 81). This
chapter uses this distinction as a stepping stone for the analyses that fol-
low, and with the note that in discourse theoretical terms, this distinction
would be a distinction between the particular and the universal.
13.3 Discourse Theory as a Social Ontology

This chapter mainly builds on Ernesto Laclau and Chantal Mouffe’s dis-
course theoretical framework (Laclau, 1990, 1993, 1996a, 1996b, 2001,
2005; Laclau & Mouffe, 1985, 1987) that has its starting point in the
poststructuralist idea that discourse constructs the social world in mean-
ing, and that, owing to the fundamental instability of language, meaning
can never be permanently fixed (Jørgensen & Phillips, 2002, p. 6).
Different discourses are engaged in a constant discursive struggle with
one another to fix the meanings of language in their own way, that is, to
achieve hegemony, the dominance of one particular perspective (Jørgensen
& Phillips, 2002, p. 7). A key advantage of Laclau and Mouffe’s theoreti-
cal framework is its sensitivity to conflict and struggle over meaning and
identity (Walton & Boon, 2014, p. 352). Laclau and Mouffe (1985)
assume ‘the primacy of the political’, the concept of politics referring to
the manner in which we constantly constitute the social in ways that
exclude other ways (Jørgensen & Phillips, 2002, p. 36). Overall, their
approach assumes a focus on ‘the social’ as a horizon of discourse.
Discourse theory draws on the work of Saussure and structural linguis-
tics, and a post-Saussurian ontology of signification (Glynos & Howarth,
2007, p. 143). The unit of analysis is the sign, which, according to
Saussure, consists of two sides: the form (signifier) and the content (signi-
fied). The relation between the two is arbitrary (Chandler, 2002; de
Saussure, 1981/2011). One of the basic principles in the analysis of lin-
guistic signs is that in language there are only differences, with no positive
terms (de Saussure, 1981, p. 120). ‘All identities within the linguistic
system of signs are therefore conceived in terms of relational and differ-
ential values’ (Torfing, 1999, p. 87). As noted by Walton and Boon
(2014, p. 359): ‘At the heart of any Laclau and Mouffe informed dis-
course analysis is the premise that a full and complete matching of a sig-
nifier (the word) to a signified (meaning) is impossible.’ In poststructuralist
theory, signs still acquire their meaning by being different from other
signs, but the structure becomes changeable, and the meanings of signs
can shift in relation to one another (Jørgensen & Phillips, 2002, p. 11).
For the analysis of concrete discourses, important focal points that char-
acterize Laclau and Mouffe’s discourse theory that are important for ana-
368 D. Madsen
lyzing floating and empty signifiers are: the unifying effect of nodal points
(Torfing, 1999, p. 96), the relations of difference and equivalence, and
the workings of different kinds of overdetermination, together with the
concept of articulation. These points will be addressed in turn below.
Meaning is fixed around certain nodal points (Laclau & Mouffe, 1985,
p. 112). A nodal point is a privileged sign around which the other signs
are ordered; the other signs acquire their meaning from their relationship
to the nodal point (Jørgensen & Phillips, 2002, p. 26). For example
‘interdisciplinary research’ in the quote from NAS in Sect. 13.2, acquires
its meaning from ‘integration’ as a nodal point. That is, an entity (dis-
course, identity or social space) is always established relationally. In our
example, according to the definition above, the identity of ‘interdisciplin-
ary research’ is established in relation to ‘integration’. When exactly this
meaning is ascribed to ‘interdisciplinary research’, all other possible
meanings that the sign could have had are excluded (in terms of this spe-
cific discourse). Therefore, as Jørgensen and Phillips explain, ‘a discourse
is a reduction of possibilities. It is an attempt to stop the sliding of the
signs in relation to one another and hence to create a unified system of
meaning’ (2002, p. 27). Relations can be of difference or equivalence.
The logics of equivalence and the logics of difference (Howarth, 2000)
variously refer to the construction of identities as equivalent and/or dif-
ferent. According to Laclau and Mouffe, the logic of equivalence is a logic
of the simplification of the social space, while the logic of difference is a
logic of its expansion and increasing complexity (Laclau & Mouffe, 1985,
p. 117).
Signifying elements, i.e. signs, can be either moments or elements. And
it is only moments that can be privileged signs in the form of nodal points.
Elements, as opposed to moments, are the signs whose meanings have
not yet been fixed; they are polysemic (Jørgensen and Phillips (2002,
p. 27), and have the status of floating signifiers which can also be described
as elements that are particularly open to different ascriptions of meaning
(Laclau, 1990, p. 28, 1993, p. 287; Jørgensen & Phillips, 2002, p. 28),
and are therefore the signs that ‘different discourses struggle to invest
with meaning in their own particular way’ (Jørgensen & Phillips, 2002,
p. 28). The empty signifier, in turn, is ‘a pure signifier without the signi-
fied’ (Žižek, 1989, p. 97). Howarth (2015) describes how the notion of
an empty signifier supplements the role of nodal points in the partial fixa-
tion of meaning and argues: ‘… if floating signifiers are ideological ele-
ments that are not securely fixed in a particular discourse and can thus be
constructed in diverse ways, then empty signifiers are points of fixation
that can hold together multiple and even contradictory demands in a
precarious unity (Laclau, 1995, 1996a, 1996b, 2004)’ (Howarth, 2015,
p. 12).
Articulation is the practice of creating and temporarily fixing meaning
(Laffey & Weldes, 2004). It is articulation that establishes a relation
among elements (Torfing, 1999, p. 298). Laclau and Mouffe define artic-
ulation as ‘any practice establishing a relation among elements such that
their identity is modified as a result of the articulatory practice. The struc-
tured totality resulting from this articulatory practice, we will call dis-
course’ (Laclau & Mouffe, 1985/2014, p. 91). This ‘modification of
identity’ means that as soon as an element is articulated to a signifying
chain in a discourse, it is transformed into a moment. In our example,
articulating ‘interdisciplinary research’ to ‘integration’ reduces the ambi-
guity of ‘interdisciplinary research’ in that it is partially fixed to the nodal
point ‘integration’. As described by Jørgensen and Phillips (2002, p. 28),
the discourse establishes a closure, but it is only a temporary stop to the
fluctuations in the meaning of the signs. A discourse can always be under-
mined by articulations that place the signs in different relations to one
another (Jørgensen & Phillips, 2002, p. 39).
The field of discursivity (Laclau & Mouffe, 1985, p. 111) is constituted
by all the possibilities that the discourse excludes; according to Laclau
and Mouffe (1985), a discourse never exhausts all the possibilities for the
ascription of meaning. Jørgensen and Phillips explain:
The field of discursivity is considered a reservoir for the ‘surplus of mean-

ing’ produced by the articulatory practice—that is, the meanings that each
sign has, or has had, in other discourses, but which are excluded by the
specific discourse in order to create a unity of meaning. (2002, p. 27)
It is ‘the surplus of meaning’ in the field of discursivity that prevents

the fixing of the floating signifier because it renders possible the struggles
over meaning; in our example ‘interdisciplinarity’, it not only hinders
370 D. Madsen
consensus about definitions of ‘interdisciplinarity’ within the scholarly

literature, but the ‘surplus of meaning’ may be one of the reasons why we
also see the overflowing of meaning from general perceptions of ‘interdis-
ciplinarity’ as ‘all things modern and creative and progressive about sci-
ence’ (Caruso & Rhoten, 2001, p. 5), that is, interpretations that may
either be only remotely relevant to the scholarly endeavour, or, for
instance, interpretations needed for legitimation purposes.
13.3.1 Discourse-as-Representation—
Discourse-as-Language
Carpentier and De Cleen (2007) make an illustrative overview of macro-

and micro-approaches to discourse as a starting point for exploring differ-
ences between Critical Discourse Analyses (CDA) and what they term
Discourse Theoretical Analysis (DTA) based on Laclau and Mouffe. They
stress the distinction between discourse-as-language and discourse-as-
representation (Carpentier & De Cleen, 2007, p. 277). CDA’s approach
is characterized as discourse-as-language because it remains closely con-
nected to linguistic textual analysis, whereas the approach based on
Laclau and Mouffe’s work is defined as discourse-as-representation, as it
is a macro-level approach in that they do not regard discourse ‘merely as
a linguistic region within a wider social realm’, but offer a more encom-
passing conceptualization of discourse that ‘insists on the interweaving of
semantic aspects of language with the pragmatic aspects of actions, move-
ments and objects’ (Torfing, 1999, p. 94; Carpentier & De Cleen, 2007,
p. 277). As Laclau and Mouffe (1987, p. 100) phrase it: ‘This totality
which includes within itself the linguistic and the non-linguistic, is what
we call discourse.’ As also noted by Phelan and Dahlberg (2011), ‘within
the horizon of a discourse theoretical ontology, it is important to empha-
size how representation is understood as constitutive, rather than merely
reflective, of social practice’ (p. 5).
DTA’s inclusion of the non-linguistic is the major difference from
CDA’s conception of discourse as dialectical to an extra-discursive reality,
and leads to different methodologies, with detailed linguistic analysis of
actual instances of discourse by CDA, whereas Carpentier and De Cleen

note for DTA:
DTA’s all-encompassing approach to discourse-as-representation, sup-

ported by its social ontology, has left it rather blind for the specificity of
language and form, but has provided theoretical support for the in-depth
analysis of the construction of political identities, embedded in the sociol-
ogy of conflict and antagonisms. (Carpentier & De Cleen, 2007, p. 278)
Glynos et al. (2009) provide an overview of and analysis of six different

approaches to and techniques in the study of discourse and highlight dif-
ferences and similarities between the approaches in terms of ontology,
focus and purpose. A detailed account of their analysis is beyond the
scope of this chapter, but it is important to note for Laclau and Mouffe’s
discourse theory that it takes ‘its principal objects of investigation to be
practices or regimes of practices’ with the aim to ‘critically explain their
transformation, stabilization, and maintenance’. Glynos et al. (2009)
state that ‘[i]n this perspective, discourse functions as an ontological hori-
zon, and this means that practices—and any other object which can be
qualified as meaningful—are by definition discursive in character’ (p. 9).
See also Phelan and Dahlberg’s (2011) discussion of the distinction
between approaches that assume a focus on ‘the social’ as a horizon of
discourse compared to those that primarily focus on the analysis of lin-
guistic and semiotic detail.
13.3.2 Signifying Practices Are Empirical Data
Generally, the conceptual framework of Laclau and Mouffe’s discourse

theory is perceived as abstract and difficult to operationalize (Uldam,
2010; Walton & Boon, 2014; Zienkowski, 2012, 2017). As described by
Dahlgren (2011, p. 226), discourse theory does not immediately offer
much in terms of a handy methodology, and generally its methodological
side remains a challenge. Zienkowski (2012) addresses ‘the post-
structuralist methodological deficit’ that does not provide a linguistic
toolbox for analysis. Walton and Boon (2014, p. 356) refer to ‘the pau-
372 D. Madsen
city of discussion around the translation of discourse theory into dis-

course analytic practice’. Discourse analysis refers to ‘the practice of
analyzing empirical raw materials and information as discursive forms’
(Howarth & Stavrakakis, 2000, p. 4). Empirical data are viewed as sets of
signifying practices through which a particular ‘reality’ is discursively
constituted, ‘thus providing the conditions which enable subjects to
experience the world of objects, words and practices’ (Howarth &
Stavrakakis, 2000, p. 4). These include Laclau and Mouffe’s logics of
equivalence and difference (Howarth & Stavrakakis, 2000, p. 5).
As noted by Laffey and Weldes (2004, p. 28), discourse theory ‘reasons
backward to establish structure from its empirical manifestations. It asks
what the conditions of possibility are of this or that particular discursive
production.’ Therefore, method in this context ‘refers to the conceptual
apparatus and empirical procedures used to make possible this retroduc-
tion’ (Laffey & Weldes, 2004, p. 28, note removed). Discourse theory
may proceed empirically by investigating articulation through represen-
tational practices and identification of nodal points (Laffey & Weldes,
2004). This point of departure for analysis is in line with the tools that
Jørgensen and Phillips (2002) find useful for empirical analysis, and
which also specifically include concepts for conflict analysis: floating sig-
nifiers, antagonism and hegemony (Jørgensen & Phillips, 2002,
pp. 49–50).
13.3.3 The Approach Taken in This Chapter
The approach of conflict analysis is used in Madsen (in review) who iden-
tifies ‘interdisciplinarity’ as a floating signifier in two competing macro
discourses that vie for hegemony in the field of interdisciplinarity studies.
This chapter is an extension of the analysis of ‘interdisciplinarity’ as a
floating signifier, in that it seeks to flesh out how the floating signifier
‘interdisciplinarity’ becomes an empty signifier—and vice versa. The sig-
nifier ‘interdisciplinarity’ has an identical form whether it is fixed, float-
ing or empty, but its signifieds are not identical, and their functions are
different. The identical linguistic manifestation of the signifier’s different
functions is a problem for eliciting insights into the representation of
absence because this requires the identification of the empty signifier.

However, the ‘blindness’ of discourse theory to the specificity of language
and form, referred to by Carpentier and De Cleen (2007, p. 278), may
be somewhat alleviated by differentiating analytically between the form,
i.e. the linguistic manifestation of the signifier and its possible functions,
i.e. if it is relatively fixed, floating or empty, cf. Fig. 13.1. Added to this is
that within the discourse theoretical literature itself, the concepts of float-
ing and empty signifiers are sometimes used indiscriminately, and con-
crete examples of what may actually happen in the transitions from
floating to empty and vice versa are hard to find; to my knowledge, exam-
ples including signifiers at an empirically observable level are non-existent,
and thus there seems to be a need for an attempt at least, to concretize the
mechanisms of emptying and filling.
My empirical material is the literature on interdisciplinarity, broadly
conceived, from which two papers have been chosen for the analysis
below. Bruun et al. (2005) is taken as the point of departure, as it explic-
itly makes a distinction between ‘interdisciplinarity in the specific sense’
and ‘interdisciplinarity in a generic sense’ which lends itself to an initial
identification of logics of equivalence and difference. Besides aligning
with the NAS definition of interdisciplinary research quoted in Sect.
13.2, Bruun et al.’s (2005) work rests on a categorization of interdisci-
plinary research that is well known and frequently quoted in the scholarly
literature, namely, Aboelela et al.’s (2007) literature review Defining
Interdisciplinary Research and their ‘continuum of integration’. Aboelela
et al found that ‘In all sources there was common acknowledgement of a
continuum with respect to interdisciplinary research and the degree of
synthesis involved in the process and achieved in the outcome’ (Aboelela
et al., 2007, p. 329). This second text makes it possible to unfold in more
detail the logics of Bruun et al.’s (2005) distinction. Both examples are
within ‘the dominant, epistemologically oriented approach’, cf. Sect.
13.2, one of the two macro discourses within the field of interdisciplinar-
ity studies (Madsen, in review). The examples serve as an empirical back-
drop to the theorizing of absence.
374 D. Madsen
13.4 Analyzing Logics of Signification

13.4.1 ‘Interdisciplinarity’ Between the Particular
and the Universal
The distinction introduced by Bruun et al. (2005) between ‘interdiscipli-

narity in the specific sense’ and ‘interdisciplinarity in the generic sense’
builds on an empirical study in which a sample of research proposals was
analyzed to find out ‘what proportion of the research proposals was inter-
disciplinary, and what kinds of IDR were suggested in the proposals’
(pp. 79–80). Bruun et al. (2005) developed a taxonomy for interdisci-
plinary research (p. 87) based on ‘the common distinction between
multi- and interdisciplinary* research’; they explicitly qualify ‘interdisci-
plinarity in the specific sense’ with an asterisk (*) to address ‘the double
meaning of the concept of interdisciplinarity’ (p. 81). ‘Interdisciplinarity
in the specific sense’ is contrasted to multidisciplinarity, whereas ‘inter-
disciplinarity in the generic’ sense is contrasted to disciplinarity.
Multidisciplinary research is characterized as ‘not being integrative in the
sense of producing a shared understanding or synthesis’ (p. 80), whereas
interdisciplinary research in the specific sense ‘integrates separate disci-
plinary data, methods, tools, concepts, and theories in order to create a
holistic view or common understanding of a complex issue, question, or
problem’ (Bruun et al., 2005, p. 80). They specify that ‘[r]esearch becomes
interdisciplinary in the generic sense—that is, either multidisciplinary or
interdisciplinary*—whenever the research activity involves several fields
in some more or less loosely coupled way’ (p. 81).
From the perspective of meaning construction within a discourse theo-
retical framework, where meaning is constructed either in terms of differ-
ence or equivalence (Torfing, 2005, p. 14), we see how Bruun et al.
construct the identity of three entities: multidisciplinarity,
‘interdisciplinarity in the specific sense’, and ‘interdisciplinarity in the
generic sense’, respectively. These identities are established relationally, in
relation to something they are not: Multidisciplinarity is not interdiscipli-
narity; and ‘interdisciplinarity in the generic sense’ is not disciplinarity.
Thus, according to the logic of difference, it is the differential relations
between the discursive moments that are constitutive of their identity

(Torfing, 1999). At the same time, we see how the logic of equivalence cre-
ates dichotomizations, two opposing camps: Multidisciplinarity is not
‘interdisciplinarity in the specific sense’, and ‘interdisciplinarity in the
generic sense’ is not disciplinarity, respectively.
Whereas the logic of difference creates identity, the logic of equivalence
functions by splitting a system of differences and creating a frontier
between two opposed camps. This means that meanings are condensed
around two antagonistic poles that mutually exclude each other (Howarth,
2000). In our two related examples, opposition is constructed in two dif-
ferent ways:
13.4.1.1 M
ultidisciplinarity Is Not ‘Interdisciplinarity
in the Specific Sense’
For multidisciplinarity is not ‘interdisciplinarity in the specific sense’,

meaning is articulated to <integration> as the nodal point, i.e. the mean-
ings of the two entities are partially fixed by reference to the nodal point.
For ‘interdisciplinarity in the specific sense’, the nodal point has a posi-
tive value, whereas for multidisciplinarity, the value is negative. Put
another way, it is the nodal point that creates and sustains the identity of
multidisciplinarity (defined by no integration) and ‘interdisciplinarity in
the specific sense’ (defined by integration), respectively.
13.4.1.2 ‘Interdisciplinarity in the Generic Sense’ Is Not

Disciplinarity
When Bruun et al. (2005) specify that ‘[r]esearch becomes interdisciplin-

ary in the generic sense—that is, either multidisciplinary or
interdisciplinary*—whenever the research activity involves several fields
in some more or less loosely coupled way’ (p. 81), the opposition between
multi- and ‘interdisciplinarity in the specific sense’ is collapsed in that
what differentiates the two (plus/minus integration) is subsumed into
‘interdisciplinarity in the generic sense’. This is because the operation
376 D. Madsen
from specific to generic involves the logic of equivalence, i.e. what multi-
disciplinarity and interdisciplinarity have in common is accentuated, at
the expense of what the differences between them are. Thus, Bruun et al.
(2005) are explicit in their differentiation between ‘interdisciplinarity in
the specific sense’ and ‘interdisciplinarity in the generic sense’, and they
emphasize ‘the double meaning of interdisciplinarity’ to pursue their goal
to find out ‘what proportion of the research proposals was interdisciplin-
ary, and what kinds of IDR were suggested in the proposals’ to create a
taxonomy (pp. 79–80). This is their reason for excluding ‘interdisciplinar-
ity in the generic sense’ from their study, and in this way they refrain from
addressing the consequences of ‘interdisciplinarity in the generic sense’.
From a discourse theoretical perspective, Bruun et al.’s (2005) distinc-
tion is between particularity and universality. According to the logic of
difference, it is the differential relations between the discursive moments
that are constitutive of their identity (Torfing, 1999, p. 300), but in the
operation of subsuming both multidisciplinarity and interdisciplinarity
as a particular entity under interdisciplinarity in a universal sense, their
differential aspect—what creates the identity of the particular entities—is
collapsed. There is a loss of meaning (Torfing, 1999, p. 97), and it is not
clear what the identity of interdisciplinarity in a universal sense is, other
than it is not disciplinarity. It is important to note that the emptying of
meaning in the process is not reflected at the level of the signifier. It is the
signified that is emptied of its differential character, and ’interdisciplinar-
ity’ is the signifier of both a particular signified and a universal signified.
And in the process of emptying the signified of meaning, ’interdisciplin-
arity’ becomes an empty signifier.
13.4.2 ‘Interdisciplinarity’ Comes to Represent

the Whole Chain of Equivalences
The taxonomy developed by Bruun et al. (2005) is based on what within

‘the dominant, epistemologically oriented approach’ is often referred to
as the common distinction between multi- and interdisciplinary research.
Aboelela et al. (2007) identify three qualitatively different modes of inter-
disciplinary research represented by different degrees of synthesis along a
continuum from the least degree of synthesis (multidisciplinary), pro-
ceeding to a moderate degree (interdisciplinary) and finally arriving at

the greatest degree of synthesis (transdisciplinary)—a continuum of inte-
gration. Their typologies are shown in Table 13.1.
Table 13.1 Typologies of interdisciplinary research (Aboelela et al., 2007, p. 337)

Typology
Author Lattuca (2001) Klein (1996) Rosenfield (1992)
Degree of synthesis
Least Informed Instrumental Multidisciplinary:
disciplinarity: interdisciplinarity: teams work in
disciplinary bridge building parallel or
questions may be between fields sequentially from
informed by problem-solving their specific
concepts or theories activity, does not disciplinary base to
from another seek synthesis or address a common
discipline fusion of different problem
Synthetic perspectives
disciplinarity:
questions that link
disciplines (question
either belongs to
both or neither
disciplines)
Moderate Synthetic disciplinarity Epistemological Interdisciplinary:
interdisciplinarity: teams work jointly
restructuring a but still from a
former approach discipline-specific
to defining a field base to address a
common problem
Greatest Transdisciplinary: Transdisciplinary: Transdisciplinary:
the application of a movement teams work using
theories, concepts, toward a a shared
or methods across coherence, unity, conceptual
disciplines with the and simplicity of framework,
intent of knowledge drawing together
developing an discipline-specific
overarching theories, concepts,
synthesis and approaches to
Conceptual address a common
interdisciplinarity: problem
questions without a
compelling
disciplinary basis
378 D. Madsen
Aboelela et al. (2007) articulate ‘interdisciplinary research’ to <integra-

tion> as the nodal point and further specify different degrees of integra-
tion which they all articulate to <a continuum of integration>. Within a
discourse theoretical framework, the continuum of integration can be
seen as a chain of equivalences. The operation of subsuming all varieties
of degrees of synthesis into a continuum with the label ‘Typologies of
Interdisciplinary Research’ (Aboelela et al., 2007, p. 337) collapses the
differences between the modes of interdisciplinary research and, thus,
their identity. Accordingly, the meaning ascribed to each identity is being
emptied. By labelling the typologies ‘interdisciplinary research’, the logic
of equivalence works to stress what each mode of ‘interdisciplinary
research’ has in common, at the expense of what the differences between
them are, as we saw in the Bruun et al. (2005) example above. Thus, all
the varieties of ‘interdisciplinary research’ in Aboelela et al.’s table
(Table 13.1), become ‘interdisciplinary’ because they are united by a con-
tinuum of integration that is represented as ‘Typologies of Interdisciplinary
Research’.
As we saw in the Bruun et al. (2005) example above, the difference
between multidisciplinarity and ‘interdisciplinarity in the specific sense’
was collapsed into equivalence, so that ‘interdisciplinarity in the generic
sense’ serves as opposition to disciplinarity. Following the logic of equiva-
lence, all the different entities in Aboelela et al.’s continuum of integra-
tion also collapse their differences into equivalence. According to Torfing:
‘The differential character of social identities collapses as they become
inscribed in chains of equivalence that construct them in terms of a cer-
tain “sameness”’ (1999, p. 124). This ‘sameness’, however, only relates to
one aspect. As Torfing explains (1999, p. 96): ‘There is no simple identity
between the equivalential identities since they are only the same in one
aspect while being different in others.’ This means that it can be only
their negation of disciplinarity that the different types of ‘interdisciplin-
ary research’ have in common in Aboelela et al.’s typology. And it is
exactly this ‘certain sameness’ that is emphasized to create a frontier to
what it is not—‘the common enemy’.
As Torfing notes, as a chain of equivalence is expanded to include all
the various groupings, ‘it becomes clear that it does not possess a positive
content of its own’ (1999, p. 175). In this case, the different entities of
the typology are the same in that together they serve as opposition to
disciplinarity—they are ‘equivalent in their common rejection of the
excluded identity’ (Laclau, 2005, p. 70). Therefore, ‘interdisciplinarity’ is
nothing other than the opposite of disciplinarity and as such it becomes
an empty signifier. Thus, we see how the logic of equivalence works as a
logic of simplification (Laclau & Mouffe, 1985, p. 117). But before we
can continue the discussion of the empty signifier, it is necessary to briefly
look at the floating signifier.
13.4.3 ‘Interdisciplinarity’—One Signifier, Different

Functions
Overall, within the dominant approach to ‘interdisciplinarity, ’interdisci-

plinary research’ is based on the premise of integration of knowledge. If
we look at ‘interdisciplinary’ and ‘interdisciplinarity’ as signifiers within
this discourse on the basis of the two texts analyzed above, <interdisci-
plinary research> and <integration> are the nodal points that confer par-
tially fixed meaning on these signifiers. The focus in this analysis is
<integration> as the privileged sign around which the other signs are
ordered. This means that <interdisciplinary research>, <interdisciplinary
>/<interdisciplinarity> and <integration> constitute a ‘knot of meanings’
united in their construction of disciplinarity as ‘the other’ following the
logic of equivalence. As previously mentioned, this dominant discourse is
struggling with another macro discourse that contests the premise of
integration; this competing discourse is sceptical of integration as a prem-
ise but does not exclude <integration> in <interdisciplinary research>,
neither does it dichotomize ‘interdisciplinarity’ and disciplinarity. Thus,
the ‘knot of meanings’ common to the two texts just analyzed should also
be seen against the backcloth of the competing discourse, because what is
seen as nodal points within the context of the dominant discourse
analyzed so far, namely <interdisciplinary research>, <interdisciplinary>/
<interdisciplinarity> and <integration> are—when compared to the
competing discourse—also floating signifiers:
380 D. Madsen
Nodal points are floating signifiers, but whereas the term ‘nodal point’
refers to a point of crystallisation within a specific discourse, the term
‘floating signifier’ belongs to the ongoing struggle between different dis-
courses to fix the meaning of important signs. (Jørgensen & Phillips, 2002,
p. 28)
The NAS definition of ‘interdisciplinary research’ (Laffey & Weldes,

2004) quoted above, can be seen as an attempt to stabilize the dominant
discourse. Thus, as we have seen, within this discourse, it is <integration>
that creates and sustains the identity of ‘interdisciplinarity’, and as previ-
ously noted, this definition is an attempt at fixing the meaning of the
signifier ‘interdisciplinarity’, which within the context of this dominant
discourse, is a fixation of meaning. But once another discourse rearticu-
lates ‘interdisciplinary research’, the signifier becomes floating. Floating
signifiers are polysemic, which is a challenge to the practice of articula-
tion, because polysemy disarticulates a discursive structure (Laclau &
Mouffe, 1985). This means that once nodal points have been constructed,
they can be rearticulated or disarticulated. In the two texts analyzed,
‘interdisciplinary research’ is articulated to <integration> as a nodal point,
but a competing discourse disarticulates this relation which means that
the identity of ‘interdisciplinary research’ is no longer related to <integra-
tion> as a nodal point, and ‘interdisciplinary research’ becomes floating
also.
To recapitulate, in the Bruun et al. (2005) text, the signifier ‘interdis-
ciplinarity’ is used to represent both a particular signified and a universal
signified. In Aboelela et al.’s (2007) typology, all varieties of ‘interdisci-
plinary research’ become ‘interdisciplinary’ because they are united by a
continuum of integration that works as a chain of equivalences, subsum-
ing their differential character under the label ‘Typologies of
Interdisciplinary Research’. In Aboelela et al.’s (2007) typology, the signi-
fiers ‘interdisciplinary’ and ‘interdisciplinarity’ variously represent spe-
cific instances of the least, moderate and greatest degree of synthesis as
well as a signifier which comes to represent the common identity which
sustains an opposition to what it is not, namely disciplinarity. And assum-
ing this function, the signifier becomes empty. Further, if we include the
wider context of the field of interdisciplinarity studies, and the struggle
between two macro discourses, in the dominant discourse that Bruun

et al. (2005) and Aboelela et al. (2007) subscribe to, ‘interdisciplinarity’
functions as a floating signifier. In one discourse ‘interdisciplinarity’ is
articulated to <integration>, in the competing discourse it is not. For
these different functions or signifieds: the (partially) fixed signified in the
NAS definition, the floating signifier in the struggle between two macro
discourses, the particular signified in Bruun et al. (2005), and the univer-
sal signified in Bruun et al. (2005) and Aboelela et al. (2007), we only
have one linguistic manifestation of the signifier: ‘interdisciplinarity’.
13.5 How Empty Signifiers Signify Absence

13.5.1 T
ensions Between Particularity
and Universality
So far, the discussion of the signifier has mostly focused on the different
functions of the signifier. We have seen in the analysis above how the
logic of equivalence works in the continuum of integration to collapse
the different entities into equivalence, and how the signifier ‘interdiscipli-
narity’ at the same time represents each particular identity that articulates
its relevant differential character along the continuum of integration, as
well as its equivalential aspect that represents the negation of disciplinar-
ity. In this latter function we see an empty signifier, but it is only tenden-
tially empty because it does not as such lose all difference (Thomassen,
2005, p. 309). As Torfing notes:
The relation between difference and equivalence is … undecidable. The

discursive identities are inscribed both in signifying chains that stress their
differential value, and in signifying chains that emphasize their equiva-
lence. The tension between the differential and equivalential aspects of
discursive identities is unresolvable, but political struggles may succeed in
emphasizing one of the two aspects. (Torfing, 1999, p. 97)
This relation between equivalence and difference requires some

unpacking, as does the hegemonization of one of the two aspects. First,
382 D. Madsen
an attempt is made in the model, Fig. 13.2, to concretize the complexity

and tensions between particularity and universality in a discursive
system.
I first lay out the possibility in a discursive system of partially fixing the
relationship between a signifier and its signified(s) between its two impos-
sibilities: To the left: the ‘impossibility of ultimate fixation of meaning’
and to the right: the ‘impossibility of absolute non-fixation of meaning’
(Zienkowski, 2012, pp. 506–507). Moving from left to right in the
model, indicated in the ‘emptying arrows’, implies a loss of meaning; and
‘the filling arrows’ above indicate the opposite direction—when signifiers
gain content. Next, to the left, the ‘relative fixation’ of meaning to indi-
cate that some fixations of meaning may become so conventionalized or
sedimented ‘that we think of them as natural’ (Jørgensen & Phillips,
2002, p. 26). An example would be the NAS definition of interdisciplin-
ary research discussed above, if there was no competing discourse.
In Fig. 13.2, the ‘relative fixation’ is an indication of ‘the particular’ to
show how ‘meaning is intrinsically linked to the differential character of
identity’ (Torfing, 1999, p. 97). As formulated by Laclau (1996a, 1996b),
‘[t]his emptying of a particular signifier of its particular, differential signi-
fied is, …, what makes possible the emergence of “empty” signifiers as the
signifiers of a lack, of an absent totality.’ But the signifier can be only
tendentially empty, because ‘a remainder of particularity cannot be elimi-
nated’ (Laclau, 2001, p. 11), that is, ‘[signifiers] retain some of their
mutual differences’ (Thomassen, 2005, p. 293). Therefore, the unresolv-
able tension between the differential and equivalential aspects that is
indicated in Torfing’s quote above (1999, p. 97), is included in the lower
part of the model to conceptualize the tension between particularity and
universality. And ‘absence’ is underlying the universal dimension.1
13.5.2 Hegemonization
According to Howarth (2015, p. 12) ‘empty signifiers provide the sym-

bolic means to represent these essentially incomplete orders.’ Their func-
tion is to ‘incarnate the “absent fullness” of an essentially incomplete
discursive system’. This ‘incarnation’ is at the heart of the hegemonic
operation, indicated in the upper right corner of Fig. 13.2 as it is within

the logic of hegemony that empty signifiers emerge (Howarth & Griggs,
2006, p. 31). The ‘incarnation’ means that an empty signifier borrows its
content from one of the entities constituted within the equivalential
space (Laclau, 2001, p. 9), i.e. ‘interdisciplinarity, in our case, may bor-
row its content from any of the differential signifieds in the chain of
equivalences, and becomes a name for an absence. An absent fullness
cannot have any positive content or form of representation of its own
(Laclau, 1996a, 1996b, p. 42), and as Laclau notes (2001, p. 11) there is
‘a movement of mutual contamination between the universal and the
particular’. Accordingly, ‘the presence of empty signifiers … is the very
condition of hegemony’ (Laclau, 1996a, 1996b, p. 43), and Laclau goes
on to note that various forces ‘can compete in their efforts to present their
particular objectives as those which carry out the filling of that lack. To
hegemonize something is exactly to carry out this filling function’ (Laclau,
1996a, 1996b, p. 44). As formulated by Torfing (2005, p. 15) ‘[a]rticula-
tions that manage to provide a credible principle upon which to read
past, present, and future events, and capture people’s hearts and minds,
become hegemonic.’
13.5.2.1 ‘Interdisciplinarity’ Becomes the Name for ‘All

Things Modern and Creative and Progressive
About Science’
Going back now to Weingart’s conclusion, cf. Sect. 13.2, it is suggested

that it is interdisciplinarity as ‘the promise of innovation and surprise’
that, within the dominant discourse analyzed above, is filling the lack
using the empty signifier ‘interdisciplinarity’ to hegemonize the ‘absent
universality’. This hegemonic operation depends on a previous emptying
of the differential character of the signifieds according to the logic of
equivalence, as analyzed in the Aboelela et al. (2007) typology above, so
that the ‘absence’ can be filled with ‘the promise of innovation and sur-
prise’. If we consider the widespread use of ‘interdisciplinarity’ to mean
everything and nothing at the same time, and, following Laclau, it is
further suggested that it is the emphasis on the equivalential aspects that
384 D. Madsen
makes it possible for the signifier to become a name for ‘all things modern
and creative and progressive about science’ (Caruso & Rhoten, 2001,
p. 5), quoted in Sect. 13.2.
This means that, following Torfing (1999, p. 97) quoted above on the
unresolvable tension between the differential and equivalential aspects,
that the success of the empty signifier is a result of political struggles, and
thus it is the emptying process that makes ‘interdisciplinarity’ prone to
appropriation by other purposes than those that can be accounted for
within a scientific field. This further suggests that the logic of significa-
tion and the tension between difference and equivalence may be impor-
tant tools for theorizing the borderland between rigorous scholarship in
interdisciplinarity studies and the surrounding ideological and political
forces that emanate from other agendas. This chapter, in seeing the field
as politicized because of the widespread use of ‘interdisciplinarity’ every-
where in academia, has taken an important step in identifying the func-
tion of ‘interdisciplinarity’ as an empty signifier, as equivalential
constructions appear to be of marginal interest to the scientific field,
whereas the logic of difference is a more complex articulation of elements
that are incorporated into an expanding order, which seems to be more in
line with the ideals of academic discourse. There seem to be no previous
studies focusing on the logics of equivalence and difference, or the pro-
duction and use of empty signifiers, in academic discourse.
13.5.3 E
mpty Signifiers Urge Us to Look
for Something that Is Absent
Although discourse theory is, in essence, political, and would not imme-
diately suggest itself for analyses of academic discourse, the exploration of
‘interdisciplinarity’ functioning as a (partially) fixed, floating and empty
signifier, has served as an empirical example of how an empty signifier
representing an absence emerges within the logic of hegemony.
Typical central objects of investigation for discourse theory are issues
of identity formation, the production of novel ideologies, the logics of
social movements (Howarth & Stavrakakis, 2000), political engagement
in activism (Uldam, 2010) and activist discourse and minority debates
(Zienkowski, 2012, 2017). As formulated by Zienkowski (2017) empty

signifiers ‘propel us towards action and political mobilisation’ (p. 54), and
he analyzes how the notion of ‘integration’ has functioned as an empty
signifier in Flemish debates about minorities for decades, considering
how ‘the integrated society’ frequently operates as a mythical ideal, argu-
ing that lack of integration supposedly leads to a lack of society (p. 55).
And he emphasizes how ‘[e]mpty signifiers urge us to go look for some-
thing that is absent or lacking in our identity as well as in our society.
They simultaneously promise meaning and withhold it (Glynos &
Howarth, 2007, p. 131).’ (2017, p. 55).
According to Howarth and Stavrakakis (2000), societies are organized
and centred on the basis of ‘such (impossible) ideals. What is necessary
for the emergence and function of these ideals is the production of empty
signifiers’ (2000, p. 8). Laclau (1996a, 1996b) generalizes the argument
of the signifier indicating that ‘any term which, in a certain political con-
text becomes the signifier of the lack, plays the same role’. ‘Politics’, he
continues, ‘is possible because the constitutive impossibility of society
can only represent itself through the production of empty signifiers’
(1996a, 1996b, p. 44).
13.6 Conclusion
The aim of this chapter has been to theorize absence and its representa-
tion in the empty signifier. However, as it is only the empty signifier that
represents absence, the challenge remains how to distinguish the empty
signifier from a floating or (partially) fixed signifier. The form of the signi-
fier is identical in all three cases, but their function is not. The conceptual
framework of discourse theory is quite abstract and does not immediately
offer much in terms of a methodology. This also applies for the identifica-
tion of an empty signifier. As suggested by Zienkowski (2017) ‘[e]mpty
signifiers urge us to go look for something that is absent or lacking in our
identity as well as in our society. … (2017, p. 55), but this does not
answer the question of what to look for to identify the empty signifier in
the first place. Still, the model in Fig. 13.2, may help to initially identify
a signifier’s function between fixation and non-fixation, between particu-
386 D. Madsen
larity and universality, and to explore the signifying practices and logics
at play.
There is nothing conspicuous about absence in the discourse theoreti-
cal framework of Laclau and Mouffe as it is part of their ontology. This
also means that absence is given, and not a phenomenon that can arise or
be identified per se. But what can be identified is its possible representa-
tion by an empty signifier, in which case absence may be seen as con-
spicuous by its presence. This chapter’s contribution to existing literatures
on absence in discourse has been to introduce ontological absence and its
representation in empty signifiers that emerge within the logic of hege-
mony. It is hoped that this chapter’s differentiation between the function
of the signifier and its linguistic manifestation may provide a stepping
stone to further research into the linguistic aspects of discourse theory
and the development of methods to connect this conceptual framework
with the use of text corpora.
Notes
1. Absence is also described as an ‘absent universality’, ‘absent totality’ or
‘absent fullness’ (Thomassen, 2005, p. 309), ‘inherent lack’ (Glynos et al.,
2009) or ‘incomplete order’ (Howarth, 2015), and the ‘universal’ is vari-
ously described as an ‘empty space’ or ‘empty place’ (Torfing, 1999). For
a detailed discussion, see Zerilli (1998), Torfing (1999), Norval (2000),
Laclau (2000, 2001). And specifically for a discussion of how new spaces
of representation may be constituted by myths and imaginaries, see Norval
(2000) and Madsen (2016).
References
Aboelela, S. W., Larson, E., Bakken, S., Carrasquillo, O., Formicola, A., Glied,
S. A., et al. (2007). Defining interdisciplinary research: Conclusions from a
critical review of the literature. Health Services Research, 42(1), 329–346.
Bourdieu, P. (1999). The specificity of the scientific field. In M. Biagioli (Ed.),
The science studies reader (pp. 31–50). New York: Routledge.
Bruun, H., Hukkinen, J., Huutoniemi, K., & Klein, J. T. (2005). Promoting
interdisciplinary research: The case of the academy of Finland. Publications of
the Academy of Finland, Series No. 8/05.
Carpentier, N., & De Cleen, B. (2007). Bringing discourse theory into media
studies: The applicability of discourse theoretical analysis (DTA) for the
study of media practises and discourses. Journal of Language and Politics, 6(2),
265–293.
Caruso, D., & Rhoten, D. (2001). Lead, follow, get out of the way: Sidestepping
the barriers to effective practice of interdisciplinarity. Hybrid Vigor White
Chapters.
Chandler, D. (2002). Semiotics: The basics. New York, NY: Routledge.
Chiapello, E., & Fairclough, N. (2002). Understanding the new management
ideology: A transdisciplinary contribution from critical discourse analysis
and new sociology of capitalism. Discourse & Society, 13(2), 185–208.
Chouliaraki, L., & Fairclough, N. (1999). Discourse in late modernity. Rethinking
critical discourse analysis. Edinburgh: Edinburgh University Press.
Committee on Facilitating Interdisciplinary Research, National Academy of
Sciences, National Academy of Engineering, Institute of Medicine. (2004).
Facilitating interdisciplinary research. Washington, DC: The National
Academies Press.
Dahlberg, L. (2011). Discourse theory as critical media politics? Five questions.
In L. Dahlberg & S. Phelan (Eds.), Discourse theory and critical media politics
(pp. 41–63). London: Palgrave Macmillan UK.
Dahlgren, P. (2011). Mobilizing discourse theory for critical media politics:
Obstacles and potentials. In L. Dahlberg & S. Phelan (Eds.), Discourse theory
and critical media politics (pp. 222–249). Springer.
Fairclough, N. (1992). Discourse and social change. Cambridge: Polity Press.
Foucault, M. (1975). I, pierre rivière, having slaughtered my mother, my sister, and
my brother: A case of parricide in the 19th century. Iowa City, IA: University of
Nebraska Press.
Frickel, S., Albert, M., & Prainsack, B. (2016). Introduction: Investigating
interdisciplinarities. In S. Frickel, M. Albert, & B. Prainsack (Eds.),
Investigating interdisciplinary collaboration: Theory and practice across disci-
plines. New Brunswick, NJ: Rutgers University Press.
Glasze, G. (2007). The discursive constitution of a world-spanning region and
the role of empty signifiers: The case of francophonia. Geopolitics, 12(4),
656–679.
388 D. Madsen
Glynos, J., & Howarth, D. (2007). Logics of critical explanation in social and
political theory. London: Routledge.
Glynos, J., Howarth, D., Norval, A. J., & Speed, E. (2009). Discourse analysis:
Varieties and methods (No. NCRM/014). National Centre for Research
Methods, University of Essex.
Howarth, D. (2000). Discourse. Buckingham: Open University.
Howarth, D. (Ed.). (2015). Ernesto Laclau: Post-Marxism, populism and critique.
Howarth, D., & Griggs, S. (2006). Metaphor, catachresis and equivalence: The
rhetoric of freedom to fly in the struggle over aviation policy in the United
Kingdom. Policy and Society, 25(2), 23–46.
Howarth, D., & Stavrakakis, Y. (2000). Introducing discourse theory and politi-
cal analysis. In D. Howarth, A. Norval, & Y. Stavrakakis (Eds.), Discourse
theory and political analysis identities, hegemonies and social change (pp. 1–23).
Manchester: Manchester University Press.
Jacobs, J. A. (2013). In defense of disciplines: Interdisciplinarity and specialization
in the research university. Chicago, IL: University of Chicago Press.
Jørgensen, M. W., & Phillips, L. J. (2002). Discourse analysis as theory and
method. Los Angeles, CA: Sage.
Klein, J. T. (1996). Crossing boundaries. Knowledge, disciplinarities, and interdis-
ciplinarities. Charlottesville, VA: University Press of Virginia.
Klein, J. T. (2010). Creating interdisciplinary campus cultures: A model for strength
and sustainability. John Wiley & Sons.
Laclau, E. (1990). New reflections on the revolutions of our time. London: Verso.
Laclau, E. (1993). Power and representation. Politics, Theory, and Contemporary
Culture, 36, 277–296.
Laclau, E. (1995). Subject of politics, politics of the subject. Differences: A
Journal of Feminist Cultural Studies, 7(1), 146–165.
Laclau, E. (1996a/2007). Emancipation(s). London: Verso.
Laclau, E. (1996b). Why do empty signifiers matter to politics? In E. Laclau
(Ed.), Emancipation(s) (pp. 36–46). London: Verso.
Laclau, E. (2000). Constructing universality. In J. Butler, E. Laclau, & S. Žižek
(Eds.), Contingency, hegemony, universality: Contemporary dialogues on the left
(pp. 281–307, at pp. 296–307). London: Verso.
Laclau, E. (2001). Democracy and the question of power. Constellations, 8(1),
3–14.
Laclau, E. (2004). Glimpsing the future. In S. Critchley (Ed.), Laclau: A critical
reader (pp. 279–328). London: Routledge.
Laclau, E. (2005). On populist reason. London: Verso.
Laclau, E., & Mouffe, C. (1985). Hegemony and socialist strategy: Towards a radi-
cal democratic politics. London: Verso.
Laclau, E., & Mouffe, C. (1987). Post-marxism without apologies. New Left
Review, 166, 79.
Laffey, M., & Weldes, J. (2004). Methodological reflections on discourse analy-
sis. Qualitative Methods, 2(1), 28–31.
Lattuca, L. (2001). Creating interdisciplinarity: Interdisciplinary research and
teaching among college and university faculty. Nashville, TN: Vanderbilt
University Press.
Madsen, D. (2016). Liberating interdisciplinarity from myth: An exploration of
the discursive construction of identities in information studies. Journal of the
Association for Information Science and Technology, 67(11), 2697–2709.
https://doi.org/10.1002/asi.23622
Madsen, D. (in review). Understanding ambiguities in interdisciplinarity stud-
ies through discourse theory. A critical analysis of two discourses—or why
the ‘silo rhetoric’ is so pervasive. Manuscript submitted for publication.
Norval, A. J. (2000). The things we do with words—Contemporary approaches
to the analysis of ideology. British Journal of Political Science, 30(2), 313–346.
ing absences. International Journal of Corpus Linguistics, 19(1), 118–146.
Phelan, S., & Dahlberg, L. (2011). Discourse theory and critical media politics:
An introduction. In L. Dahlberg & S. Phelan (Eds.), Discourse theory and
critical media politics (pp. 1–40). Basingstoke: Springer.
de Saussure, F. (1981/2011). Course in general linguistics. Suffolk: Fontana.
Rosenfield, P. L. (1992). The potential of transdisciplinary research for sustain-
ing and extending linkages between the health and social sciences. Social
Science & Medicine, 35(11), 1343–1357.
Schröter, M., & Storjohann, P. (2015). Patterns of discourse semantics: A
corpus-assisted study of financial crisis in British newschapter discourse in
2009. Pragmatics & Society, 6(1), 43–66.
Stone, D. A. (2013). Beyond common ground: A transdisciplinary approach to
interdisciplinary communication and collaboration. In M. O’Rourke,
S. Crowley, S. D. Eigenbrode, & J. Wulfhorst (Eds.), Enhancing communication
& collaboration in interdisciplinary research (pp. 82–102). Los Angeles: Sage.
Taylor, C. (2013). Investigating absence in press discourses of migration (poster).
Retrieved from https://sussex.academia.edu/CharlotteTaylor/Talks
Taylor, C. (2014). Investigating the representation of migrants in the UK and
Italian press A cross-linguistic corpus-assisted discourse analysis. International
Journal of Corpus Linguistics, 19(3), 368–400.
390 D. Madsen
Thomassen, L. (2005). Antagonism, hegemony and ideology after heterogene-

ity. Journal of Political Ideologies, 10(3), 289–309.
Torfing, J. (1999). New theories of discourse: Laclau, Mouffe and Zizek. Oxford:
Blackwell.
Torfing, J. (2005). Discourse theory: Achievements, argument, and challenges.
In D. R. Howarth & J. Torfing (Eds.), Discourse theory in European politics:
Identity, policy and governance (pp. 1–32). Houndmills: Palgrave Macmillan.
Uldam, J. (2010). Fickle commitment: Fostering political engagement in ‘the flighty
world of online activism’. Department of Intercultural Communication and
Management, Copenhagen Business School. PhD Series 35.2010.
Walton, S., & Boon, B. (2014). Engaging with a Laclau & mouffe informed
discourse analysis: A proposed framework. Qual Research in Orgs & Mgmt,
9(4), 351–370.
Weingart, P. (2000). Interdisciplinarity: The paradoxical discourse. In P. Weingart
& N. Stehr (Eds.), Practising interdisciplinarity (pp. 25–41). Toronto:
University of Toronto Press.
Zerilli, L. M. (1998). This universalism which is not one. Diacritics, 28(2),
3–20.
Zienkowski, J. (2012). Overcoming the post-structuralist methodolocial deficit-
metapragmatic markers and interpretive logics in a critique of the Bologna
process. Pragmatics, 22(3), 501–534.
Zienkowski, J. (2017). Articulations of self and politics in activist discourse: A dis-
course analysis of critical subjectivities in minority debates. Cham: Springer.
Žižek, S. (1989). The sublime object of ideology. London: Verso.
Index1
0-9, AND SYMBOLS 208, 229, 230, 233, 235,

9/11, 68, 69, 83, 87 236n13, 245, 252, 253, 281,
282, 285–289, 298, 315
Broadcast media, 125, 308, 310
A
Absence (absences, absent), 1, 5, 26,
65–87, 95–97, 126, 159–188, C
191, 215, 241, 281, 305, 329, CA, see conversation analysis (CA)
359, 377, 378 CDA, see critical discourse analysis
Agency, 1, 4, 6–10, 12, 18, 37, 38, (CDA)
45, 61, 80, 82, 83, 139–141, Censorship, 9, 10, 29, 85, 142,
143, 154, 192, 202, 207, 208, 306–310, 314, 316, 319,
227, 257, 260, 264, 265, 310 324
China, 34, 86, 191–193, 197,
199–202, 207
B Climate change, 12, 241, 243, 244,
Backgrounding, 3, 4, 11–15, 75, 249, 250, 253–270, 271n2,
165, 192, 196, 198–200, 207, 271n3, 271n8
Note: Page numbers followed by ‘n’ refers to notes.

1

Postdisciplinary Studies in Discourse, https://doi.org/10.1007/978-3-319-64580-3
392 Index
Concealment, 5, 8, 10, 11, 27, Discursive absences, 2, 4, 7, 12, 18,

209n3, 217, 220, 294, 67, 127, 140, 153
296–299 Dominant discourses, 132, 133, 148,
Content analysis, 19, 66, 73, 137, 151, 152, 218, 264, 366,
195, 306 379–381, 383
Conventional silence, 8, 10
Conversation analysis (CA), 19, 334,
336, 337, 352 E
Corpus analysis, 15, 195, 362 Egypt, 99, 102, 103, 106, 109, 114,
Corpus assisted discourse studies 115
(CADS), 27, 95–121 Election manifestos, 12, 26, 39–42
Corpus linguistics (CL), 19, 27, 58, Empty signifier, 13, 359, 377–386
72, 95, 97, 119, 359 Environmental refugees, 12, 241,
Corpus tools, 121, 194 252, 260
Corpus, corpora, 15, 16, 25–28, Europe, 26, 34, 35, 37, 39–42, 44,
30–38, 43, 48, 50, 51, 53, 55, 47–52, 54, 56, 58, 59, 88n1,
56, 59, 61, 96–102, 110, 111, 113, 233, 234, 256
120, 192, 194–207, 209n2, Evasion, 10, 11, 106, 334
333, 335, 362, 386 Exclusion (exclude, excluded), 7, 15,
Critical discourse analysis (CDA), 1, 27, 28, 30, 39, 48, 68, 80, 83,
2, 4, 14, 16, 28, 72, 95, 96, 85, 86, 127, 131–133, 138,
119, 216–218, 282, 287, 288, 139, 144, 152, 153, 161, 164,
361, 370, 371 165, 196, 201, 203, 205, 207,
Critical discourse studies (CDS), 6, 236n13, 242, 245, 252,
19, 37, 306, 307 268, 288, 324, 367–369,
Cross-Cultural Discourse Analysis 375, 379
(CCDA), 223, 224
F
D Facebook, 26, 134, 135
DA, see discourse analysis (DA) Foregrounding, 3, 4, 11, 13, 14, 26,
Digital silence, 29–30 81, 169, 192, 207, 208, 229,
Discourse analysis (DA), 3–5, 27, 242, 249, 252, 253, 256, 259,
119, 193–195, 232, 281–299, 262, 282, 285, 291, 294, 295,
306, 367, 372 298
Discourse theoretical approach, Frame analysis, 241, 252, 260
13 Framing, 11, 12, 16, 19, 27, 29,
Discourse theory, 19, 361–363, 35–38, 43, 55–57, 59, 61, 79,
367–373, 384–386 126, 129, 131, 137, 143,
Index
393
147–149, 151, 153, 160, 161, M

163–174, 179, 180, 186, 188, Meaningful absences, 6, 7, 14,
192, 208, 222, 241–270, 215–235
284–286, 291, 292, 307, 316, Media discourses, 10, 12, 98, 126,
317, 323 128, 130, 131, 138, 139, 191,
French discourse analysis, 216, 208, 268, 281, 287, 288
218–223 Mediterranean, 26, 32, 34, 35, 38,
50, 51
Metadiscourse, 14, 18, 209n3, 329,
H 337
Hegemony, 3, 4, 6, 7, 130–133, 135, Metalinguistic(s)/ally / comments/
137, 142, 145, 147, 153, 218, reference, 10, 13, 14, 17, 18,
234, 242, 246, 361, 363, 372, 221, 316, 335
382–384, 386 Metaphor/ic/al/s, 4, 11, 12, 16, 55,
217, 223, 250, 251, 256, 269,
285, 298, 310
I Middle East (MENA), 97
Ideology, 3, 17, 18, 78, 81, 82, 84, Migration (migrants, immigration,
219, 220, 236n10, 241, 246, emigration, emigrants), 12, 25,
253, 281–283, 285, 294, 299, 243, 249–251, 253–255,
305, 307, 311, 313, 317, 319, 257–259, 261–267, 270, 271n5
323, 325, 369, 384 Misrepresentation(misrepresent/ed/
Implication, 11, 126–128, 134, 139, ing/s), 28, 131, 153, 161–163,
172, 217, 220, 221, 253, 268, 186
306, 312, 313, 317, 340, 344 Multimodal/ity/DA, 15, 32, 305, 324
Intention, 7–11, 18, 27, 28, 61, 96,
125, 143, 152, 216–221, 241,
260, 299, 307, 316, 330, 335 N
Iran, 12, 65, 112, 113, 117 Narrative, 4, 65, 126, 127, 131, 132,
136–138, 145, 151, 152, 154,
201, 234, 265–268, 318, 332,
K 350, 351, 360
Keyword(s), 26, 28, 32, 35–37, 48, 50, Narrative analysis, 136
56, 112, 195, 203, 207, 245 Newspaper articles, 12, 16, 112,
194–196, 199, 242, 250, 306,
323
L Newspaper discourse, 12, 13
Libya, 35, 53, 102–105, 107, 109, Nigeria, 282, 286, 288, 290, 291
110, 114–117, 119 North Africa (MENA), 97
394 Index
O S
Omission (omit, omitted, omitting), Sayable/sayability, 9, 70
8, 10, 11, 13, 16–18, 161, Secrecy, 8
179, 182, 186, 193, 204–206, Self-censorship, 8–10, 131, 310
247, 253, 281–284, 286, 290, Semiotic, 1, 19, 66–68, 71–73, 80,
292, 293, 298, 306, 307, 313, 81, 86, 220, 371
319, 335 Sexual minority/ies, 12, 15, 125
Signified, 13, 18, 362, 367, 368,
372, 376, 380–383
P Signifier, 5, 13, 18, 359–378
Pakistan, 12, 65 Silence(s), 1–14, 16–19, 26,
Police interview(s), 10, 13, 329–353 125–154, 191–208, 216–221,
Political discourses, 18, 25, 27 224, 226–228, 232, 234, 235,
Pollution, 12, 192, 197, 198, 200, 247, 265, 285, 287–289, 293,
201, 203, 207 295, 298, 299, 305, 306, 319,
Presupposition(s), 11, 12, 16, 134, 320, 322, 325, 329, 330,
225, 247, 253, 268 333–340, 342, 345, 348, 349,
Print media, 125, 308, 310 352, 353, 355n2
Silencing(to silence, silenced), 9, 10,
12, 16, 25–30, 37, 60, 61,
R 126, 127, 130–134, 139–141,
Refugee crisis, 34, 38, 61 144, 152, 153, 218, 219,
Refugee(s), 12, 26, 32, 34–38, 42, 236n1, 266, 306, 307, 315,
49–52, 54–56, 58, 59, 61, 320, 322–324, 329
241, 252, 260 Smog, 12, 191
Representation(represent/ed/ing/s), Social actor(s)/ (analysis), 4, 15–17,
3, 4, 11, 12, 15–17, 26, 31, 77, 125, 126, 168, 178,
35, 61, 67, 68, 70–87, 96, 98, 182–186, 192, 196, 202, 208,
103, 109, 110, 115, 117, 236n13, 281, 282
126–128, 130, 131, 133, 134, Social media, 17, 25, 28, 31, 132,
136, 144, 150–152, 154, 162, 133, 135, 136, 152, 154, 188
189n8, 193, 196, 197, 199, Sociolinguistics, 2, 119
200, 202–207, 216, 218, Spain, 25
223–235, 236n9, 241, 252, Strategic silence, 28, 194
260, 282, 283, 287, 293–295, Syria, 32, 34, 35, 37, 38, 47, 50–53,
305, 307, 308, 310, 311, 55–57, 59, 61, 102–108, 110,
313–315, 322–324, 359–386 113, 114, 116, 117
Index
395
T U
Taboo, 8–10, 18, 218, 226, Uganda, 127–130, 133, 135, 138,
305–315, 317, 319, 321–325, 139, 143, 144, 146, 149, 150,
326n1, 326n3 153, 154
Taboo language, 307, 308, 310–313, United States/US/USA, 16, 68, 69,
315, 317, 319, 321–323, 325, 88n3, 101, 103, 106, 107,
326n1, 326n3 112, 113, 115, 142, 166, 172,
Terror/ism/ist/ists, 32, 34, 36–39, 177, 194, 251, 305–311, 315,
50, 56–59, 67–69, 80, 106, 320, 329
178, 183, 185, 255, 288, 291, Unsaid, 9, 16, 28, 128, 215–228, 232,
293 233, 235, 268, 288, 294, 306
Textual silence, 12, 16, 193, 209n4,
247
Twitter/ tweet/ing/s, 25–61, 127, V
132, 134–138, 145–149, Visual communication, 72
151–154, 188 Visual grammar, 71, 86

Exploring Silence and Absence in Discourse: Empirical Approaches

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Exploring Silence and Absence in Discourse: Empirical Approaches

Uploaded by

Copyright:

Available Formats

Exploring Silence and

More information about this series at

Exploring Silence and

Postdisciplinary Studies in Discourse

Library of Congress Control Number: 2017962073

© The Editor(s) (if applicable) and The Author(s) 2018

Cover illustration: © fanjianhua / Getty Images

Printed on acid-free paper

This Palgrave Macmillan imprint is published by Springer Nature

Part I Comparison as Means to Identify Silence and Absence 23

2 Not for Twitter: Migration as a Silenced Topic in the 2015

3 Absence in Visual Narratives: The Story of Iran and

4 Intimations of ‘Spring’? What Got Said and What Didn’t

5 Cross-media Studies as a Method to Uncover Patterns of

6 Critically Illuminating Relevant Absences in Public

7 Silence and Absence in Chinese Smog Discourses 191

8 Theoretical and Methodological Challenges in

9 What’s Not in a Frame? Analysis of Media

10 A Discourse Analysis of Absence in Nigerian News Media 281

Part III Analysing Surface Indicators of Silence and Absence 303

11 What the f#@$!: Policing and Performing the

12 The Use of No Comment by Suspects in Police Interviews 329

13 Conspicuous by Presence: The Empty Signifier

Manuel Alcántara-Plá is an Associate Professor in the Department of Linguistics

also discussed linguistic devices used in Nigerian media discourse to project to

Kieran O’Halloran is a Reader in Applied Linguistics at King’s College,

Discourse (Benjamins 2013) and has published in international journals and

l’enfant paraît… Le discours des guides parentaux en France et en Allemagne (Presses

Fig. 3.1 Diachronic patterns for Gaze (Eye Contact): Iranian

Fig. 6.1 Major lexical cohesive chains in Monbiot’s argument 169

Table 2.1 Migration in the election manifestos 39

Table 9.1 Overview of four issue-frames 251

© The Author(s) 2018 1

understood as resources for interpretation for the interactants themselves,

[t]he emphasis on linguistic analysis implies an emphasis on available dis-

The question of power and hegemony is closely related to this since

This is relevant for silence and absence since “[s]ignificance involves

It seems that only when we can hold non-occurrence of speech against

investigate development and emergences of discourse and to look at pro-

Concealment and omission can also be assigned to individual agency.

because we normally adhere to norms of sayability either through self-

A major function of silencing is to contain […] opposition by identifying

absence phenomena above in order to vary expression, but it does not

1.2 Manifestations of Silence and Absence

contributions that propose comparison as a way to identify absence in dis-

1.3 Methodological Approaches

The previous work on silence is somewhat eclectic. While a large body

1.3.1 Comparison as a Means to Identify Absence

If we accept that meaningful absences require a possible presence against

1.3.2 Exploring Means That Produce Absence

Verbalised references to silence and absence by discourse participants

and breadth of conclusions that can be drawn about perceptions of silence

As all these studies illustrate, successful methods of identifying absence

Blommaert, J. (2005). Discourse: A critical introduction. Cambridge: Cambridge

Jensen, V. (1973). Communicative functions of silence. ETC: A Review of

2.1 Silencing in the Digital Debate

This work is part of a project focused on the political discourse in the

M. Alcántara-Plá • A. Ruiz-Sánchez (*)

© The Author(s) 2018 25

should not be crucial or relevant to the intended communication. When

Both statistics and qualitative analysis help us to describe the way

2.1.2 Silence and Power

This chapter follows a critical discourse analysis perspective, where “dis-

Part I Comparison as Means to Identify Silence and Absence 23

2 Not for Twitter: Migration as a Silenced Topic in the 2015

3 Absence in Visual Narratives: The Story of Iran and

4 Intimations of ‘Spring’? What Got Said and What Didn’t

5 Cross-media Studies as a Method to Uncover Patterns of

6 Critically Illuminating Relevant Absences in Public

7 Silence and Absence in Chinese Smog Discourses 191

8 Theoretical and Methodological Challenges in

9 What’s Not in a Frame? Analysis of Media

10 A Discourse Analysis of Absence in Nigerian News Media 281

Part III Analysing Surface Indicators of Silence and Absence 303

11 What the f#@$!: Policing and Performing the

12 The Use of No Comment by Suspects in Police Interviews 329

13 Conspicuous by Presence: The Empty Signifier

Fig. 6.1 Major lexical cohesive chains in Monbiot’s argument 169

Table 2.1 Migration in the election manifestos 39

Table 9.1 Overview of four issue-frames 251

1.2 Manifestations of Silence and Absence

1.3 Methodological Approaches

1.3.1 Comparison as a Means to Identify Absence

1.3.2 Exploring Means That Produce Absence

2.1 Silencing in the Digital Debate

2.1.2 Silence and Power

2.1.3 Digital Silence

2.2 The Corpus

2.2.2 Electoral Manifestos

2.3 Immigration as an Expected Topic

2.3.1 Political Context

2.3.2 Migration in the News

2.3.3 Migration in the Electoral Manifestos

2.4 Migration as a Silenced Topic in Twitter

2.4.3 Borders and the Mediterranean

2.5 Aunque eso no dé votos

a nalysis of what might be missing. Qualitative lenses provide an effective