IDENTITY

STUDIES
IN THE
SOCIAL
SCIENCES

Identity, Violence
and Power
Mobilising Hatred, Demobilising Dissent
Guy Elcheroth and Stephen Reicher
Identity Studies in the Social Sciences

Series Editors
Margaret Wetherell
The Open University
Milton Keynes, United Kingdom
Valerie Hey
University of Sussex
Brighton, United Kingdom
Stephen Reicher
School of Psychology and Neuroscience
University of St. Andrews, St. Andrews
United Kingdom
Identity brings together work on core social categories such as social class,
race, ethnicity, gender, disability and sexuality. This series investigates the
ways in which social and personal identities are lived and performed in
spaces and contexts such as schools, work places, clinics, homes, com-
munities, streets, politics and public life, and explores a range of theo-
retical, methodological and epistemological debates over, for example,
the demise of essentialist models, the rise of ‘identity politics’ and the
relationship between psychological and social processes. Identity research
has been the vehicle for some profound reflections on the nature of new
and emerging social and cultural forms and the impacts of globaliza-
tion, transnationalism, postcolonialism and multiculturalism. This series
welcomes critically and theoretically-informed work in a variety of areas
including nationhood, family, gender and class, as well as on issues of
identity and space, media representations of identity, social inclusion and
exclusion and social identity theory.

More information about this series at

http://www.springer.com/series/14859
 Guy Elcheroth • Stephen Reicher

Identity, Violence
and Power
Mobilising Hatred, Demobilising Dissent
Guy Elcheroth Stephen Reicher
Institute of Social Sciences School of Psychology
University of Lausanne University of St Andrews
Lausanne, Switzerland St Andrews, United Kingdom

Identity Studies in the Social Sciences

ISBN 978-0-230-27260-6    ISBN 978-1-137-31728-5 (eBook)
DOI 10.1057/978-1-137-31728-5

Library of Congress Control Number: 2016962652

© The Editor(s) (if applicable) and The Author(s) 2017

The author(s) has/have asserted their right(s) to be identified as the author(s) of this work in accordance
with the Copyright, Designs and Patents Act 1988.
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.

Printed on acid-free paper

This Palgrave Macmillan imprint is published by Springer Nature

The registered company is Macmillan Publishers Ltd.
The registered company address is: The Campus, 4 Crinan Street, London, N1 9XW, United Kingdom
 Preface

An Invitation to Perplexity

This book was not written in a day. The idea of writing a joint book
which examined intergroup relations through the prism of collective
mobilisation (or else demobilisation) first crossed our minds in the spring
of 2008. This was a period that we spent together in the seemingly time-
less environment of St Andrews—a small, beautiful and ancient town
perched on the far edge of the European continent which was at the
epicentre of the Scottish religious wars of the sixteenth century, where
the wars have been largely forgotten since but nothing else seems to have
changed much.
We were meeting about once a week to compare patterns from our
respective research on ethnic violence in the former Yugoslavia and in
India, and to design a new study on the local repercussions of a global
event—the rise of the Scottish independence movement in the aftermath
of the Iraq invasion. Early on in our discussions, it appeared to us that
these cases are telling examples of dynamics that do not yet receive the
full attention they deserve in social psychology texts. Both in the former
Yugoslavia and in India, there is a striking contrast between, on the one
hand, grand narratives of entrenched ethnic or religious conflict, and,
on the other, the reality of fluid social relations and malleable identity
constructions on the ground. But it also struck us that we need to beware
v
vi Preface

of making too stark a distinction between fixity and fluidity. There are
periods when social categories and social relations do indeed remain con-
stant over extended periods and periods where they go through rapid
changes. Which one discovers is a matter of timing. In general terms,
it would be wrong to treat either as the norm. Rather, we need a more
historicised approach which allows us to identify and analyse the turning
points where social relations lose (or gain) their fluidity and where iden-
tity constructions become frozen (or unfrozen).
When we do take an historical view—whether in relation to our own
studies of the former Yugoslavia and of India, or indeed of elsewhere—it
becomes abundantly clear that, at different times, the cleavages in social
relations have been based on different social categories: sometimes people
confront each other in terms of nation and sometimes in terms of class,
or caste, or religion or ethnicity. Multiple differences have therefore to be
forgotten and multiple conflicts have to be put on hold before a critical
mass of people can think or act together in terms of any one of these, still
more before one can take it as self-evident that there is a long history of
conflict between Hindus and Muslims, Serbs and Croats, or whatever
other groups are seen as trans-historical entities.
So how are categories reconfigured? How could social relations become
frozen along ethnic lines and ethnicity become a matter of life and
death—as they did in Vukovar or in Sarajevo, in Ayodhya or in Gujarat,
or in many other places at different times in history?
Back in 2008 an impression arose which evolved into a critical thesis
that led us to re-assess the existing evidence and re-analyse our own. Such
dramatic disruption of the normal fluidity of collective identities cannot
just happen spontaneously. It requires such a violent shock to the system
that the fundamental ways in which we are able to relate to others are
changed and hence we begin to think and talk of the cleavages between
self and other, ‘us’ and ‘them’, in new ways. And not only must the shock
be violent but also violence is a powerful means of achieving such a
shock. Even if you never previously saw yourself and others in terms of,
say, religious identities, if you are attacked as a Muslim—or even hear
of others attacked as Muslims—can you afford to ignore the possibil-
ity that you, your family, your friends might be positioned as a Muslim
in future encounters? Can you therefore afford to avoid acting as if you
  Preface  vii

were a Muslim, thereby confirming others view of you as such and hence
fuelling a vicious spiral? Unwillingly, perhaps, but no less effectively, the
awful realities of violence affect the identities through which we organise
our everyday world.
This insight might seem relatively modest and mundane. But the more
we thought about it and the more we considered the implications, the
more the fabric of received wisdom on conflict and intergroup relations
began to unravel. Traditionally, conflict is seen as an output of identity
processes. There are two broad variants of this. On the one hand it is
argued that longstanding identities generate animosities that, given the
power to act, will result in violence. So, notably in the former Yugoslavia,
it is argued that the state may have temporarily suppressed the ancient
hatreds between groups in the region. But once that state dissolved,
people could express their hatreds in ways that still haunt the imagina-
tion. On the other hand it is argued that those with power and influence
deliberately invoke antagonisms between groups and deliberately incite
violence which ordinary people are incapable of resisting.
The three terms here—identity, violence and power—are ones that we
retain and which are central to our analysis. But we reconceptualise their
relationship. Instead of putting either identity or power at the start of
the process—as inbuilt features of our psychological make-up which all
too easily generate antagonism towards outsiders or conformism towards
cynical leaders—we also treat them as outcomes. Equally, instead of put-
ting violence at the end of the process we also place it at the start. Instead
of conceptualising a simple linear relationship between terms, we exam-
ine the multiple configurations of identity, violence and power. Violence
thereby becomes much more than the tragic end of the play. Above all, it
is the shock that serves to create and consolidate identities and thereby to
transform power relations during the next act.
However, one cannot alter the way in which identity, violence and
power relate to each other without rethinking these core constructs them-
selves. Our approach to each construct lays much more stress than usual
on meta representation and communication. That is, who we are, what
we think and what we do is as much a matter of what we think others are
thinking as of what we think ourselves, and also of what we think others
will allow us to do. It follows that reconfiguring identities may be best
viii Preface

achieved by altering our knowledge of others’ thoughts and intentions

rather than trying to change our internal beliefs. Our ability to gain that
knowledge—the relations of communication between people—thereby
gains centre stage.
We thus understand and analyse identities, violence and power as
social facts. That makes them no less real for those who experience
them and no less consequential in their effects. But it does mean that
they arise out of human activity and they shape human activity from
the inside. Indeed, identity, violence and power are the most power-
ful of tools for crafting our social worlds. By occluding old identities,
violence makes it impossible to mobilise people on the basis of those
erstwhile constituencies. By creating new identities, violence makes
possible new forms of mobilisation and thereby creates the power to
sustain new social worlds.
One way of seeing our model, then, is as a psychological example of a
mobilisation approach to intergroup relations and social conflict (some-
thing that may be common in other social sciences, but which is some-
thing of a departure for psychologists). It is a model that recognises the
importance of manifold forms of social practice and of social commu-
nication (not only from leaders to followers). As many scholars before
us, we don’t presuppose that ingroup identities and outgroup hatreds
are stable realities and we examine how and why people are mobilised
around particular identities to hate particular others. But we also exam-
ine the processes whereby, under extraordinary circumstances associated
with specific historical periods, these identities and these hatreds evolve
towards a more stable form. While we argue that these hatreds arise
within a contingent social order, we hope to elucidate how sometimes
they come to be seen as part of an inevitable order.

Writing While the World Goes By

We have described how we set out to write in 2008. Back then, we thought
that our argument might be pushing at an open door. Our focus was to
be on a set of intriguing conundrums that arise once one follows through
a constructivist view of identity: How can such fluid and malleable things
  Preface  ix

as collective identities sometimes become hard social facts? Why do people

sometimes actively support leaders whose politics have disastrous effects on
their own interests? What makes political violence such a dreadful vehicle
to re-pattern collective identities, impacting even on those who under nor-
mal circumstances would forcefully resist? Our assumption was that estab-
lishing a critique of essentialist approaches to identity would be the easier
and less original part of the work. It could function as a discrete backdrop
to the main stories of the book. After all, who in the early twenty-first
century could still seriously believe that differences in language, religion
or culture are sufficient to explain why people who are different from each
other would violently clash with each other?
But much has happened in the ensuing years, and looking back, early
2008 seems almost an age of innocence. It was before the financial crash,
the ensuing recession, unemployment, austerity and chronic global sense
of economic insecurity. It was before an explicitly ethnic party—the
Bharatiya Janata Party (BJP) and its leader, Narendra Modi, who pro-
mote the notion of India as a Hindu nation—won power in the world’s
largest democracy. It was before the Arab Spring, the ongoing violence
in Egypt, the conflict in Yemen, the collapse of the Libyan state, the civil
war in Syria, the rise of ISIS and the return of ISIS militants to their
European countries of origin. It was before people displaced by these and
other conflicts began to flee to Europe, before the spectacular rise of anti-­
immigrant populisms in even the most liberal of democracies, and before
the resurrection of borders and barriers between European nations. It
was before the return of war to European soil in the clashes between
Ukrainian forces and Russian-backed separatists, and before the Russian
annexation of Crimea.
In today’s more fractious world, it is all too easy to believe that that
ghosts of ancient and ineradicable conflicts have come back to haunt
us—or even that we are being punished for our temerity in denying their
existence. As we write now, in the spring of 2016, even the Eurovision
Song Contest (ordinarily that most innocent frippery) has become politi-
cally and ethically charged. The winning song, entitled ‘1944’, is about
the deportation of Crimean Tatars by Stalin with obvious reference to
the annexation of 2014. The lyrics begin “When strangers are coming/
They come to your house/They kill you all/and say/We’re not guilty/not
x Preface

guilty.” More prosaically, the notion that conflict arises out of a clash of
engrained identities has gained renewed legitimacy in public discourse
and policy decisions concerning war and peace. It has become a readily
available grid through which to understand and respond to contempo-
rary conflict.
Accordingly, what we had thought to be easily dealt with in terms
of a background to our studies has now become far more controversial.
What might once have been an open door has become stiff with age and
far more effort is needed to shift it. We could no longer just state a set of
assumptions about the constructed nature, and hence the contingency,
of identity, violence and power. We were in danger of using a language
to explain the world just at the point that the world was moving beyond
that language. The fear that our words would not even speak to a new
generation of students and scholars, for whom the book was primarily
intended, began to haunt us.
This fear led us to revise our manuscript to be more explicit about the
problems with fatalistic conflict theories. Accordingly, the first part of
the book spells out our criticisms of models that treat group identities as
immutable, conformity within groups as natural, and hostility between
groups as unavoidable. As we have already intimated, there are two vari-
ants of this view and we devote a chapter to each: Chapter 1 looks at the
idea that ordinary people are doomed to hate those from different groups
and leaders can, at best, mitigate the worst excesses; Chap. 2 looks at the
idea that ordinary people are doomed to follow leaders who incite them
to hate those from different groups.
The second part then provides a systematic outline of our own position
based on three key constructs: identity, violence and power. This involves
rethinking both how each of these relates to the others and also how the
construct itself should be understood. So, in Chap. 3, we start with i­ dentity
as something that is created through shared social practices and hence is
transformed through the disruption of existing practices. In Chap. 4 we
turn to violence as a particularly potent means of achieving such disrup-
tion and hence of reconfiguring the map of social identities and social
groups. Finally, in Chap. 5, we address power—more specifically, the role
of leadership and power politics in managing violence and identity.
  Preface  xi

In the third part, this conceptual model is applied to three case stud-
ies. In Chap. 6 we dissect the dynamics of religious violence in parts of
India. In Chap. 7 we examine ethnic violence in the former Yugoslavia.
In Chap. 8 we analyse anti-war protest in the UK, and more specifically
in Scotland, in reaction to the military invasion of Iraq. Since parts of
these case studies have been published before as journal articles, we were
able to present the findings here without disrupting the flow of the book
with too much methodological detail. We also opted to integrate each
case study with new, so far unpublished, material that extends both the
empirical and the conceptual reach of our analysis.

Beyond the Prophet Motive

The changes of recent years have led us not only to alter the structure and
stress of our argument but also to reconsider its impact and implications.
As lines of public debate shifted so that discourse that had been perceived
as extremist became normalised, we were led to become more reflexive
about the way that we—as social scientists—refer to the mainstream and
to the margins in public controversies. As, like everyone else, we were
continuously taken by surprise by the new givens and developed an ever
more uncomfortable feeling of running behind events, we were forced to
confront a foundational question: What exactly is our analysis for? After
all, the conventional justification of scientific analysis is that it allows us
to predict what will happen. If we (like everyone else) so obviously fail in
prediction, then what on earth is the point of what we are doing?
To address such deeply troubling questions, let us consider a further
case which is, as we write, still unfolding. We refer to Burundi, whose past
and present plights are largely ignored despite the fact that something like
a quarter of a million people were killed in past atrocities, c­ ulminating in
1972 and 1993, and that, over the last year alone a further quarter of a
million people have fled the country in fear of further violence.
At one level, the latest bout of violence was clearly foreseen. When the
first author travelled to Burundi in early 2014, the country seemed safe.
It was perfectly possible to talk in a relaxed manner with local researchers,
aid workers and activists. Many people spoke openly. Some of them were
xii Preface

happy to air their differences about the challenges facing Burundian soci-
ety in its attempts to deal with a legacy of violence. It was easy to travel
around the country and to visit commemoration sites, each of which told
a different story about the nature of this past violence. But among well-
informed locals, few expected this to last. With surprising consistency
and great precision many conversations alluded to the prospect that the
current period of calm and tolerance would come to an end within 18
months, to be replaced by a new period of heightened tension.
They were right. After President Pierre Nkurunziza decided to ignore
the two-term limit on office and, in April 2015, announced his inten-
tion to seek a third term, there were massive protests in the streets of
Bujumbura (the Burundian capital), a radio crack-down, a failed coup
attempt, and, eventually, vicious repression of the anti-third-term oppo-
sition. That is also when the flow into exile began. But if many people
foresaw these tensions starting from a long way off, there was a limit to
their prophetic powers. Once conflict had started, no one could tell how
it would develop. Everyone was perplexed as to how the next day might
turn out.
It was as if an understanding of the simple power calculus between the
main actors, a glance at the electoral calendar, and an awareness of past
events (elections in Burundi have repeatedly been tainted by violence)
was enough for well-informed observers to predict that a crisis would
occur, and even when it would occur. Just like clockwork, the president
would try to cling on to power, his rivals would cry foul and an almighty
stooshie (to use an evocative Scottish word for conflict) would break out.
But once events were in motion—once social forces had been moved
out of a stable equilibrium into a state of volatility—then the smallest of
causes could produce the largest of effects. It then became all but impos-
sible to spot what was coming and where things were going.
Strangely enough, as knowledgeable locals became more perplexed,
international observers became more certain about the focus of concern.
They had one question: Would the crisis lead to an outbreak of ethnic
violence? So, on 13 May 2015, while the coup attempt was still unfold-
ing, CNN splashed the headline “Amid fears of ethnic violence, coup
attempt reported in Burundi”. Meanwhile, the International Business
Times announced, “Africa watches Burundi coup to see if conflict spreads,
  Preface  xiii

reignites Hutu-Tutsi ethnic conflict”. The following day the UK Daily

Telegraph newspaper entitled its report: “Burundi shares ethnic balance
that led to Rwanda genocide—but this conflict is different”.
Here was prophecy. At the time these and suchlike headlines were
written, they didn’t describe the reality on the ground. If anything, they
were ahead of that reality. A Hutu general was leading a coup against a
Hutu president, with whom he previously fought together in the rebel-
lion movement, built up the powerful CNDD-FDD party, and served
as his head of intelligence services until a mere two months before. Both
loyal and putschist army factions were ethnically mixed, as was the state
administration and the protest movement in the streets. It was only after
the failure of the coup that things began to change. Some members of the
ruling party started to spread rumours that opposition to the president
was a Tutsi plot against a Hutu ruler and they combined this with fierce
repression against various Tutsis alongside their alleged anti-­presidential
‘sympathisers’. Would this ethnic frame succeed? Would it reignite past
horrors? Many Tutsis were not prepared to take the risk that it wouldn’t
and so started to flee. This exodus in turn seemed to validate the ethnic
frame. It certainly broke the back of the protest movement. As people
became afraid to gather together, many returned to the private sphere of
their homes. An uneasy calm returned to the streets combined with anx-
ious expectations of further violence in the neighbourhoods of Bujumbura.
This crisis has taught us two important lessons. The first concerns the
radical inadvisability of making predictions. In part, this is because of the
likelihood of getting things wrong. Beyond being able to posit in general
terms that a conflict might occur, we can see how, without the benefit
of hindsight, it is hard to know what forms a conflict will take, how the
conflict might develop or how it will turn out. We are not dealing with
mechanistic or deterministic processes here. The nature and course of
events depend upon how critical actors make sense of them, and sense
making is a non-deterministic, slippery, infinitely variable and infinitely
malleable process. So, by posing as prophets and by making predictions
we run a serious risk of making fools of ourselves.
But there is another, and possibly greater, problem: occasionally, our
predictions could turn out to be right, not because they were particu-
larly clear-sighted, but because they feed into an interpretation of events
xiv Preface

that increases the odds of the predicted outcome. Our prophecies may
end up as self-fulfilling in themselves or, more plausibly and more prob-
lematically, condoning the prophecies made by the most powerful con-
flict actors (those who are in the best position to make their prophecies
become true). The underlying issue here is that, if we accept that identi-
ties and conflicts are constructed and contingent rather than natural and
inevitable, and if we also accept that identities can be rooted in the way
we will be seen by others as much as in the way we see ourselves (like
those who fled for fear of being apprehended as Tutsis) then we can never
be certain of providing an innocuous definition of events. We are not like
taxonomers defining reality from the outside. We are, whether we like it
or not, insiders who are part and parcel of the conflict process. If we take
ethnic categories as givens, and if anyone takes us seriously, then people
will increasingly expect to be seen in ethnic terms and act accordingly.
One reaction is to insist that no one takes academics seriously or even
notices our obscure scribblings. Quite apart from the ironic nature of a
defence of academia based on the fact that academics are useless, the claim
tends to deny certain historic realities. Over the last two centuries, social
scientists have been at the very core of various national projects and cer-
tain disciplines—history, of course, but also anthropology, archaeology
and others—arguably arose precisely in order to sustain such projects. But
also, as we shall see at various points in the book, politicians have explicitly
drawn on academic analyses (e.g., ethnic conflict is inevitable) to draw
policy conclusions (e.g., there is no point in intervening to try and stop it).
The second lesson we learn from Burundi takes us from the war of
words to the war on words and on those who spread the word. Much
of the early conflict surrounded control of the radio, the country’s only
mass media. Indeed mass protests in Bujumbura were triggered when the
government decided to close down the privately owned station, Radio
Publique Africaine, which had been supporting opposition to Nkurunziza’s
third-term presidential bid.
The subsequent 13 May coup against Nkurunziza began with an
attempt to seize the capital’s main radio station, where the rebellious
soldiers were met by loyalist troops. The two army factions exchanged
gunshots before the rebels, realising their inability to take control of the
station, decided to surrender. The radio station was damaged during the
  Preface  xv

short bout of fighting. State forces then took advantage of the resulting
confusion in order to destroy Bujumbura’s remaining four main radio
stations. While the domestic media were forcibly silenced, the interna-
tional media remained silent out of disinterest. Little space for collective
discussion remained. Burundians were left without appropriate means of
coordinating their understanding of the new realities. Because the gov-
ernment had declared that there was no crisis, it became difficult even to
talk about the events without being open to accusations of treason. In
terms which we will develop further in due course, people were left in a
state of epistemic isolation.
The situation brings to mind Foucault’s use of Bentham’s panopticon as
a metaphor for the operation of power. This refers to a structure whereby
an authority at the hub can see all those arranged around the rim, but
these people cannot see each other. They are therefore held in the gaze of
that authority without being able to draw upon their fellows for support.
In such a situation, where communication between people is excluded, it
becomes impossible to counter the voice of authority. Authority thereby
retains the unique capacity to define events, to construe identities and to
shape collective action. If only it were possible to break down the parti-
tions between those around the rim, things would be very different.
And with these thoughts in mind we can begin to return to the point
of writing this text. The first purpose of any words should be to open up
conversations—in this case concerning the nature of conflict. If there is
one thing we would hope to achieve with this book, it would be to bring
people together in new configurations to address received truths and con-
sider new perspectives. It is for this reason that we have ­deliberately ­written
our book in a way that transcends traditional d ­ isciplinary ­boundaries. We
have also sought to make the book as clear and as lively as possible in
order to be accessible for those outside the academy who have to handle
and live with the many conflicts which cleave our world: practitioners,
politicians and the interested public. We have tried to avoid jargon where
possible, to define it where not, and to use theory only as far as it sheds
light on phenomena of broader interest. Our hope is that, even if readers
disagree with some of our analyses, they will feel that they have at least
found out something about conflicts in the world and something that
helps them in questioning the available explanations of conflict.
xvi Preface

That leads on to another—central—reason for our book in particu-

lar, and for critical new theory in general. More and more, psychologists
tell us that human beings have an engrained craving for a world that is
known and certain. The academic enterprise is often justified in terms
of answering that call: by promising to increase our certainty about the
world, how it works and how it will work out. However this is a danger-
ously one-sided perspective, because sometimes the value of academic
work is (and should be) to disrupt existing certainties, to provide new
ways of seeing and thinking, to raise questions where there were none
before. Indeed one of the main purposes of new theory is to raise new
questions and not simply to provide new answers. That is certainly how
we want our own contribution to be judged.

Back to Perplexity

To be more concrete, our ambition for this book is as follows. Having

listened to the news in print, on television or online, and having heard
about Sunnis fighting Shias, Kurds fighting Turks, Hutus fighting Tutsis
or whatever—and maybe even having heard some pundit explaining that,
back to time immemorial, Sunnis have always fought Shias, Kurds have
always fought Turks, and Hutus have always fought Tutsis—we hope that
our readers will experience a heightened sense of dissatisfaction.
More specifically, we hope that readers will be left with two niggling
questions. The first is to ask ‘Why Sunni and Shia, and why right now’?
Since it obviously cannot be true that every Sunni has hated and aggressed
against every Shia at every point in history (or does so today), how has
reality come to be seen in terms of these groupings in a specific time and
place? Why have these particular categories been chosen over possible
alternatives—who is promoting these categories and why are they doing
so? In the same way that a word, endlessly repeated, can come to sound
strange to the ear, so the categories endlessly repeated in the media may
eventually begin to sound absurd—and reading this book might speed up
the process. How mad that, out of a huge range of possibilities, we should
have ended up dividing and treating people in this specific way. And how
much more absurd to suppose that this must ever be so.
  Preface  xvii

The second question is how did we get from ‘Sunni and Shia’ to ‘Sunni
against Shia’? Is a mention of identity sufficient to account for conflict? Is
invoking identity and past conflicts between those of different identities
enough to trigger present violence—at least when there was sufficient
power to turn malevolent intentions into actions? Is violence a tragic but
unintended outcome of identity dynamics or is it a critical component
of those dynamics?
In sum, our aim (like the court jester or the poet) is principally to pro-
vide new angles for looking at the familiar and making it seem surprising.
Where there was complacency we wish to bring unease. Where there was
certainty, we wish to bring perplexity. It is not simply that we have no
aspiration to play prophets. If anything, we aspire to escape from an age
of prophecy. Certainty closes down the future. It keeps us on the straight
and narrow. We prefer to think about the future as open, to envision mul-
tiple possibilities, to facilitate debate and, hence, extended choice over
which of these possibilities are worth pursuing. To the same end, we aim
to expose the means by which academics, activists or autocrats seek to
close down our options. Violence, we contend, is prime amongst these.

Lausanne, Switzerland Guy Elcheroth

St Andrews, UK  Stephen Reicher
 Acknowledgements

One of our core arguments in this book is that human understanding

and action derive more from what happens between us than from what
happens within us. Who we are, what we think and what we do are all
functions of what others think of us and act towards us, and, yet more
recursively, what we think others think of us and will do to us. Our own
ideas are inextricably intertwined with those of others. They are impos-
sible to separate out. And what is true of the argument in the book is
equally true of the book itself.
Although there are parts of the text which derive more clearly from the
concerns from one or the other of the two authors, over time the conver-
sations between us have shaped and reshaped these thoughts to the extent
that we can no longer draw clear lines and claim ‘this is mine’ and ‘this is
his’. It is truly a collaborative enterprise which neither of us could have
completed alone and both of us have benefited immeasurably by those
many conversations with each other.
But the web goes wider. Many people, both wittingly and unwittingly,
have contributed to this project. We have gained immeasurably from
their work and from our dialogues with them. As with each other, it is
impossible to determine exactly which idea belonged to whom. In this
way, although the book was actually written by the two of us (and we take
full responsibility for all the arguments it contains) it is also a collective
accomplishment.
xix
xx Acknowledgements

Of those others who contributed, first and foremost come Sandra

Penic and Rakshi Rath. The Croatian case study wouldn’t have been pos-
sible without Sandra and the Indian case study wouldn’t have been pos-
sible without Rakshi. The two relevant chapters (6 and 7) largely draw on
material that Sandra and Rakshi collected as part of their thesis work. We
are grateful to both of them for allowing us to reuse this material here and
also for their subtle and fine-grained insights. It was only through these
that we were able to understand the material and appreciate its signifi-
cance for our broader argument. Another person who provided invalu-
able practical assistance is Nicole Repond, who checked and completed
the references.
More broadly, many of our colleagues in both Lausanne and St Andrews
have been generous with their criticisms, their comments and their sug-
gestions both on our general arguments and on preliminary drafts of the
chapters. And there are colleagues beyond our home institutions who
have played an equally valuable role. The material on obedience and lead-
ership in Chaps. 2 and 5 and also the Indian research in Chap. 6 comes
out of many years of productive collaboration with Alex Haslam and
Nick Hopkins. The research presented in Chap. 7 is rooted in a long and
fruitful collaboration with Dario Spini, and several concepts introduced
in Chap. 3 originate in joint work with Willem Doise.
Without doubt, the book would have been different and poorer without
regular, tough but cordial exchanges over the last three years with friends
and colleagues from all five continents gathered within the Pluralistic
Memories Project. These exchanges were an invaluable tool in temper-
ing our ethnocentrism and for helping us to consider how the problems
discussed here might look from different parts of the worlds. Visits to Sri
Lanka, Burundi and Palestine have been eye-opening in many regards.
They provoked a whole set of revisions. They also generated the seeds for
what we hope to become the next generation of case studies.
Finally, we could not have completed this book without generous
material support. The Universities of Lausanne and St Andrews pro-
vided time, space and stimulating environments for us to write. The UK
Economic and Social Research Council (ESRC) funded two separate
research projects—one on Milgram and obedience and the other on col-
lective action in India—which fed much material into our text. The Swiss
  Acknowledgements  xxi

National Science Foundation (SNSF) provided financial support at the

start of our journey, through a fellowship which, in 2007–2008, allowed
the first author a one-year stay as a visiting research fellow in St Andrews.
Towards the end of the journey, the SNSF has helped again, through
the research for development scheme (r4d), by funding the Pluralistic
Memories Project since 2014.
 Contents

Part I Critique   1

1 Hateful Groups and Weak Powers?   3

2 Evil Leaders and Obedient Masses?  39

Part II Model  71

3 Identity: The Group as a Collective Performance  73

4 Violence: How Collective Shocks Transform Social

Practices  99

5 Power: The Role of Leadership at Critical Junctures 129

Part III Case Studies 153

6 Riots, Religion and the Mobilisation of Communal Hatred

in India (with-Rakshi Rath)   155
xxiii
xxiv Contents

7 Ethnic Violence in the Former Yugoslavia: From Myth

to Reality (with Sandra Penic)   183

8 British Warriors and Scottish Voters: When ‘Rallying

the Nation’ Backfires   215

Conclusion   243

Index   261
 List of Figures

Fig. 8.1  Moral principle positions and the timing of debates (above)

as well as social cleavage positions (below) according to their
coordinates along two dimensions defined by an MCA of
their joint occurrences within parliamentary interventions.
Arguments related to the ‘home front’ appear in bold.223
Fig. 8.2  The positions of individual interventions (marked by party
affiliation) according to their coordinates along two dimensions
defined by the MCA (see Fig. 8.1) of the joint occurrences of
moral principles, social cleavages and the timing of debates.
Interventions from members of separatist parties appear
as filled grey triangles. 225

xxv
 List of Tables

Table 7.1  Rates of acceptance of mixed marriage before and after

the war, across eight political entities (sorted by decreasing
level of pre-war ethnic diversity) 185
Table 7.2  Perceived sources of inter-group inequality in pre-war
Yugoslavia188
Table 7.3  Popular support for future institutional scenarios
in pre-war Yugoslavia 189
Table 7.4  Perceived ethnic conflict in pre-war Yugoslavia 189
Table 7.5  Affiliation to different territorial entities, before and
after the war 191
Table 7.6  Estimate of the population share for three types of
identification in Croatia and Serbia; with group means of
attachment, glorification and collective guilt acceptance 195
Table 7.7  Summary of the results of the thematic coding of
the parliamentary debate held on 13 and 14 December 2005 202
Table 8.1  A structured inventory of moral arguments: generic moral
judgements and nested concrete arguments invoked by the
‘external front’ and the ‘home front’ as well as frequencies
before and after the invasion 221

xxvii
xxviii  List of Tables

Table 8.2  A structured inventory of categorical arguments: generic

social cleavages and nested concrete arguments invoked
by the ‘external front’ and the ‘home front’ as well as
frequencies before and after the invasion 222
Table 8.3  Multivariate predictors of the separatist opposition
(above) versus the Labour majority (below) vote:
partial logistic regression coefficients 235
Part I
Critique
 1
Hateful Groups and Weak Powers?
Ancient Hatreds (And Other Fables)

In the Orson Welles film Mr. Arkadin, the eponymous hero recounts a
fable:

And now I’m going to tell you about a scorpion. This scorpion wanted to
cross a river, so he asked the frog to carry him. No, said the frog, no thank
you. If I let you on my back you may sting me and the sting of the scor-
pion is death. Now, where, asked the scorpion, is the logic in that? For
scorpions always try to be logical. If I sting you, you will die. I will drown.
So, the frog was convinced and allowed the scorpion on his back. But, just
in the middle of the river, he felt a terrible pain and realized that, after all,
the scorpion had stung him. Logic! Cried the dying frog as he started
under, bearing the scorpion down with him. There is no logic in this! I
know, said the scorpion, but I can’t help it—it’s my character. Let’s drink
to character.

This fable applies well to the topic of our book. The simplest story about
violence between groups is that violence is just what groups do. It might
not be palatable. It might not seem logical. Often, indeed, those who
attack others are themselves consumed by the ensuing violence. But

© The Author(s) 2017 3

G. Elcheroth, S. Reicher, Identity, Violence and Power,
DOI 10.1057/978-1-137-31728-5_1
4  Identity, Violence and Power

groups can’t help it. It is in their character—the dark side of humanity

divided into groups—whether we want to ‘drink’ to it or not.
The simple story does not imply constant violence between groups,
and is therefore immune against refutation by the historical reality of
extended periods of peace. Groups will not always act on their hatreds,
especially when there is a powerful centre to counteract such centrifugal
tendencies from spiralling out of control. But take away such dampening
forces and the violence will once more flare into the open. Such positions
have gained contemporary momentum in two subsequent waves: first,
since the fall of the Berlin Wall and the ensuing bout of conflict both
within the old Soviet Union and the former Yugoslavia; second, in the
wake of the post-nine-eleven wars in Afghanistan and Iraq, the subse-
quent rise of the Islamic State group and the ideological framing of the
wars fought against them all in terms of ‘clashes of civilisation’.
When, after the end of the Cold War, conflicts arose between such
groups as Armenians and Azeris, Bosniaks and Serbs, Kyrgyz and Uzbeks,
an often expressed view was that Communist rule may have suppressed
fundamental ethic identities and ethnic antagonisms—but it never did
eradicate them, nor could it have. As the authoritarian regime crumbled,
so the identities and the conflicts re-emerged—if anything, more virulent
than ever for having been suppressed for so long. From this perspective,
after decades of the iron rule of ‘ideology’, the stage was finally cleared for
the historic revenge of ‘ethnicity’.
As illustration, one only has to turn to two statements made within
months of each other in 1993 by the leaders of the USA and the UK. On
20th January, in his Inaugural Address, President Clinton warned that:

Today, a generation raised in the shadows of the cold war assumes new
responsibilities in a world warmed by the sunshine of freedom but threat-
ened still by ancient hatreds and new plagues.1

On 23rd June, Prime Minister Major addressed the British House of

Commons, referring specifically to conflict in the former Yugoslavia, but
spelling out the argument subsumed in Clinton’s brief statement:

 The full text is available at http://www.presidency.ucsb.edu/ws/?pid=46366.

1
 1  Hateful Groups and Weak Powers?  5

The biggest single element behind what has happened in Bosnia is the
collapse of the Soviet Union and of the discipline that that exerted over the
ancient hatreds in the old Yugoslavia. Once that discipline had disappeared,
those ancient hatreds reappeared, and we began to see their consequences
when the fighting occurred. There were subsidiary elements, but that collapse
was by far the greatest. (Hansard, 23rd June 1993, col. 324)

Both comments were made against the backdrop of the war raging in
Bosnia–Herzegovina, whose atrocities alarmed Western public opinion
and urged Western intellectuals to re-work their analytic concepts. The
year following the end of the war, 1996, Samuel Huntington published
his famous article on the Clash of Civilizations, to which we shall return in
detail later in this chapter. The language used in this article has provoked
much controversy since its publication, in academic and non-­academic
circles alike. But despite all sceptic voices, Huntington appeared to have
generated a widely available frame, to which many have come to refer
again in more recent years, in a global context of mounting tensions.
Prominent voices have come to claim that the clash of civilisations is now
materialising.
Following the shooting of French cartoonists in Paris in January 2015,
and the subsequent international stigmatisation of religious censorship in
the name of Islam, Turkish President Tayyip Erdogan issued the following
warning: “Despite all our efforts to prevent it, the clash of civilisations
thesis is being brought to life”. The year before that, a Huffington Post
essayist had argued that “We should have seen it all coming”, alluding
to the sombre prophecies spelt out in Huntington’s article: “The future
(the article) describes has become our present and the challenges it raises
will continue to define the global order for decades to come” (Johnson,
2014).
The same year, former French Minister of foreign affairs Dominique
de Villepin gave an address at the World Cultural Forum. He took this
opportunity to comment on the way reality had surpassed the fears he
had prominently expressed more than a decade earlier, when France
did not follow the United State’s call to invade Iraq. In February 2003,
Villepin had warned against the consequences of an invasion of Iraq in a
resounding speech given at the UN Security Council, wondering whether
6  Identity, Violence and Power

“such intervention (would not) be liable to exacerbate the divisions

between societies, cultures and peoples, divisions that nurture terrorism?”
The title of De Villepin’s 2014 talk—“Will the clash of identities con-
sume us?”—was as evocative as its opening phrase: “Since 2001, we have
entered a vicious clash of civilizations. We have not seen the bottom of
it yet”. His description of the calamitous state of Iraq a decade after the
invasion names a clear culprit, the cult of identity:

In Iraq today, the national feeling almost disappeared behind ethnic and
religious identities. But such identities exist only as differences from each
other. They become hysterical, incompatible and intolerant of all diversity.
The cult of identity is a selfish and brutal vision of the world than can lead
to the most terrible crimes.

Epistemologically, the most troubling aspects of the clash of civilisations

debate are the thin lines between comment and warning, between pre-
diction and prescription, along which it evolves. These ambiguities are
well illustrated by the following words used by New York Times editorial-
ist Roger Cohen in a controversial opinion piece, published in February
2015:

To call (the Islamist) movement, whose most potent manifestation is the

Islamic State, a “dark ideology” is like calling Nazism a reaction to German
humiliation in World War I: true but wholly inadequate. There is little
point in Western politicians rehearsing lines about there being no battle
between Islam and the West, when in all the above-mentioned countries
tens of millions of Muslims, with much carnage as evidence, believe the
contrary.

There are actually two remarkable things about Cohen’s argument. The
first is that it introduces a notion of “inadequate truth”: the idea that
even when complex accounts are analytically accurate, it can still be mor-
ally wrong to utter them. The second lies in the implied logic that their
hate constrains our reality. Cohen perceives (on whatever basis) “tens of
millions of Muslims” who believe there is a battle between Islam and
the West, and he therefore infers there is no other choice left than to
accept the battle as a fact. Doing so, the editorial appears to call for a
 1  Hateful Groups and Weak Powers?  7

self-imposed restriction of epistemic freedom, elevated to the status of a

moral duty. The rationale seems to be that complex truths are a danger-
ous luxury in a world where many angry people believe in simple stories.
It is rather easy to see how such calls, were they followed diligently, can
lead to a vicious spiral of rapidly reducing freedom of interpretation and,
thereby, of options for action. If whenever you anticipate that certain
others see the world in a black-and-white fashion you need to adopt the
same view, then the imagination of a few will fatally end up in leading all
to see the world in black and white, and act accordingly.
What is explicit in all these quotes is the Hobbesian notion that peo-
ple have an inherent tendency to violence which will inevitably manifest
itself once state structures dissolve and people return to a state of nature.
Moreover, this bestial nature is tied to collective identities: it is pursued
and legitimated in the name of collective interests, not personal gain; it
is targeted against collective enemies, not personal foes. However, what is
perhaps more telling, precisely because it is taken for granted rather than
stated explicitly, is an assumption that these collectivities will be organised
around deep-rooted communal (ethnic, religious or cultural) fault lines.
This presupposition that there is a primordial manner in which people
define themselves down the centuries is what ensures continuity between
the past and the present, why people see events of many hundreds of years
ago relevant to them currently, and why ancient battles remain a matter of
what ‘they’ did, and continue to do, to ‘us’ in the here and now.
Whether those who take on board this assumption are right or wrong
in arguing that present conflicts reproduce past history, what is undoubt-
edly true is that their explanations of conflict reproduce past accounts
of history. Indeed, the quoted depictions of ethnic violence following
the break-up of the Soviet empire eerily echo accounts of similar vio-
lence following the break-up of the Ottoman and Austro–Hungarian
empires. Take, for instance, the words of the Nobel Prize-winning author
Ivo Andric, who wrote about the violence which erupted in Visegrad in
1914:

That wild beast, which lives in man and does not dare to show itself until
the barriers of law and custom have been removed, was now set free. The
signal was given, the barriers were down. As has so often happened in the
8  Identity, Violence and Power

history of man, permission was tacitly granted for acts of violence and
plunder, even for murder, if they were carried out in the name of higher
interests, according to established rules, and against a limited number of
men of a particular type and belief … (1945, pp. 282–283)

There are very different ways, however, of explaining the rise in eth-
nic or religious violence since the end of the Cold War. The same year
as Huntington published his Clash of Civilizations article, Brubaker and
Laitin (1998) published a theoretical paper that identified as the key fac-
tor the disappearance of macro-political cleavages between capitalism and
communism. This removed any political incentive to frame conflicts ideo-
logically and consequently enhanced the relative payoff of ethnic conflict
frames. After 1991, a rebellion fought in the name of communism would
no longer recruit international allies to the side of the insurgents. However,
if one fought in the name of overturning ethnic repression or in order to
achieve national self-determination, then it might be possible to recruit the
support of ethnic ‘kin groups’ in other states and of members of the ethnic
diaspora. It might also win the support of third-party neutrals and interna-
tional human rights organisations.
There is much to recommend such an approach. First, it doesn’t treat
conflict as inevitable, nor does it treat the framing of conflict in terms of
particular group memberships as inevitable. Rather, it sees such categories
as resources, which are actively invoked for the purposes of mobilising
support. It therefore points to the importance of leadership. It also looks
to the importance of the contemporary context (and not only events of
the past) in determining which categories are employed, and, more to
the point, which categories are successfully employed. It therefore opens
up an investigation into why and when the appeals of leaders succeed
in mobilising the masses. The question, then, is not whether a culture
provides tales in which an evil ethnic foe is set against a virtuous ethnic
ingroup. Such tales are indeed widely available. The question is rather
why and when these tales are taken up, woven into political rhetoric, and
why they momentarily work (cf. Nirenburg, 1996).
But we are getting ahead of ourselves. We shall discuss such questions
in detail in later chapters. For now our point is that approaches such as
that of Brubaker and Laitin, for all the respectful recognition they have
 1  Hateful Groups and Weak Powers?  9

gained among peers, appear not to have decisively shaped the way most
people, policymakers and media commentators think about conflict, or at
least about how they talk about conflict in public. They are still eclipsed,
in political communication and public debates, by those who read the
present as the eruption of a long and troubled past. While constructivist
thinking over the past decades undoubtedly had a profound impact on
the social sciences, and on conflict studies in particular, it does not appear
to have had a similar impact beyond academia.
At first glance, this might be understood as a consequence of aca-
demics’ lack of ability, perhaps of motivation, to get the more complex
stories out and make them relevant to those interested in dealing with
real-world problems—the ivory tower cliché. But, as many examples
throughout this book will illustrate, the problem might as well reside
in a form of ambiguity within the academic field itself. While very few
social scientists at present would enthusiastically self-declare as ‘primor-
dialists’, or argue in their theoretical writings that identity is immu-
table, many do adhere to research practices that treat their research
subjects as if they had one overarching and stable group belonging,
which orients their perspective on the world—one that informs the
research design and data analysis.
In our post-Cold War world, where the loudest and most powerful
voices describe conflict in ethnic terms, it therefore becomes ever more
critical for social scientists to ask why there has been such a turn rather
than to follow it ourselves. That is why, in what follows, we treat the three
prevailing narratives of collective identity formation as metaphors. In so
doing we express our understanding that they are better seen as frames
which guide and limit the evidence collected about the nature of identity,
rather than as a reflection of the evidence itself.

1.1 Metaphors of Identity

In this section, we shall examine three broad metaphors, which are used
to explain how people divide into different and antagonistic groups. The
first likens group identities to kinship, the second to civilisations and
the third to games. We do not suggest that these cover the full range of
10  Identity, Violence and Power

understandings of identity that have been developed in the social sci-

ences. Indeed, the main goal of Chap. 3 will be precisely to show that
alternative understandings are available, and in Chaps. 4 and 5 we will
discuss their implications for the study of collective violence and of strug-
gles over power. For the moment, our goals are more restricted. We seek
to highlight how common sense metaphors frame and constrain the way
social scientists look at group identities and the way they link to violence.
We also seek to draw attention to the social and political consequences of
these understandings.

1.1.1 Identities Defined by Descent: The Kinship

Metaphor

Descent is central to most definitions of ethnicity. For example, Fearon

(2008, pp. 852–853) states that “In ordinary English usage, the term
‘ethnic group’ is typically used to refer to groups larger than a family in
which membership is reckoned primarily by a descent rule […] That
is one is or can be a member of an ethnic group if one’s parents were
also judged members”. In this view, markers of ethnic identity such as
‘common language, religion, customs, sense of a homeland, and relatively
dense social networks’ might describe the group as a whole, but they don’t
define who belongs to it as an individual. In Fearon’s words, “Becoming
fluent in the language, manners, and customs of Armenia will not make
me ‘ethnically Armenian’. The key constitutive feature is membership
reckoned by descent”.
If people are born into ethnic groups, it is only a short step to argue
that ethnic identity is also something that derives from birth—from biol-
ogy rather than social experience. Primordialist conflict theorists read-
ily make this step. As Harvey (2000) explains, not all group affiliations
are equal. Rather, “from an ‘evolutionary’ perspective […] ethnic ties are
inherently more potent (and fit) as an organizing force than, say, ties
based on class or occupation. Individuals are bound to their ethnic group
by virtue of some ‘absolute import attributed to the very tie itself ”’. In
evolutionary language, social markers of identity, like ‘language, dia-
lect, customs, diet’, are relevant only in terms of ‘phenotypic matching’.
 1  Hateful Groups and Weak Powers?  11

They are mere cues to biological relatedness. In more technical terms,

they serve as a ‘kin recognition mechanism for humans’.
The significance of this mechanism is that it informs us with whom
we share a common reproductive interest, and hence who will be on our
side (and who won’t) when push comes to shove. This, so the evolution-
ists claim, is universally true, but they do acknowledge that the types of
situation which invoke kinship categories vary across time. Melotti, for
instance, quoted by Harvey (2000, p. 51), asserts: “when man had more
or less mastered the inimical forces of his … environment (such as hun-
ger, cold, and predatory animals), war became the main selective factor
in human evolution”.
Harvey helpfully puts the various steps of the argument together:
“humans tend to bond for evolutionary reasons, primarily to enhance
reproductive success, and enhancing kin survival through reproductive
success is the key to evolutionary explanation for ethnic conflict” (p. 41).
This is effectively the fable of the scorpion and the frog all over again, but
this time taken a step further. It is not just that groups conflict because
that is the nature of groups. It is that humans conflict along ethnic lines
because that is in the nature of human beings. Whether we like it or not,
and however much we try to work against it, in the end our racist nature
will out and ethnic mayhem will ensue.

1.1.2 Identities Defined by Opposing Worldviews:

The Civilisation Metaphor

Not everyone who explains conflict in terms of inevitable categories sees

these categories as a matter of descent. Indeed, one of the most famous
such explanations—Huntington’s (1993) ‘clash of civilizations’ thesis—
takes what seems at first glance to be a very different approach.
For Huntington, identities are not rooted in biology. He makes no
reference to evolutionary processes or to reproductive instincts in order
to explain why categories form. For him, categories relate to ‘civilisa-
tions’—what one might call ‘ways of seeing’—rather than set forms of
being. Moreover, he sees these civilisations on a very large scale. There
are, Huntington contends, no more than seven or eight such civilisa-
12  Identity, Violence and Power

tions in total, whose boundaries largely overlap with those of the main
world religions: “Western, Confucian, Japanese, Islamic, Hindu, Slavic-
Orthodox, Latin American and possibly (sic) African civilisation” (p. 25).
One might think that such an approach would open the way to a
rather more flexible approach to categories. After all, our biology and our
biological relatedness to others may be fixed, but, especially in an increas-
ingly globalised world with massive movements of populations from con-
tinent to continent, one’s ‘civilisation’ would seem more open to change.
But that is not Huntington’s own view. He suggests that we can no more
overcome barriers based on civilisation than those based on biology. This
is because of the very long history that produced them—a history that
reaches back much farther than particular political belief systems which
prioritise alternative categories—such as class. In part because of this his-
tory, civilisational identities act as fundamental filters which affect all
aspects of the ways in which people perceive and experience social reality.
Huntington provides a long list to underline his point:

The people of different civilisations have different views on the relation

between God and man, the individual and group, the citizen and the state,
parents and children, husband and wife, as well as differing views on the
relative importance of rights and responsibilities, liberty and authority,
equality and hierarchy. These differences are the product of centuries. They
will not soon disappear. They are far more fundamental than differences
among political ideologies and political regimes. (1993, p. 25)

Such unbridgeable differences make agreement and co-operation across

civilisations difficult, if not impossible. They make misunderstandings,
disagreements and conflicts likely, if not inevitable. But, as Huntington
argues next, it is not simply that civilisational identities themselves are
fundamental, fundamentally opposed and resistant to change. It is also
that individuals cannot change their civilisational identity:

In the former Soviet Union, communists can become democrats, the rich
can become poor and the poor rich, but Russians cannot become Estonians
and Azeris cannot become Armenians. In class and ideological conflicts,
the key question was ‘Which side are you on?’ and people could and did
 1  Hateful Groups and Weak Powers?  13

choose sides and change sides. In conflicts between civilisations, the ques-
tion is ‘What are you?’ That is a given that cannot be changed. (p. 27)

At this point, for all their seeming differences, the civilisational and the
kinship accounts of identity become all but indistinguishable. Although
they use different means to do so, both essentialise identity as necessary,
eternal and inescapable. In the end, the story is the same. You can mask
them for so long, you can seek to replace them with more contingent
categories, but in the end people will revert to fundamental cleavages.
Hence, Huntington uses the past tense to refer to those ‘class and ideo-
logical conflicts’ which are now consigned to the dustbin of Soviet histo-
ries. He uses the present tense for ‘conflicts between civilisations’ which,
because they have very old roots, are far more relevant for the present and
future. On this basis he is able to present a map of Europe dating back
to 1500, which marks the “eastern boundary of Western Christianity”
(p. 30), and use it both to explain conflicts at the end of the twentieth
century and also to prophecy that the “next world war, if there is one, will
be a war between civilizations” (p. 39).

1.1.3 Identities Defined by Competition: The Games

Metaphor

The third metaphor, the games metaphor, may seem to sit very oddly
with the two we have just discussed: ethnicity as kinship and ethnicity
as civilisation. To start with, whereas these others root conflict in the
differences between groups (and the differences in what they believe in,
care for or aspire to), the games metaphor roots conflict in similarities of
belief, value and aspiration. It is because we want the same thing—but
can’t all have it—that we fight others for the commonly desired prize.
Moreover, whereas kinship and civilisation approaches explain inter-
group relations in general, and conflict in particular, through the past,
games approaches situate their explanations firmly in the present. It
is not the trans-historical essence of the group that matters—it is the
particular set of circumstances they find themselves in which matter.
As ­circumstances change (and it is in the nature of circumstances to do
14  Identity, Violence and Power

so), a games approach suggests that group relations will change. Indeed,
this was the core point that the classic ‘games’ studies sought to show. In
effect, then, the core challenge we all face is not the primordial ethnic
or cultural group but rather (to borrow the famous quote which Harold
Macmillan may or may not have actually uttered) “events, my dear boy,
events”.2
According to the games metaphor, intergroup relations are essentially
structured like sports competitions. The whole point of a football team
in a football league, say, is to play against and to win over other teams.
Players could try to get along with each other and avoid competition, but
at that point they would cease playing football and being league teams.
So, it is argued, once competition is part of the very definition of the
group context, then rivalry and distrust and even hostility seem derived
almost as a logical consequence. Such theories are often referred to as
‘realistic’ conflict theories because they assert that intergroup hostility is
neither irrational nor accidental but rather based on an objective reality
of conflicting interests. The only way our team can win the game is to
make their team lose. We therefore want them to lose, we try to make
them lose and we rejoice at their misfortune.
The foundational and prototypical studies in the ‘realistic conflict’ tra-
dition are those conducted by Muzafer Sherif and his colleagues in the
late 1940s and early 1950s using American boys at summer camps—most
famously, the 1954 ‘Robbers Cave’ study (Sherif, Harvey, White, Hood
& Sherif, 1961/1988). The whole point of these studies was to use the
camps as a blank canvas, creating groups that had never existed before,
which had no history at all (let alone a history of antagonism), and then
creating different relations between them to see what would ensue. So, in
the 1954 study, the boys were divided into two groups: the Rattlers and
the Eagles. At first, they didn’t know of each other’s existence, but then
they were brought together in competitive games: baseball, tug-of-war
and so on. This immediately created animosity. Friendship choices across
groups were strikingly rare; the boys began to stereotype members of the
other group in derogatory terms. They began to denigrate them and their

 For a history of the phrase, see http://www.telegraph.co.uk/comment/personal-view/3577416/

2

As-Macmillan-never-said-thats-enough-quotations.html.
 1  Hateful Groups and Weak Powers?  15

achievements. Most dramatically, they began to attack each other’s huts,

stealing and vandalising property.
Sherif ’s most fundamental message was that antagonisms between
groups can be turned on and turned off by circumstances and, as a con-
sequence, even the most peaceful people can be made violent. This is
summed up in one of the most important quotes in the history of social
psychology, written to describe the point at which the conflict reached
its zenith:

If an outside observer had entered the situation after the conflict began …
he could only have concluded on the basis of their behaviour that these
boys (who were the ‘cream of the crop’ in their communities) were either
disturbed, vicious or wicked youngsters. (Sherif & Sherif, 1969, p. 254)

This work and these words have had considerable influence across the
social sciences. Somehow ironically, they have been taken on board by
ethnic theorists when they formulated ethnic competition models. These
view interethnic conflict as grounded in objective conflicts of interest,
typically over the share of economic rewards (e.g., Olzak, 1992). The
scarcer a resource, and the more a resource is valued, so the more com-
petition and the more antagonism there will be between ethnic groups.
Hence, prejudice, hate crimes and interethnic violence flourish where
there is economic scarcity and/or high immigration flows.
But even if these theories borrow the games metaphor to explain when
groups conflict, they remain fundamentally at odds with games theo-
ries in explaining what groups conflict. So, whereas Sherif saw groups as
defined and constituted by the game, ethnic competition theorists intro-
duce a notion of groups as prior to the game, as defined long before any
game has ever started.
That is, Sherif saw group boundaries, memberships, solidarities and
antagonisms as essentially arbitrary. They are created by the ways in
which people are put in competitive (or co-operative) relations within
the situation itself. Thus, when the social structure is such that a gain for
the Rattlers is a loss for the Eagles (and vice versa), people see themselves
and others in terms of those groups. But when, later in the study, Sherif
16  Identity, Violence and Power

altered the structure such that the Rattlers gained when the Eagles gained
(and vice versa), an inclusive group emerged and antagonism diminished.
By contrast, ethnic competition theorists presuppose that group
boundaries, memberships, solidarities and antagonisms will always be
along predefined ethnic lines. All competition does is determine whether
relations between groups will be more or less toxic. But prejudice and
conflict, when they do occur, won’t be between any old (or rather, any
young) groups, but, in the first place, between natives and immigrants or,
more generally, between people born into different nations or culturally
defined groups.
On the whole, such theories (unlike realistic conflict theories) don’t
explicitly address why co-operation and competition would operate on
predetermined ethnic groups. They just take it as a given that this is the
case. When the question is posed in practice, the answer falls back on
assumptions about biological relatedness or about long-standing cultural
processes. That is, in practice, the games metaphor tends not to be used
in its ‘pure’ form, but in a more or less implicit amalgam with one of the
other two metaphors that we have been discussing.

1.2 Why the Metaphors Are Misleading

Our previous discussion begins to touch on some of the unsettling impli-
cations of the various approaches to violent conflict that we have been
discussing. To the extent that they present ethnic or cultural categories
as fixed across biological or historical time, and to the extent that they
suggest different groups are prone, if not doomed, to fight, then the pes-
simistic conclusion is that human beings are bound to be ethno-centric
and group conflict cannot be avoided. Such conflicts will occur all by
themselves. No intervention is needed. They might be suspended, but
they cannot be stopped. In the long run, most intervention will be futile
at best and may even make things worse.
We will discuss these matters in more detail in the next section. But
first, we want to address the validity of the three approaches. After all,
we cannot reject arguments simply because they are uncomfortable. If
people were naturally ethno-centric, we would have to face up to the fact.
 1  Hateful Groups and Weak Powers?  17

We can, however, reject arguments because they don’t account for the
evidence—either suggesting things we know not to be so or else denying
things that we know to be the case. When it comes to the three meta-
phors, there are at least four such bases on which to express doubt.

1.2.1 Errors of Generalization

If one is looking for them, it is easy to find examples of conflicts between

different ethnic or civilisational groups. Huntington, for instance, bases
his argument on multiple examples of such conflicts. The examples are
graphic and they make his position seem compelling. But the real test of
an argument is not whether you can find examples to support it (hypoth-
esis confirmation) but what happens when you try to find examples to
undermine it (hypothesis disconfirmation). Billig (1987), for instance,
counsels that if ever anyone makes a strong claim (ethnic groups are
inherently prone to conflict), then consider the exact opposite (ethnic
groups are inherently prone to co-operate) and run with that as hard
as you can. Only if you get nowhere should you give credibility to the
original claim.
In this case, it is as easy to find situations where different ‘civilisa-
tions’ co-exist harmoniously as it is to find examples where they conflict.
To take just one of countless examples, consider Mazower’s (2005) his-
tory of the co-existence of Jewish, Muslim and Christian communities in
Salonica. So to claim that civilisations inevitably clash is simply untrue—
or at least it is a dramatic over-generalization. Moreover, to make such a
claim doesn’t help us understand when such conflicts break out and when
they don’t. After all, even where there is a history of antagonism between
groups such as Jews, Muslims and Christians, it remains true that most of
the time people live together peacefully. Violence is always sporadic and
requires close attention to the immediate context in order to understand
how, why and when tolerance turns to violence and vice versa (see, for
instance, Nirenburg’s 1996 analysis of relations between the three reli-
gions in medieval Aragon).
Kaufmann tries to address this problem by arguing that, perhaps, not
all ethnic groups clash in all circumstances, but once they do, the violence
18  Identity, Violence and Power

cannot be forgotten. Ethnic grudges last forever and hence peaceful

co-­existence between the relevant groups becomes impossible. Leaving aside
the problem that this still leaves the question of why violence occurs in
the first place unanswered, and also that it fails to explain when violence
will and won’t occur between the said groups, there are still two major
problems with such an account. The first is that, even if it were true that
a memory of past ethnic antagonism feeds into future ethnic antagonism,
the very fact of memorising presupposes that one views oneself and oth-
ers in ethnic terms. But sometimes it is possible that we entirely forget
about past groups and hence the violence associated with them, such that
both become entirely irrelevant to our past and future.
The great historian Ernest Renan made this point eloquently when he
asserted that modern social forms, such as the nation, depend upon for-
getting the older social categories which needed to be brought together
to make a country. In his words:

No French citizen knows whether he is a Burgundian, an Alan, a Taifale, or

a Visigoth, yet every French citizen has to have forgotten the massacre of
Saint Bartholomew or the massacres that took place in the Midi in the
thirteenth century. (1990, p. 11)

If Renan is talking about forgetting over an extended historical period,

there is evidence that one’s sense of ethnic identification can also change
in the shorter term. Thus, Bhavnani and Miodownik (2009) analyse
the two first rounds of the Afrobarometer survey, in 1999–2001 and in
2002–2004. They show how, in post-apartheid South Africa, the propor-
tion of people who answered that their ‘first and foremost’ identification
lies with their ethnic group was almost halved within three years, falling
from 42 % to 22 %. In neighbouring Zimbabwe, the decline was even
more dramatic, going down from 47 % to a mere 13 % in the same
period. Unlike diamonds, ethnic groups are not forever.
The second problem with Kaufmann’s account is the converse of this.
If ethnic categories can decline to irrelevance, hence removing the basis
for ethnic conflict, so such categories can become important and lead to
conflicts in contexts where none existed before. Many years ago, a young
Bosnian woman was interviewed on the BBC’s flagship morning radio
 1  Hateful Groups and Weak Powers?  19

news programme ‘Today’. She explained how she had fled the violence in
Sarajevo and now lived in London. Even so, she still always had a suitcase
packed under her bed in case she had to flee. The interviewer expressed
surprise. With the casual ethno-centrism of the English, he pointed out
that London is not Sarajevo and we don’t engage in barbarities such as
ethnic cleansing here. The woman replied by explaining how, when she
grew up, Sarajevo was a byword for cosmopolitanism. It was modern
and vibrant and effortlessly diverse. Ethnicity meant nothing to her; she
didn’t even know if most of her friends and schoolmates were Muslim or
Orthodox or Catholic. That is why, when the conflict and the divisions
came, they were all the more devastating. They came as lightning from a
clear blue sky. If it could happen in Sarajevo, she concluded, it can hap-
pen anywhere.
We will examine exactly how such ethnic division was produced in
Sarajevo, seemingly out of nothing, in future chapters. For now our point
is that the assumption that ethnic/civilisational conflict is inevitable—or
at least, once it starts, it is bound to continue—leads not only to over-­
generalizations but also to under-generalizations. It breeds not only
pessimism or even fatalism about the inevitability of conflict in some
circumstances, but also complacency about the impossibility of conflict
in other circumstances. It is hard to say which is worse.

1.2.2 Errors of Association

It is very rare to find unconditional relationships in the social sciences.

There are always so many factors at play. So perhaps we are setting the
bar a bit high. Perhaps it would be fairer to ask whether there is an overall
association between ethnic or cultural divides and violence: where there
is one, the other becomes more likely. After all, this is the sort of thing we
usually look for. If one exception would damn a theory, no theory would
ever last long in the social sciences.
But even when addressed in more moderate terms, the ethnic account
does not fare well. In two independent analyses of all state dyads, by year,
over most of the nineteenth and twentieth centuries both Henderson
and Tucker (2001) and Chiozza (2002) reached the same conclusion:
20  Identity, Violence and Power

violent conflict is more likely between two states of the same ‘civilisation’
(according to Huntington’s criteria) than between two states of different
civilisations. Remarkably, this relationship even holds when controlling
for geographic proximity (Henderson and Tucker) or for the existence of
a common border between the two states (Chiozza). The point is clear
and deserves stressing: throughout the era of nation states, a country is at
less risk of becoming involved in a war with another state from a ‘different
civilisation’ than with an equally remote state from the ‘same civilisation’.
This may seem decisive enough to justify scepticism about the civilisa-
tion metaphor, but there are data to buttress the argument more gener-
ally. If one turns one’s attention from relations between states to relations
within states, the kinship metaphor would suggest that increased ethnic
diversity would lead to more conflict. But again, the evidence points in the
opposite direction. Østby (2008) reports that diversity does not increase
the risk of civil violence. It is only where there are social inequalities
between ethnic groups that there is more violence. Wimmer, Cederman
and Min (2009) make a similar point. Analysing data from all indepen-
dent states of the world since the end of World War II, they show that,
once one controls for the magnitude of exclusion from political power of
certain ethnic groups, ‘linguistic fractionalisation’ (which is often used as
an indicator of ethnic diversity in comparative analyses) is only margin-
ally related to the outbreak of violence.3
Finally, the relationship between migration rates and anti-immigrant
prejudice, central to the game metaphor in ethnic competition theories,
appears to be inconsistent at best. It is true that Semyonov, Raijman and
Gorodzeisky (2006) found cross-sectional evidence that anti-foreigner
sentiment was higher in European countries with high rates of non-­
European immigration in 1988, 1994 and 1997. But in 2000 the rela-
tionship had disappeared. In the United States, Scheve and Slaughter
(2001) looked at three different time points and found no difference in

3
 One could add that the indicator of ethnic diversity is less exogenous and the relationship more
circular than it might look at first glance. If most people in France speak French today this is pre-
cisely an outcome of the fact that linguistic identities that were important until the nineteenth
century have been forgotten together with the grievances that opposed the groups. In the former
Yugoslavia, Serbo-Croat has been replaced by four new languages after the war, that is, violence and
nationalism created the languages, and not linguistic diversity the violence.
 1  Hateful Groups and Weak Powers?  21

attitudes towards immigrants between those living in areas of high immi-

gration and those who did not. In Canada, Wilkes, Guppy and Farris
(2008) conducted an extensive time-series analysis spanning the last
quarter of the twentieth century and also found no relationship between
immigration rates and attitudes towards immigrants in Canada.
So, whether one looks at international conflict or intranational con-
flict, whether one looks at settled ethnic populations or the flow of ethnic
migrants, however much one gives them the benefit of the doubt, the evi-
dence in support of the argument that ethnic or cultural diversity makes
conflict more likely is, at best, hard to find.

1.2.3 Errors of Group Psychology

Let us turn, now, from the empirical evidence (or lack thereof ) for con-
flict theories to their conceptual underpinnings. As we have seen, the
three metaphors rest on the general claim that groups have an inherent
tendency to conflict with each other. The point is actually often taken as
so self-evident as to require no justification. However, on those occasions
where the argument is warranted, two authors in particular tend to be
referenced: Sherif for his boys camp studies and the resultant realistic
conflict theory, and Tajfel for his minimal group studies and the resultant
social identity theory (Tajfel, Billig, Bundy & Flament, 1971; Tajfel &
Turner, 1979). In both cases though, there is a problem of partial findings
being presented and of specific claims being extracted from the broader
theoretical framework of which they form a part. This way, caricatures of
the seminal studies and theories have replaced the real thing, and argu-
ments are advanced in the name of these theories which are at odds with
their original intentions.
In the case of Sherif, the focus is so exclusively on how competition
produces conflict that it is forgotten how his key point was to show that
conflict is contingent, dependent on how people are set against each other
in particular social systems and how conflict can therefore be ­overcome.
It is also forgotten that, even where there is competition and conflict,
Sherif didn’t see this simply as a matter of relations between the two
groups. He also examined how intergroup competition can create new
22  Identity, Violence and Power

opportunities and incentives for struggles within groups. In fact three

out of four of Sherif ’s original research hypotheses regarding the stage
of competitive games referred to the changing ingroup dynamics, and
only one to intergroup orientations. Finally, it is forgotten that Sherif did
not see competition between groups as a simple objective given. He was
well aware that it is important to examine the frame of reference through
which people understand social relations, the nature of their groups and
the relations between them. Indeed, as he discovered at his own cost in
an abandoned 1953 boys camp study, you can put groups in competitive
games, but they won’t always see themselves as in competition and fight
over the spoils (see Reicher & Haslam, 2014).
Overall, Sherif ’s aim was to show that ‘the dark side’ of human con-
duct—hatred, war, genocide—was not inevitable. It didn’t reflect some
unchangeable essence of our social or biological nature. Rather, it reflected
something about the ways in which we are put into competition within
a contemporary market society (Kayaoglu, Batur & Asliturk, 2014).
Ethnic competition theories, then, are not an application and extension
of the realistic conflict approach. They are its nemesis.
Turning now to Tajfel, the use of his work to argue that ‘discrimination
between groups is inevitable’ is, if anything, even more of a distortion
and a sign of sloppy scholarship. In his minimal group studies, boys are
divided into two groups on trivial or even random grounds. They then
have to divide points between two individuals about whom all they know
is that one is a member of their own group and the other is a member of
the other group. The key finding is that they show a small but significant
tendency to give more points to the ingroup member even if this comes
at the cost of the absolute level of reward to the ingroup. Tajfel then
explains this finding by arguing, first, that people define their identity
in terms of membership of the group; second, that they then evaluate
their group membership through comparison with relevant other groups;
and third, that in order to achieve a positive evaluation (and the positive
esteem that flows from it) they differentiate themselves from the other
group, seeking to come out better in the relevant comparison.
In the minimal group studies, the only group one has to identify with
is that to which one is assigned, the only group one has to compare with
is the one to which one is not assigned, and the only dimension along
 1  Hateful Groups and Weak Powers?  23

which one can compare oneself positively is the number of points one’s
group is allocated. In the minimal group studies, then, differentiation
necessarily means giving more points to one’s own group. However, Tajfel
has always been very clear about the dangers of generalising from the
specific behaviours in a study rather than examining how the underlying
process might operate differently in different settings:

What is, however, important is a clear realization that the ‘general’ case is
an impossible myth as long as human beings behave as they do because of
the social expectations with which they enter an experiment—or any other
social situation. If these expectations are shared—as they always are by defi-
nition to some degree in any social context—I shall obtain data from my
experiment which are neither ‘general’ nor ‘individual’. The observed regu-
larities of behaviour will result from the interaction between general pro-
cesses and the social context in which they operate. Without the knowledge
of this context the data may be irrelevant to the confirmation or the falsifi-
cation of a hypothesis. (Tajfel, 1972, p. 74)

In the case of the minimal group studies, this argument plays out at
two levels. First, whereas the process of differentiation may play out as
financial discrimination in the highly constrained setting of the experi-
ments, it may play out in many different ways in different group con-
texts. Indeed, the substantive outcomes of the differentiation process will
always depend on the natures of the groups involved, the things that they
value and therefore the dimensions along which they seek to compare
themselves. In some cases that might result in behaviours that bring the
other group down (we want to be harder, stronger, richer), but equally, in
other cases it might result in actions which benefit the outgroup (we want
to be more generous, kinder, more loving). So whereas differentiation
means discrimination in the minimal group studies, this will not always
be so and groups will not always discriminate against each other.
At the second level, social identity theory does not suggest that members
of groups always differentiate themselves from outgroups. Whether they
do or not depends upon a series of structural and ideological factors such
as the possibility of m ­ ovement between different social groups (perme-
ability), and the legitimacy and the stability of intergroup relations. Social
identity theory, then, is definitely not a theory of (inevitable) intergroup
24  Identity, Violence and Power

discrimination. It is not even a theory of (inevitable) intergroup differentia-

tion. It is actually a theory of how the differentiation process plays out in
different social settings (see Reicher, 2004).

1.2.4 Errors in the Description of Conflicts

Now, let us consider still more closely the assumption that conflicts are
driven by ethnicity, and that violence is motivated by ethnicity. Ethnic
conflict accounts are fundamentally rooted in a way of representing the
protagonists in a conflict as Serb and Croat, Hutu and Tutsi, Hindu and
Muslim—whether or not they are, whether or not they see themselves
as such, and whether or not this is relevant to what they do. Thus, when
the conflict is between different groups it is often assumed that religious
differences are the true underlying cause of conflict. When, say, a Muslim
worker confronts a Hindu moneylender, it is assumed that it is the reli-
gious (rather than the economic) category that counts. When a Hindu
attacks a Muslim, it is seen to be about the fact that they don’t like that
the other is of a different religion rather than about more immediate tem-
poral concerns. Once you have found your generic Hindu and generic
Muslim, enquiry can stop, for that in itself is enough to explain violence.
However, as our previous examples illustrated, people do not always
view the world and themselves through ethnically tinted spectacles.
We all have many group-based identities—as a Catholic, as Swiss, as a
socialist, as a football fan, say—and these will become salient in different
contexts. It follows equally that the way I categorise others will change
according to context: the same individual who I may see as an outgroup
Protestant in one context, I may see as an ingroup Swiss in another. No
one can be classed as ingroup or outgroup in general terms. We can only
make these judgements in specific contexts.
But even if people do see themselves in ethnic terms, and even if they
were to feel antagonism towards each other because of ethnic differ-
ences, that still would not get us very far. It certainly would not explain
any ­violence that occurred. This takes us back to a point we have made
before. Even in the most extreme cases of ethnic conflict, violence is never
a constant. It doesn’t happen all the time; it doesn’t happen in every place.
 1  Hateful Groups and Weak Powers?  25

Simply invoking universal ethnic antagonisms doesn’t tell us where vio-

lence happens, when violence happens, why it starts and how it stops.
Take India as a case in point, where so-called communal violence
between Hindus and Muslims has claimed, and continues to claim, hun-
dreds, thousands, even millions of lives (if one includes the events of
partition). As Brass (2006) argues, with particular power, riots are not
a simple chain of action and reaction. They do not simply repeat them-
selves without cease. Waves of rioting come and go. Any single wave lasts
for a few days at most. It may be true that particular massacres, as in 1992
(following the destruction of the Babri Masjid and 2002 (in Gujarat), are
so horrific and so command our attention as to make the time between
them seem insignificant and make us feel that violence is going on all the
time. But, in fact, most of the time passes without riot. So why did the
dispute come to a head in 1992 and why was it reinvigorated in 2002?
The same questions can be asked spatially as well as temporally. Again,
riots so command the field that we can fail to notice that even the most
intense ‘wave’ of violence leaves most places untouched. Why, in particu-
lar, was the most intense rioting that followed the sacking of the Babri
Masjid in Bombay—a city nearly 1000 miles away, one of Asia’s most
cosmopolitan cities and a most unlikely site for such carnage (Tambiah,
1996)? Once more, if one assumes that riots simply reflect the universal
antagonism of all Hindus towards all Muslims, then there is no way of
knowing except to invoke chance.
But there is still a last point to consider, in order to fully grasp the
problem. Collective violence is never an inchoate explosion in which
anything goes. However extreme the actions, they always have a pattern.
Some targets are attacked. Some are left alone or even defended (e.g.,
Davis, 1973; Thompson, 1971). The Indian evidence tells a similar story.
Violence is always expressed in culturally meaningful forms and pat-
terns, and these are different in different events (e.g., Brass, 1997, 2006;
Wilkinson, 2005).
Putting these various points together we reach the conclusion that
the greatest problem with the three metaphors is not only that are they
­misleading, but even if they were right they wouldn’t tell us very much.
They are simply irrelevant to the explanation of ethnic (or any other)
26  Identity, Violence and Power

conflict. But this is not to say that they are irrelevant full stop. For what
they lack in explanatory power they make up for in pragmatic impact.

1.3 Accounts and Accountability

Any account of any event necessarily has implications for whether and
how the outcome could have changed, for who was responsible and hence
for who should (either literally or metaphorically) pay the consequences.
As discourse analysts remind us, accounts always manage accountabil-
ity (Edwards & Potter, 1992). If, for instance, my car ploughs into the
back of yours, was it ‘just one of those things’ which you can’t get rid of
without getting rid of driving, was it because you slammed your brakes
on too hard, because I was inattentive, because the garage didn’t service
my brakes properly, because the council didn’t maintain the road surface
properly … each different account points the finger in different direc-
tions and opens (or closes) the door to different solutions.
What is true of everyday accounting is also true of academic explana-
tions. And as the stakes grow higher, so the stamp of academic credibility
on any particular account becomes all the more important. This is par-
ticularly relevant to our concerns here: how one explains deadly violence
can itself be a matter of life and death. More specifically, accounts of
conflicts as historic fatalities can serve to exonerate those who would oth-
erwise be seen as war criminals—or at least accessories to murder. They
can also serve to warrant policy options which, we shall suggest, at best
leave existing antagonisms intact, and at worst reproduce the conditions
that bred them. We shall address each of these concerns in turn.

1.3.1 Letting Leaders Off the Hook

In order to clarify our point about exoneration, let us ask the questions:
Who according to the three metaphors, are the perpetrators? Who is
responsible for doing the killing and for letting it happen? These seem
like obvious questions, but actually they are rather hard to answer. If pas-
sionate hatreds and violence flow from ethnicity or civilisation, then all
 1  Hateful Groups and Weak Powers?  27

members of the ethnic group, or even civilisation, are responsible. But if

all ethnic groups, let alone civilisations, are similar in their inbuilt hostil-
ity for others, then everyone is responsible—which is the same as saying
no one can be picked out as responsible since we are all as guilty as each
other. All we can do, in our more sanguine moments, is bemoan this
critical flaw in our evolved human nature.
By contrast, there is a linked, but distinct, question that is much easier
to answer: Who is not responsible? Who is off the hook? This is not to
deny that many try to fudge or hide the answer. But political scientist
Roger Petersen (2002) is more open. Following his bold aphorism that
“it is better to be clear and wrong than to be unclear” (p. 35) he clearly
argues in his influential treaty ‘Understanding ethnic violence’ that lead-
ership was not necessary and not to blame for the ethnic violence that
swept across Eastern Europe in the twentieth century.
Petersen asks, “How does a collective body, such a violent mob, come
to act as a coherent unit in terms of specifying an ethnic target?” He
instantly replies: “Emotion can coordinate motivations and effectively
point a legion of individuals in one particular direction. Emotion can
substitute for leadership” (p. 4, our emphasis). In other words, leaders did
not form people into groups, they did not specify targets, they did not
incite violence. In fact they did not—they could not—lead at all. All they
could do was follow, echoing what the masses had already understood
and done. In Petersen’s own words, leadership elites were “responding
to structural change and mass emotion rather than shaping it” (p. 252).
In such circumstances, these elites neither needed nor had any par-
ticular abilities or skills, and Petersen is accordingly scathing about them,
particularly the political leader of the Bosnian Serb forces during the
1992–1995 Bosnian war, Radovan Karadzic: “it is difficult to see how this
bad poet, average psychiatrist, and convicted embezzler could become a
leader so easily if he did not tie into some existential motivational force
among Bosnian Serbs” (Petersen, 2002, p. 35).
At the time when Peterson was writing these lines, Karadzic was
Europe’s most wanted fugitive. Eventually he was arrested, and from 2009
he has had to answer for his deeds during the war before the International
Criminal Tribunal for the Former Yugoslavia (ICTY). To observers of his
28  Identity, Violence and Power

trial, it is intriguing to see how close his own line of defence sometimes
comes to the analysis of scholars like Petersen.
In the hearing of 16 October 2012, Karadzic declared:

In the 1990s, the Serb community in Bosnia-Herzegovina in 1992 became

very anxious when the HDZ appeared in Croatia and when Izetbegovic
made his announcement in Bosnia-Herzegovina. The Serbs were fright-
ened of the Ustasha rhetoric and there were also suspicions that Mr.
Izetbegovic, the author of the Islamic Declaration, would apply this Islamic
Declaration as a political platform of his party. There is not a single Serb or
any man anywhere in the world who could convince Serbs that there is or
isn’t a threat of genocide. It is only when they saw where things were going
in Yugoslavia that they understood what’s the writing on the wall before I
even said a word.

The critical point of this defence is, of course, that the Serbian masses
perceived that they were facing a “threat of genocide” before Karadzic
himself “even said a word”. They spontaneously “saw where things are
going”, when other ethnic groups supported ‘Ustasha’ (i.e., fascist) or
‘Islamic’ parties. They understood the dangers posed by these other
groups, and acted to pre-empt those dangers without needing any guid-
ance from Karadzic.
Far from being a war-monger or genocidaire, then, Karadzic presented
himself to the court as a peace-loving group therapist, who had writ-
ten poetry for children, and cherished personal friendships with Muslims
and Croats. If there was any violence perpetrated by Bosnian Serbs, it
occurred in self-defence against the threats and aggression from the other
communities. It occurred despite, and not because of, Karadzic’s leader-
ship, so he was not responsible for his people’s deeds. If one believes his
words, collective violence did not occur as a consequence of the power
that Karadzic exerted over his people, but as a consequence of the limits
of his power.
A second example makes the relationship between social scientific
accounts and (legal) accountability even more explicit. Before the ICTY
turned its attention to Karadzic, Dario Kordij and Mario Cerkez (the
former a Bosnian Croat political leader, the latter a military commander)
 1  Hateful Groups and Weak Powers?  29

were indicted for crimes against humanity committed against Muslims in

the Lašva Valley in 1993. Part of the prosecution’s case was a sociological
report arguing that the Croatian leadership had promoted a climate of
excessive nationalism supportive of war crimes. In response, the defence
invited Stepan Mestrovic, a Croat-born American sociologist, to address
the Tribunal. In the audience of 26 June 2000, he stated that:

Contrary to the assertions made by Dr. Allcock in his expert report, who
claimed that (…) Franjo Tudjman somehow single-handedly engineered or
produced ethnic tension in Croatia, what the data suggests is that prior to
Franjo Tudjman even being on the scene, the sense of ethnic distance arose
of its own accord, from the bottom up, and whatever factors led to it cer-
tainly they were not and cannot be attributed to Franjo Tudjman.

Mestrovic went on to argue that, to the extent that the Croatian elites
did have anything to answer for, it was a sin of omission, not commis-
sion. It was not that they communicated a nationalist vision, but rather
that, as the old certainties were collapsing, they failed to communicate an
alternative vision. As a result “a vacuum was created in which formerly
communist nations were looking to the west that was not prepared to
give them guidance and nationalism was, so to speak, the logical alterna-
tive. But my point is that this nationalism arose from a bottom up as a
way to fill this vacuum precisely because there was no top down or centre
guidance”.
At the end of the trial, Dario Kordij and Mario Cerkez were found
guilty of crimes against humanity. In this particular case the judges did
not appear to be convinced by the attempt to shift accountability away
from individual leaders and towards the masses. But the very fact that the
attempt was made and that the ICTY had become a theatre of sociologi-
cal controversy provides a telling illustration of how scientific debates can
become a life-and-death matter for leaders.
If such bottom-up arguments serve to deny the immediate responsibil-
ity of local leaders in creating antagonism and violence, they also serve
to absolve more distant leaders from the responsibility of doing anything
to stop the violence. We have already shown how John Major and Bill
30  Identity, Violence and Power

Clinton subscribed to the ancient hatreds narrative. Recall Major’s speech

to parliament in June 1993 where he claimed that conflict in Bosnia
was the resurgence of ancient hatreds, no longer constrained by Soviet
discipline. The key point about that speech was that it was used to refuse
British intervention. Indeed, the passage we cited was preceded by the
words “I do not envisage that any further British troops will be sent to
Bosnia in the near future”.
Clinton also drew the link between ancient hatreds and non-­
intervention, albeit less publicly, but even more explicitly. Journalist
Richard Reeves recounts how Clinton had decided on a policy of actively
siding with the Bosnian Muslims, dubbed ‘Lift and Strike’ (lift the arms
embargo on Bosnian Muslims and strike at Bosnian Serbs). Then, one
day, the President came to a meeting carrying Robert Kaplan’s best-
selling book Balkan Ghosts—a book which explicitly associates eth-
nic diversity and recurrent atrocity and which puts the ferocity of the
conflict in Bosnia down to the nature of its ethnic patchwork. Reeves
quotes Clinton as saying, “My wife read this and I read some of it too.
And it says that we can’t succeed doing anything in that society. They’ve
been killing each other for thousands of years and they’re going to keep
doing it”. Reeves then quotes the reaction of the Secretary of Defense,
Les Aspin, who was in the room: “Aspin said later, he was sitting there
thinking, ‘He’s going to go south on Lift and Strike’ … And in fact the
President did”.4

1.3.2 Legitimating Ethnic Segregation

As should be clear by now, the question of who is responsible for starting

violence or else for stopping it cannot be separated from the question of
what can be done to ensure violence doesn’t re-occur. If violent conflict
is just part of the character of groups and an inevitable consequence of
our natural tendency to prefer people from our own kind or civilisation,
this implies a fatalistic pessimism, which does not just relieve leaders of
responsibility, but also radically reduces the policy options open to them.
4
 
Retrieved on 15th September 2015 from http://www.pbs.org/wgbh/pages/frontline/shows/
choice/bill/reeves.html.
 1  Hateful Groups and Weak Powers?  31

For if, as Huntington suggests, history is written in advance, and if, as

a consequence, groups are doomed to fight, then, as Bill Clinton con-
cluded, there is precious little one can do about it even if one wanted to.
However, just as there are more or less radical variants of theoretical
fatalism, so there are weaker and stronger forms of policy pessimism.
Thus, ethnic competition theorists, even if they assume that conflict nec-
essarily occurs across ethnic lines, still argue that the level of conflict will
be moderated by levels of competition. If there is enough cake for every-
body, different ethnic groups can co-exist in the same society. But if the
cake is not big enough (normally measured by looking at rates of growth
and employment) or there are too many people (normally measured from
the perspective of the majority by looking at rates of migration or minor-
ity birth rates), then there will be trouble. In particular, when there is an
economic downturn and the cake shrinks, then one’s options for stop-
ping conflict reduce to clamping down on migration or even expelling
minorities.
The argument is taken a step further by those who argue that eth-
nic diversity is doomed to remain an irreversible source of disharmony,
once a first instance of ethnic violence has occurred. This is the position
adopted by Kaufman, who asserts that “competition to sway individual
loyalties does not play an important role in ethnic civil wars, because
ethnic identities are fixed by birth” (p. 140). He does, however, suggest
that ethnicity can vary in its significance. In particular, “war hardens eth-
nic identities” (p. 139) and also ‘shrinks the scope for individual iden-
tity choices’ (p. 143). Once this has happened, change becomes difficult.
Ethnic groups become impossible to reconcile. Policies that might once
have had a chance—power-sharing arrangements, state-building, recon-
struction of ethnic identities—become redundant. “Even if ethnic hos-
tility can be ‘constructed”’, states Kaufman, “there are strong reasons to
believe that violent conflicts cannot be ‘reconstructed’ back to ethnic har-
mony” (p. 153). The only remaining means of breaking a cycle of ethnic
violence then would be ethnic separation.
Kaufman retains a tactful silence as to whether ethnic diversity causes
conflict in the first place. His call for segregation has been limited to situ-
ations where conflict has already happened. Lim, Metzler & Bar-Yam,
however, have shown no such constraint. Natural scientists by training,
32  Identity, Violence and Power

these authors embarked on a slippery journey when they applied not only
the analytic techniques but also the causal models of their own disciplines
to nothing less than the explanation of ethnic violence in India and the
former Yugoslavia. From an academic perspective, the journey appeared
successful since their findings made it into a Science article in 2008. In
this article, Lim, Metzler & Bar-Yam, used an analysis of concurrent spa-
tial patterns between local ethnic mixing and ethnic violence to propose
that conflict reflects “the natural dynamics of type separation, a form
of pattern formation also seen in physical or chemical phase separation
(p. 1541)”. That is, with the same regularity as a stone falling through
water, proximity means violence and separation means peace. There is as
much sense in trying to fight this natural law as there would be in leaping
off a cliff and hoping gravity will not work. Ethnic segregation is neces-
sary to prevent violence and not just as a response to it.
We are not suggesting that any of these authors are cynical or politi-
cally motivated in what they propose. Indeed, our point is that their
conclusions (which certainly are congenial to others who are politically
motivated) are the logical conclusions which flow from the fatalistic
premises inherent in the kinship, civilisations or games metaphors used
to think about groups. If Kaufmann, Lim and others were right about the
nature of groups, then they would also be right that we face a stark choice
between endless ethnic war and apartheid.
If that was the choice, many might be tempted to choose apartheid,
as the lesser of two evils. But let us not lose sight of the fact that ethnic
separation in places such as the former Yugoslavia and India is a terrible
evil in many ­different ways. It institutionalises the outcome of ethnic
cleansing. It durably disrupts multi-ethnic social ties, including between
close friends and relatives. It prevents the contacts between ethnic group
members, which might otherwise reduce the prejudices between them.
It grants no right of return to people who have been illegally and vio-
lently expelled. It officially endorses and structurally supports definitions
of identity promoted by the architects of mass killing and persecution. It
renders the presence of ‘ethnic outsiders’ intolerable.
This is a long litany, to which we need to add the fact that, even if
the rhetoric of separation is ‘separate but equal’ we know from apart-
heid South Africa that, in practice, separation means consigning the less
 1  Hateful Groups and Weak Powers?  33

powerful to radically impoverished economic and social circumstances.

The alternative therefore has to be truly awful for separation to have any
justification. There can only be any moral grounds for separation if the
consequences of a policy of ethnic mixing are quite literally MAD: mutu-
ally assured destruction. But what if—as we have argued—such assump-
tions are wrong? What if it is far from assured that diversity will lead to
conflict? Then the balance shifts radically. When the kinship, civilisation
or games metaphors feed into policies, they engender a series of nega-
tives and mitigate against nothing. Indeed, they can only inspire policies
which create the very things they claim to protect against: entrenched
ethnic categories, unbridgeable civilisational divides, indelible fears and
eternal resentments. Once such a fatalistic viewpoint spills over from ana-
lysts to actors, we really are in trouble.

1.4 Beyond Fatalism

In this chapter, we have been critically addressing the idea that violence
is a product of immutable group processes based on immutable identi-
ties. We have seen that this idea can come along in slightly different
guises: while some authors see us humans as programmed by our genes
to stick together with those who are similar to us and compete with those
who are different (the kinship metaphor), others see us as programmed
by a long history that has divided us into cultures of incommensurable
worldviews and values (the civilisation metaphor), while still a third set
of authors sees hostility as rooted in zero-sum competitions for social
and economic goods, which display the peculiarity that the competing
teams were made in long-forgotten times and everyone since has had
to play on the team of his remote ancestors (the games metaphor). For
all the subtle nuances that separate them, one over-arching argument
unites the three metaphors: we cannot escape our past.
What the proponents of all three metaphors have in common is a read-
ily available, simple answer to the puzzle as to why (from a contemporary
perspective) collective behaviour often looks so disturbingly irrational:
it does not respond to the logic and stakes of current reality, because it
is locked into the logic of an (imagined) distant past—when our genes
34  Identity, Violence and Power

formed, when civilisations emerged or when the teams were made. The
notion of ancient hatreds typically expresses nothing else than the fact
that the origin of conflict is projected into the same mythical era as the
foundation of groups themselves. From that time on, groups are imag-
ined to be prisoners of absurdly tragic spirals of violence, whereby each
group sees itself as responding to previous offences of the other: ‘we see
our attack on you as a response to your attack on us and a way of pre-­
empting your future attacks, although we can anticipate that you might
see our attack as an unwarranted provocation revealing our vicious nature
and necessitating a defensive reaction …’ Like Sisyphus rolling his rock
up the hill, we are all fated to repeat this futile pattern forever.
Our arguments against such fatalistic conflict theories have been both
analytic and normative. Analytically, we have shown how they only gain
credibility by selectively focusing on the few cases where they seem to
apply. But this ignores the fact that there are many more cases where they
don’t apply and that, overall, there is little to suggest that ethnic diversity
makes conflict more likely. Moreover, even in the cases that are cited as
support, fatalistic conflict theories explain very little about the phenom-
ena—not when they happen, not where they happen, not the forms that
violence takes, not even the choice of who is and who isn’t targeted.
Normatively, we have shown how these approaches serve to limit
responsibility for extreme crimes and how they point to policy options
that are literally worse than useless—they make future crimes more likely.
How do we deal with ethnic violence? By structuring the world ever more
in terms of ethnicity and thereby creating the world that ethnic cleansers
dream of?
One of the places where the analytic and the normative come together
most obviously has to do with the issue of leadership. Leaders are in the
business of guiding how we interpret the world and hence how we should
act in it. So, by presupposing or obscuring the issue of how violence is
interpreted, fatalistic conflict theories deny or obscure the role of leaders
in violence. If we are programmed to see the world in particular ways,
then leaders become simultaneously innocent and redundant.
As we will show in the next chapter, there certainly are those who do
so emphasise the importance of leadership that they reduce followers to
 1  Hateful Groups and Weak Powers?  35

mere ciphers. But a stark opposition of leadership and followership—of

the influence of the elite and the active choices of the population—still
misses the main point. If we want to understand violence, we need to
address both how leaders frame situations of violence and how the popu-
lations embrace, adapt or reject, and constrain what leaders do. There
are active agents at all levels. Agency—and hence responsibility—is not
reserved to one class of people.
Let us then conclude by summarising the key analytic constructs which
underpin the fatalistic conflict approaches that we have critically exam-
ined in this chapter, and which need to be re-considered if we want to
move beyond the position that conflict between groups is immutable and
violence between groups a normal consequence of conflict. The first of
these is identity. Fatalistic conflict theories assume that people see them-
selves and others in terms of fixed categories. They also assume that rela-
tions between those of different categories are bound to be fraught, prone
to be violent, and that once violence has started there is no way back.
The second construct is power. This is seen as an exogenous factor. It is
the ability of other groups to stop people acting according to their iden-
tity or expressing their hostility towards others. It can therefore dampen
the expression of violent tendencies between people of different ethnicity
or civilisation, but it can never remove them. When the power of the
third group passes (as it always does), then ‘nature’ will reassert itself.
That is why the end of empire is always a bloody affair.
The third construct, which we have already invoked, is violence. In the
view of fatalistic conflict theories, violence is driven by identity, although
its concrete occurrence can be moderated by the presence or absence of
external power.
So, their argument is about identity, violence and power. So will ours
be throughout the book—however, we will plead for a rather different
way of articulating the three constructs. But before we can do that, we
still need to take our critical review a step further and extend it to the
theories that appear to claim or to imply the exact opposite of theories
examined so far: that all evil comes from evil leaders, and that all we need
to know about ordinary people is how easily they can be led to obey any-
one on anything.
36  Identity, Violence and Power

References
Andric, I. (1945). The bridge on the Drina. Chicago: University of Chicago Press.
Bhavnani, R., & Miodownik, D. (2009). Ethnic polarization, ethnic salience,
and civil war. Journal of Conflict Resolution, 53(1), 30–49.
Billig, M. (1987). Arguing and thinking: A rhetorical approach to social psychology.
Cambridge: Cambridge University Press.
Brass, P. R. (1997). Theft of an idol: Text and context in the representation of collec-
tive violence. Princeton, NJ: Princeton University Press.
Brass, P. R. (2006). Collective violence, human rights, and the politics of curfew.
Journal of Human Rights, 5(3), 323–340. doi:10.1080/14754830600812324.
Brubaker, R., & Laitin, D. D. (1998). Ethnic and nationalist violence. Annual
Review of Sociology, 24, 423–452.
Chiozza, G. (2002). Is there a clash of civilizations? Evidence from patterns of
international conflict involvement, 1946–97. Journal of Peace Research, 39(6),
711–734. doi:10.1177/0022343302039006004.
Davis, N. Z. (1973). The rites of violence: Religious riot in sixteenth-century
France. Past & Present, 59, 51–91.
De Villepin, D. (2014, August 19). Will the clash of identities consume us? The
Huffington Post (Worldpost). Retrieved from http://www.huffingtonpost.com/
dominique-de-villepin/clash-of-identities-impact_b_5509538.html
Fearon, J. D. (2008). Ethnic mobilization and ethnic violence. In D. A. Wittman
& B. R. Weingast (Eds.) (pp. 852–868). Oxford: Oxford University Press.
Hansard. (1993). Commons Sitting of 23 June 1993—European Council (Copenhagen)
(Official Report Series 6, Vol. 227). Retrieved from http://hansard.millbanksys-
tems.com/commons/1993/jun/23/european-council-copenhagen
Harvey, F. P. (2000). Primordialism, evolutionary theory and ethnic violence in
the Balkans: Opportunities and constraints for theory and policy. Canadian
Journal of Political Science/Revue Canadienne de Science Politique, 33(01),
37–65. doi:10.1017/S0008423900000032.
Henderson, E. A., & Tucker, R. (2001). Clear and present strangers: The clash
of civilizations and international conflict. International Studies Quarterly,
45(2), 317–338. doi:10.1111/0020-8833.00193.
Huntington, S. P. (1993). The clash of civilizations? Foreign Affairs, 72(3),
22–49. doi:10.2307/20045621.
Johnson, T. (2014, October 6). The clash of civilizations: Our new global order.
The Huffington Post (UK edition). Retrieved from http://www.huffingtonpost.
co.uk/tom-johnson/the-clash-of-civilizations_b_5656201.html
 1  Hateful Groups and Weak Powers?  37

Kayaoglu, A., Batur, S., & Asliturk, E. (2014). The unknown Muzafer Sherif.
The Psychologist, 27, 830–833.
Mazower, M. (2005). Salonica, city of ghosts. London: Harper Perennial.
Nirenberg, D. (1996). Communities of violence. Princeton, NJ: Princeton
University Press.
Olzak, S. (1992). The dynamics of ethnic competition and conflict. Stanford, CA:
Stanford University Press.
Østby, G. (2008). Polarization, horizontal inequalities and violent civil conflict.
Journal of Peace Research, 45(2), 143–162. doi:10.1177/0022343307087169.
Petersen, R. D. (2002). Understanding ethnic violence: Fear, hatred, and resent-
ment in twentieth-century Eastern Europe. Cambridge: Cambridge University
Press.
Reicher, S. (2004). The context of social identity: Domination, resistance, and
change. Political Psychology, 25(6), 921–945.
Reicher, S. D., & Haslam, S. A. (2014). Camps, conflict and collectivism. The
Psychologist, 27, 826–829.
Scheve, K. F., & Slaughter, M. J. (2001). Labor market competition and indi-
vidual preferences over immigration policy. Review of Economics and Statistics,
83(1), 133–145. doi:10.1162/003465301750160108.
Semyonov, M., Raijman, R., & Gorodzeisky, A. (2006). The rise of anti-­
foreigner sentiment in European societies, 1988–2000. American Sociological
Review, 71(3), 426–449. doi:10.1177/000312240607100304.
Sherif, M., Harvey, O. J., White, B. J., Hood, W. R., & Sherif, C.W. (1961/1988).
The robbers cave experiment: Intergroup conflict and cooperation. Middletown,
Connecticut: Wesleyan University Press.
Tajfel, H., Billig, M. G., Bundy, R. P., & Flament, C. (1971). Social categoriza-
tion and intergroup behaviour. European Journal of Social Psychology, 1(2),
149–178. doi:10.1002/ejsp.2420010202.
Tajfel, H., & Turner, J. C. (1979). An integrative theory of intergroup conflict.
The Social Psychology of Intergroup Relations, 33(47), 74.
Tambiah, S. J. (1996). Leveling crowds: Ethnonationalist conflicts and collective
violence in South Asia. Berkeley: University of California Press.
Thompson, E. P. (1971). The moral economy of the English crowd in the eigh-
teenth century. Past & Present, 50, 76–136.
Wilkes, R., Guppy, N., & Farris, L. (2008). “No thanks, we’re full”: Individual
Characteristics, national context, and changing attitudes toward immigra-
tion1. International Migration Review, 42(2), 302–329. doi:10.1111/j.1747-
7379.2008.00126.x.
38  Identity, Violence and Power

Wilkinson, S. I. (2005). Communal riots in India. Economic and Political Weekly,
40(44/45), 4768–4770.
Wimmer, A., Cederman, L.-E., & Min, B. (2009). Ethnic politics and armed
conflict: A configurational analysis of a new global data set. American
Sociological Review, 74(2), 316–337. doi:10.1177/000312240907400208.
 2
Evil Leaders and Obedient Masses?

Let us now turn things around—in more senses than one. If, as we argued
in the previous chapter, one cannot explain violent conflict in terms of the
deep-seated hostility that members of groups have for each other, then
perhaps it has more to do with people doing what others tell them to do.
Perhaps conflict is a reflection of the will of elites as channelled through
the masses, rather than the will of the masses themselves. Perhaps the
perpetrators of violence are simply obeying orders, which only reflect
the motives of those who give the orders—and do not reveal more about
those who carry them out than their propensity to obey, albeit sometimes
in a shockingly thoughtless way.
In such a perspective, the role of leadership appears under a radically
different light to what we’ve seen so far. If most people are reduced to the
role of subservient followers, leaders command the masses rather than
accommodate to them. They do not merely adapt to the course of histori-
cal events, shaped itself by complex factors. They make history.
A strong version of history as a tale of either great or greatly evil leaders
can be read in Stoessinger’s (2007) Why Nations Go to War. In this book,
Stoessinger discusses the reasons for many past wars, including the 1990s

© The Author(s) 2017 39

G. Elcheroth, S. Reicher, Identity, Violence and Power,
DOI 10.1057/978-1-137-31728-5_2
40  Identity, Violence and Power

wars in the former Yugoslavia. Stoessinger’s account is the exact opposite

of the mass-level approaches discussed in the previous chapter: “The story
of the dismemberment of Yugoslavia is the story of one man’s destruc-
tive hubris: Slobodan Milosevic, who destroyed his country and died in
a jail cell while on trial on charges of genocide” (XV, our emphasis).
Such a view opposes the agency of decision-makers to the determinism
of contingencies:

(The) theme of inevitability is a haunting and pervasive one. Most of the

statesmen who made the crucial decisions behaved like actors in a Greek
tragedy. The terrible dénouement was foreseen, but somehow it could not
be prevented (…). Historians too have been affected by this fatalistic atti-
tude. As one leading scholar has summed up his analysis of the outbreak of
the war (WWI): ‘All the evidence goes to show that the beginning of the
crisis (…) was one of those moments in history when events passed beyond
men’s control.’ (…) such a view is wrong: Mortals made these decisions.
(p. 4)

But if history is driven by a handful of leaders, what does that imply

regarding the agency of armies of followers? If millions of people can
be ordered into collective catastrophes like the two world wars, the
most dangerous aspect of human psychology lies no longer in our
ability to hate, envy or resent people different from us, but in our
ability to obey our own leaders—even when their orders are bluntly
misguided, morally outrageous or potentially calamitous in their
consequences.
Over the last half century and more, this idea has exercised an
extraordinary hold over social psychologists, social scientists and
indeed our society at large. The notion that people are natural con-
formists, and that their conformity explains the worst excesses of
human behaviour, has been used to explain all sorts of phenomena—
most particularly the worst of the worst, the Nazi Holocaust. To a
large extent, the prominence of this assumption originates in Stanley
Milgram’s studies on human obedience—studies that have become the
most famous ever conducted in the discipline of psychology (Banyard
& Grayson, 2000).
 2  Evil Leaders and Obedient Masses?  41

2.1 S
 tanley Milgram and the Study
of Obedience
Before Milgram had started his studies in December 1961, the strong
and simple assumption was that perpetrators must be peculiarly vicious
people. They could not be just like the rest of us but rather had to have
severely twisted profiles, idolising the powers above them and brutalising
those below them; craving rigid order and hating ambiguity; demanding
conformity and determined to eliminate deviance (and deviants). In the
terms of the most influential such theory, they had to have authoritarian
personalities (Adorno, Frenkel-Brunswik, Levinson, & Sanford, 1950).
Milgram’s studies marked an irrevocable break with the previous com-
mon sense that extreme brutality must be related to extreme personalities.
His studies showing how people with very ordinary personalities were led
to disturbing levels of cruelty are amongst the few which have gone beyond
the disciplinary boundaries of psychology to have influence across the aca-
demic world. They are amongst even fewer which have impacted on public
consciousness, for even if people do not recognise Milgram’s name, many
do recall those studies in which people delivered massive electric shocks to
innocent victims under the orders of an experimenter.
Being so well known, there is no need to provide more than a brief
outline of the experimental set-up (see Milgram, 1974 for a full account,
and Blass, 2004, for an account of the history leading up to the stud-
ies). Participants were invited to take part in a set of studies ostensibly
about the role of punishment in learning. Once at the laboratory, they
drew lots with another participant (actually a confederate of Milgram’s)
to determine who would be allocated the position of teacher and who
that of learner—although the draw was rigged so that the real participant
was always the teacher. After this, the learner was strapped into an electric
chair. The teacher then read out a series of word pairs which the learner
had to remember. Following this, the teacher gave the first word of a pair
followed by four possible options and the learner had to respond with
the correct match. Each time an error was made, the teacher delivered
an electric shock, the level of shock escalating by 15 volts for each subse-
quent error, all the way up to 450 volts.
42  Identity, Violence and Power

In fact, of course, the electric shock machine was bogus. But Milgram
choreographed the studies to make them realistic. The learner responded
consistently at different shock levels, expressing pain at first, then com-
plaining of a heart condition and demanding to be released, then escalat-
ing these complaints and demands and finally falling into an ominous
silence. Throughout, the experimenter impassively asked the partici-
pant to continue the study and, should the participant show resistance,
employed a pre-scripted set of four ‘prods’ to try to get them to con-
tinue. Only if resistance continued after the fourth prompt was the trial
terminated.
So would people deliver what they believed to be large and potentially
lethal electric shocks to a victim whose sole offence was to make an error
on a memory task? Certainly very few of those who Milgram questioned
believed so in advance. Of a sample of 110 people (39 psychiatrists, 31
college students and 40 middle-class adults), none believed that they
would go all the way, no one believed that they would go beyond 300
volts (labelled, on the ‘shock machine’ ‘intense shock’) and, on average,
they believed that they would go up to around 135 volts (somewhere
between ‘moderate shock’ and ‘strong shock’). When questioned about
the behaviour of others, they predicted that few people would go beyond
150 volts and only a tiny pathological fringe of about one person in a
thousand would continue to the end point.
But when it came to the actual results on what came to be known as
the ‘new baseline study’ (actually experiment 5 out of the 18 variants that
Milgram describes in his 1974 book), fully 65 % of participants went all
the way, and the mean shock level was between 360 and 375 volts. To
invoke a much over-used metaphor, these were findings which shocked
the world. But if Milgram had uncovered a phenomenon of great con-
sequence, finding an explanation of that phenomenon proved altogether
more troubling.
In the first publication of results from his studies in the Journal of
Abnormal and Social Psychology (1963), Milgram discusses 13 possible
contributory factors—and in later publications he adds yet more. This
discussion acknowledges that the participant is conflicted, torn between
two competing voices (the experimenter and the learner) and two com-
peting obligations (to heed authority and to avoid harming our peers).
He considers the various factors which increase the prestige of ­authority
 2  Evil Leaders and Obedient Masses?  43

and our obligation to authority. He also notes the fundamental uncer-

tainties surrounding the study and what can be demanded of the par-
ticipant (something we will return to presently). Eleven years later, when
Milgram published Obedience to Authority (his seminal summary of the
studies: Milgram, 1974), he again notes all these various considerations.
However, they are overshadowed by an altogether different approach: the
agentic state account.
Milgram argues that when exposed to a legitimate authority people
enter into an agentic state, a radically different state of consciousness
which is underpinned by shifts in neural functioning. It is a state in
which people see themselves “as an agent for carrying out another per-
son’s wishes” (p. 133). He then explains how entry into the agentic state
affects the behaviour of his participants:

the entire set of activities carried out by the subject comes to be pervaded
by his relationship to the experimenter; the subject typically wishes to per-
form competently and to make a good appearance before this central fig-
ure. He directs his attention to those features of the situation required for
such competent performance. He attends to the instructions, concentrates
on the technical requirements of administering shocks, and finds himself
absorbed in the narrow technical tasks at hand. Punishment of the learner
shrinks to an insignificant part of the total experience, a mere gloss on the
complex activities of the laboratory. (p. 143)

In this treatment, extreme abuses become a matter of unawareness. It isn’t

that participants uncritically endorse the harm done to victims—they
simply don’t consider it. It isn’t that perpetrators unthinkingly endorse a
‘morality’ that allows harm, it is that morality is reduced to how well they
have done the bidding of authority.

2.2 Entrenching Conformity Bias

There is probably only one other study in psychology that comes close
to matching Milgram in terms of notoriety and impact. That is Philip
Zimbardo’s Stanford Prison Experiment, conducted a decade later in
1971. This experiment is again so well known that it only needs m
­ inimal
44  Identity, Violence and Power

introduction. Ordinary young college students were randomly divided

into the role of Prisoners and Guards and incarcerated in a simulated
prison environment (actually the basement of the Psychology Department
at Stanford University), with Zimbardo himself acting as the ‘Prison
Superintendent’. According to the received account, to be found in vir-
tually any psychology textbook, the Guards rapidly became so brutal, and
the Prisoners so passive and disturbed, that the study—scheduled for two
weeks—had to be terminated after only five days (see Haney, Banks &
Zimbardo, 1973 or Zimbardo, 2007 for fuller accounts). Like Milgram,
Zimbardo filmed much of his study, and the images remain shocking to
this day (see Zimbardo, 1989). Guards wielding billy clubs can be seen
abusing the prisoners, sexually humiliating them and imposing harsh
physical punishments.
For Zimbardo, these events were eloquent testimony to the power of
the situation over human behaviour. People simply cannot escape the
demands placed upon them. Dress them in a uniform, put them in a role
and they are virtually programmed to enact the associated role require-
ments. Or, as Zimbardo and his co-researchers had it, the aggression of
the Guards was “emitted simply as a ‘natural’ consequence of being in
the uniform of a ‘guard’ and asserting the power inherent in that role”
(Haney et al., 1973, p. 12).
As Stanley (2006) recognised, this argument takes us a step further
than the obedience studies:

Zimbardo’s prison study was even more shocking [than Milgram’s research],
if only because the students assigned to play guards were not instructed to
be abusive, and instead conformed to their own notions of how to keep
order in a prison.

We would take slight issue with this formulation. The suggestion is not
that people simply enacted their idiosyncratic notions of Guard (or
Prisoner) behaviour. Rather, they carried out socio-cultural notions of
what the role entails. The point is that, according to Zimbardo, you don’t
even need an authority figure to stand over you and police conformity
to the existing social order. Human psychology ensures that people will
police themselves. We are all born conformists. We act in ways that repro-
 2  Evil Leaders and Obedient Masses?  45

duce the status quo. Human beings have an inner conservatism which
keeps the social system going.
The view of people as inherently conservative goes hand in hand with
a view of people as essentially passive—and the more conservative, the
more passive. This is true both of masses and of elites. Thus, the whole
point of Milgram’s misleadingly named ‘agentic state’ analysis is that peo-
ple lose agency. They simply enact the will of others. But, ironically, the
fact that people are held to enter this state in the presence of authority
means that the authority figure doesn’t need to do anything in particular
in order to secure compliance. The business of leadership is reduced to
simply ‘being there’. And, as we have just argued, Zimbardo’s formu-
lation extends the argument still further: people are driven by external
forces even in the absence of authority, and authorities don’t even need to
be there in order for the systems and norms they represent to be upheld.
In such a world, social change is rendered all but impossible.
When, over three decades later, Zimbardo (2007) published The
Lucifer Effect, he re-asserted this perspective, and went even further in
his stance to seek agency—and hence, accountability—elsewhere than
among the immediate perpetrators. Interestingly, when Zimbardo
explains the methodology underlying his book, he emphasises the need
to build charges against “senior military officers” and their accomplices
in the “civilian command structure” (rather than against low-ranked sol-
diers)—but mainly for pragmatic reasons pertaining to the “limits of our
legal system”, which demands that individuals and not situations or sys-
tems be tried for wrongdoings:

it is time (…) to go up the explanatory chain from person to situation to

system. Relying on a half dozen of the investigative reports into these
abuses and other evidence from a variety of human rights and legal sources,
I adopt a prosecutorial stance to put the System on trial. Using the limits
of our legal system, which demands that individuals and not situations or
systems be tried for wrongdoing, I bring charges against a quartet of senior
military officers and then extend the argument for command complicity to
the civilian command structure within the Bush administration. The
reader, as juror, will decide if the evidence supports the finding of guilty as
charged for each of the accused. (p. XIII)
46  Identity, Violence and Power

In recent years, a number of models have emerged which extend this

logic. These include system justification theory, which, as the name sug-
gests, proposes an inherent bias, even amongst the disadvantaged, towards
accepting and rationalising the status quo (Jost & Hunyadi, 2002). In a
line of thought similar to Zimbardo’s, system justification theory goes a
step further in teaching us why ‘the system’ is so powerful: because we
all want it to be powerful, almost regardless of whether it is functional
or dysfunctional, just or unfair, beneficial or harmful to our own inter-
ests. There are particular implications for how people react—and why
they react the way they do—in times of crisis and instability: by embrac-
ing conformity and respect for authority even more when the system is
shaken:

An additional hypothesis that may be derived from system justification

theory is that people should be motivated to defend the existing social
system against threats to the stability or legitimacy of the system. If there is
a defensive motivation associated with system justification, then it should
be more pronounced under circumstances that threaten the status quo […]
Thus, we hypothesized that situations of crisis or instability in society will,
generally speaking, precipitate conservative, system-justifying shifts to the
political right, but only as long as the crisis situation falls short of toppling
the existing regime and establishing a new status quo for people to justify
and rationalize. (Jost, Glaser, Kluganski, & Sulloway, 2003, p. 351)

In other words, people would generally oppose social change as long

as it does not happen—a view which clearly does not locate the forces
that (sometimes do) bring about social change in the agency of ordinary
people.
But most influentially perhaps among the contemporaneous theories
that see humans as generally more skilled in conforming to social hier-
archies than in toppling them is social dominance theory. The classic
statement of the theory (Sidanius & Pratto, 1999) opens with a counsel
of despair:

Despite tremendous effort and what appear to be our best efforts stretching
over hundreds of years, discrimination, oppression, brutality, and tyranny
 2  Evil Leaders and Obedient Masses?  47

remain all too common features of the human condition. Far from having
escaped the grip of human ugliness in the civil rights revolutions of the
1960s, we seem only to have increased the overall level of chaos, confusion,
and intergroup truculence during the post-civil rights era and the resolu-
tion of the cold war. (p. 3)

The message couldn’t be clearer: resistance is futile. At best it is an illu-

sion and most likely it will make things worse. This is because it goes
against the grain of human nature. Sidanius and Pratto spell out the three
‘primary assumptions’ on which social dominance theory is based: first,
that all social systems have hierarchies based on age and gender and other
categories besides; second, that “most forms of group conflict (e.g. rac-
ism, ethnocentrism, sexism, nationalism, classism, regionalism) can be
regarded as different manifestations of the same basic human predisposi-
tion to form group-based social hierarchies” (p. 38); third, that there may
be countervailing tendencies to attenuate hierarchies, but, still, these will
always be weaker than the tendencies to enhance hierarchies and, at best,
will moderate rather than eliminate inequality.
It would be unfair and misleading to leave our description at this. For
social dominance theory is a rich and multi-dimensional account which
addresses the role of ideologies (or ‘legitimating myths’) in promoting
or challenging hierarchy and also the role of institutions in structuring
the lived experience of inequality (see, for instance, Sidanius & Pratto,
1999, p. 40). Nonetheless, for all this sophistication, the theory still puts
the universal urge for inequality and domination at the start of a one-­
way process; it still conceptualises this universal as driving the process,
and while ideologies and structures may affect how the urge manifests
itself, they are nonetheless fated to transmit it in one way or another.
Once again, we are reminded of Sisyphus: we cannot roll away the bur-
den of our hierarchical nature, and if we try it will come back to crush us.
Sidanius and Pratto aim to spare us the pain and the disillusion involved.
As they say at the end of the book: “we hope our work has helped reveal
some of the consensually approved social practices and beliefs that pre-
vent us all from realizing our collective democratic and inclusionary ide-
als” (p. 310)
48  Identity, Violence and Power

This, then, takes the conservatism argument to a new level. If Milgram

suggests that our psychology inclines us to obey authority figures and
Zimbardo proposes that we do what is expected of us even without bid-
ding, system justification theory and social dominance theory propose
that, even if the local contingencies temporarily steer us towards social
instability or even rebellion, our natures will inexorably lead us back
towards recreating the status quo ante.

2.3 A Mass-Level Excuse?

In the previous chapter, we wondered whether bottom-up theories of
collective violence—incidentally or deliberately—serve a function of
exonerating elites. We now need to address the reverse question: Does
framing collective violence in terms of simple obedience absolve the
masses? Might it even be that, in part, that is precisely why these theories
have become so popular?
Zimbardo is the most explicit about his motivation to seek “mitigating
circumstances” for rank-and-file soldiers involved in grave human rights
breaches. In the introduction to The Lucifer Effect he describes his feelings
provoked by his participation as an expert witness for the defence of one
US soldier, eventually condemned and sanctioned because of his role in
torturing Iraqi prisoners in Abu Ghraib: “I was (…) frustrated and angry,
first by the military’s unwillingness to accept any of the many mitigating
circumstances I had detailed that had directly contributed to his abusive
behavior and should have reduced his harsh prison sentence” (Zimbardo
p. X).
Not only does Zimbardo describe the Abu Ghraib trial as an unsettling
and formative experience, he also uses his observations to sustain the
conclusions from the Stanford Prison Experiment regarding the pre-emi-
nence of situational forces over individual will, in a real-world context:

One of the dominant conclusions of the Stanford Prison Experiment is

that the pervasive yet subtle power of a host of situational variables can
dominate an individual’s will to resist. With this set of analytical tools at
 2  Evil Leaders and Obedient Masses?  49

our disposal, we turn to reflect upon the causes of the horrendous abuses
and torture of prisoners at Iraq’s Abu Ghraib Prison by the U.S. Military
Police guarding them. The allegation that these immoral deeds were the
sadistic work of a few rogue soldiers, so-called bad apples, is challenged by
examining the parallels that exist in the situational forces and psychological
processes that operated in that prison with those in our Stanford prison.
(Zimbardo, p. XII–XIII)

In the context of the Abu Ghraib trial, the fact that only a very small
number of soldiers were judged provoked a debate as to whether they
were really exceptionally ‘bad apples’ or just chosen scapegoats—whether
through their trial, all ordinary soldiers who had been similarly perverted
by the system were symbolically tried. By contrast, there has been one
recent judicial experiment where the masses were literally put on trial.
Consistent with Rwanda’s political option not to leave participation in
the genocide unpunished, between 2002 and 2012 an estimated 700,000
suspects have been tried by 250,000 specially elected judges (McKnight,
2014) in local Gacaca proceedings, whose ad hoc procedures were created
when the country had to face the unprecedented challenge of bringing
to justice about a tenth of its entire population. In that context, Strauss
(2006) collected his own testimonies among perpetrators. Among the
reasons for taking part in the genocide, Strauss highlighted obedience:

First, men participated in the killing because other men encouraged,

intimidated, and coerced them to do so in the name of authority and ‘the
law’. Many respondent described situations where they believed that they
faced a choice between being punished or committing violence and many
choose the latter. (p. 141)

Finally, Strauss concluded that “intra-ethnic coercion and pressure appear

to have been greater determinants of genocidal participation than inter-
ethnic enmity”. This conclusion was commented on in the following way
by one of Strauss’ colleagues, a direct observer at the Gacaca trials:

This explanation is consistent with many of the testimonies I have heard in

gacaca trials, but more systematic analyses of those testimonies and more
micro-level studies are needed. A constant refrain that Straus hears from
50  Identity, Violence and Power

confessed perpetrators is that they were following orders and that disobedi-
ence would have led to punishment or even death. This sounds like egre-
gious self-absolution from admitted killers, but Straus makes us take it—and
them—seriously. (Waldorf, 2007, p. 268)

Shall we take perpetrators seriously when they tell us that they only fol-
lowed orders—knowing that this is their best chance to be granted miti-
gating circumstances? And how shall we judge the theories that do take
seriously perpetrator’s narratives of their own role in atrocities? These
questions take on yet another dimension when the person claiming to
have only followed orders is in effect a high-ranking official holding a key
position in a genocidal chain of command.

2.4 The ‘Banality of Evil’

Adolf Eichmann was in exactly such a position when the Nazi regime
implemented its ‘final solution’. He is widely regarded as the chief func-
tionary of the Holocaust. He is the man who ensured that millions of Jews
and others were delivered to the death camps. Not surprisingly, when he
was captured in Argentina, smuggled to Israel and scheduled to stand trial
for his immense crimes, there was a sense of fervent anticipation. Those sit-
ting in the Jerusalem courtroom on the day the proceedings started—the
German-Jewish political philosopher Hannah Arendt amongst them—
were fascinated by the prospect of finally encountering the individual who
bore responsibility for so many deaths. Surely a man who had done such
monstrous things would have the aspect of a monster. He would display
a “dangerous and perverted personality” (Arendt, 1963/1994, p. 21). He
would be “a man obsessed with a dangerous and insatiable urge to kill”
(ibid). The shock, then, was all the greater when Eichmann finally stepped
into his glass booth, flanked by two security guards.
The man they saw appeared anything but monstrous. He seemed
mild and inoffensive: small, thin and balding, slightly stooped, peering
out from thick glasses, fastidiously taking notes on the proceedings. Far
from being exceptional—a creature apart—he seemed typical. A typi-
cal bureaucrat. And this image, this paradox, the contradiction between
 2  Evil Leaders and Obedient Masses?  51

what the man had done and how he seemed, lay at the core of Arendt’s
reports from the trial and her famous book Eichmann in Jerusalem first
published in 1963. It generated what, for Arendt, was the key lesson
of the trial and the key message of her analysis, that is, “the lesson of
the fearsome, word-and thought defying banality of evil” (1963/1994,
p. 252, emphasis in the original). It was the only time the phrase ‘banal-
ity of evil’ appeared in the text, but the phrase has come to dominate our
understanding of human atrocities ever since.
Arendt does not only argue that ordinary people do have the capacity
to do extraordinary harm, she also provides an account of the processes
which can lead to such harmdoing. The problem, she argues, is cogni-
tive rather than moral. Perpetrators do what they do through thought-
lessness. They become so fixated on the process of doing their jobs—on
being trustworthy bureaucrats—that they lose sight of the consequences
of their actions. Their focus is on how well they fulfil the demands put
upon them as opposed to how they impact others. Their aim is to be
good followers rather than good human beings. In Arendt’s own words,
Eichmann was someone who obeyed orders and who obeyed the law: he
“had no motives at all. He merely, to put the matter colloquially, never
realized what we was doing” (1963/1994, p. 287).
However, it was no accident that at his trial, Eichmann appeared as an
insignificant, mild and fastidious bureaucrat. This was a deliberate strat-
egy, agreed with his lawyer, to confound the expectations of the prosecu-
tion. As one analyst acerbically observes: “in suggesting that [Eichmann]
was ‘merely thoughtless’ [Arendt] in fact adopts the very self-presentation
he cultivated” (Vetlesen, 2005, p. 5; see also Cesarani, 2004). Moreover,
Eichmann was not the first to adopt such a strategy. Neitzel (2007) has
unearthed a fascinating archive containing secret recordings of German
prisoners of war in British hands during World War II. In their unguarded
conversations, senior officers are aware that they may be held culpable for
their part in the Holocaust and they discuss ways of avoiding responsibil-
ity—including the argument that they were merely ‘following orders’.
Moreover, they realise that, for the argument to stick, all have to agree to
the same line.
The interpretation of a text is not always in the hands of the author.
In Arendt’s case, the interpretation of Eichmann in Jerusalem and the
52  Identity, Violence and Power

meaning of ‘the banality of evil’ was irrevocably influenced by its conjunc-

tion with Milgram’s investigation of the human capacity for harmdoing,
which started exactly one week before Eichmann’s trial finished. Milgram
acknowledges his debt to Arendt in the early pages of the 1974 book,
which at the same time foreshadows his account of the ‘agentic state’.
Referring first to the Holocaust, and then to his own studies, Milgram
writes:

After witnessing hundreds of ordinary people submit to the authority in

our own experiments, I must conclude that Arendt’s conception of the
banality of evil comes closer to the truth than one might imagine. The ordi-
nary person who shocked the victim did so out of a sense of obligation—a
conception of his duties as a subject—and not from any peculiarly aggres-
sive tendencies. (1974, p. 6, emphasis in the original)

It is arguable that Arendt’s position has been much misunderstood (see,

for instance, Newman, 2001). Certainly her notions of thoughtlessness
and of responsibility (e.g., Arendt, 2005) are richer and more nuanced
than what the Milgram–Arendt conjunction suggests. They are less to do
with crude unawareness of one’s acts than with a lack of reflection about
what one does. They are to do with an uncritical acceptance of a tradition
or an authority, more than a failure to heed them.
The publication of Eichmann in Jerusalem was met with a furious
controversy, which had less to do with Arendt’s comments regarding
Eichmann himself than her critique of the role of the Jewish authorities
during the Holocaust. At one point, Arendt was accused of being a self-­
hating Jew, one who was ashamed of her Jewishness and wished to ­disown
it. She retorted that being Jewish is an indisputable fact, one which she
cannot change and would not disclaim, but vigorously disputed that this
means she must love the Jewish people or love what they do. If anything,
her responsibility is to be critical of other Jews. She asserts that “there can
be no patriotism without permanent opposition and criticism” and that
“wrong done by my own people naturally grieves me more than wrong
done by other people” (Arendt, 2007, p. 467).
Nonetheless, it is the synthetic reading of the ‘banality of evil’, in
which Arendt is buttressed and read through the prism of Milgram,
 2  Evil Leaders and Obedient Masses?  53

which has endured. It is a reading which sees obedience as almost natu-

ral, neurochemical and inevitable: people cannot help but obey orders,
however extreme. This reading actually undermines Arendt’s insistence
on the choice and responsibility of those who commit ‘crimes of obedi-
ence’. It is a reading which portrays the feelings, motives and beliefs of
the masses as altogether irrelevant to the explanation of conflict and its
excesses. Indeed, it renders the mass as a mere cipher of elite intentions.
And (to repeat) it is a reading which has proved remarkably resilient and
popular—for after all, if the historical and the psychological records con-
cur, surely there must be something in it? To quote Novick (2000):

‘From the sixties on, a kind of synergy developed between the symbol of
Arendt’s Eichmann and the symbol of Milgram’s subjects, invoked in dis-
cussing everything from the Vietnam War to the tobacco industry, and, of
course, reflecting back on discussions of the Holocaust’. (p. 137)

2.5 Beyond Obedience

There are several problems with the assumption that human beings are
natural conformists, that they are destined to reproduce existing social
relations of authority and inequality, and that this explains the human
capacity to pursue the most vicious acts against others. This assumption
radically underplays the level of resistance to authority, to the roles into
which we are cast and to systems of inequality. What is striking is that,
when one goes beyond the received accounts and looks closely at the
primary materials, even those studies which have been used to entrench
the ‘conformity account’ are replete with examples of people refusing to
conform or to heed authority.
In the case of Milgram these examples are easy to find, because he
was so systematic in documenting his findings, including a substan-
tial archive which now resides in the Sterling Memorial Library at Yale
University. As we have already outlined, Milgram conducted many vari-
ants of his basic paradigm—18 detailed in his 1974 book and several
others, including some characterised as pilots, which are fully available
in the archive (see, for instance, Rochat & Blass, 2014). This begs the
54  Identity, Violence and Power

question of why one condition, as described above, has been picked out
and described as a ‘baseline’. In fact, in the different conditions, the
percentage of people who are fully compliant with the experimenter
and who continue shocking to the end of the scale varies from 0 % to
100 % (Reicher & Haslam, 2012). There is no principled reason to pick
out one of these studies and claim it is more foundational than the rest,
and therefore to characterise the level of compliance in that study as
more characteristic than the others. That is, the studies provide no basis
for privileging conformity over resistance (or vice versa). Rather, seen in
the round, they raise the question of when people comply in the face of
authority and when they resist.
The same goes for the Stanford Prison Experiment—although here the
argument is made more difficult by the fact that we only have a partial
record of what exactly happened. Nonetheless, even from the materials
that are available in the public domain (e.g., Zimbardo, 1989, 2007) it is
clear that the received account is, once more, very partial. Far from being
universally passive, the Prisoners at first acted together to challenge the
authority of the Guards. Indeed, by the end of the first day they were
dominant and until the end of the study some Prisoners continued to
rebel, albeit now they were more isolated. Equally, far from accepting
their role, many of the Guards were deeply uneasy about their authority.
Some actively sided with the Prisoners, some sought to be scrupulously
fair, and there is only one clear example of a Guard being systematically
brutal (see Zimbardo, 1989; Reicher & Haslam, 2006).
Once again, this leads us to shift from asking ‘why do people conform?’
to ‘when do people conform and when do they resist?’ In order to address
this latter question, Reicher and Haslam ran a study using a system of
Guards and Prisoners in a simulated prison setting. The study was not a
replication of Stanford, but rather introduced a number of interventions
which were designed to inform the question of when people rebel. Even
more than in Zimbardo’s study, the Guards became divided over their
use of authority and the Prisoners united in rejecting their subordination.
This led to a reversal of the original power relations and ultimately to a
collapse of the Prisoner-Guard system.
When the study was published, Zimbardo critically reacted, pointing
out the artificiality of the setting: “what is the external validity of such
 2  Evil Leaders and Obedient Masses?  55

events in any real prison anywhere in the known universe? In what kind
of prisons are the prisoners in charge? How could such an eventuality
become manifest?” (2006, p. 49). This is a fair challenge, but it is actually
remarkable how easy it is to find evidence of resistance in prisons, even
to the extent of prisoners effectively running the system (e.g., Carroll,
2006; McEvoy, McConnachie, & Jamison, 2007; Mariner, 2001). Even
in the most repressive of settings (perhaps particularly in such settings)
prisoners are able to be in control. Take, for example, the case of Robben
Island where Nelson Mandela was for long imprisoned. Mandela writes:
“ultimately we had to create our own lives in prison. In a way that even
the authorities recognized, order in prison was preserved not by ward-
ers but by ourselves” (1994, p. 464). Moreover, Mandela provides some
insights into how such control was achieved—insights which, as we shall
see, will prove very helpful in our later discussions:

Our survival depended on understanding what the authorities were

attempting to do to us, and sharing that understanding with each other. It
would be hard, if not impossible for one man alone to resist (…) The
authorities’ greatest mistake was to keep us together, for together our deter-
mination was reinforced. We supported each other and gained strength
from each other. Whatever we knew, whatever we learned, we shared, and
by sharing we multiplied whatever courage we had individually. (p. 463)

But, if we are talking of repressive settings and oppressive systems, we are

necessarily led back to the context which generated our contemporary
obsession with conformity and obedience. We are led back to the Nazi
era and, in particular, to the carceral system of camps. The system was
distinctive not only in its brutality but also in the way it was designed
to keep people in a state where resistance would be impossible—divided
from each other, set against each other, starved and humiliated (Sofsky,
1997). And yet, despite this, resistance did occur (Langbein, 1994;
Unger, 1986). It took many forms, including open revolt. At the apex of
the system were six ‘death camps’ designed for the systematic extermina-
tion of inmates: Auschwitz, Belzec, Chelmno, Majdanek, Sobibor and
Treblinka. At three of these (Auschwitz, Sobibor and Treblinka) there
were uprisings, and there are unconfirmed reports of another at Belzec.
56  Identity, Violence and Power

The best documented uprising was at Sobibor where, on 14 October

1943, inmates seized arms, killed guards and broke free. Of 600 inmates,
300 escaped and 50 survived the war. Here, as elsewhere, most of those
who resisted the Nazis died in the attempt. This was inevitable given the
grossly unequal relations of power. As Mais reflects, summing up the
current consensus on Jewish resistance during the Holocaust in general,
“true, the Jews were slaughtered, but clearly not as sheep” (2007/2008,
p. 19).
It is important not to swap from naturalising conformity to naturalis-
ing resistance. Resistance is not a universal phenomenon, and still less is
resistance universally successful. However, even in the most oppressive
settings that malignant human imagination has managed to devise, the
potential for resistance remains present. There is nothing either in the
world of psychological research or in the world beyond to sustain the view
that people are natural conformists, inherently incapable of disobeying.
Obviously, by stressing resistance we stress the active nature of the
subject: people do not just go along with a world made by others but also
seek to shape their own social worlds. It is possible to take the argument
further, and to address a second problem with the argument that humans
are natural conformists: it radically overplays the degree of passivity of both
elites and masses.
Actually, people are active when they conform as well as when they
resist. To stay with the case of the Holocaust, let us now reconsider the
biography of Adolf Eichmann. Unlike the way Eichmann presented him-
self during his trial, he and others were not just doing as they were told.
Eichmann rose to prominence through the creative and innovative ways
he found to deport the Jewish population in Vienna (Cesarani, 2004).
The apogee of his impact came in 1944 when he was sent to Budapest to
oversee the deportation of some half a million Jews to the death camps
(and Eichmann, who had visited them, knew very well what went on
beyond the barbed-wire). At that point, Himmler (Eichmann’s superior)
was well aware that Germany was losing the war. He tried to do a deal
with the allies, asking for military materials in return for Jewish lives.
Eichmann, however, was fiercely opposed and he challenged Himmler’s
proposals. Far from ‘merely’ following orders, Eichmann believed in the
extermination of Jews, and he worked assiduously to achieve it—even
 2  Evil Leaders and Obedient Masses?  57

against orders. The extent of this belief was revealed after the war when,
speaking to a fellow Nazi, he insisted that:

My innermost being refuses to say that we did something wrong. No—I

must tell you in all honesty, that if of the 10.3 million Jews shown by [the
statistician] Korherr, as we now know, we had killed 10.3 million, then I
would be satisfied. I would say ‘All right. We have exterminated an enemy’.
(Cesarani, 2004, 219)

Nor is Eichmann distinctive in this regard. Kershaw (1993) makes

the point that to argue that Nazis were blindly obedient is to mis-
understand the nature of the Nazi State. Hitler certainly indicated
broad objectives, but he rarely gave specific instructions. Hence, his
underlings had to be creative in determining how to achieve these
objectives and each competed to outdo the other in how far their
achievements went. This, Kershaw describes as “working towards the
Fuhrer”. It is as true of the functionaries of the Holocaust as anyone
else. As Lozowick puts it:

Eichmann and his ilk did not come to murder Jews by accident or in a fit
of absent-mindedness, nor by blindly obeying orders or by being small cogs
in a big machine. They worked hard, thought hard, took the lead over
many years. They were the alpinists of evil. (2002, p. 279)

There may be something very specific about the structure of the Hitler
State—and the Fuhrer’s aim to set his underlings to compete against each
other rather than unite against him—which required Nazi functionaries
to be such active conformists. Still, there are intriguing similarities with
observations from psychological research made in completely different
contexts. A conversation between two of the participants in the Stanford
Prison Experiment, one the most brutal of the Guards, dubbed ‘John
Wayne’ for his aggressive swagger, the other one of the Prisoners that he
tormented, illustrates this point. The conversation occurred after the end
of the study and ‘John Wayne’ (or rather, David Eshelman, to give him
his real name) asked, “what would you have done if you were in my posi-
tion?” The erstwhile Prisoner replied:
58  Identity, Violence and Power

I don’t know. But I don’t think I would have been so inventive. I don’t think
I would have applied as much imagination to what I was doing. Do you
understand? … If I had been a guard I don’t think it would have been such
a masterpiece. (Zimbardo, 1989)

This tallies with Eshelman’s own account where he claims that he was
running his own ‘experiment’, trying out new forms of abuse and see-
ing how people would respond. In other words, he could be said to have
been ‘working towards the experimenter’—not mechanically doing the
bidding of another but actively creating innovative ways of carrying out
his task. Albeit on a totally different scale of brutality, this tallies with
contemporary accounts of Nazi conformity. Yet, of course, it is based on
a single anecdote and it would be rash to hang too much explanatory
weight upon it. So it makes sense to turn again to Milgram’s studies, both
because of the systematic nature of the evidence and because of the sig-
nificance of these studies in underpinning the ‘banality of evil’ account.
There are many indications from Milgram’s own data which give the
lie to his claim that “punishment of the learner shrinks to an insignificant
part of the total experience”. Perhaps the most dramatic evidence derives
from Milgram’s filmed record of the sessions.1 In these one can see the
participants agonise over what to do. They certainly show awareness of
the learner’s predicament. They alert the experimenter to his expressions
of pain, his demands to be released, his silences. They argue and remon-
strate with the experimenter and seek reasons to justify terminating the
study (see Gibson, 2011, 2014 for a detailed analysis of their discursive
strategies). They sigh and sweat and giggle nervously. And when, at the
end, the learner emerges to reveal that he has not suffered, they show
massive relief. It lacks all credibility to suggest that the learner is insig-
nificant to them.
If such descriptive evidence is deemed insufficient, it is complemented
by quantitative evidence that points to exactly the same conclusions.
Thus, an analysis of the points at which participants are most likely to defy
the experimenter points to the 150- and 315-volt points—­respectively
1
 Many of these can be found online. See, for instance, Milgram’s own film Obedience at http://
veehd.com/video/4751627_Obedience-The-Original-Milgram-Experiment-1962-nYx64—espe-
cially the segment from 22 mins to 39 mins which shows a trial almost in its entirety.
 2  Evil Leaders and Obedient Masses?  59

the points at which the learner first asks to be let out of the study and
first states categorically that he is “no longer part of the experiment” (see
Milgram, 1974, pp. 56–7 for a full transcript of what the learner says at
the various shock levels). Evidently, participants are paying attention to
the learner.
But equally, the participants are listening carefully to what the experi-
menter says. They are not simply in thrall to anything he says. Thus, the
different prods used to urge people on are differentially efficacious. Most
obviously, three of the prods can be read as requests (“please go on”) or
as justifications (“the experiment requires that you continue”). Only one,
the fourth and final prod, is unambiguously an order (“you have no other
choice, you must go on”) (see Milgram, 1974, p.  21, emphasis in the
original). The evidence that is available suggests that the use of the fourth
prod in particular led to heightened disobedience. Moreover, in a recent
replication (Burger, 2009), every time that the fourth prod was used,
disobedience ensued. Now, it is arguable that this has nothing to do with
the content of the prod, but simply an order effect—that is, by the time
it comes to the experimenter having to urge people on for a fourth time,
nothing they say would be effective. So, in a recent study, Haslam, Birney
and Reicher (2014) have untangled order from content, using different
prods in different conditions of the study. Still we find that prod 4 incites
greater disobedience. In the light of this evidence it seems that, what-
ever the Milgram studies do show, they certainly don’t demonstrate that
people always obey orders. Quite the contrary. People are making active
choices between the experimenter and the learner based on precisely how
each addresses them.
This argument has implications for the experimenters as well as the par-
ticipants: once the former acknowledge that the latter are discerning and
that they actively weigh what the experimenters do, then experimenters
can no longer rest on their laurels. Once it is clear that just ‘being there’
is not enough, we need to attend to what authorities have to do and say
in order to secure compliance. We have to attend to their activity as well
as that of the participants. A number of recent studies have done precisely
this. They unpick all the careful work that Milgram undertook—and had
his experimenter undertake—in order to ensure that participants kept
shocking. This includes the bureaucratic structure of the study (Russell,
60  Identity, Violence and Power

2014), the design of the shock machine and the wording used to describe
different shock levels (Russell, 2011), the careful choice of personnel to
act as the experimenter and the learner, respectively (Russell, 2013), the
various forms of contractual obligation that participants had to accept
before they began the study, the careful scripting of the prods, the ways
that, at times and seemingly with Milgram‘s consent, the experimenter
would depart from the script during the study (Gibson, 2011, 2014),
as well as the broad moral justification for the studies, couched within a
discourse of scientific progress.2
In sum, whether we are dealing with conformity or resistance, with par-
ticipants/masses or experimenters/authorities, everyone is actively involved
in making sense of the situation and determining what should be done.
Moreover, the activity of the mass and of the authorities is interdepen-
dent. It is because ordinary people actively weigh what is said to them that
those who aim to influence them need to couch their words and actions so
carefully. Equally, it is because different sources of influence address them
with plausible and powerful arguments that ordinary people must work at
choosing between them. So how do they decide? What determines which
of the many voices that surround them people eventually heed?
A third and final problem with the natural conformist assumption is
then that it does not allow these simple but fundamental questions to be
addressed or answered. As we will now see, the reasons for this failure largely
stem from the assumption’s one-sidedness in relying solely on internal psy-
chological mechanisms to explain why and when people do conform.

2.6 C
 onformity, Resistance and the Problem
of Epistemic Isolation
When Milgram’s experiment was moved from the prestigious Yale labo-
ratories to an office building in downtown Bridgeport and the experi-
menter was introduced as working for an unknown private organisation
2
 The study was introduced as an important investigation into the topic of learning and of how
much punishment is best to aid learning. When participants came to the laboratory it was stressed
that: “we know very little about the effect of punishment on learning, because almost no truly sci-
entific studies have been made of it in human beings” (Milgram, 1974, p. 18).
 2  Evil Leaders and Obedient Masses?  61

rather than for the University, the percentage of people who were fully
obedient decreased from 65 % to 47.5 %. When the ‘experimenter’ was
introduced as another ‘ordinary person’, obedience decreased further
to 20  % and when there were two experimenters who disputed as to
whether shocks should be given, no one fully obeyed. Indeed, no one
went beyond point 11 (165 volts) of the 30-point scale (450 volts). That
is, as the experimenter’s scientific authority becomes less clear and more
contested so people follow his instructions less. Or, in other words, peo-
ple’s willingness to obey depends in part on the epistemic capital of the
authority.
But equally, obedience depends upon the epistemic capital of the par-
ticipant. Consider the following interchange between the experimenter
and Jan Rensaleer, an industrial engineer and a participant in one of the
early studies:

MR. RENSALEER: The man, he seems to be getting hurt.

EXPERIMENTER: There is no permanent tissue damage.
MR. RENSALEER: Yes, but I know what shocks do to you. I’m an
electrical engineer, and I have had shocks.
(Milgram, 1974, p. 51)

Or again, read this interchange with Getchen Brandt, a medical tech-

nician and a participant in the only study involving women:

GRETCHEN: He has a heart condition, I’m sorry. He told you

that before.
EXPERIMENTER: The shocks may be painful but they are not
dangerous.
GRETCHEN: Well, I’m sorry, I think when shocks continue like
this, they are dangerous. You ask him if he wants
to get out. (Milgram, 1974, p. 85, emphasis in the
original)

In both cases, the participants have technical expertise and technical

knowledge which allows them to contest the claims of the experimenter
and hence reject his authority to define the situation.
62  Identity, Violence and Power

But it is important not to reduce epistemic certainty to attributes of

the various parties involved in the study. Critically, it is important to con-
sider also the relationships between them. Perhaps the best known subset
of studies within Milgram’s programme was the one where he varied the
closeness of the participant to the learner. In the first study the partici-
pant is with the experimenter in a different room to the learner and gets
no feedback apart from banging on the wall after 300 volts. All partici-
pants go up to this level, and 65 % are fully obedient to the end. In the
second study, there is more voice feedback such that the participant hears
the learner expressing complaints and making demands throughout the
study. Here participants start breaking off earlier, from the 135-volt level,
but much the same percentage (62.5 %) continue to the end. In a third
study, the experimenter, participant and learner are in the same room so
that the participant can see and hear all that goes on. Again people break
off earlier and many less (40 %) continue up to 450 volts. Finally, in a
fourth study, the participant is not only in the same room as the learner
but also has to push his hand onto a metal plate in order (ostensibly) to
deliver shocks. In this condition, full obedience reduces to 30 %.
The largest discontinuity, then, is less to do with precisely what the
participant hears and does, but between a situation in which he is iso-
lated with the experimenter in a separate room or not. Milgram himself,
in one of his many early explanations that we have alluded to, suggests
that these physical arrangements may be important in terms of group
­formation (see Milgram, 1965). Indeed, he briefly reprises the argument
in his 1974 book:

placing the victim in another room not only takes him father from the
subject, it also draws the subject and the experimenter relatively closer.
There is an incipient group formation between the experimenter and the
subject, from which the victim is excluded. The wall between the victim
and the others deprives him of an intimacy which the experimenter and the
subject could feel. In the Remote condition, the victim is truly an outsider,
who stands alone physically and psychologically. (1974, p. 39)

The quote is worth reproducing in full because it introduces a construct

that will prove crucial as a thread that runs throughout our analysis. This
 2  Evil Leaders and Obedient Masses?  63

is the notion that the extent to which people are in contact or else isolated
from each other—and whose communications are privileged over oth-
ers—is of central importance to how they understand who they are and
what they should do. Alone with the experimenter, participants are sub-
ject overwhelmingly to their relationship with him. It is not simply that
this has the cognitive effect of rendering their common inclusion within
a scientific enterprise more salient. It is also that this has the pragmatic
effect of shielding the participant from any alternative perspective which
challenges the experimenter’s account of scientific progress and what that
justifies. In a phrase, epistemic isolation lends epistemic certainty to the
perspective of authority.
When it comes to epistemic isolation, there is another relationship
which is equally important, if not more so, as that between participants
and learners. It is one that is largely hidden in Milgram’s studies because
the design generally involves just one participant at a time, whereas, out-
side the laboratory, there are often many of us together in the face of
authority. That is, in these studies, participants are isolated from their
fellow participants. They have no knowledge of what their peers think
and do. They don’t know if it is normal to privilege the interests of sci-
ence or of the ordinary person. As Milgram acknowledges, they also
don’t know if shocking is bizarre behaviour or if not shocking would be
bizarre. In this state, they are again confronted only with the position of
the experimenter and the way they are positioned by him. They have no
counterweight with which to challenge that positioning. Except in one
condition.
In the so-called two peers rebel condition, the subject is in a room with
the experimenter and (as the name suggests) two peers who co-operate in
administering the task—these supposed peers being yet more confeder-
ates of Milgram. At the 150-volt point, one of these peers withdraws from
the study. The others are told to continue without him. At the 210-volt
point the other peer withdraws. Again, the remaining (authentic) partici-
pant is told to continue. In this condition only 10 % of these participants
are fully obedient. Significantly, perhaps, only three pull out when the
first peer rebels. Twelve pull out when the second peer rebels. One way of
interpreting this is to say that in order to defy authority it is not enough
to witness examples of defiance; it is necessary to establish a consensus
64  Identity, Violence and Power

that challenges the position of authority. Whether this precise interpreta-

tion is accepted or not, there is nonetheless clear support for the idea that
the opinions of peers is critical to the way one sees one’s own self, and
that the impact of authority depends upon its ability to monopolise com-
munication and interaction with members of the population.
Great harm is generally done in the name of a great cause. Those
who we might condemn as doing ill tend to see themselves as doing
good. However, critically, the underpinnings of identity and authority
are always bound up with what others think and people’s knowledge of
what they think. Even if every single participant in Milgram’s studies had
doubts about the significance of the cause and the probity of administer-
ing shocks, these often came to nothing as long as they were isolated from
each other and exposed only to the experimenter’s narrative. The shift
from conformity to resistance, then, may not be so much a matter of
changing the minds of individuals as of overcoming the isolation between
them. To draw on Foucault’s famous use of Bentham’s panopticon as a
model of power in general—a structure in which the central authority
is at the hub and gazes out at his subjects arranged as if on the rim of a
wheel, each visible to the authority, each divided from and invisible to
the other (see Foucault, Burchell, Gordon, & Miller, 1991)—change is a
matter of overcoming the partitions between us more than the attributes
within us.
There is much more that could be said here about the relationship
between how we see ourselves and how others see us. But these are issues
that we shall return to and elaborate throughout the book. For now, and
before we conclude this chapter, let us briefly move for one last time
between psychological studies and historical instances of conformity and
resistance, and let us ask whether our overall argument is plausible in the
latter as well as the former domain.
In a telling analysis, Einwohner (2014) makes a telling case that epis-
temic isolation was at the heart of Nazi extermination policies and was
critical in determining levels of Jewish resistance. The Nazis did all they
could to hide the fact that everyone was to be killed and that deporta-
tions meant death, and those ghettos that were more isolated from others
and where rumours of the death camps were less likely to reach were also
less likely to revolt (see also Gutman, 1971; Tiedens, 1997). The twisted
 2  Evil Leaders and Obedient Masses?  65

charades at the death camps themselves—camp signs that read ‘Arbeit

Macht Frei’ (Work Sets You Free), gas chambers disguised as showers—
were designed to maintain the deceit to the very end and to allow the
killings to be more orderly. And where people lived within the camps and
the reality could no longer be hidden, other strategies were used to ensure
that those who were concentrated physically remained isolated psycho-
logically. That is, as Sofsky puts it, the aim was to ensure that “their ori-
entation is not to each other but past one another” (1997, p. 154). One
way of doing this was to set up a hierarchy so that some prisoners became
accomplices of the Nazis. This blurred the boundaries between guards
and inmates such that people did not know who were genuine peers and
became reticent about saying what they thought for fear that an appar-
ent ally could turn out to be the enemy. More crudely, the system was so
brutal that few survived long enough to get to know each other, develop
trust and thereby overcome their mutual isolation.
On rare occasions, however, inmates were able to overcome this isola-
tion, often because groups who had prior links of solidarity were allowed
to stay together in the camps (see Haslam & Reicher, 2012). This was the
case in Sobibor where the roots of the revolt lay in the arrival of some
80 Jews who had served in the Red Army and who benefited from the
leadership of a former lieutenant, Alexander Pechersky (see Arad, 1987;
Pechersky, 1975). This tallies with the description we have provided of
Robben Island where the authorities unwittingly empowered the inmates
by facilitating communication and interaction between them. It is, per-
haps, appropriate to draw our discussion to a close by reprising Mandela’s
words: “together our determination was reinforced. We supported each
other and gained strength from each other. Whatever we knew, whatever
we learned, we shared, and by sharing we multiplied whatever courage we
had individually” (1994, p. 463).

2.7 Conclusion
In this chapter we have considered another simple but powerful expla-
nation of intergroup conflict—and more particularly of the harm that
people are capable of inflicting in such conflicts. As with the explanations
66  Identity, Violence and Power

considered in the previous chapter, these reduce conflict to the expression

of something inherent in the human condition. But whereas in Chap.
1 this ‘something’ had to do with hostile urges within the actor, here in
Chap. 2 it has to do with our supposed tendency to do as others urge.
We need not hate our victims, we need have no motivations as concerns
them. Our fatal flaw is that we are ‘natural conformists’—in different
variants, conformists to authority figures, conformists to roles or con-
formists to a hierarchical status quo.
We have examined how this view gained strength by weaving together
historical and psychological strands of enquiry. And yet, we have sought
to demonstrate that both the historical and the psychological evidence
does not support the view that perpetrators have no will except to satisfy
the will of others. In fact, it tells a very different story.
First, even in those seminal studies which are used to support accounts
of our inherent conformity, resistance is rife. This is not to replace the
notion that we are natural conformists with the idea that we are natural
rebels. However, it does mean that we can’t take conformity for granted—
we must ask when it occurs and what produces conformity (or resistance)
in particular contexts.
Second, once conformity can no longer be taken for granted, it becomes
clear that people must make an active choice between as to whether they
conform or not. Whether people decide to heed or to defy authority,
they do not do so inattentively or by default. Rather, they make informed
decisions, and whatever they choose, it is something that they consider to
be the right thing to do. Equally, authorities must work hard to influence
that choice and to produce conformity.
Third, the effectiveness of authorities depends upon their ability to
shape the way in which people are in communication with or else isolated
from various potential audiences. The more the authority can monopo-
lise communication and isolate their target from others, the more they
can control how people represent their world and hence choose to act
within it.
This argument brings us back to three core concepts. First, to identity:
people’s decision to conform or to resist is a function of their sense of
identity, and the extent to which authorities are able to speak for and to
that identity. Second, to violence: the underlying concern of our whole
 2  Evil Leaders and Obedient Masses?  67

discussion, the reason why conformity has been of such concern to those
within social psychology and those beyond—indeed to all who live in
the shadow of the Nazi Holocaust and succeeding genocides. Third, to
power: the power to shape identities, in particular by creating settings in
which people are led to meet each other as foes or friends, and the power
that derives from shaping identities. Once again though, it is important
to examine the reciprocal relationship between constructs: how violence
derives from the ways that identity is defined and controlled, but also the
way in which violence serves as a means of gaining power over identity
and defining it in ways that establish particular people as authoritative.
In the following chapters, we will interrogate identity, violence and
power, and the links between them, in much more detail. For now, we
can conclude our discussion so far on a note of qualified optimism. There
is no part of the human substance which impels us, either directly or
indirectly, towards harming others. Rather, as Todorov has argued, “good
and ill are of ‘one substance’ with human life because they are the fruits
of our freedom, of our ability to choose at every point between several
courses of action” (2004, p. 26). Nonetheless, we can examine the fac-
tors that constrain our choices. Even if these do not determine the out-
come, they can make it easier or more difficult to conform or rebel; they
can strengthen or weaken, reinforce or undermine different orientations.
While these factors become consequential by the way that they work on
our psychology, they themselves exist beyond us, between us, in the ways
that we are dispersed and organised in the social world.

References
Adorno, T. W., Frenkel-Brunswik, E., Levinson, D. J., & Sanford, N. R. (1950).
The authoritarian personality. Oxford: Harpers.
Arad, Y. (1987). Belzec, Sobibor, Treblinka: The Operation Reinhard death camps.
Bloomington: Indiana University Press.
Arendt, H. (1963/1994). Eichmann in Jerusalem. New York: Penguin Books.
Arendt, H. (2005). Responsibility and judgement. New York: Shocken Books.
Arendt, H. (2007). The jewish writings. New York: Shocken Books.
Banyard, P., & Grayson, A. (2000). Introducing psychological research: Seventy
studies that shape psychology (2nd ed.). Basingstoke: Palgrave Macmillan.
68  Identity, Violence and Power

Blass, T. (2004). The man who shocked the world: The life and legacy of stanley
milgram. New York: Basic Books.
Burger, J.  M. (2009). Replicating milgram: Would people still obey today?
American Psychologist, 64(1), 1–11. doi:10.1037/a0010932.
Carroll, R. (2006, September 30). Bars, brothels and a regime of terror: Inside
the jail run by its inmates. The Guardian. Retrieved from https://www.the-
guardian.com/world/2006/sep/30/rorycarroll.mainsection
Cesarani, D. (2004). Eichmann: His life and crimes. London: Heinemann.
Retrieved from http://gataxe.iblogger.org/w/the-left-and-the-jews-the-jews-
and-­the-left-by-david-cesarani.pdf
Einwohner, R. L. (2014). Authorities and uncertainties: Applying lessons from
the study of jewish resistance during the holocaust to the milgram legacy.
Journal of Social Issues, 70(3), 531–543. doi:10.1111/josi.12076.
Foucault, M., Burchell, G., Gordon, C., & Miller, P. (1991). The foucault effect:
Studies in governmentality. Chicago: University of Chicago Press.
Gibson, S. (2011). Social psychology, war and peace: Towards a critical discur-
sive peace psychology. Social and Personality Psychology Compass, 5(5),
239–250. doi:10.1111/j.1751-9004.2011.00348.x.
Gibson, S. (2014). Discourse, defiance, and rationality: “Knowledge work” in
the “obedience” experiments. Journal of Social Issues, 70(3), 424–438.
doi:10.1111/josi.12069.
Gutman, I. (1971). Youth movements in the underground and the Ghetto revolts.
Jerusalem: Yad Vashem.
Haney, C., Banks, C., & Zimbardo, P. (1973). A study of prisoners and guards
in a simulated prison. Naval Research Review, 30, 4–17.
Haslam, S. A., Birney, M. E., & Reicher, S. D. (2014). Nothing by mere author-
ity: Evidence that in an experimental analogue of the Milgram paradigm
participants are motivated not by orders but by appeals to science. Journal of
Social Issues, 70, 473–488.
Haslam, S. A., & Reicher, S. (2012). When prisoners take over the prison: A
social psychology of resistance. Personality and Social Psychology Review, 16,
154–179.
Jost, J. T., & Hunyady, O. (2002). The psychology of system justification and
the palliative function of ideology. European Review of Social Psychology, 13,
111–153.
Kershaw, I. (1993). “Working towards the führer.” Reflections on the nature of
the Hitler dictatorship. Contemporary European History, 2(02), 103–118.
doi:10.1017/S0960777300000382.
 2  Evil Leaders and Obedient Masses?  69

Langbein, H. (1994). Against all hope: Resistance in Nazi concentration camps

1938–45. London: Constable.
Mariner, J. (2001). No escape: Male rape in U.S. prisons. New  York: Human
Rights Watch.
McEvoy, K., McConnachie, K., & Jamieson, R. (2007). Political imprisonment
and the ‘war on terror’. In Y. Jewkes (Ed.), Handbook of prisons (pp. 293–
323). Cullompton: Willan.
McKnight, J. (2014). The anatomy of mass accountability: Confronting ideol-
ogy and legitimacy in Rwanda’s Gacaca courts. Conflict Trends, 1, 35–42.
Milgram, S. (1965). Some conditions of obedience and disobedience to author-
ity. Human Relations, 18(1), 57–76.
Milgram, S. (1974). Obedience to authority: An experimental view. New  York:
Harper & Row.
Neitzel, S. (2007). Tapping Hitler’s generals: Transcripts of secret conversations,
1942–45. Barnsley, South Yorkshire: Pen & Sword Books.
Newman, L. S. (2001). The banality of secondary sources: Why social psychologists
have misinterpreted Arendt’s thesis. Unpublished manuscript, Syracuse
University, Syracuse, NY.
Novick, P. (2000). The Holocaust in American life (First Mariner Books.).
New York: Houghton Mifflin Company.
Pechersky, A. (1975). Revolt in Sobibor. In Y. Suhl (Ed.), They fought back: The
story of Jewish resistance in Nazi Europe (pp. 7–50). New York: Shocken Books.
Reicher, S. D., & Haslam, S. A. (2006). Rethinking the psychology of tyranny:
The BBC prison study. British Journal of Social Psychology, 45, 1–40.
Reicher, S. D., & Haslam, S. A. (2012). Obedience: Revisiting Milgram’s obedi-
ence studies. In J. R. Smith & S. A. Haslam (Eds.), Social Psychology: Revisiting
the classic studies (pp. 106–125). London and Thousand Oaks, CA: Sage.
Rochat, F., & Blass, T. (2014). Milgram’s unpublished obedience variation and
its historical relevance. Journal of Social Issues, 70(3), 456–472. doi:10.1111/
josi.12071.
Russell, N. J. C. (2011). Milgram’s obedience to authority experiments: Origins
and early evolution. British Journal of Social Psychology, 50(1), 140–162.
­doi:10.1348/014466610X492205.
Russell, N.  J. C. (2013). Gina Perry. Behind the shock machine: The untold
story of the notorious milgram psychology experiments. Journal of the History
of the Behavioral Sciences, 49(2), 221–223. doi:10.1002/jhbs.21599.
Russell, N.  J. C. (2014). The emergence of Milgram’s bureaucratic machine.
Journal of Social Issues, 70(3), 409–423. doi:10.1111/josi.12068.
70  Identity, Violence and Power

Sidanius, J., & Pratto, F. (1999). Social dominance: An intergroup theory of social
hierarchy and oppression. Cambridge: Cambridge University Press.
Sofsky, W. ([1993]1997). The order of terror: The concentration camp (trans:
William Templer). Princeton, NJ: Princeton University Press.
Stoessinger, J. (2007). Why nations go to war (10th ed.). Cengage Learning (first
published in 1978).
Strauss, S. (2006). The order of genocide. New York: Cornell University Press.
Tiedens, L. Z. (1997). Optimism and revolt of the oppressed: A comparison of
two polish Jewish Ghettos of World War II. Political Psychology, 18(1), 45–69.
doi:10.1111/0162-895X.00044.
Unger, M. (1986). The prisoner’s first encounter with Auschwitz. Holocaust and
Genocide Studies, 1(2), 279–295. doi:10.1093/hgs/1.2.279.
Vetlesen, A. J. (2005). Evil and human agency: Understanding collective evildoing.
Cambridge: Cambridge University Press.
Waldorf, L. (2007). Ordinariness and orders: Explaining popular participation
in the Rwandan genocide. Genocide Studies and Prevention: An International
Journal, 2(3), 267–270.
Zimbardo, P. (1989). Quiet rage: The Stanford prison study (video). Stanford, CA:
Stanford University.
Zimbardo, P. (2007). The Lucifer effect. New  York: Random House. Retrieved
from  http://www.goodreads.com/work/best_book/964959-the-lucifer-effect-
understanding-how-good-people-turn-­evil
Part II
Model
 3
Identity: The Group as a Collective
Performance

In the previous chapters, our argument was developed in two steps. First,
we discussed the challenges arising when public discourse or scientific
analysis of political violence takes an ethnic or ‘civilisational’ turn. We
added our voice to those who plead for vigilance such that we avoid tak-
ing descriptions of events in terms of (say) ‘ethnic conflict’, ‘ethnic vio-
lence’ or inferred ‘ethnic hatred’, as an explanation of these events. We
concluded that a non-circular analysis needs to take a step back and,
instead of taking ethnic or cultural categories for granted as the primary
units of analyses, we need first of all to explain how these categories came
into being. Or rather, we need to show what makes people accept such
categories as ‘real’ in the pragmatic sense that they provide a grid for inter-
preting social experience and for giving a direction to social behaviour.
In the second chapter, we then criticised a pervasive model of why
people accept and act on particular versions of category and category rela-
tions—that is, the notion that human beings are somehow programmed
to obey authority, no matter how cruel those authorities might be and
how brutal their instructions.
The main problem with both approaches is that they transform spe-
cific instances into general rules. Certainly, ethnic hostility occurs at

© The Author(s) 2017 73

G. Elcheroth, S. Reicher, Identity, Violence and Power,
DOI 10.1057/978-1-137-31728-5_3
74  Identity, Violence and Power

some times and in some places. But it doesn’t always happen and it takes
specific political forces and social processes to make it occur. Equally,
toxic obedience occurs, but it is far from ubiquitous and one can find
ample evidence of dissent events in the classic studies used to highlight
obedience phenomena.
In Chap. 2, we introduced the notion of epistemic isolation as a key
pre-requirement for toxic obedience: people conform unconditionally to
a particular authority when the understandings of that authority stand
uncontested and they are sealed off from any opportunity to act jointly on
the basis of alternative understandings. In this chapter, we will develop a
similar argument about hostility between groups: before a set of people will
act together as a coherent ‘us’ against a hostile ‘them’, the range of interpre-
tations of social reality must be radically curtailed so as to exclude anything
which questions who ‘we’ are, who ‘they’ are, and how the two interrelate.
In the next chapter, we will go on to discuss the specific role of vio-
lence in bringing about situations of extreme epistemic isolation where
alternatives to ingroup conformity and outgroup hostility are difficult
to conceive and impractical to act upon. For now, let us examine more
closely how epistemic isolation or else epistemic co-ordination affect the
terms in which conflicts are experienced, and why the balance between
the two is at the heart of the matter.

3.1 Defining Conflict, Defining Identity

So what, concretely, is involved in incidents of violence? What, in par-
ticular, has to happen for an incident to become critical and feed into a
spiral of violence—say ethnic violence?
First, both the victims and the wider community must be defined in
ethnic terms. Thus, an attack on a neighbour is less of an attack on an
individual than an attack on a ‘Serb’, a ‘Croat’, and so on. And, to the
extent that I also define myself as a ‘Serb’ or a ‘Croat’, the attack becomes
an attack on ‘me’. Indeed, once people are defined as interchangeable
members of a common category, it is a matter of mere chance and hence
of diagnostic irrelevance as to which individuals happened to be attacked.
It might just as well have been me.
 3  Identity: The Group as a Collective Performance  75

Second, the perpetrators also have to be defined in ethnic terms, and,

more specifically, as members of the ethnic outgroup. They cannot be
‘the poor’, ‘the desperate’ or ‘the criminal’. They must be the Serb to my
Croat, or the Croat to my Serb. This then means not just that we are
under attack, but that we are under attack by them. And just as every
member of the ingroup is potentially under attack, so every member of
the outgroup is potentially an attacker.
Third, it is necessary to attribute violence to an enduring hatred that
members of the ethnic outgroup have for the ethnic ingroup. If the vio-
lence were to be explained in terms of other causes—say people were
provoked by agitators or they were goaded by specific frustrations—then
one might be able to deal with the violence by removing these specific
causes. If the hatred were to be temporary, then one might not have to
fear for the future. But if the cause is ingrained hatred, and hence the
hatred is continuous, then they will always be a threat to us. Any of them
living amongst us are fifth columnists, and we can only be safe by getting
rid of them.
As we can see, then, the spiral of violence depends upon a complex
set of representations, explanations and attributions. Violence will only
breed violence if it is interpreted in a very particular way. So where do
these interpretations come from and how, at some point, do they trump
others? Those who subscribe to fatalistic models of conflict eschew these
questions. They assume that such interpretations just arise naturally (e.g.,
we have evolved to see things in ethnic terms, to be mistrustful of eth-
nic others and, presumably, to see them as mistrustful of us). This is
problematic because it describes the phenomena in ways that presuppose
precisely what is at issue. It involves imposing the ethnic lens of the ana-
lyst on the participant, and projects ethnic categories forged by violent
conflict into the past, thereby providing anachronistic explanations of the
onset of conflict and violence.
To avoid a similar pitfall, we proceed in a different way and rather than
portraying enduring ethnic conflict as something so familiar as to barely
require comment, we seek to render it as something strange and rare. We
see it as an anomaly to the normal pattern of social relations which are
characteristically fluid and, from moment to moment and site to site,
are based on many different social categories. We then discuss how this
76  Identity, Violence and Power

anomaly can arise—how, more precisely, epistemic isolation freezes the

definition of categories into a rigid (ethnic) pattern under specific and
extreme circumstances.
To fully grasp what is at stake, let us consider a related issue, which
fatalistic models of conflict and fatalistic models of obedience both
struggle to account for, that is, variability across space and time. If our
psychic apparatus makes us naturally defiant or even hostile towards eth-
nic outgroups, why are there then about as many types of reactions to
‘diversity’ as there are culturally diverse societies? How can the same ter-
ritory accommodate the highly integrated multi-cultural society that was
pre-war Bosnia-Herzegovina, and equally the ethnically genocidal society
of the war years? Similarly, if our psychic apparatus renders us chroni-
cally inclined to obey authority, how can we possibly account for the fact
that the seemingly quiescent societies of Eastern Europe and the Arab
world surprised everyone with the rise of mass revolutionary movements
in 1989–1990 and 2011–2012, respectively?
When so many people almost simultaneously engage in actions they
have never endorsed before, it seems obvious that these behavioural changes
cannot be caused by previous changes in everyone’s core personality—that
is, in the intimate beliefs and values forged by a singular life history. The
rapid diffusion of very specific social behaviours, which leads to riots, revo-
lutions or collective violence, must have something to do with changes
in the way people relate to one another, more than with changes in their
personalities. When certain behaviours spread like wildfire, where initially
moderate actions and modest demands give rise to an escalating series of
ever more radical actions and demands, where historic turning points are
reached, we have to concentrate on the mechanisms that enable or dis-
able a set of people from communicating, interacting and co-ordinating
their practices. The conceptual framework that we want to develop needs
to address these mechanisms. It should show both how a sense of shared
identity emerges through such mechanisms and how it facilitates them in
turn. Given that large-scale collective action presupposes that many indi-
viduals who do not know each other personally are still able to imagine
and recognise each other as members of one group, their capacity to reach
a shared interpretation of palpable markers of identity—defining who is ‘in’
and who is ‘out’—becomes critical in the process.
 3  Identity: The Group as a Collective Performance  77

A few years ago, together with Willem Doise, we tackled the theo-
retical problem of epistemic co-ordination, focusing on how separate
individual representations of the world become integrated into social
representations (Elcheroth, Doise & Reicher, 2011). We argued then,
and still argue today, that the general mechanisms governing how peo-
ple co-ordinate to understand and transform their social worlds can be
conceptualised adequately by articulating the two traditions of thought
that have developed under the banners of social representation theory,
originating in the works of Serge Moscovici (1961, 2008), and of social
identity theory, originating in the work of Henri Tajfel (1975, 1981)
and developed by John Turner, Hogg, Oakes, Reicher & Wetherell
(1987).
We identified four defining characteristics of social representations:
shared knowledge, meta-knowledge, enacted communication and world-­
making assumptions. In the following passages, we now propose a revised
version of the critical passages in the 2011 article, elaborated for present
purposes to make more explicit the indivisible relationship between social
representations and social identities. We show how these same four char-
acteristics relate to the way identities are made, un-made and re-made,
through co-ordinated knowledge, experience and behaviour. That is, we
explain identity processes by reference to social practices—practices that
speak to each other, become possible by communication and constitute
acts of communication themselves.

3.2 S
 hared Knowledge: Identity as Epistemic
Co-ordination
The first premise of our approach is that what shapes social behaviour
is shared social knowledge. This is true in two connected senses. On the
one hand, what counts is not our idiosyncratic experience but our knowl-
edge of things that are experienced at a collective level. There is evidence,
for instance, that belief and action are less a function of whether ‘I am
unemployed’ or else ‘I have suffered from discrimination’, but more of
whether ‘we suffer high levels of unemployment’ or ‘we are the subjects of
discrimination’ (Elcheroth, 2006; Kinder, 1998; Mutz, 1998).
78  Identity, Violence and Power

On the other hand, experience impacts our knowledge through the

way we make sense of it in terms of shared bodies of knowledge. These
exist not only in our own minds but are offered to us by peers, pundits
and politicians. Critically, they are also instantiated in material culture:
books, films, newspapers, statues, museums and so on. Moreover, there
is an ‘intertextuality’ to these representations such that, when new phe-
nomena come along, we achieve knowledge by anchoring them in already
existing stocks of knowledge (see Moscovici, 2008). For instance, when
Saddam Hussein first came to the world’s attention after the invasion
of Kuwait, he and his acts were interpreted by rooting them in widely
shared understandings of Hitler and Nazism. Indeed, on one notori-
ous Time magazine cover, Saddam’s moustache was manipulated to look
more like Hitler’s (see Herrera & Reicher, 1998).
Critically, the importance of shared knowledge is not just that it is
broader, but also that it is deeper. That is, understanding undergoes a
qualitative epistemic transformation by being shared. An individual view-
point is always contingent. If, say, I am refused a job, it might have been
because I performed badly or the individual interviewer took against me
personally. But if we are consistently denied jobs, then that constitutes
discrimination. In other words, opinion is transformed into social fact
through the accomplishment of common interpretations of shared expe-
riences. And, if opinions are an insecure basis for undertaking potentially
costly actions, social facts are a firm foundation from which one can act
in the world. Take, for instance, a study conducted by Wright (1997)
in which participants faced strong discrimination. When they faced the
situation alone, without any clue as to how others interpreted the situa-
tion, they reacted with resignation. However, as soon as there was a mini-
mal breach in their epistemic isolation—that is, if they heard what they
supposed to be a fellow participant expressing anger at what he named as
‘discrimination’—they began to mobilise and enlist others in a collective
boycott of the experiment. As the study demonstrates, a sense of mean-
ing, of justice and injustice, of mastery and agency, along with all their
consequences, derives from participation in collective meaning-making
practices.
This focus on knowledge as shared logically leads us to ask how it
comes to be shared. The obvious answer is through communication. But
 3  Identity: The Group as a Collective Performance  79

the critical qualification is that much of this communication is implicit.

We therefore need to look at what a set of people have, jointly, to take
for granted in order to be able to communicate at all. Often, to state one
thing explicitly, dozens of other things need to be implied at the same
time.
To take an example, all the more powerful for being banal, a discus-
sion of whether the weather will be good tomorrow presupposes, first,
an agreed definition of ‘good’ which privileges particular communities
and particular practices (good normally means calm, hot and sunny,
which is pleasant for the leisure of those who work mainly indoors but
may be less desirable for farmers needing rain for their crops or sail-
ors becalmed without wind). Second, it presupposes an agreed spatial
frame of reference—where do we want the weather to be good: in the
local town, the region or the country? (as Billig, 1995, points out, often
terms like ‘the weather’ assume a national frame of reference.) What is
true of ‘will the weather be good?’ would be all the more true were we to
discuss more explicitly social and political issues: ‘is the economy doing
well?’, ‘is the President a good leader?’ and so on.
For the most part, we don’t have the opportunity to discuss or even
to think consciously about all these assumptions and all their conse-
quences. The sum of these never-fully-spelt-out ideas continuously spins
an invisible web of meanings and associations, which shape what we do,
think and say. In this way, implicit communication creates the conditions
through which different practices of sharing knowledge become possible.
The invisible web allows us to interact seamlessly with others who
subscribe to the same assumptions, and it functions as the discrete con-
text within which all new claims will be interpreted. That also implies
that communication feels much more like hard work between individu-
als whose semantic associations overlap more weakly. As de Toqueville
described in his Democracy in America, this is one of the burdens of being
a stranger in a new country. As a corollary, we are often able to intuit
similarities or differences of identity through the daily experience of ease
or effort in communicating with others.
Hence, people’s ability to articulate their own understanding with that
of others will be constrained by the ways they are able to relate to oth-
ers in the world, and by the unspoken assumptions they share or don’t
80  Identity, Violence and Power

share with others. This is all the more important as people need to make
their social environments intelligible if actions are to be organised into
meaningful sequences. Where the world cannot be rendered intelligible,
the psychological consequences are generally severe: stress, strain, a sense
of estrangement, a feeling of helplessness.
But even with these various practical, discursive and ontological limits
to the types of representations we produce together, there is still consid-
erable room for manoeuvre. The primary task is then to investigate the
complex and slippery processes by which people jointly produce specific
meanings.

3.3 Meta-Representations: Identity

as Collective Awareness
To grasp these processes, we first need to clarify the statement that people
act on the basis of shared knowledge. As this statement stands, it could
be misinterpreted. Take, for instance, Wright’s study on discrimination
that we invoked earlier. In the ‘isolation’ condition, it may be that every
single participant believes that they are experiencing discrimination.
In this sense, they already share the same knowledge. However, as we
saw, this was not enough to invoke resistance. What is important, as
shown in the ‘communication’ condition, is that they become aware of
the thoughts that they share with others. It is only in this condition that
people gain the certainty and the confidence to resist. The message, then,
is that what counts in social representations is not only what we think,
but what we think that other people think: what we will refer to as their
meta-knowledge.
But as we have already suggested, not all such meta-knowledge is
equivalent. We respond very differently to what different others think
and do. If, for instance, I am on the left and I see people at a political
rally laugh and applaud, I am likely to react very differently as a function
of whether I have categorised them as on the left or on the right. In both
cases, what they do is likely to influence how I interpret the message to
which they are responding. But in the former case (where they are part of
 3  Identity: The Group as a Collective Performance  81

my political ingroup) it is likely to make me more supportive, whereas in

the latter case (where they are outgroup) it is likely to make me more crit-
ical. It is here that the representation of social categories becomes critical
in organising whether we embrace the beliefs of others or eschew them
(cf. Platow, et al. 2005).
The psychological underpinning of identity therefore resides in the
human capacity to simultaneously process information from two per-
spectives. On the one hand, we analyse information from an egocentric
perspective—in terms of what it means to us and for us. On the other,
we analyse it from an allocentric perspective—in terms of what it might
mean for relevant others, for our relation to them, and for how they
might react to our own different reactions to that information. In other
words, we are inherently reflexive beings who operate on our knowledge
of our own minds, our knowledge of other minds, and even our knowl-
edge of other mind’s knowledge of our minds.
This approach also suggests that those who wish to change how people
interpret the world (i.e., who wish to achieve influence), might best do
so by working on the assumptions people hold about the interpretations
that are shared by others. After all, on the whole we have greater epistemic
certainty about what we think than about what others think. Hence, it is
generally easier to shift the latter than the former.
The power of the mass media is particularly pertinent here. Their influ-
ence derives, not least, from the fact that as people surf the net, listen to
the radio, read the paper or watch TV, they are aware that many others
are accessing the same website, listening to the same radio programme,
reading the same paper or watching the same television programme as
them. So, even if every single consumer remains sceptical about the infor-
mation they are exposed to (and will readily express such scepticism when
asked), they can still be influenced by virtue of their inferences about
the impact of the media on others. This in turn can lead individuals
to ­incorporate media messages into their own personal communication
strategies, thus resulting in a multiplicity of interpersonal conversations
which sustain the impression that the message is relevant and hence rein-
force its impact. Conversely, the impression that certain interpretations
of the world are not shared can generate a self-fulfilling prophecy by gen-
erating multiple acts of individual self-censorship.
82  Identity, Violence and Power

This long-standing intuition is supported by Paluck’s (2009) work in

post-genocidal Rwanda. She examined the impact of a radio soap opera
specially designed to promote reconciliation on relations between ethnic
groups. The results were highly encouraging. Those who listened together
in small local communities did indeed bond more with others across eth-
nic boundaries and did speak out against sectarianism. But when it came
to just how these beneficial effects were produced, the findings were far
more nuanced. Contrary to the expectations of those who created the
soap (Staub & Pearlman, 2009), listening had no effect on people’s own
opinions and beliefs about the conflict. What did change, however, was
what they thought others exposed to the programme now believed—the
expectation was that there would be a shift towards greater support for
reconciliation. And, as we are suggesting as a more general mechanism,
it was beliefs about the beliefs of others in their communities which pro-
duced the changes in their own actions.

3.4 E
 nacted Communication: Identity as Joint
Performance
While Paluck’s study was about the impact of discourse diffused through
the mass media on consciously shared representations, words are not
always required to change beliefs about how relevant others experience
the world, and position themselves in the world. There are indeed a num-
ber of illuminating examples of how we will misunderstand the nature
of representations if we look at what people say to the exclusion of what
they do. A case in point is Jodelet’s (1991) classic work on representations
of madness. As she showed, people may not say that they think mental
illness is contagious, but the ways in which they separate their own crock-
ery from that of sufferers suggests otherwise.
But it is not just that practices are important in terms of communicat-
ing the perspective of others; it is arguable that they are more powerful
than explicit discourses to the extent that they are more silent. To hear
someone state things overtly (‘this is what people think’) always opens up
at least the possibility of disagreement (‘oh no they don’t). To see some-
 3  Identity: The Group as a Collective Performance  83

one act requires us to infer their position without inviting a challenge.

It draws on implicit assumptions which we have already seen to be so
important to social representations and social identities.
Take, for instance, Falasca-Zamponi’s (1997) analysis of how the fas-
cist salute operated in Mussolini’s Italy. Her point is precisely that the
impact of this practice did not primarily occur through the act of chang-
ing individual beliefs. Anyone who gave the salute could retain a sense
that he or she was doing it reluctantly, pragmatically, without being a
‘true believer’. However, each person, seeing everyone else give the salute,
could not take the risk of believing (or acting on the belief ) that they
were insincere. They had to infer belief from the silent act. The salute was
therefore a particularly powerful means of changing perceptions of shared
beliefs; it created the illusion of a consensus and it thereby discouraged
dissent.
The relevance and importance of practices goes further. We previously
mentioned that the creation of shared knowledge operates through the
dynamics of intertextuality—prior representations supporting subse-
quent representations. However, invoking prior understanding is some-
thing active. It is a matter of practices of remembering, of celebrating
and of commemorating together in ways which enact particular under-
standings of the world. National days, anniversaries of famous victories,
birthdays of past and present leaders are all ways of bringing us together
and getting us to act together on the basis of selective beliefs about who
we are.
But, powerful as such spectacular practices might be, they tend to be
more powerful to the extent that they are more mundane and inscribed
into the textures of everyday life. It might be a matter of different treat-
ment of different groups at airport security checkpoints, the maintenance
of traditional practices or else the establishment of different schools for
different ethnic and religious minorities; it might be checks on certain
groups to ensure that they are entitled to social benefits, to the use of
public services and so on. Such practices don’t only point to these groups
as ‘other’, but also to the nature of their otherness—‘they’ are dangerous,
culturally alien, un-trustworthy and so on. That is, they tend to trans-
form people into threatening others.
84  Identity, Violence and Power

All in all, our understanding of social reality, of other people and of the
ways they relate to us, is not only constructed through social communi-
cation (at both interpersonal and mass media levels) but equally derives
from the accumulation of concrete experiences that fill an ordinary life.
These experiences provide us with a sense of interdependence with other
people. All of us have concretely experienced the fact that others can con-
firm or challenge our viewpoint, can support or impede us, can sustain or
harm us. Correspondingly, we have all developed interpretative strategies
for distinguishing between who might do the one and who might do the
other. Obviously, those interpretative activities draw upon shared narra-
tives that render our idiosyncratic experiences meaningful. But, equally,
these narratives are rendered relevant and plausible through recurrent
patterns in the concrete organisation of social interactions. That is, the
social narratives which are offered to people as frames of interpretation
need to make sense of mundane experiences. They need to help us to act
appropriately in various social situations. In Gramsci’s formulation, they
need to have practical adequacy (see Sayer, 1979).
The key point here is that narratives and practices are not in opposi-
tion. Each is powerful to the extent that it is complemented by the other:
what is said highlights what is done, and what is done makes sense in
terms of what is said. And what binds narratives and practices together
are social institutions. These both tell stories about how the world needs
to be organised and also organise the social world. They structure social
interactions in particular ways and they create regularities in collective
experience which leads people to gain a common feeling that particular
forms of social interdependency are authentic realities. They thereby give
credibility to accounts of social relations which presuppose such forms of
interdependency. In other words, institutionalised social structures allow
narratives about collective identities to sound plausible and become rel-
evant in the light of concretely experienced patterns of interdependence.
To be somewhat more concrete and to continue with the themes intro-
duced above, the Nazi ideology of Adolf Hitler as the supreme leader
of a homogenous ethnic German nation from which Jews (among oth-
ers) were totally excluded was both the subject of a relentless ideological
assault (Kershaw, 1987) and also inscribed in a series of institutionalised
practices, from the mundane realities of the ‘Heil Hitler’ salute (which
 3  Identity: The Group as a Collective Performance  85

all ‘ethnic’ Germans were required to give, but which Jewish people were
explicitly prohibited from giving—see Allert, 2009) to the exclusion of
Jewish people from trades and professions, and to the expulsion of Jewish
people from their homes and homelands—ultimately to the death camps.
There is one final point we wish to make about narratives, practices
and the nature of/relations between social groups. That is, these exist in a
dynamic and developing relationship which depends not only upon dif-
ferences of understanding but also upon commonalities of understanding
between these groups. Take, for instance, the Hindu who deliberately
provokes a riot by throwing a dead pig on the steps of a mosque in India.
This act is performed on the basis of understanding the narrative lens
through which it will be viewed by the Muslim community (a funda-
mental act of desecration), the consequent practices to which it will give
rise in this community (collective anger and violence), and how this in
turn will be viewed by the Hindu community (the barbarity and threat
of the Muslim ‘other’) and feed into their practices (retaliatory violence
and communal retrenchment).
On the one hand, then, we see how practices feed into narratives
which generate new practices, which in turn affect narratives … and so
on. On the other hand, we see (ironically) that the ability to create vio-
lence in this way occurs not despite but because of a shared heritage
and accurate presumptions about interpretative activities across religious
groups. In this sense, common understandings and collective awareness
are organised as sets of dialogues enveloped in practices (see Marková,
2003; Gillespie, Cornish, Aveling and Zittoun, 2008), where opposite
positions and antithetical ‘themes’ are enacted all the more effectively
when both sides are able to understand the core of both lines of argument.

3.5 W
 orld-Making Assumptions: Identity
as Collective Agency
Common understandings of the world and collective awareness hence
not only arise from social practices, they also often make possible those
social practices (and only those social practices) that then sustain them.
Effective nationalism creates the national categories that it assumes
86  Identity, Violence and Power

(see Reicher & Hopkins, 2001). Seeing someone as an enemy can lead
us to treat them in ways that make them behave as an enemy. Therefore,
we need to be precise about the meaning of ‘social context’. Any read-
ing of the problem solely in terms of the question ‘how do people give
meaning to what is already out there?’ misses the most interesting point.
Social and historical contexts are not just sets of external background
factors that impact shared representations, but are themselves realities
brought into existence through such representations.
This ontological stance is easily misunderstood or else misrepresented.
Hence, we must be clear about what is meant by this and what it implies.
The starting point is to appreciate that the day-to-day reality in which we
live is largely constituted by what Searle (1995) labelled institutional facts.
All aspects of our everyday lives—from the time we get up in order to get
to work, the traffic regulations which govern our drive to the office, the
rules which govern what we do once there, to the value of the money that
we earn—are part of a human-made world. Such institutional facts can
be defined by two properties. On the one hand, they exist only as a con-
sequence of human agreement; on the other, from an individual perspec-
tive, they appear as objective facts: (at least part of ) their consequences
are independent of subjective cognition.
Things like money, citizenship, degrees, classes, mortgages and crimes
would not exist if no one believed that they existed. Or, to be more pre-
cise, they would not exist if there were no storekeepers, border guards,
students, bankers or police officers acting on the basis of the belief that they
do exist. But then, it becomes important to make a key distinction. What
we are not implying is that, were an individual to deny these institutions,
they would go away. Changes in individual representations do not alter
the existence or essence of specific institutional facts. Were you to drive
on the wrong side of the road, try to use conch shells as currency or else
claim to live in your own independent republic with its own laws, you
would soon discover this. We are, therefore, not proposing an extreme
form of philosophical solipsism which is easily caricatured. What we do
assert, however, is that changes in shared representations can and fre-
quently do lead to changes in the institutional world.
The relationship between shared representations and institutional real-
ities is therefore bi-directional. On the one hand, formalised ­regulations,
 3  Identity: The Group as a Collective Performance  87

especially when they are perceived as legitimate by a critical mass of peo-

ple, are generally powerful tools for clarifying and thus stabilising mutual
expectations. Moreover, fostering the creative nature of normative expec-
tations by formalising them is not limited to the scope of classical legisla-
tors operating within classical nation-states. As Doise (2002) claimed, the
international diffusion of human rights has genuinely transformed social
relations across manifold spheres of lives and places (see also Elcheroth,
2006; Gély & Sanchez-Mazas, 2006; Spini & Doise, 2005). On the other
hand, giving institutional support to a particular position by making it
a law does not inoculate that position against change. One pertinent
example, in some countries and states at least, concerns the privileging of
heterosexual marriage through various forms of legislation. As alternative
forms of relationship have gained greater legitimacy, the law has fallen
out of step. In the end, the law has had to change in order to accommo-
date actual social practices.
Our argument so far, then, is that social realities are created, maintained
and transformed by collective practices that generate and uphold shared
systems of meaning and shared expectations. There is a corollary to this:
although individual dissent is insufficient to change institutional realities,
once an individual’s discontent is articulated with his or her belief that it
is shared with others, action for social change becomes a viable option.
Similarly, once a person’s unease with the way in which a valued group
identity is defined becomes articulated with his or her belief that others
share this unease, a re-definition of the identity becomes possible.
Such mobilising beliefs draw on the invisible web of meanings we
referred to earlier on, composed, for example, of common sense truisms,
widely diffused political slogans or basic legal prescriptions. The aware-
ness that a significant proportion of community members have access to
this background knowledge can create a dynamic of escalation whereby
individuals presuppose the support of others and, hence, act in ways
that elicit support from others—such as when one or two members of
an audience start clapping after a speech, driven by the confidence that
others will follow, and then others infer that clapping is an appropriate
response, and, soon, individual acts are transformed into the collective
practice of ovation. Or else, and perhaps more consequentially, when
a few people turn up to a protest, believing that they will not be alone,
88  Identity, Violence and Power

which leads yet others to turn up to subsequent events and culminates

in mass demonstrations with genuinely revolutionary potential. Such a
process was critical to the Arab Spring (Ghonim, 2012).
To take our argument a little further, we can see that the effectiveness
of an individual actor in shaping the world lies not so much in his or
her own actions per se, but more in the way in which he or she is able,
through those actions, to shape the expectations and, hence, the actions
of others. Understood in this way, agency comes down to the capacity to
shape mutual expectations within a group, in such a manner as to enable or
impede co-ordinated actions directed towards a given purpose.

3.6 Demystifying Identity

The theoretical approach to identity that we have introduced in this chap-
ter changes our way of looking at its role in producing ‘hatred’ as well as
‘conformity’. Let us now try to summarise the insights gained, proceed-
ing in two steps: first, by showing how some of the most powerful and
intriguing aspects of collective identity become conceivable with no need
to assume immutable binding forces related to common descent, destiny
or other essential factors; and second, by inferring why we need to look
more closely at the mobilisation of collective identity to advance in our
understanding of both intergroup conflict and ingroup conformity.
To put it briefly, many approaches to identity lay their emphasis on
an examination of people’s intimate sense of self and other. We wish to
complement that with an equal emphasis on an examination of visible,
sometimes ostentatious, expressions of identity. We focus on the way in
which certain markers of identity become shared, highlighted as impor-
tant, and hence become pragmatically relevant.
In practice, the question of whether individuals consider their own or
another person’s language, accent, religion, skin colour or place of birth as
a relevant piece of information or not will often depend on whether they
anticipate that such markers of identity will make a difference in the way
other people will act towards them or a third person. In Germany under
the Nuremberg race laws, as in South Africa under the Apartheid regime,
the question whether anyone truly believed that ‘race’ matters—that
 3  Identity: The Group as a Collective Performance  89

Jewish descent or skin colour was an appropriate criterion to judge

a person—was of limited pragmatic value. By contrast, what mat-
tered was the knowledge that these markers were publically defined as
foundational, and hence anyone who ignored them when engaging in
romance, friendship or work relationships would be in deep trouble.
So too in less extreme contexts, identity markers matter because of what
we understand others to make of them. Employers treat class and racial
markers seriously because of the way they believe customers will react.
Adults might worry about the political identity of their acquaintances
for fear of alienating their friends, while adolescents might be more
concerned by cultural markers (what music someone likes, for instance)
for similar reasons.
To be clear, our argument here is not to deny the importance of our
intimate sense of self—we know that to be important and that both our
mundane sense of self and our passionate sense of self drive many key
behaviours. What we are denying is the idea that identity is solely or even
primarily about such a sense of self. Only when we complement this with
the insight that identity is equally to be understood as a set of concrete
markers of identity that are perceived as being shareable in a given social
context can a number of seemingly mysterious attributes of identity
become intelligible. And the first of these is precisely why people do invest
so much in their identities—why we spend so much energy in categorising
ourselves and others (even in the cultural context of Western-capitalist-
post-modernist societies that so highly valorise individual distinctive-
ness); why we are so attached to our identities and why we agonise so
much about the loss of identity.
The explanation is that identity connects us to other people—some-
thing that can only happen because of public agreement about the mark-
ers of identity such that I can be confident that the way I define myself
will be accepted by others and will be the way they define me. This con-
nection is not only something sentimental or perceptual but also highly
concrete and pragmatic. A common understanding of identity leads peo-
ple to assume a common understanding of the world and a common set
of goals. It thereby allows people to co-ordinate their behaviour, to take
joint action, and to achieve such social power as makes them more able
to actually achieve their goals.
90  Identity, Violence and Power

The second question which can be better addressed by combining the

intimate and public dimensions of identity is why policymakers talk so
much about identity. In contrast to a common view in political science,
we propose that ‘identity’ in political rhetoric is more than an attempt
to circumvent more weighty political issues like economic management,
social policy or foreign affairs. The capacity of policymakers (or those
who challenge them) to create results in any or all of these fields is criti-
cally contingent upon the enlistment of public support. In a democracy
(and even in autocracies), you can’t mobilise troops to go to war unless
you can also mobilise the population in favour of war. Now, insofar as
the creation of shared identities brings masses of people together with a
shared understanding of the world and a shared understanding of what is
in their interests—in other words, identities create constituencies—then
successful mobilisation will be dependent upon the ability to define iden-
tities in ways that make the politician’s project mesh with the population’s
interest. That is, the effectiveness of a politician in any area—economic,
social, military, cultural—will depend upon their capacity to act as an
authoritative “entrepreneur of identity” (Reicher & Hopkins, 2001).
It is all very well to explain why politicians will talk about identity and
will seek to entrench particular definitions of identity. But, of course,
this would be of little consequence if these definitions were without con-
sequence. So our second question leads irresistibly to a third—one to
which we devote most attention in this book: Why is identity so vulnerable
to manipulation? If it were true that an identity (say national identity)
were entirely an internal and intimate matter—the sum of millions of
citizens’ answer to a lifelong quest to discover who they truly are—then
it is most unlikely that something as ephemeral as a few mass-mediated
speeches could have any impact on it at all. How could a few words
re-define the content or meaning of that identity? How could passing
rhetoric trigger shifts in public opinion leading a national majority to
see their nationhood as impelling them towards building a welfare state
or else cutting social benefits, embracing those displaced by war or else
policing migrants, going to war or else suing for peace? Once again, mal-
leability and change is the best possible argument against taking identity
as an essence within the person. It is the best argument possible for relat-
ing what we think and feel on the inside to what happens between us on
 3  Identity: The Group as a Collective Performance  91

the outside. It is at this latter level that political rhetoric largely exerts its
effect, altering what we believe about what others believe and specifically
believe about us; consequently, altering how we act towards those others
and hence how they act towards us; ultimately re-framing the organisa-
tion of the social world and hence the social categories which are able to
make sense of it.

3.7 A Mobilisation Perspective

Underpinning our arguments about identity is an assumption about the
nature of understanding which is at odds with much (but certainly not
all) of psychological theorising. That is, we eschew a perceptual approach
which assumes that people come to comprehend their world through a
silent neutral process of contemplation: we individually look at what is
out there, we process the information through a combination of generic
processes and personal biases, and we thereby generate a picture of the
way things are. Rather, when it comes to any issue of significance—how
do we deal with immigrants, what do we think of people of a different
religion or ‘race’ or ethnicity, what do we think of those who are job-
less and on welfare—we are subject to a cacophony of different voices
providing us with different definitions, explanations and solutions. We
are enjoined to see the world in different ways. Our understandings are
actively mobilised.
Drawing on our analysis of identity, there are multiple reasons why it
is important to look at the ways that identities are mobilised, crafted and
used. The first is that the definition of identities is not the background
to intergroup conflicts but always part of these conflicts. Different parties
to the conflict will advance different definitions. Much of the struggle is
actually over the representation of the conflict itself—what categories are
involved (who are ‘we’ and who are ‘they’?), what is the category content
(what does it mean to be ‘us’?), what are the category boundaries (who
belongs to ‘us’ and who to ‘them’?), what are the category relations (what
do ‘they’ portend for ‘us’?). It is also about the way these definitions have
moral consequences (what sort of actions are right or wrong, and what is
the legitimacy of different actors?) and practical consequences (what sorts
92  Identity, Violence and Power

of courses of action are possible?). Once these representations are deter-

mined, most of the work is done, and many other things follow.
Take, for instance, the Arab–Israeli conflict—or is it the Palestinian–
Israeli conflict, or even a conflict between Palestinians and Western
imperialism? This problem of description is emblematic of the issues.
The Israeli state tells a story in which they are a small, vulnerable and
historically oppressed people threatened with annihilation by the might
of the combined Arab countries. In this context, the moral responsibility
of outsiders in the international community is to protect them against
the shadow of another Holocaust (see Gamson & Herzog, 1999). The
Palestinian leaders tell a story where the small, vulnerable and histori-
cally oppressed Palestinian people are threatened with annihilation by the
might of the Israeli Defence Force, which is backed by the even greater
imperial might of the United States. So here, the moral responsibility of
the rest of the world stands against Israel. How much credit is given to
each of these narratives by various actors, both locally and in the global
arena, is highly consequential for the dynamics of the conflict, the bal-
ance of forces and the way things are likely to play out.
The next reason for looking at mobilisation is that it is critical to under-
standing the relationship between identity definition and context. On
the one hand, context shapes rhetoric. More specifically, the ways that
categories are defined depends upon the nature of the audience one seeks
to mobilise. Thus, Klein & Licata (2003) show that the nationalist leader
Patrice Lumumba altered his representation of Belgians and Congolese as
a function of whether he was addressing the former or the latter. In front
of the colonialists, and in order to demobilise their ­opposition to inde-
pendence, he would describe Belgians as benevolent and the Congolese
as pacific. In front of the colonised, and in order to mobilise them for
independence, he would describe the Belgians as oppressors and the
Congolese as victims.
On the other hand, rhetoric shapes context. More specifically, context
is invoked in order to sustain particular versions of category relations.
So, as Stevenson, Condor and Abell (2007) showed, the answer to the
question of whether Irish Catholics or Irish Protestants are the minority
group depends on the framing of the relevant context as either ‘Northern
Ireland’, ‘Ireland’ or the ‘United Kingdom’. Different political leaders
invoke these different contexts in order to render plausible their own
 3  Identity: The Group as a Collective Performance  93

version of who is the minority and who the majority. Sometimes even
the same leaders refer to different contexts at different moments, when
political opportunities and strategies change.
Once again, we can see how these matters are the very stuff of the rel-
evant conflicts. They should therefore be the focus of analysis, not only its
backdrop. Furthermore, while sometimes the goal of expressing a given
version of identity is to mobilise the collective toward a form of action, at
other times it can be ‘just’ to consolidate the identity. That is, rather than
being aimed at promoting a specific and immediate instrumental pur-
pose, identity is invoked to achieve, maintain or deepen a shared under-
standing that the group exists and that it has a specific culture (see Klein,
Spears and Reicher, 2007).
Finally, we need to look at the mobilisation of identity in order to
appreciate the critical role of mass communication, and hence of mass
media, in shaping collectivities and collective action. In political affairs in
general, and in large-scale conflicts in particular, the collective experience
of events is necessarily mediated, since no one can have a complete pic-
ture of the conflict by drawing only on his or her immediate perceptions,
or even those of his or her personal contacts. The media circulate explicit
narratives and discourse about conflicts, but also images and perspec-
tives that sustain (or contradict) these narratives. For example, Lipson
(2009) provides a systematic analysis of camera shots broadcast by the
BBC and by CBS during one week in the early stage of the invasion of
Iraq in 2003. He shows how ‘embedded journalism’ meant that British
and American troops were pictured up close. We see their facial features
and their emotions, making it all the easier to identify with them, with
their hopes and their fears. By contrast, shots of Iraqis tended to be rare
and at a distance. Only 20 % of pictures coming from Iraq displayed
locals and even then they were generally only of people in the mass, of
crowds, of soldiers running and shouting.
At this point, the critic might respond that people are media-savvy.
They are well aware of these devices and biases and are not swayed by
them. But that is of little comfort if the media work by affecting what we
think others think and, hence, what positions can reasonably be expected
to be shared. The fact that mass communication does not so much affect
what each of us feels and thinks in private does not imply that its impact
on what we are capable of doing collectively is not critical.
94  Identity, Violence and Power

Therefore, if you want to escape from the influence of the mass media,
it is not enough to switch your TV off. You also need to let your neigh-
bours know that you are doing so. This is precisely what inhabitants of
the small Polish city of Swidnik did in February 1982 (Crawshaw &
Jackson, 2010), during the early years of the ‘Solidarnosc’ protest move-
ment. Exhausted by the pro-regime news coverage of the state-controlled
television, which either entirely ignored or unilaterally delegitimised the
protest movement, an increasing number of Poles decided not to watch
the daily news broadcasts any more. In Swidnik though, people started
to realise that their private boycott would have much more impact if they
found a way to express it publicly. At that point, an increasing number of
residents started to go for a walk at exactly the time when the news was
transmitted. Some went so far as to take their televisions with them on
a pushchair or other improvised vehicle. This made it very clear to any
observer that ‘I am going for a walk at 19:30’ actually meant ‘I am not
watching the news’. The movement soon spread to other cities and the
regime eventually became so nervous about it as to impose a daily curfew
from seven o’clock, thereby obliging people to stay at home during the
news. At this point, the ruling elite was obviously not in a position to
make people watch the state-controlled news, let alone to make them
trust the news, but at least they could make it as difficult as possible for
each individual to be confident that other individuals had also switched
their TV off.

3.8 Conclusion
Our analysis of identity in this chapter has four key elements.
First, we argue against the notion that identities are fixed, set, immuta-
ble either for all time or at any particular point in time. Across time, iden-
tities are always fluid and always contingent on what they enable a set of
people to do together. As social practices change, old categories become
obsolete and new categories are formed. If, for a period of time, identities
do become frozen in a particular configuration, that is the exception to
the rule. That is what requires explanation. In short, identities are always
performed and our task is to elucidate the processes of their performance.
 3  Identity: The Group as a Collective Performance  95

Second, when it comes to this task, it is important to appreciate that

our sense of self is not down to us alone. Who we are is also a matter of
how we think others will see and treat us, how we act in consequence
and how in turn that impacts the perceptions of others. It follows that
performing identity in new ways is as much, if not more, a function of
change in people’s awareness of how others see them as of change in how
they see themselves.
Third, important as it is to investigate how identities come into being,
it is equally important to investigate what particular identities make pos-
sible or impossible. We have seen how identities don’t just reflect the ways
that people are organised in the world but that they also organise people
in the social world. Identities mobilise people into collective actors. They
thereby create the social power to change the world: identities are world-
making resources (Reicher & Hopkins, 2001).
Fourth, precisely because identities do produce social power, those
who wish to wield such power (politicians, leaders and other activists)
will actively seek to construct versions of identity that sustain their practi-
cal projects. On the one hand, then, identities will always be contested.
On the other, identities will always be actively mobilised. Understanding
identity is therefore a matter of asking how particular versions of identity
gain traction and ultimately come to be seen as self-evident.
In the next chapter, we examine the role of violence in this process.
We examine how the threat, the actuality and the memory of violence
serve to reconfigure our relations to others, what we can know of others
(and what they know of us), and thereby serve to reconfigure identity and
power in society.

References
Allert, T. (2009). The Hitler salute. New York: Metropolitan Books.
Billig, M. (1995). Banal nationalism. London: Sage Publications Ltd.
Crawshaw, S., & Jackson, J. (2010). Small acts of resistance: How courage, tenac-
ity, and ingenuity can change the world. New York: Union Square Press.
Doise, W. (2002). Les représentations sociales : leçons du passé et défis
d’aujourd’hui : Social representations (Vol. 41, pp. 101–110). Presented at the
Conférence internationale sur les Représentations Sociales, Sage Publications.
96  Identity, Violence and Power

Elcheroth, G. (2006). Individual-level and community-level effects of war

trauma on social representations related to humanitarian law. European
Journal of Social Psychology, 36(6), 907–930.
Elcheroth, G., Doise, W., & Reicher, S. (2011). On the knowledge of politics
and the politics of knowledge: How a social representations approach help us
rethink the subject of political psychology. Political Psychology, 32(5),
729–758.
Falasca-Zamponi, S. (1997). Fascist spectacle: The aesthetics of power in Mussolini’s
Italy. Berkeley: University of California Press.
Gamson, W. A., & Herzog, H. (1999). Living with Contradictions: The taken-
for-­granted in Israeli political discourse. Political Psychology, 20(2), 247–266.
Gély, R., & Sanchez-Mazas, M. (2006). The philosophical implications of
research on the social representations of human rights. Social Science
Information, 45(3), 387–410.
Ghonim, W. (2012). Revolution 2.0. London: Fourth Estate.
Gillespie, A., Cornish, F., Aveling, E. L., & Zittoun, T. (2008). Conflicting
community commitments: A dialogical analysis of a British woman’s World
War II diaries. Journal of Community Psychology, 36(1), 35–52.
Herrera, M., & Reicher, S. (1998). Making sides and taking sides: An analysis
of salient images and category constructions for pro- and anti-Gulf War
respondents. European Journal of Social Psychology, 28(6), 981–993.
Jodelet, D. (1991). Madness and social representations: Living with the mad in one
french community. Berkeley: University of California Press.
Kershaw, I. (1987). The ‘Hitler Myth’: Image and reality in the Third Reich.
Oxford: Oxford University Press.
Kinder, D. R. (1998). Opinion and action in the realm of politics. In D. T.
Gilbert, S. T. Fiske, & G. Lindzey (Eds.), The handbook of social psychology,
Vols. 1 and 2 (4th ed., pp. 778–867). New York: McGraw-Hill.
Klein, O., & Licata, L. (2003). When group representations serve social change:
The speeches of Patrice Lumumba during the Congolese decolonization.
British Journal of Social Psychology, 42(4), 571–593.
Klein, O., Spears, R., & Reicher, S. (2007). Social identity performance:
Extending the strategic side of SIDE. Personality and Social Psychology Review,
11(1), 28–45.
Lipson, M. (2009). “If it wasn’t rolling, it never happened”: The role or visual ele-
ments in television news. In L. Harman & L. Lombardo (Eds.), Evaluation and
stance in war news: A linguistic analysis of American, British and Italian television
news reporting of the 2003 Iraqi War (pp. 140–169). London: Continuum.
 3  Identity: The Group as a Collective Performance  97

Marková, I. (2003). Dialogicality and social representations: The dynamics of mind.

Cambridge: Cambridge University Press.
Moscovici, S. (1961/2008). Psychoanalysis: Its image and its public. Cambridge:
Polity Press.
Mutz, D. C. (1998). Impersonal influence: How perceptions of mass collectives
affect political attitudes. Cambridge: Cambridge University Press.
Paluck, E. L. (2009). Reducing intergroup prejudice and conflict using the
media: A field experiment in Rwanda. Journal of Personality and Social
Psychology, 96(3), 574–587.
Platow, M. J., Haslam, S. A., Both, A., Chew, I., Cuddon, M., Goharpey, N., et al.
(2005). ‘It’s not funny if they’re laughing’: Self-categorization, social influ-
ence, and responses to canned laughter. Journal of Experimental Social
Psychology, 41, 542–550.
Reicher, S., & Hopkins, N. (2001). Psychology and the end of history: A cri-
tique and a proposal for the psychology of social categorization. Political
Psychology, 22(2), 383–407.
Sayer, D. (1979). Marx’s method. Brighton: Harvester Press.
Searle, J. R. (1995). The construction of social reality. New York: Simon and
Schuster.
Spini, D., & Doise, W. (2005). Universal rights and duties as normative social
representations. In J. Finkel & F. Moghaddam (Eds.), The psychology of human
rights and duties: Empirical contributions and normative commentaries
(pp. 21–48). Washington, DC: American Psychological Association.
Staub, E., & Pearlman, L. A. (2009). Reducing intergroup prejudice and con-
flict: A commentary. Journal of Personality and Social Psychology, 96(3),
588–593.
Stevenson, C., Condor, S., & Abell, J. (2007). The minority-majority conun-
drum in Northern Ireland: An orange order perspective. Political Psychology,
28(1), 105–125.
Tajfel, H. (1975). The exit of social mobility and the voice of social change :
Notes on the social psychology of intergroup relations. Social Science
Information, 14(2), 101–118.
Tajfel, H. (1981). Human groups and social categories: Studies in social psychology.
Cambridge: Cambridge University Press.
Wright, S. (1997). Ambiguity, social influence, and collective action: Generating
collective protest in response to tokenism. Personality and Social Psychology
Bulletin, 23(12), 1277–1290.
 4
Violence: How Collective Shocks
Transform Social Practices

We now come to the heart of our concerns: the matter of violence. Where
prior research has tended to focus on the question ‘how is violence pro-
duced’ we argue that this needs to be complemented by asking ‘what is
produced by violence?’ Correspondingly, the core question in this chap-
ter is how violence serves to transform identity.
The performative model of collective identity, outlined in the previ-
ous chapter, proposes that a set of people will only perceive themselves
as being bound together by a common identity if they can concretely act
together in the terms defined by that identity. That is, identity is funda-
mentally about doing, not just about thinking. It follows that anything
which alters what people can do together will likewise alter their sense of
identity. Violence is just such a thing—in fact, it is a dreadfully effective
way of reshaping shared action.
To be slightly more formal, violence affects identity to the extent that
it re-patterns the social practices through which a group of people per-
form and uphold their common identities. In part, such transformation
might be achieved through the creation of new practices, or else by mak-
ing previously rare practices more common. For instance, Angus Calder
(1992) shows how, due to processes like the evacuation of children and

© The Author(s) 2017 99

G. Elcheroth, S. Reicher, Identity, Violence and Power,
DOI 10.1057/978-1-137-31728-5_4
100  Identity, Violence and Power

the use of public shelters during the Blitz, Britons in World War II were
able to come together across previously impenetrable boundaries of class.
Nonetheless, above all, violence operates in a negative way. By making it
more difficult for people to do certain things, or impossible to interact
with certain others, it contributes to the obsolescence and oblivion of the
identities sustained by these activities and interactions.
Collective identities in the aftermath of violence therefore represent
the bonds of solidarity and sociality that are left over once many, if not
most of, people’s ordinary social connections have been broken. They
are radical reductions of identity. As a consequence, one typical feature of
identities re-shaped by violence is their rigidity: by giving up alternative
ways to define themselves, people also lose—sometimes temporarily,
sometimes chronically—their capacity to navigate flexibly between a
variety of relevant identities. Therefore, violent turning points do not
simply provoke shifts from one prevalent form of identity to another;
they also transform the nature of identity, from something fluid into
something frozen.
This point about the loss of fluidity and the freezing of identities
through violence is sufficiently foundational for us to spend a consid-
erable portion of this chapter in illustrating and explicating it through
a concrete example. This concerns the siege of Sarajevo, to which we
already alluded in our first chapter. We choose this example because, if
identities came to be reduced to one single overarching (ethnic) dimen-
sion in the previously vibrant and cosmopolitan Sarajevo of the early
1990s, there is no fundamental reason why war could not produce the
same sombre outcome anywhere else.
Having worked through the Sarajevan case, we will then develop our
argument in two different ways. First, we will argue that the re-patterning
of practices and identities does not depend upon the actuality of war.
Things don’t have to be as bloody as they were in Bosnia for violence to
make a difference. Indeed they don’t have to be bloody at all. The mere
anticipation of violence can be sufficient to generate processes of trans-
formation. That is, believing that we might plausibly come under attack
from others can be enough to corrode our everyday practices—where we
go, who we talk to, who we interact with. We will show how our social
world and social identities begin to close down as soon as people behave
 4  Violence: How Collective Shocks Transform Social Practices  101

collectively as if war is real, as if previous practices are now dangerous,

as if erstwhile friends are now enemies. The collective imagination of
conflict can be every bit as toxic and produce just as powerful a collective
shock to the system as the collective experience of conflict.
Second, we will consider a critical implication of the realisation that
violence induces profound qualitative alterations in collective identity:
by transforming the identity of the groups involved, it transforms the
very meaning and purpose of collective conflict. As a consequence, what
set conflict in motion in the first place (its so-called root causes) tends
to become increasingly irrelevant for understanding the conflict—what
drives it, what is at stake, and what can be done to contain it—as it
unfolds.

4.1 Performing Ethnic War

Sarajevo might be an overused symbol. History textbooks tell us that
World War I started here and with it the first act of the short twentieth
century, which Hobsbawm (1994) dubbed the ‘age of catastrophe’. Much
later, between 1992 and 1995, as the city’s siege unfolded in front of the
eyes of the world’s press (unlike the mass killings committed in Bosnia’s
hinterland), it became the dramatic epilogue to the selfsame century. The
sinister images of senseless sniper-fire terrorising a starving civilian popu-
lation appeared to epitomise the end to an illusion of socialist ‘brother-
hood and unity’.
When French President François Mitterrand landed at Sarajevo air-
port on 28 June 1992—the same date on which Archduke Ferdinand
of Austro-Hungary was assassinated in 1914—he made a dramatic con-
nection between the first and the last great European tragedies of the
twentieth century. Hobsbawm (1994) would later comment on the event
as revealing the historic amnesia of his contemporaries. The significance
of the date—and hence of Mitterrand’s gesture—went largely unnoticed
by younger citizens. But one might also wonder whether those who did
perceive the relationship read the gesture as intended. Instead of seeing
a plea for solidarity with the Bosnian people, could Mitterrand’s linking
of conflicts across time ironically have fed into the narrative of ‘Balkan
102  Identity, Violence and Power

ghosts’? Could he have reinforced the idea, so insistently spun by Robert

Kaplan, that Sarajevo was simply the latest in a long line of conflicts that
derive from the primordial and tribal nature of the region. As Kaplan
formulated it, with his unerring capacity for dramatising and for using
simplistic clichés, “Twentieth-century history came from the Balkans.
Here men have been isolated by poverty and ethnic rivalry, dooming
them to hate. The politics have been reduced to the level of near anarchy
that from time to time in history has flowed up the Danube into central
Europe” (Kaplan, 1993, p. XXVII).
The problem with this claim, of course, is that it involves an arbitrary
selection of conflictual moments from across history in order to sustain
the claim of a conflictual essence. But, moving beyond the clichés and the
bloody drama, what was life really like in pre-war Sarajevo and how did
it change? What did the siege actually mean for ordinary Sarajevans in
their daily struggle for survival? And what can we learn from their expe-
riences about the way collective violence transforms collective identity?
Thankfully, we have the work of anthropologist Ivana Maček (2009),
whose meticulous ethno-graphic observations of Sarajevan life behind
the headlines allow us to address these questions.
The first surprise that awaits readers of Maček’s book Sarajevo Under
Siege, especially those who were exposed to the relentless ethnic framing
of the world’s media, concerns the composition of the city’s improvised
defence forces. In the first stages of the war, Sarajevans from families
of different religious or cultural backgrounds—‘Muslims’, ‘Croats’ and
‘Serbs’ as others might label them but not how, initially at least, they
labelled themselves—fought together on the frontlines. At that stage, the
dominant interpretation of the war was not in terms of conflict between
ethnic groups, but more in terms of an urban/rural divide. Sarajevo was
defending its sophisticated and cosmopolitan way of life against brutal
attacks bred in the backward and parochial countryside.
As a consequence, the large numbers of mainly Muslim refugees who
came to the city in order to flee fighting in the neighbouring villages and
towns were viewed with suspicion by city-dwellers—even Muslim city-­
dwellers. Indeed, if the ethnicity of the refugees had any significance to
Sarajevans, it lay in the contrast between their mono-ethnicity and the
sophisticated multi-ethnic identity of the city.
 4  Violence: How Collective Shocks Transform Social Practices  103

We can go further: it wasn’t just that many people eschewed simple

ethnic categories, but that they actively subordinated these categories
to what they considered to be more fundamental lines of divide. Thus,
rather than employ a category such as ‘Serb’, Serbs were divided accord-
ing to whether they stood for or against the city: ‘good Serbs’ or ‘ortho-
dox people’ were friends; ‘Chetniks’ (a reference to World War II Serbian
paramilitaries with a decidedly chequered history) were the enemy.
Yet, for all that the city inhabitants were uneasy with the ethnic fram-
ing of the conflict and found it to be in contradiction with their own
experiences, they were well aware that others did see events through an
ethnic lens and that in some situations their own religious affiliation
might become critical. Maček reports a joke that circulated in Sarajevo
during the siege, as a reflection of the opportunistic approach to religion
which spread through a largely secular population:

How do people manage to leave Sarajevo? When they pass Croatian snipers
they raise two fingers (which is the Catholic way to cross oneself ), when
they pass Serbian snipers they raise three fingers (the Orthodox way to
cross oneself ), when they pass Muslim snipers they raise five fingers, the
whole hand (the Muslim way of praying), and when they finally get out
they raise one finger, the middle one (an expletive gesture). (Quoted by
Maček, 2009, p. 168)

It wasn’t only in the world of jokes that people re-discovered religion as

a means of leaving the city. Croatia offered citizenship and a road out of
the Bosnian war zone to inhabitants who could document that they were
baptised Catholics—an opportunity many were happy to seize. Equally,
religion wasn’t only relevant to getting out of Sarajevo; it was also a means
of surviving in Sarajevo. Life became increasingly hard. Shortages became
endemic. Access to humanitarian aid became vital. However, distribution
of food supplies was in the hands of religious associations who delivered
aid only to their co-religionists, assuming that other associations would
take care of ‘their’ communities. Caritas, for example, provided food to
those who accepted having their houses blessed by a Catholic priest. Since
Caritas was the most efficient food supplier in the city, objective inequali-
ties arose between those who could claim some link to Catholicism and
those who could not.
104  Identity, Violence and Power

The worst off in town were those people who were not linked to any of
the main religious communities. They had to face the shortages, dangers
and uncertainties of life during the siege without being able to count
on the solidarity of an organised community—a critical resource in the
chaos of war. Against this backdrop, it is not surprising that, during the
war years, an increasing number of Sarajevans attended religious services,
learnt how to pray in public as Catholics, as Orthodox, as Muslims or as
Jews, and started to celebrate religious holidays together with their fami-
lies, friends and neighbours. In this context, acts of devotion that might
initially seem foolish and irrational (such as gathering in front of the
Catholic cathedral after mass even though the space was openly exposed
to sniper fire) can be seen to make sense.
In short, the contingencies of survival under siege increasingly made
Sarajevans behave as if religion was a central part of their identity—
and exposed them to others acting as if religion mattered (possibly
more than life itself ). They thereby became increasingly aware of their
friends’, neighbours’, colleagues’ and comrades’ religious identity. They
also became more expert about the relevant markers of identity, allowing
them rapidly to tell an Orthodox Serb apart from a Catholic Croat, or a
Muslim.
More critically, perhaps, these developments in town had their repercus-
sions amongst those in the trenches defending the town. As the ethnic/reli-
gious polarisation proceeded apace, and as Sarajevan Muslims learnt of the
ethnic massacres in Eastern Bosnia, they began to consider that they too
might be at risk—again, not because of what it meant to themselves to be
‘Muslim’, but because of what it might have meant to others who catego-
rised them as ‘Muslims’. Moreover, without knowing exactly who might
see them as such and attack them as such, it began to make sense to view
others as ‘Serbs’. Solidarity in the trenches began to give way to distrust.
For the Serbs such distrust, and its implications, was equally corrosive.
Rather than remain under constant suspicion, many young Serbs chose
to quit the city’s defence forces and to move to the other side of the front-
line, to the ‘Serb-held’ neighbourhoods of the town, where they expected
to find a more accepting environment. But this only made the situation
worse, especially for those Serbs who remained behind.
 4  Violence: How Collective Shocks Transform Social Practices  105

To flit overnight across the line was seen by Muslims as an act of

betrayal. How could you trust people who might desert in the midst of a
war and who, still worse, might then shoot at you from enemy trenches?
Rather than suffer a potential fifth column in one’s midst, was it not bet-
ter to exclude such potentially disloyal elements? The distrust of Serbs
grew; the screw was tightened. More Serbs left and so the spiral grew
more vicious.
A second wave of ethnic homogenisation among the city’s defence
forces was initiated in Spring 1993 when a new front opened in the Lašva
Valley, northwest of Sarajevo, which pitted Croat against Muslim fight-
ers. This led Muslim commanders to conclude that it was too risky to
leave the defence of Sarajevo in Croat hands. So they decided to disarm
Croat combatants. One of them recorded his own sense of betrayal at
this policy: “They woke you in the dormitory with the gun aimed at your
head, the comrades who were in the trench with you” (quoted by Maček,
2009, p. 196).
Eventually, despite (and not because of ) the will of those involved, an
ethnic grid was superimposed on the war and the Sarajevan defence forces
mutated de facto into a Muslim army. As that happened, non-Muslims
in Sarajevo began to fear for their own safety from erstwhile neighbours.
As this chronicle shows, the siege of Sarajevo represents a textbook case
of a violent turning point in the trajectory of collective identities. Ethnic
identity was not a pre-eminent frame to Sarajevans before violence broke
out—it was largely irrelevant to most aspects of their lives. But it became
an inescapable dimension in their lives through the social practices that
developed under conditions of the siege. Sarajevans then had to live with
this new social fact. It became all but impossible to ignore their own and
their neighbours’ ethnic identities during the siege, and difficult to sim-
ply go back and forget these identities after the siege. But if the Sarajevo
case lends itself particularly well as a starting point to a discussion of how
violence impacts upon identity—as violence was so obviously present in
daily life and its consequences so clearly discernable—it should not lead
us to think of the impact of violence as necessarily conditional upon such
dramatic conditions, where an entire civilian population live in a war
zone and could, at almost any point, be the target of snipers or shells.
106  Identity, Violence and Power

To get a sense of the actual boundary conditions of the psychosocial

processes at play, we rather need to consider the diversity of the circum-
stances under which violence can be real in its consequences, and these
circumstances go far beyond the frontlines of war zones. This diversity
is what we will explore next, using a deliberately broad set of examples.
Indeed, to illustrate just how broad the applicability of the analysis might
be, we will start our analysis in Switzerland, a country renowned for
peace not war; for neutrality not partisanship; for the mundane not the
dramatic.

4.2 T
 he Second World War Did happen
in Switzerland
In 2011, viewers of a French-language Swiss TV show elected General
Henri Guisan, who led the Swiss army during World War II, ‘Romand
of the century’. At one level, this might seem an entirely unsurprising
choice. Similar TV shows in other places had asked audiences to elect
their greatest figure of all time. De Gaulle was chosen as the greatest
Frenchman of all time. Churchill was elected as the greatest Briton of all
time. Being a World War II leader seems to convey a distinct advantage.
But, on reflection, the choice of Guisan is rather different to that of
De Gaulle or Churchill and might give us pause for thought. The first
difference is anecdotal and concerns the use of categories. Whereas De
Gaulle was a proud Frenchman who achieved for France and Churchill
was an emblematic Briton who achieved for Britain, Guisan was selected
as a great Romand (the western francophone region of Switzerland that
corresponds to the boundaries of the TV station’s audience). But his fame
is entirely due to the fact that he commanded the federal army, as a Swiss
general.
Once again, we see here the contingency and flexibility of social iden-
tities at work: how the same person or people can be defined in terms
of different categories (national, regional, etc.), how different definitions
of social categories are used as a function of different ways of organis-
ing social practices (since Switzerland has no national TV channel, but
rather different language channels for the different linguistic regions of
 4  Violence: How Collective Shocks Transform Social Practices  107

the country, the category ‘Romand’ makes sense insofar as it corresponds

to the boundaries of this particular channel’s audience); and how the
ways others define us may differ from the ways we define ourselves (we
don’t know what it meant to Guisan himself to be a Romand even if he was
defined as such in this exercise). Further, it hints at the conditions under
which an individual may be categorised in different ways. Would there
have been such eagerness to claim Guisan as a Romand if he had been less
successful in his career? We are reminded of Einstein’s wry comment that
“if my theory of relativity is proven successful, Germany will claim me as
a German and France will declare me a citizen of the world. Should my
theory prove untrue, France will say that I am a German, and Germany
will declare that I am a Jew”.
The second difference between the choice of Guisan and De Gaulle/
Churchill is more central to the argument in this chapter. An outsider
might reflect that the choice of a World War II leader makes sense in
France and Britain because of the huge impact of the war in those two
countries. France was under occupation. Britain was all but invaded.
Moreover, the war made fundamental changes to British society: it trans-
formed social relations in the country and it transformed Britain’s place in
the world. In addition, it is arguable that Britons continue to obsess about
the war and continue to cherish it as ‘our finest hour’. But Switzerland?
World War II self-evidently did not take place in Switzerland. Apart from
a few bombs that allied countries accidentally dropped on the wrong side
of the Swiss–German border, the country was not exposed to combat. So
why would a war exert such a hold in a country where it didn’t happen?
How could a commander of an army that did not actually fight become a
collective hero, and be recognised as such even among a generation that,
in its overwhelming majority, was born well after the war in question?
To the outsider, the more one reflects on the choice of Guisan, the
more surprising it becomes. But for the insider, things are rather dif-
ferent. The premise that generates such surprise—the observation that
World War II did not take place in Switzerland—doesn’t seem quite so
self-evident. It may well be true that the Swiss army didn’t engage in
­combat with enemy troops but that didn’t mean that it was inactive. The
army may not have fought in a conventional sense, but under General
Guisan’s leadership it was prepared and daily preparing for combat:
108  Identity, Violence and Power

it was ‘mobilised’, according to the terminology used by Swiss officials and

people.
Nor was the general public unaffected by the war. The fact that a
German invasion was deemed possible at any moment produced a great
deal of sympathy and solidarity with the soldiers who were ready to fight
(and die) for their country. To some, the fact that Nazi Germany never
attempted to invade Switzerland was evidence of how well they prepared
and how effective they were. The lack of combat, then, did not indicate a
lack of Swiss military involvement. To the contrary, it was evidence of their
involvement, their prowess and the brilliance of their leader. From this per-
spective, Guisan won the battle (and the war) by not having to fight it.
Just like the French, the British and others, the Swiss therefore feel
entirely justified in remembering World War II as the most dramatic
chapter in their own national history. So when, in more recent years,
critical historians argued that Hitler’s decision not to invade Switzerland
had more to do with his appreciation that a neutral Switzerland better
served Nazi Germany’s financial and economic interests than an occupied
Switzerland, and less with his anticipation of strong military resistance
(see Bergier, 2002), they provoked a highly emotional public debate.
Many accused these historians of betraying the bravery and devotion of
the Swiss troops during World War II.
Even in Switzerland, that quintessential land of peace, war can be
experienced as real, violence can be seen as imminent, and this can lead
to both new forms of mobilisation in the moment and powerful memo-
ries which dictate what can and can’t be said long after. More generally,
in arguing that violence re-defines social identities through re-patterning
social practices, we need to be clear that the boundaries of violence do
not equate to the boundaries of the combat zone. Rather, we need to look
at the ways violence is imagined—where it could occur, when it could
occur—and how this imaginary of violence impacts what people do. In
other words, we must explore the spatial and temporal boundaries to
the ways that communities live with violence rather supposing we know
them in advance.
The temporal boundaries to imagined violence are particularly impor-
tant: the impact on practices and identities does not necessarily come to
an end with the end of violence. Kusturica’s film masterpiece Underground
 4  Violence: How Collective Shocks Transform Social Practices  109

provides a metaphorical illustration of this insight: after the end of World

War II, an isolated community lives on in an underground bunker, as if
the war was still in progress. For more than a generation, they entirely
subordinate their individual lives to a collective cause—the liberation of
Yugoslavia by partisan fighters. They continue to believe that the weap-
ons that they manufacture in the bunker are being used by these partisans
in their fight.
But we don’t need to look to fiction to see how past violence continues
to shape societies in the present. Similarly, we don’t need to infer that
people are deluded, that they don’t realise that the old war is formally over
and that the shooting and killing has stopped. In contemporary South
Africa, no one lives in a shelter believing that the structures of apartheid
are still in place. But despite the dismantling of segregation laws, an end
to forcible removals and the formal opening of public spaces to all, those
previously designated as ‘Blacks’ and ‘Whites’ still tend to avoid proximal
interaction, whether on the beaches, in residential neighbourhoods or
even in University cafeterias (Clack, Dixon, & Tredoux, 2005; Dixon &
Durrheim, 2003; Durrheim & Dixon, 2001).
In the South African case, this enduring segregation is not explicitly
justified by reference to apartheid and apartheid-era conflicts. After all,
apartheid is a spoilt ideology and the official discourse has been struc-
tured around truth and reconciliation as means of moving beyond the
past (O’Brien, 2001). Where segregation is argued for, it is in terms of
different, seemingly more liberal discourses such as environmentalism.
That is, poor black squatter camps in lush white neighbourhoods are
criticised as a ‘blot on the landscape’ (Dixon, Reicher & Foster, 1997).
However, in many other cases, past conflicts are explicitly invoked to
maintain contemporary social practices. Enemies and threats are not
invented from scratch at the point when violence begins to escalate. That
is, there is generally a long-term context to short-term crises—one which
makes representations intelligible, practices familiar and hence facilitates
the production of violence. If, as we have seen in Sarajevo, a siege has
profound effects on social relations, the same is true of a siege mentality.
Israel is a case in point. Present-day relations with Palestinians are incom-
prehensible without considering the way that Jewish history is represented.
In his great Social and Religious History of the Jews, Salo Baron (1969)
110  Identity, Violence and Power

­ elineates (and challenges) what he terms the ‘lachrymose’ version of Jewish

d
history. According to this viewpoint, from biblical times onwards, Jewish
experience equates with suffering and this is linked to the view that Jews
are always surrounded by enemies who seek to destroy them—an ongo-
ing experience of which the Holocaust is only the latest and most virulent
manifestation, but certainly the most salient. In a 2009 survey, an astound-
ing 98 % of Jewish-Israeli adult respondents stated that remembering the
Holocaust is a guiding principle in their life, even more important than
having a family (Arian, 2012, cited in Klar, Schori-­Eyal, & Klar, 2013).
Such a view of history feeds into psychological orientation which Klar
and colleagues dub a ‘perpetual ingroup victimhood orientation’, which
in turn frames the way current conflicts are viewed. For instance, amongst
those who subscribe to this victimhood orientation, outgroup members
are perceived as having more hostile intentions and there is less guilt at
extreme actions taken against them—especially when they are actively
reminded of past catastrophes. The argument has been eloquently for-
mulated by Bar-Tal and Antebi (1992) who referred to the Jewish siege
mentality as a set of collective beliefs that “come and go, riding the crests
of associations incessantly flowing through the individual’s mind, espe-
cially primed by contextual objects and events” (p. 635).
This notion of ‘come and go’ is crucial, for it implies that such beliefs
are not self-evident, that they do not automatically stay alive (as ‘ancient
hatreds’ arguments imply) and that they have to be kept alive—or at least
invoked at particular points in time. So how does this happen?
The Israeli case provides us with some insights. To start with it is impor-
tant to stress that the ‘lachrymose’ version of history is a very selective
account. As Baron memorably observed, “suffering is part of the destiny”
of the Jewish people; “so is repeated joy as well as ultimate redemption”
(cited in Eckhardt, 1992, p. 135). Moreover, even if one subscribes to
this version, one can draw different lessons from it—not only “never
be a passive victim again” but also (albeit less prominently) “never be
a perpetrator” (Klar et al., 2013). So the notion of Jews as ­victims who
must strike first to ensure that they survive is one of many narratives of
Jewishness and it has to be actively invoked—as, for instance, by Prime
Minister Benjamin Netanyahu when addressing the nation on Holocaust
Memorial Day, 2010:
 4  Violence: How Collective Shocks Transform Social Practices  111

in every generation there are those who stand against us. And in this generation
we must fortify our strength and independence so that we will be able to prevent
the current enemy from carrying out its plan. (Cited in Klar et al., 2013,
p. 135)

The important thing about this example is not just what is said, but the
context in which it is said: Holocaust Memorial Day. The Holocaust is
woven into Israeli society in a plethora of different ways: memorial days,
memorial sites, museums, statues, textbooks, films, trips. To take just two
telling statistics (both from Klar et al., 2013), in just one (relatively lib-
eral) newspaper, the term Holocaust (Shoah in Hebrew) appeared as often
as Israeli–Arab conflict. Every year, 16 % of the entire high school cohort
go on trips to the death camps, mainly in Poland. As Liebman and Don-­
Yihya noted in 1983, the memory of the Holocaust is still omnipresent
in Israel, cutting across age, education and country of origin.
Another way of putting this—invoking the work of Billig, 1995—is
to say that memories of the Holocaust have become banal in Israel—not
in the sense of trivialising the event of course, but rather in the sense of
being so pervasive as to become embedded in all areas of life, as part and
parcel of what it means to belong to the Israeli nation. Billig gives many
powerful examples of the banality of nationhood—how it is presupposed
in terms of the way we talk about the weather, sporting results, what is
newsworthy (because it matters for the nation) and whether the news
is good or bad (because it does or does not serve the national interest).
When this occurs insidiously and continuously, and particularly when
the national interest is seen to be threatened, it can be used to far less
banal effect. To cite Billig himself:

As the Gulf and Falkland Wars indicated, forces can be mobilized without
lengthy campaigns of political preparation. The armaments are primed,
ready for use in the battle. And the national populations appear also to be
primed, ready to support the use of those armaments. (Billig, 1995, p. 7)

As in Israel, though, it is not enough to expose people to banal nation-

alism. One doesn’t become a banal nationalist simply by sitting in an
armchair and absorbing banal nationalist rhetoric. One has to be actively
involved. One has to be involved in rehearsing nationhood. While in
112  Identity, Violence and Power

many US schools children literally rehearse around the flag every morn-
ing (singing their national anthem together), there are many other ways
that people can practise their banal nationalism in daily routines: by
expressing their joy when a fellow national wins a sports competition, by
commenting on the national weather forecast or by passing on news that
is important for the nation.
While it is perfectly possible (although potentially self-isolating) to
choose not to rehearse nationhood when it comes to sporting events,
to the weather or to the news, there are other areas of life where people
will find it much harder not to play their active part. Billig gives the
example of national currencies: Who could afford the luxury of not using
them (and why exactly would anyone do that)? When Croatia became
a sovereign nation-state in 1991, its government chose to label the new
national currency the ‘kuna’. As this was the name given to the currency
during World War II, when the ruling fascist regime perpetrated mas-
sacres against the Serb minority, many saw the new currency as a fascist
and anti-Serb symbol. Yet, inhabitants of the new Croatian state quickly
became used to enacting the reality of their (soon to be war-ridden)
national state several times a day by the small act of taking the kuna out
of their wallet. By the same token, they quickly became accustomed to
the ambiguities surrounding the definition of their national identity that
were materialised in the coins and banknotes they used.
This example, like the Israeli example, shows that the way in which
people rehearse their national identity not only serves to make the nation
real but also gives substance to that reality. It helps define the values and
the ideology of the nation, who is included and who is excluded, how
the nation relates to others and how others relate to it, who is seen as an
ally and who a threat. The same is true the other way round. The way a
nation prepares for potential threats from an outgroup serves to define the
nation itself. To act as if our country could come under existential threat
at any time is a powerful means of constraining social relations. That is, if
the recollection and re-enactment of past violence can r­ econfigure identi-
ties as much as actual violence, the same is true of rehearsing for future
violence. Cold War experiences are emblematic of this. They show how
communities can be profoundly transformed by a credible threat of mass
destruction—even when that threat never materialises.
 4  Violence: How Collective Shocks Transform Social Practices  113

Extensive archival research conducted across the USA, Canada and the
UK by Davis (2007) shows how the nuclear threat by the Soviet Union
became a tangible reality for the Western public during the 1950s and
1960s through recurrent large-scale emergency civil exercises. Across the
USA, Canada and the UK, casualties of nuclear radiation, burning or
physical injury were made up with great attention for detail, in order to
increase the realism of these exercises.
In the USA, systematic civil defence training found its way into an
overwhelming majority of classrooms between 1950 and 1952. A gen-
eration of school pupils learnt and practised how to ‘duck and cover’ as
soon as a nuclear flash appeared, and thereby to protect themselves from
shockwaves in school, at home or outdoors. Most states implemented
curricular reform in the early 1950s to make sure that pupils acquired
the necessary knowledge to cope with the challenges of nuclear age, from
the chemistry of heat to international relations. But education was not
limited to children. In the early 1970s, American test families dug them-
selves into their own home-made shelters, thus demonstrating that it was
possible to protect yourself under your own steam in less than a day.
Apart from behaviour increasing the chances of physical survival fol-
lowing a nuclear strike, particular emphasis was laid on practising skills
that were deemed essential to prevent a breakdown of social order: obe-
dience to orders, not starting rumours and dispelling myths that could
cause panic. In the UK, Anna Freud advised families to “make quite a
confident ritual of air-raid precautions” (quoted by Davis, 2007, p. 109),
where everyone has a clearly assigned role—for example, children were to
take their teddy bears to the shelter—in order to foster everyone’s sense
of orderliness and security.
Overall, then, the goal of such mass dramas was not only to instruct
people how to survive individually in the aftermath of a nuclear explo-
sion, but also to instil in them a sense of civic responsibility. By rehearsing
for nuclear war, people were not only taught what an attack would con-
cretely mean to them in terms of physical survival, they were also given
an education in how they must be pro-active if they were to survive as a
community, to avoid group collapse and social paralysis. There was a clear
moral imperative not to let a social catastrophe—the breakdown of the
social fabric—add to the nuclear catastrophe and the destruction of the
114  Identity, Violence and Power

physical environment. The discipline enacted during the mass rehearsals

sent out a clear message: no defeatism or dissent was tolerated on the
social front. In that sense, rehearsals for nuclear war must have been a
paradoxical experience for those who took part in them, making salient
their moral duty to be prepared to cope with the announced apocalypse.
To cite Davis:

Role-play and acting in repeated rehearsals during peacetime was integral

to governments’ recommended preparations for their citizen’s survival in
wartime. Thus, acting was not only the method but also the ontology of
the populace to preserve life, the ethos of their nation, and the fabric of
their culture. It was the bulwark against nihilism, the motivation for belief,
and the insurance of survival. Acting was the way to buy into the idea that
civil defence could be efficacious; it was also, at the same time, the means
to see how it did not work. (p. 219)

The key moral lesson, then, was that conformity is the key to survival
because individual fate depends upon the fate of the nation as a whole
and on state policies. This is exemplified in a script prepared by British
contingency planners for the BBC in November 1964, as part of a sim-
ulated emergency programme. The script has never been broadcast; its
main purpose was to raise awareness within the media on their role in
case of a nuclear attack. In the imagined scenario, the following text was
to be read immediately following the announcement that several regions
across Britain had come under attack by nuclear weapons:

Serious fires are raging in these places and there are very many casualties. Civil
Defence and other rescue services are doing everything they can to rescue sur-
vivors. Immediate retaliatory measures were taken by our own forces and there
have been no further attacks since 3 o’clock. (Quoted by Davis, 2007, p. 191)

Through imagined broadcasts such as this, the journalists could proj-

ect themselves into a concrete nuclear war scenario. While the impact
is depicted as very severe (‘serious fires’ and ‘very many casualties’ are
mentioned), state services are still functioning: they are rescuing survi-
vors and retaliating against the aggressor. The fact that the latter piece of
 4  Violence: How Collective Shocks Transform Social Practices  115

information is immediately followed by the announcement that ‘there

have been no further attacks’ suggests that the state’s retaliatory actions
have been effective in stopping further destruction. So the authorities are
seen to apply effective retributional justice as well as effective protection
for the national community. Whether all this is plausible does not need
to be explained. Emergency rehearsals leave no time for critical reflection
or political argument. They need to focus on the—imagined (!)—facts.
Taken together, the highly diverse examples discussed in this section—
from Switzerland to Israel, and from the Cold War to the wars in the
former Yugoslavia—both clarify the argument that violence transforms
collective identity and also broaden it. They show that there are very
different ways in which people can ‘experience violence’: on the front-
line or far away from it; as something that has already happened, that is
currently happening or as something that might plausibly happen in the
future; as something directly experienced, or as something heard about
from others who were directly involved, or reported by the media, and
made real through communication, imagination and empathy. Whatever
is the exact nature of people’s experience of violence, the critical factor
determining whether and how the experience will affect their collective
identity is the active part they all play in it and which they see each other
as playing. By performing their identities differently, people transform
these identities. By adapting to violence as a shared social reality, people
create new social facts that make it increasingly difficult to return to old
ways and eventually impossible to live as if violence had never happened.
Ironically, this pertains even if violence never did happen but is merely
imagined in the past or in the future.
The fact that violence changes the very terrain on which it was built
makes life tremendously difficult both for the academic and the activ-
ist, as we shall now see. On the one hand, it becomes more difficult, if
not impossible, to make confident predictions about the way violence
will unfold. On the other hand, it becomes impractical to assume that
the best way out of conflict is back through the way in. Drawing these
various points together, we cast doubt on the optimism of those who
believe that we can both explain and resolve violence by uncovering its
root causes.
116  Identity, Violence and Power

4.3 The Elusive Hunt for Root Causes

In 1962, the American sociologist James C. Davies published an article
in which he made an ambitious promise: to pave the way “toward a the-
ory of revolution”. In concluding his seminal reflections, Davies readily
acknowledged that social scientists “are still not at the point of being able
to predict revolution” (p. 19, our emphasis). But he was no less confident
that they “should eventually be able to escape the embarrassment that
may have come to Lenin” (p. 19) who, six weeks before the upheavals
of February 1917, expressed doubts as to whether he would ever witness
revolution in his own lifetime.
How did Davies hope to accomplish such an ambitious intellectual
endeavour? At the core of his predictive approach was an attempt to iden-
tify structural patterns that would systematically precede the outbreak of
revolution. According to Davies, it was “the dissatisfied state of mind rather
than the tangible provision of ‘adequate’ or ‘inadequate’ supplies of food,
equality, or liberty which produces the revolution” (p. 6, our emphasis).
However, this did not mean that his model was primarily psychological,
for Davies aimed to specify the objective conditions which gave rise to
such dissatisfaction. His core thesis was that when deprivation is con-
stant, we find it hard to imagine any other state of being and hence learn
to live with the burden. But when levels of deprivation change—and
more particularly, when we experience a rapid downturn after a period
of steady improvement—then we find the burden unbearable. So people
will fight against their regimes when a gap opens between rising expecta-
tions and declining fulfilment of these expectations, when restrictions
are imposed on them at a time when previous progress led them to see
deprivation as no longer inevitable.
Since Davies developed his ideas, the social sciences might have become
less interested in economic class-based clashes and more in conflicts
across ethnic or cultural cleavages. However, it is less certain that they
have become more humble about their ability to predict the future. After
the Cold War ended, Fukuyama (1992) promptly proclaimed the end of
history, before (as discussed in Chap. 1) Huntington (1996) announced
the coming clash of civilisations. The ensuing years, culminating in the
post-nine-­eleven wars, gave impetus to Huntingdon’s thesis. In this con-
 4  Violence: How Collective Shocks Transform Social Practices  117

text, political scientist Roger Petersen (2002) seemed to offer the aca-
demic book market exactly what it has been waiting for: a treatise on
ethnic violence, proposing a predictive theory about when one ethnic
group will violently target another.
In his book, Petersen argues that ethnic violence follows changes in
the objective relationships between ethnic groups brought along by new
state boundaries and/or shifts in the balance of resources available to the
respective groups. People will fight against ethnic outgroups when struc-
tural changes create new threats or frustrations incarnated by the out-
group target, or else when new opportunities to attack a specific target
group arise.
For Petersen, as for Davies, changing structural circumstances ulti-
mately drive collective behaviour, while collective emotions mediate the
process. To engage in joint aggressive action, all those involved need to
be driven by similar emotions, which in turn are provoked by similar
circumstances. These circumstances can take different forms. Sometimes
it is a matter of breeding antagonisms. Petersen argues that loss of relative
status breeds resentment. So, for instance, it was the fact that Bosnian
Croats and Bosnian Serbs became minorities in the new Bosnian nation-­
state following the dissolution of the former Yugoslavia which would have
led to their attacks on the Bosnian Muslims. Sometimes it is a matter
of removing the constraints upon the expression of pre-existing resent-
ments (as in the ‘ancient hatreds’ narrative we have discussed in a number
of places). Thus, Petersen suggests that the reason why Serbs aggressed
against Kosovo Albanians in the wake of the dissolution of the former
Yugoslavia was basically because they (now) could.
Despite all the differences between Davies’ theory of revolution and
Petersen’s theory of ethnic violence, both seek to provide a predictive
framework for violent unrest which is concretised through the analy-
sis of structural factors, seen as the ultimate determinants of collective
behaviour. Both theories build upon the assertion that similarly changing
circumstances provoke similar motives among large enough groups of
people to account for mass mobilisation, be it against state authorities or
ethnic outgroups. In both cases, the structural circumstances are seen to
gain effect by the way that, in predictable terms, they unleash a kind of
convergent collective will. In this line of thinking, revolution or ethnic
118  Identity, Violence and Power

violence is conceptualised as a sequence of events that only occur when

a large mass of people want them to occur—when a collective is driven
forward by some intrinsic “existential motivational force”, to borrow
Petersen’s own terminology.
There are various problems with such thinking, some of which we have
already discussed in some detail. So, for instance, much of what we do
stems less from what we ourselves think, intend or desire, and more what
we believe others think, intend or desire. So, I may aggress against you
not because I want to hurt you but because I think you want to hurt me
and so need to deter you from the start. This raises the very obvious point
that violence is a product of interaction between two parties. Once one
invokes this interactive context, the inadequacies of an approach which
reduces violence to the will of any single party become even more obvi-
ous. Thus, the consequences of my actions depend upon the way they are
apprehended by you and how you choose to respond. This may often be
in ways that I hadn’t intended or anticipated. So, by acting on the basis
of one understanding of reality, I provoke you to create a new reality,
which then provides a different context for my subsequent actions. I may
therefore end up doing things I never dreamt of at the start.
Rather than being atypical, that is the way that things work in general.
Therborn (1982), for instance, argues that no revolutionary movement
sets out to create a revolution, but only in the way that the state responds
to reformist demands does a more radical dynamic develop. Revolutions,
and other forms of violent clashes, arise as chains of events that generate
new realities in successive waves, chains in which conflict dynamics shift
from one fragile equilibrium into the next, without a clearly discernable
linear direction or overarching ‘motivational force’ that would span over
the entire chain to link its starting point and its end point in a consistent
way.
The most dramatic historic example of such a chain of events without
a clear direction or collective motivation can be found in the few months
in summer 1914 during which the European continent descended from
an era of prosperity, progress and stability into Hobsbawm’s ‘age of
catastrophe’.
In his autobiography, the Austrian novelist Stefan Zweig describes his
stay at the Belgian seaside, in the short period between the assassination
 4  Violence: How Collective Shocks Transform Social Practices  119

of Franz Ferdinand in Sarajevo and the outbreak of World War I. His

account provides a vivid illustration of the climate in which Europeans
drifted to war, without feeling concerned about what was happening or
anticipating what was about to happen:

All imaginable nations gathered peacefully, one heard particularly many

German voices as, as every year, nearby Rhineland sent his people for sum-
mer vacation preferably to the Belgian beaches. The only interruption
came from the newspaper boys who, to boost their sales, shouted the
threatening headlines: “Austria provokes Russia”, “Germany prepares for
mobilisation”. One could see how the faces of the people, when they
bought a newspaper, became more sombre, but always just for a few min-
utes. After all we knew these diplomatic conflicts for years already; they
have always been settled at the last moment, before things would have
become serious. Why not this time again? Half an hour later one could see
the same people again cheerfully snorting and splashing in the water, the
kites rose, the gulls fluttered, and the sun was shining bright and warm
through peaceful land. (Zweig, 1944/1997, pp. 251–252, our translation)

Arguably, this lack of awareness amongst ordinary people mirrored a

similar lack of alertness amongst the European political elite who were
actually dealing with the crisis. At least this is the impression that arises
from historian Christopher Clark’s (2012) analysis of the pre-war diplo-
macy. Sleepwalkers—the evocative title of Clark’s book—alludes to his
perception of these elites on the eve of war. It is a perception repeated and
elaborated in the very final words of a 562-page chronicle of the various
twists and turns which culminated in Armageddon: “the protagonists of
1914 were sleepwalkers, watchful but unseeing, haunted by dreams, yet
blind to the reality of the horror they were about to bring into the world”.
There are three key arguments which lead Clark to this conclusion
and which are highly relevant to our argument in this book. The first is
an explicit rejection of any notion that the war expressed a firm will, and
that therefore it was inevitable. So, Clark asserts firmly: “the myth that
European men leapt at the opportunity to defeat a hated enemy has been
comprehensively dispelled” (p. 553). Instead, “for most places and for
most people, the news of mobilization came as a profound shock, a ‘peal
of thunder out of a cloudless sky”’ (p. 553).
120  Identity, Violence and Power

Second, and as a corollary, Clark takes contingency seriously. He studi-

ously avoids the danger of, retrospectively, converting the actual into the
probable. On the contrary, he asserts:

Some of the most interesting recent writing on the subject has argued that, far
from being inevitable, this war was in fact ‘improbable’—at least until it actu-
ally happened. From this it would follow that the conflict was not the conse-
quence of a long-run deterioration, but of short-term shocks to the international
system. Whether one accepts this view or not, it has the merit of opening the
story to an element of contingency. And it is certainly true that while some of
the developments I examine in this book seem to point unequivocally in the
direction of what actually transpired in 1914, there are other vectors of pre-war
change that suggest different, unrealized outcomes. (Clark, 2012, p. XXIX)

Third, what lay behind such lack of inevitability and contingency was the
fact that, repeatedly, actors were driven by their assumptions concerning
the intentions of others, that they acted in order to send a signal to oth-
ers, and that both intentions and signals were repeatedly misread. So, for
instance, when Russia finally decided to mobilise its troops it was because
it over-estimated the extent and aggressiveness of Austrian mobilisation.
This led Germany to perceive Russia as aggressive and mobilise in turn.
Thus, Clark refers to “the tendency we can discern in the reasoning of so
many of the actors in this crisis, to perceive oneself as operating under
irresistible external constraints while placing the responsibility for decid-
ing between peace and war firmly on the shoulder of the opponent”
(p. 519). Cumulatively, this produced a situation where everyone may
have been willing to fight, not because anyone wanted a war to happen
but rather due to what Clark terms “a defensive patriotism”:

the aetiology of this conflict was so complex and so strange that it allowed
soldiers and civilians in all the belligerent countries to be confident that
theirs was a war of defence, that their countries had been provoked or
attacked by a determined enemy, that their respective governments had
made every effort to preserve the peace. (Clark, 2012, p. 553)

World War I, then, constitutes a powerful argument against any attempt

to explain violence as a predictable outcome of the collective will.
 4  Violence: How Collective Shocks Transform Social Practices  121

But, however important the case, one might object that it is unique,
incomparable, unsuitable as the basis for a general argument. One might
further object that World War I is the quintessential case of a war fought
between national armies, led by a tiny elite through highly disciplined
chains of command, and in that respect very different from cases of civil
unrest, rebellion or communal violence, which rely much less on pre-exist-
ing forms of institutionalised hierarchy and order. It is therefore important
to look more systematically at multiple cases, and to look at intrastate
violence in particular, before drawing any conclusions about the respective
roles of structural factors or root causes, on the one hand, and chains of
events that develop a dynamic on their own, on the other.
In the early 1990s, Ted Gurr undertook just such a study in a quest
to identify structural factors that can predict minority group collective
action in its generality. The Minorities at Risk project involved expert
coding of the behaviour of 227 “politicised communal groups” across
90 nations, during the entire period from 1945 to 1989. This huge
and comprehensive data set allowed Gurr to ask questions such as: Are
people more likely to join a collective struggle for rights when these
rights are disrespected? Are groups that face greater disadvantage more
likely to take violent action? The conclusion can briefly be summed up
as ‘yes and no’.
Yes, minority groups that face higher economic or political disadvan-
tage are more likely to mobilise politically. They will raise grievances
more often and initiate various forms of social protest. But no, there is
no direct relationship between the magnitude of a group’s disadvantage
or experiences of discrimination, and the likelihood that this group will
ever be involved in violent rebellion:

Objective conditions (poverty, discriminatory treatment, loss of autonomy)

determine the issues around which leaders are able to mobilize collective
action. The greater the differentials between groups, the easier it is for leaders
to recruit members of disadvantaged or threatened groups. During the mobi-
lization process communal leaders give stronger voice to grievance (…) and
commit their followers to strategies of protest or rebellion. But once a group
is committed to a particular strategy, self-sustaining conflict dynamics tend
to develop: fighting groups and their opponents get locked into action-reac-
tion sequences from which it is difficult to escape. (Ted Gurr, 1993, p. 189)
122  Identity, Violence and Power

Actually, Gurr found that a group’s degree of organisation—cohesion

and leadership—not its degree of disadvantage, predicted its likelihood
of future rebellion. However, empirically, the most important predic-
tor of violence had nothing to do with the disadvantaged group itself,
but with its state environment: minorities facing more autocratic state
structures and/or repressive state reactions to their grievances were more
likely to take arms than minorities acting in circumstances offering more
space to the non-violent voicing of grievances. Furthermore, the trans-­
national environment in which groups act also appeared to play a role:
other things being equal, minorities were more likely to rebel after ‘twin
groups’ in other states had previously taken up arms or where rebellion
among similar groups occurred more frequently in the larger region.
In other words, Gurr’s monumental analysis shows that objective fac-
tors can tell us whether people are likely to set off down the path of
protest (though even here, the process is not deterministic since such
factors don’t directly drive people to take to the streets, but rather con-
stitute evidence that leaders can interpret in order to mobilise people for
action). However, objective factors cannot say where people will end up
once they have set off. And, as we have argued in general and as we also
saw from Clark’s analysis of the lead-up to World War I (which now we
can see as exceptional by the magnitude of violence involved, but not by
the process of escalation that led to this outcome), that is because of the
interactive dynamics of collective action. Context impacts on the actions
of one party, party A. But then the actions of party A are interpreted
by party B, whose responses alter the context in which party A takes its
subsequent actions—and so on (for a similar analysis of the escalation
of violence within single events of collective action, see Drury & Reicher,
2009; Reicher & Hopkins, 1996).
Gurr’s analysis therefore compels us to look at groups, states and their
international environment as systems of collective actors whose respec-
tive course of action is dependent upon the strategies of each other actor
in the system. It also highlights the fact that, when a certain point is
reached in the dynamics of escalation, violent struggle tends to become
self-sustaining: violence is driven by violence, not by the circumstances
that initially led to the conflict.
When laying out a revised and more comprehensive version of his the-
ory in People versus States, Gurr (2000) referred to the so-called chicken
 4  Violence: How Collective Shocks Transform Social Practices  123

and egg issue in explaining the causes of ethno-political conflict (p. 74):

on the one hand, group identities become ‘salient’ in part as a conse-
quence of systematic disadvantage or state repression; on the other hand,
the salience of group identities is a condition for group mobilisation and,
possible, minority–majority conflict. The identification of ‘root causes’ to
ethno-political conflicts is hence necessarily problematic. What appears
as an independent causal factor in one specific analysis might become a
dependent outcome, for example when the temporal scope of the analysis
is broadened or shifted.
A similar argument can be made with regard to spatial context. An
analysis focusing on a minority–majority dyad within a specific nation-­
state might conclude that the minority’s radicalisation preceded a major-
ity reaction. However, a broader trans-national focus might reveal that
the radicalisation itself followed repression of a kindred group in a neigh-
bouring state. More generally, Gurr (2000) re-asserted the importance of
state, international and global contexts to explain minority group collec-
tive action. Interestingly, explanatory factors identified at each of these
levels of analysis often have different effects on the empirical likelihood
of non-violent protest on the one hand and on violent rebellion on the
other. Democratic state regimes are more likely to face minority protest,
but autocratic structures are more likely to face rebellion (p. 85). Spillover
effects across trans-national communal groups increase the likely of rebel-
lion, but not of protest (p. 91). Core nations in the world system, or
nations that are closely connected to international governmental asso-
ciations, are more likely to face protest, but peripheral or more isolated
nations are more likely to face rebellion. All of these comparative findings
speak against explanatory models that treat any specific form of collective
action—notably collective violence—as explicable in similar terms to the
occurrence of group mobilisation or intergroup conflict per se. As Ted
Gurr put it:

Understanding of ethnopolitical conflict that emphasize the supposedly cru-

cial role of a single factor, such as historical animosities or cultural differences
between groups, should be avoided. Such explanations usually become sig-
nificant because they are invoked by contemporary ethnopolitical leaders
seeking to mobilize public support, not because cultural or historical differ-
ences generate a primordial urge to conflict. (Gurr, 2000, p. 95)
124  Identity, Violence and Power

4.4 Conclusion
In this chapter and the previous one, we have sought to disrupt the
standard narrative according to which identities and violence (and also
power—but we shall come to that in a moment) exist in a simple linear
relationship. To put it at its simplest, identities (somehow) produce vio-
lence. Identity is productive and violence is produced. Our argument is
that the relationship is much more nuanced, flexible and bi-directional.
Accordingly, in the last chapter we concentrated on how identity is pro-
duced (as well as being productive) and in this chapter we have focused
on how violence is productive (as well as being produced).
There have been three elements to our argument. The first has been to
show how violence re-configures identities by re-configuring the social
practices through which identities can be performed. The reason for this
is to do with the way that violence so radically alters the contingencies
of action. If I am aware that people defined in terms of membership of
another ethnic group (say a Serb) have attacked someone because they
were seen as a member of my ethnic group (say Muslim), can I take the
risk that any other Serb may not also see me as a Muslim and attack me
as such? Even if the probability is (at least initially) fairly low, for safety’s
sake, don’t I have to ignore the fact that this Serb is also a worker, a father,
a punk rocker and many other things besides, and act as if the ethnic
identity is what counts? Because the costs of not taking them as a Serb
and myself as a Muslim and getting it wrong (death) are incomparably
more severe than taking them as a Serb and myself as a Muslim and get-
ting that wrong (embarrassment).
What this example also tells us is that violence doesn’t just change the
identities we use; it closes down the possibilities of social practice and
thereby limits the identities we can enact. Violence is therefore some-
thing that limits us, which takes particular identities that correspond to
one mode of being and freezes them into our only possible ways of being.
The second element of the argument is that one doesn’t need actual
violence to freeze identities. The awareness of violence elsewhere, the
memory of violence past and the anticipation of violence in future can
become almost equally effective. It is important to clarify that this is not
a reversion to the ancient hatreds argument. Rather, the representation of
 4  Violence: How Collective Shocks Transform Social Practices  125

violence as something imminent always has to be actively mobilised by

invoking certain versions of the past and something that is sustained by
ongoing social practices.
Third, because of the way that violence transforms how people see
themselves and others—and hence changes the way they value things and
interpret events in the world—it is impossible to impose a simple narra-
tive on events such that what the involved actors believed and desired at
the start explains where they got to at the end. More simply, hunting for
the root causes of violence is a futile exercise. The notion that an analysis
of how violence came about provides us with an understanding of how to
end violence is based on a misconception of the phenomenon. It ignores
the fact that violence changes the social terrain. The path in no longer exists
as a path out.
So, if the notion of a settled collective will to cause violence is so wrong
and so unhelpful in explaining actual events, we must ask, like Gurr, why
those involved in such conflicts so often invoke such a will? And, like
Gurr, our answer is that invoking a popular will and claiming to act as
representative of the popular will is a particularly potent means by which
leaders can achieve influence. Moreover, by asserting a will to violence
and by producing violence, leaders further solidify the identities they
purport to represent. Violent conflict, as we have seen, is a particularly
dreadful vehicle through which to make masses of people behave as if
they accept the identities that are implied by the fighting. It is also a par-
ticularly effective vehicle through which to get people to stick together,
venerate their leaders and stifle dissent.
In short, having previously challenged the argument that people are
naturally inclined or programmed to always obey authority, we are now
in a position to offer at least a partial answer to the question of when and
why people sometimes obey authority. Consent has to be manufactured,
and violence is a potent instrument for manufacturing consent.
One can only understand violence by bringing the role of activists and
leaders in its manufacture out into the open. Where there is violence, we
need to make political effort visible just as it seeks to make itself invisible.
We need to bring the problem of power—how power is used to produce
violence and how violence is used to produce power—to the forefront of
our analysis. That is what the next chapter is about.
126  Identity, Violence and Power

References
Arian, A. (2012). A portrait of Israeli Jews: Beliefs, observance, and values of Israeli
Jews, 2009. Jerusalem: The Israel Democracy Institute and the AVI CHAI—
Israel Foundation.
Baron, S. W. (1969). A social and religious history of the Jews. New York: Columbia
University Press.
Bar-Tal, D., & Antebi, D. (1992). Beliefs about negative intentions of the
world: A study of the Israeli siege mentality. Political Psychology, 13(4),
633–645.
Bergier, R. (2002). La Suisse, le national-socialisme et la Seconde Guerre mondiale.
Zurich: Pendo.
Billig, M. (1995). Banal nationalism. London: Sage Publications Ltd.
Calder, A. (1992). The People’s War. London: Pimlico.
Clack, B., Dixon, J. A., & Tredoux, C. (2005). Eating together apart: Patterns
of segregation in a multiethnic cafeteria. Journal of Community and Applied
Social Psychology, 14, 1–16.
Clark, C. (2012). The sleepwalkers: How Europe went to war in 1914. London:
Penguin.
Davis, T. (2007). Stages of emergency. Cold war nuclear civil defence. London:
Duke University Press.
Dixon, J., Reicher, S. D., & Foster, D. (1997). Ideology, geography, racial exclu-
sion: The squatter camp as ‘blot on the landscape’. Text: An Interdisciplinary
Journal for the Study of Discourse, 17(3), 317–348.
Dixon, J., & Durrheim, K. (2003). Contact and the ecology of racial division:
Some varieties of informal segregation. British Journal of Social Psychology,
42(1), 1–23.
Drury, J., & Reicher, S. D. (2009). Collective psychological empowerment as a
model of social change: Researching crowds and power. Journal of Social
Issues, 65, 707–726.
Durrheim, K., & Dixon, J. A. (2001). The role of place and metaphor in racial
exclusion: South Africa’s beaches as sites of shifting racialisation. Ethnic and
Racial Studies, 24, 433–450.
Eckhardt, A. R. (1992). Sitting in the earth and laughing. London: Transaction
Publishers.
Fukuyama, F. (1992). The end of history and the last man. London: Penguin
Books.
 4  Violence: How Collective Shocks Transform Social Practices  127

Gurr, T. R. (1993). Why minorities rebel: A global analysis of communal mobi-
lization and conflict since 1945. International Political Science Review, 14(2),
161–201.
Gurr, T. R. (2000). Peoples versus states: Minorities at risk in the new century.
Washington, DC: US Institute of Peace Press.
Hobsbawm, E. J. (1994). The age of extremes: A history of the world, 1914–1991.
New York: Pantheon Books.
Huntington, S. P. (1996). The class of civilizations and the remaking of World
order. New Dehli: Penguin Books India.
Kaplan, R. D. (1993). Balkan ghosts: A journey through history. New York:
Picador.
Klar, Y., Schori-Eyal, N., & Klar, Y. (2013). The “Never Again” state of Israel:
The emergence of the Holocaust as a core feature of Israeli identity and its
four incongruent voices. Journal of Social Issues, 69, 125–143.
Maček, I. (2009). Sarajevo under siege: Anthropology in wartime. Pennsylvania:
University of Pennsylvania Press.
O’Brien, A. (2001). Against normalization: Writing radical democracy in South
Africa. London: Duke University Press.
Petersen, R. D. (2002). Understanding ethnic violence: Fear, hatred, and resent-
ment in twentieth-century Eastern Europe. Cambridge: Cambridge University
Press.
Reicher, S., & Hopkins, N. (1996). Self-category constructions in political rhet-
oric; An analysis of Thatcher’s and Kinnock’s speeches concerning the British
miners’ strike (1984–5). European Journal of Social Psychology, 26(3),
353–371.
Therborn, G. (1982). The ideology of power and the power of ideology. London:
Verso.
Zweig, S. (1944/1997). Die Welt von Gestern. Erinnerungen eines Europäers.
Frankfurt am Main: Fischer.
 5
Power: The Role of Leadership at
Critical Junctures

In one way or another we have addressed the issue of power repeat-

edly throughout the last two chapters. For while we may make analytic
distinctions between identity, violence and power in order to help lay
out our argument, and while our method of exposition so far has been
to put the focus on each term in turn, the core point is that the three
are systemically interlinked such that to address any one of them is
necessarily to invoke the other two. In the present chapter, then, power
moves from the background to centre stage. In so doing, the chapter
ties the various elements of our argument together, until we effectively
end up back at our point of departure—looking at the production of
identities.
In Chap. 3, we challenged those approaches which take identity as a
given, as fixed and as invariant, and instead developed an understand-
ing of identity as rooted in social practice. This then allows us to under-
stand how our sense of self shifts as the options and organisation of
social practice differ from place to place, and from time to time. But
equally, changing definitions of identity impact the nature and organ-
isation of social practice. When people define themselves as members of

© The Author(s) 2017 129

G. Elcheroth, S. Reicher, Identity, Violence and Power,
DOI 10.1057/978-1-137-31728-5_5
130  Identity, Violence and Power

a common social category, they are brought to act together on the basis
of common values, interests and goals. They are able to coordinate and
to support each other in reaching those goals. Identities, then, produce
social power.
Therein lies the reason why elites in particular spend so much time in
seeking to shape identities. Those who are in a position to define who we
are, what we value, what we desire and aspire to, and what we must do in
order to realise our aspirations, thereby create a world-making force and
put themselves in a position to wield it (see Reicher & Hopkins, 2001).
This makes the question of how one can make one version of categories
and identities stick, and marginalise all alternatives, a central one for both
theorists and practitioners.
In Chap. 4, our argument was that violence is one such means—and
a dreadfully efficient one. When violence divides people on the basis of
particular social identities, it becomes risky to act on the basis of any
other categories, sometimes even long after the violence has stopped. It
makes alternatives difficult in practice. Therefore, when the group is under
threat, it becomes easier for elites to eliminate dissent and demand that
the entire group rallies around them. Perhaps the most baleful example of
this is the decision of the 10th Congress of the Russian Communist Party
in 1921—in the midst of the civil war—to ban all factions and suppress
any opposition to the leadership. It is worth quoting from Lenin’s open-
ing speech to the Congress on 8th March:

discussion means disputes; disputes mean discord; discord means that

the Communists have become weak; press hard, seize the opportunity,
take advantage of their weakening. This has become the slogan of the
hostile world. We must not forget this for a moment … Our efforts
should be more united and harmonious than ever before; there should
not be the slightest trace of factionalism—whatever its manifestations in
the past. That we must not have on any account. That is the only condi-
tion on which we shall accomplish the immense tasks that confront us.1

1
 Retrieved on 28th June 2016 from https://www.marxists.org/archive/lenin/works/1921/10thcong/
ch01.htm.
 5  Power: The Role of Leadership at Critical Junctures  131

So, if it is true that identities are mobilised because they are the basis
of power, it might equally be true that the importance of violence in
producing and freezing identities is tied to the consequences in terms of
consolidating power.
In both of the previous two chapters, our argument has been premised
on the ways in which power—and particularly the power of elites—is
produced through violence and identity. Having shown this, in this chap-
ter we now look at the ways in which elites use their existing power in
order to organise, enable or incite violence, and thereby consolidate their
power for the future. Specifically, we discuss three ways in which leaders
can affect the occurrence of violence and use it for their own political
ends: by directly ordering violence, by avoiding measures to stop violence
or by creating a climate where violence appears unavoidable. While the
first two paths refer to the explicit power of leaders in an institutional
command structure, allowing them to create violent facts by commission
or by omission, the third path is special insofar as it refers to leaders’ sym-
bolic power. That is, it has to do with their capacity to invoke violence as
a plausible scenario and to do so in ways that make the actual occurrence
of violence more likely.
Before starting this discussion, it is important to identify two potential
traps. The first is the replacement of one form of fatalism with another:
to swap the claim that groups can’t help harming each other because it is
in their nature with the claim that leaders can’t help abusing their power
because that is in the nature of power. The second is to go to the oppo-
site extreme and to assume that, rather than being completely predeter-
mined, the exercise of power is completely undetermined, as if leaders
respond to each new set of events from scratch, with no preconceptions
or prior constraints. To avoid both traps, we will follow our discussion
of how leaders can promote violence with a consideration of the ways in
which long-term legacies both facilitate and constrain the ways in which
leaders can act. These histories provide a cumulative and large stock of
cards which leaders can choose to play (or not to play) when dealing with
events in the present. These choices are important, and leaders can take
different paths. But the stock of cards is still limited and so are the paths
down which leaders can take us.
132  Identity, Violence and Power

5.1 R
 epressive Power: Making Violence
Happen
When NATO started its bombing campaign against Serbia in March
1999, it led to the longest suspension of public protest against President
Milosevic for years. Far from turning a demoralised public against its
leader—as Western strategists (at least officially) expected—the bombs
actually brought the Serbian people “unity from heaven”, as the New
York Times ironically commented (quoted by Mandić, 2008, p. 25). This
unity did not simply express itself by a suspension of regime-critical col-
lective action in Serbia. On the contrary, Mandić identified an impressive
313 events (rallies, marches, riots, concerts …) during the 11 weeks of
the bombing campaign. Overwhelmingly, these were public expressions
of support for the domestic regime. Most importantly, these supportive
demonstrations were as frequent in municipalities where oppositional
parties prevailed as elsewhere. As Mandić (2008, p. 36) puts it: “The war,
it seems, suspended internal divisions and encouraged unified support
for the state, at least provisionally”. After this episode, which was the
only time during the 1990s when the war was actually fought on the ter-
ritory of Serbia proper, it took months before the opposition could again
organise effective mobilisations against the regime. Milosevic remained
in power until October 2000, benefiting from the one last moratorium
on challenges to his weakened regime which had been brought about by
the NATO intervention.
The Serbian public’s reaction to NATO bombings is far from an excep-
tion. As early as 1964, Nelson Polsby gathered anecdotal evidence for so-­
called rally effects in the USA. He concluded that, in times of war, conflict
and crisis, popular responses to the president are invariably favourable,
“regardless of the wisdom of the policies he pursued” (p. 25). Soon, this
claim was to be backed up by more systematic data (e.g., Mueller, 1970).
In 1978, Kernell showed how temporal fluctuations in US presidential
popularity from Truman to Nixon were systematically structured, among
other factors, by the wars the nation had fought. War entry typically pro-
voked substantive short-term increases in the president’s popularity—but
then this support progressively declined due to the negative effects of
mounting war fatalities on public support.
 5  Power: The Role of Leadership at Critical Junctures  133

For decades, these findings set the benchmark for the way social sci-
entists looked at rally effects. However, in 2001, the field was shaken by
the work of Baker and Oneal (2001). Using a broader set of variables
and more flexible techniques of data analyses, they dismissed the popu-
lar notion of invariable, spontaneous and almost mechanical rally effects
when a nation goes to war. Instead, they pointed to the critical impor-
tance of political communication. Rally effects are likely when the US
government actively prosecuted a foreign military campaign and when
presidential statements and/or prominent media coverage drew attention
to and supported the conflict.
Analyses of rally effects in the UK, conducted by Lai and Reiter (2005),
similarly suggested that the variability of public reactions to international
crises may have been strongly under-estimated. While these authors did
find substantial rally effects for both the Falkland and Gulf wars, they
failed to do so for the Korean, Suez and Kosovo wars, as well as for the
generality of non-violent crises in which Britain was involved. These find-
ings led Lai and Reiter to conclude that “rallies seem most likely and larg-
est after the nation has been clearly attacked or challenged and when vital
national values are at stake, although admittedly it is difficult to delineate
uncontroversially what is and is not the national interest” (p. 266). Their
conclusion therefore leaves open the key question of when ‘the national
interest’ is perceived to be under threat (and why support for the nation’s
leadership is perceived as an appropriate means of containing the threat).
In sum, the political pay-offs of conflict are not as reliable or as durable
as once thought. But nor are they negligible. With astute communication,
the outbreak of war can boost internal political support for two to three
months (e.g., Lai & Reiter, 2005). Whether that pay-off is sufficient is a
highly political question. A year after NATO bombed Serbia, Milosevic
was out of power, ending his days in a prison cell in The Hague. Did
the few months breathing space he was accorded in 2000 make any dif-
ference to his regime? Equally, did the few months after the invasion of
Iraq in 2003, in which the British public rallied around Tony Blair (as we
shall discuss in more detail in Chap. 8), count for anything compared to
the years of public outcry which eventually forced him to resign in 2007?
A tentative answer is that these periods, as fleeting as they might be,
are of particular importance because they provide leaders with formidable
134  Identity, Violence and Power

windows of opportunity to create new and irreversible facts, and that these
windows are of particular importance at times of social and political flux. In
a stable period, then, three months support for a lifetime of political exile
might seem a poor bargain. But as old regimes are disintegrating and new
entities are coming into being, a short time may be sufficient to alter the
course of history. Then, the rally effects of war may prove critical in allow-
ing political elites to create new social facts that pursue their own agenda.
In the Serbian case, Gagnon (2004) has argued that the different wars it
had been involved in during the 1990s bought the elite the time that they
needed in order to convert their privileged party position in a collapsing
socialist state to a privileged economic position in an emergent capitalist
state. In such a period of rapid transition, by the time the rally effects had
begun to fade, the old apparatchiks had emerged as the new entrepreneurs.
Just as short-term rally effects can prove valuable in times of transi-
tion, it is arguable that they are equally valuable in times of trouble.
That is, weakened leaders might be especially tempted to deliberately use
armed conflict to divert attention from their bad handling of state affairs,
increase their popularity, marginalise their opponents and generally
restore their chances of remaining in power. This diversionary war hypoth-
esis has received considerable attention among analysts of international
relations. Attempts to test the hypothesis have produced mixed findings,
however. While single cases can be found in which domestic political
motives seem to explain a rush to war, systematic comparative studies
suggest either that the magnitude of internal problems is unrelated to the
likelihood that a state will go to war (Levy, 1989, 1998), or that there is
a relationship, but it has little to do with a motivation to divert public
attention (Gleditsch, Salehyan, & Schultz, 2008).
However, there is one notable exception to this general (lack of ) pat-
tern: time-series analyses focusing on the USA have shown that the coun-
try did go to war significantly more often in periods where the incumbent
administration was economically unsuccessful or when the president was
losing popular support in public opinion polls (James & Oneal, 1991;
Ostrom & Job, 1986).
As disappointing as the lack of more consistent findings might be,
there are a series of rather simple reasons why seeking domestic support
through foreign war is an implausible phenomenon within the current
 5  Power: The Role of Leadership at Critical Junctures  135

interstate system. Given that positive rally effects are generally short-lived
and that they are generally overtaken by the negative impact of mounting
casualties, going to war against another state is only likely to bring very
short-lived political rewards to the elite, except when a rapid and suc-
cessful campaign can be expected. But very few states have the means to
attack another and count on a quick military victory, or have the political
and diplomatic resources to ensure that an attack will not provoke inter-
national sanctions or even military intervention against them.
If one adds to this the fact that even short-term rally effects are unlikely
in regimes whose leaders either have not been legitimated by general elec-
tions (Gelpi, 1997) or which are facing strong separatist tendencies (due
to the fact that much of the population may not see the conflict as their
own—see Chap. 8), then the range of state governments that are in a
position to expect rewards from diversionary interstate warfare becomes
remarkably small. The USA might be democratic, united and, above all,
militarily strong enough to be in that position, but few others are.
At this point it is worth emphasising that the arguments do not apply to
the diversionary war hypothesis in general, but specifically to diversionary
foreign wars. It is ironic, then, that most investigations of the hypothesis
are limited to such cases. One exception to this is the work of Tir and
Jasinski (2008). They have pointed out a number of good reasons why
weakened leaders might be much more tempted to direct diversionary vio-
lence against minority groups within a country, rather than against other
countries. First, there is generally no shortage of potential targets. Virtually
every country in the world has minorities and can invoke history to con-
stitute that group as a current threat of some sort. Ethnic primordialists
would probably agree with the statement that ethnic diversity and past
ethnic conflict are ubiquitous realities across all nations. Constructivists
might go even further and claim that the national ‘stock’ has no absolutely
defined boundaries, which implies that the range of groups in a nation is
theoretically infinite. All one has to do is invoke some arbitrary but share-
able marker (dress, accent, appearance ...) in order to identify a group as
such. Even the dullest of governments should have the wit to single out
some minority within the nation (in our current globalised world, immi-
grants always provide a handy candidate) and to associate them with some
kind of negatively laden collective experience within the nation.
136  Identity, Violence and Power

Second, the balance of power is almost inevitably in favour of the state.

In our contemporary world of nations, the default condition (and nor-
mative assumption) is that nation-states, not non-state groups, control
the material resources that are necessary to manage armed conflict.
Third, even if a minority finds a way to fight back, the imbalance in
political, diplomatic and symbolic resources generally gives the advan-
tage to states when it comes to the interpretation of the violent conflict,
domestically as well as internationally. Foreign powers are generally more
reluctant to intervene in a state’s ‘internal affairs’ than in interstate con-
flict. To the extent that armed resistance or rebellion can be branded as
‘terrorism’, a state engaged in internal armed struggle might even be able
to mobilise international support in its own favour.
Fourth, potential resistance to rally effects can be anticipated and coun-
tered by deliberately targeting dissenters as traitors or as ‘enemies within’.
Overall, then, a range of considerations support the contention that war
on internal enemies is more effective than war on external enemies as a means
of suppressing dissent and enforcing conformity. To those already listed we
can add one more, drawing on our theoretical discussion in Chap. 4:
intrastate violence doesn’t just create boundaries between people who
might otherwise join forces against state authorities, it also creates a bunch
of pretexts which require every citizen to behave as if the threat from the
minority group were real, and the need for state protection was pressing.
So much for the conceptual arguments. What about the evidence?
To test the domestic diversionary violence hypothesis, Tir and Jasinski
(2008) combined Gurr’s (1993, 2000) Minorities at Risk data set with
the Cross-National Time-Series Data Archive, for the years from 1996
to 2002—the only period for which Minorities at Risk comprised com-
plete enough data about the use of state force (defined in a wide sense
from limited repression to fully fledged military attacks) against internal
minorities. During this short period, the Israeli government used force
against Palestinians, the Turkish government against Kurds, the Russian
government against Chechens, the Spanish government against Basques
and the Thai government against Muslims, all in periods of economic
downturn and/or growing government unpopularity.
In an average year, across the entire world, more than one state in three
used force at least once against one of the minority groups inventoried
 5  Power: The Role of Leadership at Critical Junctures  137

by Gurr. The frequency of this phenomenon alone can be taken as an

indication of how easily state powers exploit their monopoly on the exer-
cise of violence within the boundaries of their jurisdiction. But this still
leaves the question of whether governments were more likely to use force
against minorities when they faced domestic difficulties.
They clearly were. Even when other factors known to impact the likeli-
hood of domestic violence were kept constant (the country’s overall level
of economic development, the government’s military resources, the tim-
ing of elections, the involvement in ongoing armed conflict, the demo-
graphic strength of minority groups), governments were still more likely
to use force against minorities in years during which their country was
experiencing either an economic downturn or political unrest, in the
form of protests, strikes or riots. Both factors had significant indepen-
dent net effects. That is, either economic downturn or political unrest
alone was sufficient to increase the risk of violence against minorities.
It is also noteworthy that the statistical effects are considerable but not
extreme in size: the risk of violence against minorities increased by more
than 50 % between the lowest and the highest observed values of either
economic downturn or political unrest. The effect is therefore too big to
occur by chance, but too small to allow for a deterministic interpreta-
tion of the phenomenon. Governments facing economic downturns or
political unrest will not inevitably target ethnic minorities to divert public
attention. Such violence is only one out of many possible strategies to
counter the erosion of political support and among those states that were
in such a situation between 1995 and 2002, many did opt for a different
course of action.

5.2 S
 tructural Power: Letting Violence
Happen
The use of overt state force against minority groups is the most direct
but not the only way in which calculations and decisions by state offi-
cials impact the occurrence of violence against minorities. Sometimes the
decision not to act can be just as consequential as the decision to act. The
critical question is not only to know when those who control repressive
138  Identity, Violence and Power

forces will actively intervene to create violence, but also when they will let
violence develop by choosing not to intervene in order to stop it. Studies
of so-called ethnic riots illustrate this point.
These riots are generally thought of as spontaneous outbursts of vio-
lence, fuelled by popular anger. When violent mobs appear to spontane-
ously attack, injure and kill members of different ethnic groups, this is
often used as unarguable evidence for deeply ingrained ethnic resent-
ments which lie at the base of collective violence. Certainly, it is largely
because ethnic riots are interpreted in this way that they create and freeze
ethnic oppositions for the long term. How could people continue to
engage with neighbours of different ethnicity once these people have
shown their ‘true colours’ by participating in bloody riots?
However, such a perspective overlooks three important aspects of riot-
ing. The first is simply that bloody riots are very rarely spontaneous, even
if described as such. Rioting, like any collective action that requires a
tight synchronisation of the behaviours of many individuals, presupposes
some degree of coordination and common background knowledge (e.g.,
‘they have committed an outrage against us’, ‘they have attacked us’)—
which begs the question of where that knowledge comes from and pro-
vides space for it to be manipulated.
Second, the application of an ethnic frame to a riot often occurs after
the event, and not during it. As Ramanathapillai (2006) has shown
with anti-Tamil riots in Sri Lanka, complex and chaotic events are ret-
rospectively simplified and structured throughout politicised memories:
“Particular memories are selected, kept alive, and retold as a collective
way of understanding and relating the experience (…) For example, at
the time of the riots the stories of Sinhalese atrocities were widely told,
yet the stories of Sinhalese protecting Tamils were not retold in Tamil
political narratives” (p. 4).
Third, and perhaps most critically, the assumed relationship between
the intensity of popular anger and degree of violence in ethnic riots is
largely unfounded. Whether people are harmed in such riots—and, if so,
how many—rarely depends on the size or determination of the crowd
itself. It is generally much more a function of how state forces manage
the situation. Such is the conclusion reached by Wilkinson (2007) after
extensive research into ethnic riots in many historic and contemporary
 5  Power: The Role of Leadership at Critical Junctures  139

sites, stretching from mid-nineteenth century anti-Catholic riots in the

North of Ireland to riots against minorities across post-communist States
in the late twentieth century:

In virtually all the empirical cases I have examined, whether violence is

bloody or ends quickly depends not on the local factors that caused vio-
lence to break out but primarily on the will and capacity of the government
that controls the forces of law and order. Abundant comparative evidence
shows that large-scale ethnic rioting does not take place where a state’s
army or police force is ordered to stop it using all means necessary.
(Wilkinson, 2007, p. 5)

The bulk of Wilkinson’s analytic effort then goes into disentangling the
relative contributions of ‘will’ and ‘capacity’ in the governments that
control repressive forces, with a focus on the states of twentieth-century
India. A first set of analyses led the author to rule out ‘capacity’ as a criti-
cal variable in this context:

Independent inquiries and newspaper investigations into the worst out-

breaks of Hindu-Muslim violence have found that in almost all cases local
police officers and magistrates had the forces available to prevent violence
(or could have quickly called them in) but that they failed to take preven-
tive action, either because of direct orders from their political masters or
because they feared retribution if they acted without first seeking political
approval. (Wilkinson, 2007, pp. 94–95)

That leaves ‘will’. Wilkinson’s analyses combine a variety of factors in

order to predict the number of Hindu–Muslim riots in each state of
India, for each month from 1961 to 1995. While controlling for a series
of factors—demographics, economics and also the occurrence of past
violence—ethnic riots were much more likely when state elections were
to take place in the ensuing six months compared to when there was
no ongoing electoral campaign. The risk of riots was reduced by a fac-
tor of four when the Communist party—which draws on electoral sup-
port across ethnic communities—was in power. The risk increased when
the Congress party, whose political support is concentrated within the
Hindu electorate, was in power. Furthermore, the risk of riots was only
140  Identity, Violence and Power

half as high in contexts split into multiple groups than in those split more
cleanly into two groups. Together, these various findings suggest politi-
cians were much more easily tempted to let ethnic violence occur when
the resulting ethnic divisions could bring them victory in upcoming elec-
tions. Conversely, they were more motivated to suppress potential vio-
lence when electoral success was contingent upon building cross-ethnic
coalitions with smaller minority parties.

5.3 S
 ymbolic Power: Inciting Others to Make
Violence Happen
There is a third way in which those who have (or contend for) political
power can influence the occurrence of collective violence. In addition to
deciding whether or not to order violence, or whether or not to let it hap-
pen, in their daily business they also face a seemingly much more trivial,
but sometimes just as consequential, choice: whether to invoke violence
or not. On 28 June 1989, Slobodan Milosevic chose ‘invoke’:

The lack of unity and betrayal in Kosovo will continue to follow the Serbian
people like an evil fate through the whole of its history (…) Six centuries
later, now, we are being again engaged in battles and are facing battles.
They are not armed battles, although such things cannot be excluded yet.
However, regardless of what kind of battles they are, they cannot be won
without resolve, bravery, and sacrifice, without the noble qualities that
were present here in the field of Kosovo in the days past. (Slobodan
Milosevic, 28 June 1989)

As always, the context of the speech is as important as its content. On

the 600th anniversary of the mythicised ‘battle of Kosovo’, Milosevic’s
party had bussed no fewer than a million Serbs to the historic battlefield
so as to commemorate the event. This unprecedented mass gathering was
probably meant to signify to each single participant how important the
event was to them as a whole. Whether or not the defeat previously had
any significance to those present, it now came to function as a com-
mon background knowledge that no one could ignore. According to the
 5  Power: The Role of Leadership at Critical Junctures  141

speech that they all heard together, the reason why their people had suf-
fered defeat 600 years ago was that there had been traitors in their midst.
The reason why they had to stand together now—literally and metaphor-
ically—was that the ‘lack of unity and betrayal’ that once brought an ‘evil
fate’ to them could under no circumstances be tolerated again.
United to face whom? Milosevic did not name current enemies, but in
the fourteenth century these had been foreign invaders of Muslim faith.
The medieval battle provides the context and pretext to refer to current
battles—and to specify that armed battles “cannot be excluded yet”. The
statement that “regardless of what kind of battles they are, they cannot
be won without resolve, bravery, and sacrifice” clearly signifies that Serbs
are expected to be prepared for all kinds of battles. Two years after the
speech, Serbs were fighting in Croatia against Croatian Catholics, the fol-
lowing year they were fighting in Bosnia against Bosnian Muslims and by
the end of the decade in Kosovo against Albanian Muslims. Milosevic’s
sinister prophecy had materialised more fully than the deepest pessimists
could have imagined in 1989. Would that have come to pass had there
not been a million Serbs—and beyond them, millions of people across
Yugoslavia and further afield—to hear the prophecy?
What we are saying here is that Milosevic was in the game of gen-
erating a ‘self-fulfilling prophecy’—a term initially coined by Merton
(1948), which he himself labelled the ‘Thomas theorem’, in homage to
early twentieth-century sociologist William I. Thomas: “If men define
situations as real, they are real in their consequences”. The first exam-
ple given by Merton to illustrate the theorem remains as timely in the
early twenty-first century as it was in the mid-twentieth century: a bank
becomes insolvent when its customers believe it is insolvent. Or, to be
more precise, insolvency results when a critical mass of customers does
what would be rational if the belief were to be true: they withdraw their
money from the bank. This behaviour will then lead others to believe that
the bank faces liquidity problems—or that it might be led to face them if
other people continue to behave this way—and hence to withdraw their
own money. At a certain point the assumption that the bank faces some
problems becomes true (no matter how false it might have been before).
Rumours can thereby create the realities they invoke, provided they
reach a critical mass of people and that there is a clear rationale for each
142  Identity, Violence and Power

of them to act as if the belief was true. Whenever rumours do not arise
by chance or accident, but because they serve the prophet’s interests, self-­
fulfilling prophecies can be seen as a process of effective mass manipula-
tion. In part, this effectiveness is due to the fact that it is hard to denounce
claims which seem to be proven true by the course of events:

The self-fulfilling prophecy is, in the beginning, a false definition of the

situation evoking a new behaviour which makes the originally false con-
ception come true. The specious validity of the self-fulfilling prophecy per-
petuates a reign of error. For the prophet will cite the actual course of
events as proof that he was right from the very beginning. (Merton, 1948,
p. 193)

When it comes to collective violence, however, there are two ways in

which the metaphor of the prophet can be misleading. A first problem
derives from the singular form. War-mongering policymakers are not
isolated prophets. Their prophecies become social facts only by being
repeated by those who share their interests or obey their orders. A second
problem is that policymakers do not just sit back, contemplate the world
and comment on it. Precisely because they act collectively, there can be
a division of labour between those who spread the word and those who
go further in making the prophecy become true. While one part of the
elite might refer to violence only in words, others might invoke it in more
practical ways—for example, by provoking alleged enemies into a violent
response.
Events preceding the start of the so-called second intifada vividly illus-
trate both points. Bar-Tal (2004) has documented how domestic public
opinion evolved among Israeli Jews in the period from the Camp David
peace talks in summer 2000 to the resurgence of armed conflict with
Palestinians by the end of the same year. In Israeli public discourse, the
failure of the peace talks was immediately portrayed as hard evidence for
lack of genuine will on the Palestinian side to reach an agreement with
Israel. But in summer 2000, the prophets of war had not yet achieved a
consensus around their version of reality. In public opinion polls, one par-
ticipant out of two still expressed support for the peace process. Despite
all that their elites said, 49 % of respondents still held firm and supported
 5  Power: The Role of Leadership at Critical Junctures  143

the Oslo agreement on 26–27 September. This figure would go into free
fall over the following days.
On 2nd October it had dropped already to 41 %. What had hap-
pened during the five days in between? On 28 September, Ariel Sharon,
then leader of the conservative opposition, visited the Temple Mount
in Jerusalem—a sacred site in both Muslim and Jewish traditions—and
provoked fury among Palestinian onlookers. Demonstrators started
throwing stones. The Israeli security forces responded forcefully. In
the following four days, riots took the lives of 45 people. Now the
political narrative of aggressive Palestinians could be supplemented
and seemingly confirmed by images of violent crowds. The notion
that Palestinians want to make war on Israel rather than make peace
appeared to be validated by the events. By 2002, after two years of vio-
lence, the proportion of Israeli Jews who still supported the Oslo peace
process had fallen to a mere 25 %.

5.4 Past Legacies and Present Crises

In this last example, we can see how a previous discourse (Palestinians
are dangerous) gains traction in the context of a new crisis (the outbreak
of the second intifada)—or, to put things the other way round, how the
existing discourse provides a frame for interpreting and responding to the
newer event. Here, the older and the newer refer to a short timescale of
months. But the same processes operate over much more extended tim-
escales, and indeed—thinking back to our discussion of past violence in
the previous chapter—the notion of Palestinians as dangerous after Oslo
fits into the much longer tradition whereby the Jewish people are invited
to see themselves as continuously under threat of destruction from their
foes.
In effect, the priming of public opinion rarely starts from a blank slate.
The architects of violence never invent the fears that they invoke from
scratch. They work with building blocks that are already available, and
have, in many cases, been available for a very long time. Let us therefore
complete this chapter with three examples that show how the exercise of
144  Identity, Violence and Power

power is embedded into multiple layers of history that span the years, the
decades and even the centuries.
Our first example concerns the political climate surrounding the pol-
icy that, arguably, has had the most far-reaching implications for the cur-
rent world order: the ‘war on terror’ led by successive US presidents. The
obvious turning point, which generated public support for this war, was
the attack on New York’s twin towers on 9/11/2001. For many, the events
of that day were quite sufficient to generate martial policies over the ensu-
ing years. But would these events have had quite so quick and so clear
an impact on US society and US policy had they not served to confirm
anxious expectations nourished well before?
In a public opinion poll conducted in 1986, 80 % of US citizens consid-
ered the danger of terrorism as ‘extreme’, and terrorism came out as people’s
prime concern (Zulaika, 2003). This was not only well before 9/11 but also
before the major terrorist incidents that preceded it—the bombing of the
World Trade Centre in 1993 and the Oklahoma bombing of 1995. Indeed,
only 17 people had been killed in terrorist attacks across the entire USA
over the previous five years compared to the approximately 150,000 people
who had been killed through non-terrorist violence over the same period.
To feed this concern, no less than 1322 books on terrorism were published
in the USA between 1989 and 1992 according to Zulaika (2003)—again,
before any major terrorist incident had taken place.
These figures illustrate that a ‘terrorist frame’ was highly available to
the US public long before mass terrorism actually struck the country. The
2001 attacks did not create this frame, but they validated it in the most
dramatic way. The subsequent funerals, remembrance ceremonies and
collective mourning then provided numerous opportunities to rehearse
the new patriotic script.
The rest of the story is well known: during the last months of 2001,
George Bush benefited from an outstanding rally effect, and his admin-
istration seized the opportunity to frame subsequent military campaigns
overseas, and restrictions of civil rights at home, as part of the ‘war on
terror’. When Bush announced the start of the war against Iraq to the
US public on 18 March 2003, he was able to refer to terrorism and so-­
called terror states as inescapable realities: “The cause of peace requires
all free nations to recognize new and undeniable realities (…) a policy of
 5  Power: The Role of Leadership at Critical Junctures  145

appeasement could bring destruction of a kind never before seen on this

earth”.
Our second example concerns the single most horrific outburst of
violence in recent history: the Rwandan genocide of 1994, and its
long-­term roots in colonial policies of divide and rule. In a remark-
able historic analysis, Mamdani (2001) traces back the long path from
colonialism to genocide. While acknowledging the fact that the geno-
cide was orchestrated by a small elite, as an extreme means of achiev-
ing their political ends in the present, Mamdani argues that looking
backwards is necessary to understand the responses of ordinary people
which were critical to its implementation: “If the violence from below
could not have spread without cultivation and direction from above,
it is equally true that the conspiracy of the tiny fragment of géno-
cidaires could not have succeeded had it not found resonance from
below” (p. 7).
In Rwanda, as elsewhere, European colonisers had granted a series of
privileges to a minority, as a tactic to divide the colonised and to stabilise
the power of the colonisers. In doing so, they not only relied on existing
ethnic groups, but also creatively constituted groups that echoed their
own racist ideology and which they used to legitimise social hierarchy
under colonial order. It was in this context that a myth concerning the
foreign ancestry of Tutsi people was created and associated with their
role in the colonial administration and society. The construction of an
ethnic minority group, and the combination of privilege and alien status
conferred upon it, encouraged the Hutu majority to re-direct the griev-
ances and the hostility created by colonial cruelty towards the local Tutsi
minority. This then laid the ground for future calamities, once power
relations shifted in favour of the Hutu, in the wake of decolonisation.
Mamdani explains:

To understand the logic of genocide (…) it is necessary to think through

the political world that colonialism set into motion. This was the world
of the settler and the native, a world organized around a binary preoc-
cupation that was as compelling as it was confining. It is in this context
that Tutsi, a group with a privileged relationship to power before colo-
nialism, got constructed as a privileged alien settler presence, first by the
146  Identity, Violence and Power

great nativist revolution of 1959, and then by Hutu Power propaganda

after 1990. (Mamdani, 2001, p. 14)

The third example takes us much further back into the past, to the
European medieval societies of the twelfth and thirteenth centuries.
Moore (2007) has identified this as a time when sporadic acts of perse-
cution were transformed into mechanisms that were a matter of policy,
were made systematic and became universally available. It was also a time
when a set of ideological justifications for persecution were developed.
In the process, the practice of persecution became largely independent
from the specific characteristics of any given targeted group, or from the
particular danger they were supposed to represent. The persecuting soci-
ety was born, with the “creation (…) of a single account of the victim
as enemy of God and society, which might be transferred at will to any
object, either as a class of persons already existing, such as Jews, whom
might seem desirable or convenient to persecute, or a new one, such as
sodomites or witches, which by an act of classification might be invented
for the purpose” (p. 160).
If it is independent of the characteristics of the victims, the phenomenon
has to be related rather to the motives of the persecutors and/or to the
functioning of society at large. The core of Moore’s argument is precisely
that the persecuting society came into being at the same point in time as
new elite groups arose, and sought to strengthen their power. It emerged
in a historic transition period, during which the relationships between
states and subjects were being redefined, and emerging bureaucracies con-
trolled ever more domains of people’s lives ever more closely. These societal
changes generated the need to justify more intrusive centralised powers, at
the same time as they generated a new class of people ready to do the job:

a new class of functionaries—clerics and courtiers, for whom persecution

might serve the twin purposes of providing the means to extend the power
and advance the interests of their masters, while consolidating their own
positions and undermining potential rivals. The systematic persecution of
minorities in European history had its origin in the interests and concerns
of this body of people, and not in the unregulated passions or prejudices of
the populations at large. (pp. 144–45)
 5  Power: The Role of Leadership at Critical Junctures  147

Obviously, it is not enough to refer to a tradition of persecution invented

in medieval times, or to the legacy of colonial brutality during the past
two centuries, to explain violent clashes in the twenty-first century. But
neither is it enough to refer to contextual changes in the present, deriv-
ing from the end of the Cold War and the growth of violent non-state
actors. The question is, when we face new times and new crises, how do
we make sense of them? Here, this historically stratified stock of explana-
tions constitutes a set of tools which facilitate sense-making and which
provide resources to elites in providing accounts that are instrumental in
expanding or protecting their influence.
To be more concrete, in the first chapter we referred to Brubaker and
Laitin’s (1998) argument that in the post-Cold War area, incentives for
ideological conflict frames have vanished and incentives for ethnic con-
flict frames have increased: trans-national diasporic communities have
replaced ideological blocs as reservoirs of support, and the institutionali-
sation of a new generation of human rights makes it more likely that a
plea for the respect of community rights will be heard and legitimised in
the international arena. In the process, ethnic frames have become highly
available. We have grown used to viewing the world through an ethnic
prism and, moreover, we expect others to do likewise. So ethnic histories
can be invoked as a resource by powerful elite groups as they seek to unite
the population around their leadership, claiming a common destiny and
interest with the majority (even if these do not share their interests) in
opposing a dangerous minority (who are not essential to their political
capital).
The key terms in this argument have to do with words like ‘invoked’,
‘tools’ and ‘resource’. History has to be brought into the present, and this
is done selectively as a means of advancing political projects. Moreover,
it is not entirely necessary to use accounts of the present rooted in the
existing historical repertoire. Human ingenuity and imagination always
allows us to come up with new ways of seeing. However, as with building
a house, if you have to create your own building blocks, it will necessar-
ily require more effort and take longer than using blocks that are already
available. So it will be a harder task, and in fast-moving periods of crisis,
time in particular may be of the essence. Those who use accounts that
148  Identity, Violence and Power

are more immediately intelligible and familiar will always start with an
advantage.
In sum, past legacies should be seen as part and parcel of the active
exercise of power. History does not entomb us, it does not condemn
us to repeating past mistakes, it does not substitute for agency. Rather,
invoking history is one of the principal ways in which people exercise
their agency.

5.5 Conclusion
In much analysis, antagonisms, hatreds and violence between people are
treated as errors, the result of inherent biases over which we have no
control, something we do not actively intend. Sometimes those biases
are attributed to the personalities of particular individuals, sometimes
to aspects of the psyche which we all have in common, sometimes to
histories we cannot escape. There is a plethora of such approaches for
which the worst of our actions derive from a fatal flaw in the human
condition—something we may all regret but for which no one can be
held accountable.
This chapter has challenged that viewpoint. We have started from the
premise that violence doesn’t ‘just happen’. It is made to happen. And we
look at the various ways in which elites can deliberately, knowingly and
systematically act to produce violence. Most directly, those who control
the state apparatus can use state forces to instigate violence, particularly
against internal minorities. Less directly, they can prevent state forces
from intervening when members of the majority take it upon them-
selves to attack the minority—thus both sending a signal that violence
is deemed permissible and making the expression of violence practically
possible. Least directly, leaders can incite the population to violence both
by words and acts—and most effectively by combining the two. That is,
certain groups can be portrayed as a threat and then provoked to anger
so as to create a consonance between discourse and reality. In laying out
these three paths to violence, we don’t suggest any priority between them
or indeed that in practice they can be clearly separated out. For instance
(as we will see in the case of India which forms the focus of the next chap-
 5  Power: The Role of Leadership at Critical Junctures  149

ter), one may both instigate violence in the population and then ensure
that the police stand by and let pogroms proceed.
This chapter has also addressed the tricky and nuanced issue of the
relevance of the past to the present—seeking to navigate the narrow
channel between historical determinism (past events are recapitulated
in the present) and historical amnesia (past events are irrelevant to the
present). Our argument has been that the past is important insofar
as it is made relevant to the present. This isn’t automatic: which past
events are made relevant and how they are interpreted is an active
choice.
So now, we have arrived full circle. Going backwards from the end
we return to the start of our argument: leaders use their power to create
violence because violence is a particularly effective way of freezing people
into identities, and because the construction of identity is the source of
political authority and power. We have chosen to outline our argument
in a particular order, starting with identity, then looking at violence and
finishing with power. But that is an analytic choice. In fact there is no
starting point or end point to the process. One could enter or leave at
any point, and each term is both input and output to the others. Or
at least that is true in principle. In different societies at different times,
identities will be more or less set or fluid, violence will be more or less
endemic, the power of elites will be more or less established. Equally,
then, the concrete political opportunities, pay-offs and risks associated
with mobilising along ethnic or cultural lines—and their perception by
elite groups—vary across time and place. This means that, if we want to
study them, we need to take context seriously and to examine the relevant
processes by reference to specific contexts.
Our conclusions, then, are not just theoretical but methodological as
well. We have got just about as far as we can by taking a general com-
parative approach to the analysis. To progress further we need to look
at particular cases and to develop new methods for the analysis of case
histories, which concentrate on critical junctures where the world is in
flux and can develop along several radically opposed paths. We need
to examine the role of violence in determining which path is taken.
We also need to examine the role of leadership in these processes. The
third part of the book will now be devoted to three such case studies,
150  Identity, Violence and Power

and to the development of a methodological approach suitable to their

analysis.

References
Baker, W. D., & Oneal, J. R. (2001). Patriotism or opinion leadership? The
nature and origins of the “rally “round the flag” effect”. Journal of Conflict
Resolution, 45(5), 661–687.
Bar-Tal, D. (2004). The necessity of observing real life situations: Palestinian-­
Israeli violence as a laboratory for learning about social behaviour. European
Journal of Social Psychology, 34(6), 677–701.
Brubaker, R., & Laitin, D. D. (1998). Ethnic and nationalist violence. Annual
Review of Sociology, 24, 423–452.
Gagnon, V. (2004). The myth of ethnic war: Serbia and Croatia in the 1990’s.
Ithaca, NY: Cornell University Press.
Gelpi, C. (1997). Democratic diversions governmental structure and the exter-
nalization of domestic conflict. Journal of Conflict Resolution, 41(2), 255–282.
Gleditsch, K. S., Salehyan, I., & Schultz, K. (2008). Fighting at home, fighting
abroad: How civil wars lead to international disputes. Journal of Conflict
Resolution, 52(4), 479–506.
Gurr, T. R. (1993). Why minorities rebel: A global analysis of communal mobi-
lization and conflict since 1945. International Political Science Review, 14(2),
161–201.
Gurr, T. R. (2000). Peoples versus states: Minorities at risk in the new century.
Washington, DC: US Institute of Peace Press.
James, P., & Oneal, J. R. (1991). The influence of domestic and international
politics on the President’s use of force. Journal of Conflict Resolution, 35(2),
307–332.
Lai, B., & Reiter, D. (2005). Rally ‘round the union jack? Public opinion and
the use of force in the United Kingdom, 1948–2001. International Studies
Quarterly, 49(2), 255–272.
Levy, J. S. (1989). The diversionary theory of war: A critique (pp. 259–288, Chap.
2). Boston: University of Minnesota.
Levy, J. S. (1998). The causes of war and the conditions of peace. Annual Review
of Political Science, 1(1), 139–165.
Mamdani, M. (2001). When victims become killers: Colonialism, nativism, and
the genocide in Rwanda. Princeton: Princeton University Press.
 5  Power: The Role of Leadership at Critical Junctures  151

Mandić, D. (2008). Myths and bombs: War, state popularity and the collapse of
national mythology. Nationalities Papers, 36(1), 25–54.
Merton, R. K. (1948). The self-fulfilling prophecy. The Antioch Review, 8(2),
193–210.
Moore, R. I. (2007). The formation of a persecuting society: Authority and deviance
in Western Europe 950–1250 (2nd ed.). Oxford: Wiley-Blackwell.
Mueller, J. E. (1970). Presidential popularity from Truman to Johnson. The
American Political Science Review, 64(1), 18.
Ostrom, C. W., & Job, B. L. (1986). The president and the political use of force.
The American Political Science Review, 80(2), 541.
Ramanathapillai, R. (2006). The politicizing of trauma: A case study of Sri
Lanka. Peace and Conflict: Journal of Peace Psychology, 12(1), 1–18.
Reicher, S., & Hopkins, N. (2001). Self and nation. London: Sage.
Slobodan, M. (1989, June 28). Milosevic’s Speech. Kosovo Field. Retrieved from
http://emperors-clothes.com/articles/jared/milosaid.html
Tir, J., & Jasinski, M. (2008a). Domestic-level diversionary theory of war:
Targeting ethnic minorities. Journal of Conflict Resolution, 52(5), 641–664.
Wilkinson, S. I. (2007). Explaining changing patterns of party-voter linkages in
India. In H. Kitschelt & S. Wilkinson (Eds.), Patrons, clients and policies
(pp. 110–140). Cambridge: Cambridge University Press.
Zulaika, J. (2003). The self-fulfilling prophecies of counterterrorism. Radical
History Review, 85(1), 191–199.
Part III
Case Studies
 6
Riots, Religion and the Mobilisation
of Communal Hatred in India
(with-Rakshi Rath)

For a number of years, two of us (Reicher and Rath), have, as part of a

larger research team, been studying one of the largest collective gather-
ings in the world—the Magh Mela at Allahabad in Northern India. This
gathering is an immense annual Hindu fair that lasts throughout the
lunar month of Magh (generally January–February). Large numbers of
pilgrims—mostly North Indians, predominantly elderly and from the
higher castes—come to the junction of the Ganges and Yamuna rivers
to live a devotional life. Even larger numbers converge to immerse them-
selves in the Ganges on the various auspicious ‘bathing days’ that occur
during the event. Throughout the day and night there are talks, dramas
and discussions for these devotees. But given the numbers, the Mela
attracts many other individuals and organisations. There are large shop-
ping areas. There is a funfair. Many NGOs attend to raise awareness of
issues from environmental pollution to child labour. Trades Unions hold
mass meetings. And Hindu nationalist organisations also pitch camp. Of
particular importance among them is the Vishwa Hindu Parishad (World
Hindu Council—VHP), a religious organisation close to the Bharatiya

© The Author(s) 2017 155

G. Elcheroth, S. Reicher, Identity, Violence and Power,
DOI 10.1057/978-1-137-31728-5_6
156  Identity, Violence and Power

Janata Party (Indian people’s party—BJP). Their audience consists pri-

marily of the day visitors, a cross section of the Hindu population. Their
aim is to draw on that identity, to define that identity and to politicise
that identity.
One day in 2006, we walked into the VHP tent. It was a huge space,
full of many exhibits. There were largely printed posters and charts, sup-
posedly demonstrating the high birth rates and demographic threats
posed by Christian and Muslim groups in various parts of the country.
But one set of 16 posters caught our attention. They were very graphic.
They were hand drawn and hung together. They clearly functioned as a
set and worked together to tell a story. What is more, many of the posters
had been daubed with sympathetic graffiti, showing that the audience
were engaging with the message and also how they were engaging.
The most obvious thing about the series was the presence of the cow
(depicted in 15 of the 16 posters) being subjected to various tortures
and humiliations. In many posters the animal was shackled, in some
cases stabbed, or scalded with boiling water or cut open, but most fre-
quently, its throat was being cut. The global impression was clearly one
of Hinduism—the cow has long been used as a symbol of Hindus and
Hindu identity—in deadly danger from a merciless foe.
One of the posters struck us by being particularly graphic. It was set
on the background of the Indian national flag and showed two figures,
a Muslim Arab and a Congress politician (i.e., a representative of India’s
first political party and main rival of the Hindu nationalists), both identi-
fiable by their dress. The former was holding a cow by its legs and pulling
it onto the symbolic Indian wheel of law (Ashoka Chakra), which had
been transformed into a knife-edged slaughter wheel. He was aided by
the politician, the latter pushing as the former pulls. In other posters, the
identity of the enemy varied but the complicity of Congress remained
constant. Congress politicians were depicted as either ignoring, colluding
or as actively benefitting from the acts of those who so cruelly attacked
the symbol of the Hindu nation. They either failed to protect the cow or
else, as in the Ashoka Chakra poster, actively took part in the slaughtering.
Whatever the case, they were always guilty of treason.
Two years later, we came again across the same dramatic triad composed
of the nation (represented by its sacred symbols or devoted servants), its
 6  Riots, Religion and the Mobilisation of Communal Hatred...  157

enemies and its traitors. This time, it was in a speech given in Mumbai
by the political leader of the BJP and the then Chief Minister of Gujarat,
Narendra Modi. This speech was given shortly after the city had become
the theatre of a series of train bombings on 11 July 2006 that left 189
people dead and several injured (for an account of the bombings, see Sayed &
Hakim, 2016).1 At one point in the speech, Modi asserted how he was
defeating Islamic terrorism and then quoted an anecdote comparing the
states of Gujarat and Assam, both with ‘Muslim’ neighbours (Pakistan and
Bangladesh), the former run by Modi and the latter by a Congress-led
coalition. He recalled his words to a local villager at a rally in Assam:

‘look brother, Assam’s neighbour is Bangladesh, and Gujarat’s neighbour is

Pakistan. My situation is the same as yours. They are your next-door neigh-
bours, and these are my next-door neighbours’. And I said, ‘but there is a
slight difference. You are tormented because of them, they are tormented
because of me’.

What makes this claim particularly powerful, of course, is the contrast

that Modi makes between himself and his political rivals. He defeats
the out-group enemy; the out-group enemy defeats them. Modi’s words
were met with cheers and claps from the audience. Some cried ‘well said’,
while others whistled.
At other moments of the speech Modi’s supporters went further.
When he declared that the response to terrorists should be ‘hard-hitting’,
the audience replied again with whistles and claps and with a cry of ‘drop
bombs on a Friday’, which was widely applauded. Friday, of course, is the
Muslim day of prayer. Bombs on a Friday imply an attack on the Muslim
faith itself.
This chapter will be about the posters presented at the VHP tent and
about Modi’s speech as two concrete examples of how communal vio-
lence is mobilised for partisan ends in the political and religious arena of
early twenty-first-century India. To this end, we shall start by sketching
out this context in a little more detail and then show how the posters

1
 See http://www.thehindu.com/news/cities/mumbai/all-you-need-to-know-about-711/arti-
cle7640887.ece.
158  Identity, Violence and Power

and the speech function within it. We shall then broaden our focus and
show how the very opposition between Hindus and Muslims so often
portrayed as timeless, natural and inevitable is in fact none of these but
rather arises out of and draws upon specific struggles rooted in India’s
colonial past.

6.1 T
 he Contemporary Scene: Politics
and Violence in Gujarat
In order to understand the role of religious violence and Hindu national-
ism in contemporary Indian society, it is critical to understand Modi’s
and BJP’s stronghold of Gujarat. And in order to understand the political
and social context in Gujarat, it is critical to understand the February
2002 riots.
In January 2001, the parliament of the VHP declared that the con-
struction of a temple would begin on 15 March 2002 on a highly con-
tested location in the Northern Indian city of Ayodhya. This had been
the site of the sixteenth-century Babri mosque, but, for the nationalists,
the mosque itself was built on an earlier Hindu sacred site, birthplace of
the god Ram. In 1992, a crowd of Hindu religious activists (kar sevaks)
had stormed the site and destroyed the mosque. The 2001 declaration
started a process whereby the VHP sought to mobilise kar sevaks once
more, this time to prepare the ground for a new building (International
Initiative for Justice in Gujarat, 2003).
On the 27 February 2002, one such group of volunteers was returning
home to Gujarat on the Sabarmati Express train. At Godhra station they
got into an altercation with Muslim tea vendors. Shortly after, the train
was brought to a halt outside the station and two train carriages were
set on fire. Somewhere in the region of 60 kar sevaks died. The details of
the event are highly controversial (see Nussbaum, 2007), but for Hindu
nationalists, the fire was deliberately started by a Muslim crowd.2
Over the following four days, there were attacks on Muslims in
19 districts of Gujarat. According to official figures 762 people were

 See http://www.rediff.com/news/2002/feb/27train3.htm Retrieved on 3 October 2016.

2
 6  Riots, Religion and the Mobilisation of Communal Hatred...  159

killed. Other estimates put the toll considerably higher, in the region of
2000–3000. Well over 100,000 people were displaced (Human Rights
Watch, 2002; International Initiative for Justice in Gujarat, 2003). The
violence was swift and extreme. Women-pregnant women in particular-
far from being spared, seemed to have been deliberately targeted (Amnesty
International, 2005; Human Rights Watch, 2003; Sarkar, 2002).
While most narratives present the Gujarat riots of 2002 as a response
to the train burning, there is evidence that much had been prepared in
advance. The Human Rights Watch report is blunt in its summary:

The Gujarat government chose to characterize the violence as a “spontane-

ous reaction” to the incidents in Godhra … (but) the attacks on Muslims
are part of a concerted campaign of Hindu nationalist organizations to
promote and exploit communal tensions to further the BJP’s political
rule—a movement that is supported at the local level by militant groups
that operate with impunity and under the patronage of the state. (2002,
p. 4)

The report then goes on to document the role of the police and of work-
ers and officials from Hindu nationalist organisations in various acts of
violence, such as the murder of the former member of parliament Ehsan
Jafri and some 64 others in Gulbarg Society, a Muslim neighbourhood of
Gujarat’s largest city, Ahmedabad. There is also evidence which points to
the organisation of violence prior to the 27 February. The attackers had
stockpiled weapons such as Liquefied Petroleum Gas cylinders and were
well resourced despite a general shortage. Muslim homes and Muslim
organisations had been marked out in advance (and in some areas, Hindu
houses were marked out with saffron symbols so as to be left alone).
The VHP made clear that they had lists of local Muslims and, in some
places, announced that certain areas would ‘burn’ months before they
did (Human Rights Watch, 2002; International Initiative for Justice in
Gujarat, 2003).
The nationalistic agitation that culminated in the Gujarat riots in
2002 was part of a broader movement that had shaken India over the
previous two decades and had shifted the domestic balance of power.
From the 1980s onwards, Hindu nationalist organisations like VHP
160  Identity, Violence and Power

and BJP attracted increasing support. Notably, the BJP rose from under
8 % of the national vote in 1984 to over 25 % in 1998. Moreover,
from 1998 to 2004, they were in power at the centre. Additionally,
the BJP was in power or else shared power in 14 different Indian states
including, significantly, Uttar Pradesh (the site of Ayodhya) in 1992
and Gujarat (the site of Godhra) in 2002. Alongside the rise of BJP
and allied organisations, there was a rise in communal violence: the
Muslim death toll of the 1980s was quadruple than that of the 1970s
(Ludden, 1996).
There are a number of explanations for the rise of BJP influence—and
Hindu nationalist influence more generally. One factor seems to have
been to do with changes in the political system which in turn have to
do with broader social changes. Ludden (1996) suggests that the old sys-
tem of political patronage, whereby local ‘headmen’ could deliver blocks
of votes to the Congress party, began to break down. In this context,
new ways of mobilising came to the fore, including attempts to mobilise
groups on an ethnic basis. Not only did this provide a space for the BJP,
VHP and others, but it also created a temptation for Congress and others
to act likewise. The stage was set for what has been called a ‘competi-
tive populism’, whereby different parties sought popular support through
street politics. To invoke Ludden again:

Using mass street mobilizations to topple governments became standard

practice, and organized Hindu activists turned their violence against
Muslims as part of their political strategy to dislodge the Congress. (1996,
p. 19)

As we discussed in Chap. 4, there is little doubt that communal agi-

tation was linked to electoral considerations. According to Wilkinson
(2004, 2007), hostility was much more likely in close constituencies
where it was critical to mobilise a majority constituency or in seats
where parties did not rely on minority votes (and had no need to enter
into coalitions with parties that do). Gujarat in 2002 was a clear case
in point. The BJP, which was in power, had no support from Muslim
voters in the state. For a variety of reasons, it was losing support in the
 6  Riots, Religion and the Mobilisation of Communal Hatred...  161

period 1999–2002 and was in danger of losing the state election due for
2003. Wilkinson concludes that:

At least some on the right seem to have calculated that communal tensions
and violence would reap electoral dividends for the BJP in the forthcoming
state elections. The result was that the state administration was at worst
highly partisan and at best inexcusably hesitant in preventing anti-­minority
violence and in its willingness to call in central troops and paramilitary
forces to do the job for them. (2007, p. 19)

In the Gujarat state elections shortly after the riots, which beforehand
were thought to be hanging on a knife edge, the BJP swept the board.
Overall, they won 126 seats with 49.8  % of the vote as against the
Congress Party, which won only 51 seats with 39.3 % of the vote. What
is more, a district by district breakdown of the vote shows that the BJP
fared better in areas where there had been riots. As Kumar (2003) con-
cludes, “the BJP may accept it or not, but the landslide victory for the
party in constituencies affected by the communal riots do suggest that
the polarisation between the Muslim and the Muslim voters did work
largely in favour of the BJP” (p. 272).
Two simple points emerge from this description of the Gujarat con-
text. The one is that antipathy is mobilised by portraying certain groups
(notably Muslims) as a dangerous enemy. This has been a characteristic of
Hindu nationalist organisations since their inception (Tambiah, 1996).
The other point is that antipathy is mobilised in order to gain politi-
cal advantage—polarising communities and validating the nationalists as
defenders of embattled Hindus. But these conclusions raise further ques-
tions. How precisely is it that certain others are constituted as a dangerous
enemy? When and why will this lead people to hate and destroy—and just
why should invoking such antipathies be a means of garnering support?
In order to move forward on these matters, let us turn now from a gen-
eral account of twenty-first-century Hindu nationalism to an analysis of
Hindu nationalist agitational materials. Specifically, let us look first at the
VHP posters at the Magh Mela and then consider the key speech given
by Modi just after the 2006 train bombings in Mumbai.
162  Identity, Violence and Power

6.2 H
 ow to Mobilise Intergroup Antagonism:
An Analysis of VHP Posters
Amongst the set of 16 posters we came across in the VHP tent, it was pos-
sible to distinguish a number of different types. Let us now consider these
in some detail, both how they differ and also how they work together in
telling an overall story (see Rath, 2016).
The first type of poster (of which there is just one example) portrays
India as a sacred Hindu territory. The poster depicts the God Krishna,
also known as the divine cowherd, emerging from the Himalayas into
an idealised landscape of rivers and plains. A headline proclaims that
‘the essence of India is the cow (or gau)’.3 The landscape is then made
up of multiple holy sites all of which are labelled with derivations of the
word ‘Gau’. Some of these sites are generic, like Gopuram, the intricately
carved towers at the entrance of South Indian temples. Some are spe-
cific places like the river Godavari which flows into the Bay of Bengal. A
speech bubble from Krishna draws the elements together. It reads, “the
term ‘gau’ is a symbol of sanctity, greatness and compassion in India. That
is why so many holy places in India start with the term ‘gau’.”
In this way, India is conflated with ‘Hindu’, and Indianness is conflated
with Hinduness (Hindutva). Certain groups are, by commission, sub-
sumed into the fold—thus, there is an image of the ‘Gautama’ Buddha
in the poster which represents Buddhism as part of Hinduism. Other
groups, notably Muslims and Christians, are excluded by omission.
In addition, Hindu India is represented as inherently virtuous. It
is a sacred space, a benign and bountiful space and a compassionate
space. Anything toxic must therefore be an imposition from the out-
side. Moreover, like desecrating a temple, those who threaten the land are
responsible for violating the sacred and destroying a site of virtue.
The second type of poster (of which there are three examples) depicts a
threat to India through images of violence done to the cow. In one poster,
the focus is on the threat itself. A cow is depicted tethered to a post, being
sprayed with scalding water. The headline reads, ‘Animals tortured before

 All original text is in Hindi, translations by the authors.

3
 6  Riots, Religion and the Mobilisation of Communal Hatred...  163

slaughter’. This is elaborated upon in smaller text: “The thirsty, hungry

and nearly dead cow is thrashed and dragged to the machine. The animal
is beaten incessantly with one leg shackled to the pulley. The boiling
water is unleashed onto the cow, so that its blood spreads quickly in the
body and softens the skin.”
In the other two posters, the focus is on what is under threat. That is, a
threat to the cow is not just a symbolic threat to Hindus (and not just an
attack on a symbol of Hinduism). In the one image, a cow with its throat
slit is accompanied by the words of a Hindu sage: “All spiritual activities
carried out on the land where even a drop of cow’s blood has been shed
become fruitless.” In the other, a cow is being stabbed in the back, its
blood flowing over a giant Indian rupee coin. The text reads, “Murder
of cattle is the murder of the country’s finances: As the murder of cattle
increased, there was a simultaneous increase in India’s poverty, expenses
and debts. India is the most debt-ridden country in the world.”
In other words, the threat is to all aspects of Hindu well-being, spiri-
tual and material (see Adcock, 2010 for a comparison with the arguments
of the early cow protection movement). In these posters, the source of the
threat is not made explicit although it is arguably implicit in some. Thus,
the slit throat is a sign of ritualistic killing for halaal meat as practised by
Muslims.
The third type of poster, by contrast, makes explicit just who is the
source of threat to cows and to the (Hindu) community. This is the
most frequent category, with nine examples. What is particularly strik-
ing here is the variety of sources that are identified. In some posters,
Western corporations are responsible. For example, in one, a plutocratic
figure holding a huge machete stands by the cow. The accompanying text
reads, “International conspiracy to destroy India’s natural food resources:
American Cattle Corporation advising India to slaughter 80  % of its
cows.” In other posters, Arab corporations are named as responsible for
large-scale killing in their mechanised slaughterhouses. And in yet others
a generic Muslim figure is identified as the perpetrator.
Of this last type, two examples are particularly graphic. One contrasts
the serene Hindu deity, Krishna, with two Muslim butchers. Krishna is
feeding a cow on grounds in the shape of India. Of the two butchers,
one sits on the cow as the other slits its throat. Both are grinning widely,
164  Identity, Violence and Power

clearly taking pleasure in what they do. The one says to the other, “She’s
the mother of the Hindus.” The headline denotes that this is meant as
a parable. It reads simply, “In the land of Gopal …” (another word for
Krishna, derived from Gau). The word ‘mother-fucker’ has been scrawled
as graffiti across the arms of the two butchers.
The other poster also depicts two Muslims standing over the prone
body of a cow—its legs bound and its throat slit. This time, though, one
of the figures (an attractive young boy) expresses concern, “Father, if it’s
really necessary to slaughter her, please speed it up. Look how the poor
thing is writhing in pain. I feel for her and her pain.” The father, a gross
figure with a large paunch standing with a bloodied machete in his hand,
replies, “Fool, stop talking like an infidel. If I do not torture her before
I kill her then, according to Islam, her meat would become haraam (for-
bidden) instead of halaal (edible).” A graffiti artist has again scrawled the
word ‘mother-fucker’ across his paunch.
In these posters, a peaceful, nurturant and virtuous India is beset by
the lascivious, sadistic Muslim other. Moreover, such sadism is not inci-
dental or accidental. It inheres in the very nature of Muslim culture.
Those who do not torture willingly and eagerly are not true Muslims.
This takes us to a fourth type of poster (of which, again, there is just one
example) which uses the historical figure of Shivaji to depict how enemies
should be treated. Shivaji was a seventeenth-century ruler from the cen-
tral Indian state of Maharashtra, who played a key part in the Hindu
nationalist imagination. He is represented as “a zealous Hindu warrior
fighting Muslim demons to create a Hindu nation-state” (Davis, 2004,
p. 1047), and his memory is sufficiently sacred that the translator of a
book (Laine, 2003) which simply reported historical controversies about
Shivaji’s parentage was attacked and publicly humiliated by a group of
Shiv Sena activists (a particularly militant Hindu nationalist grouping
whose name means ‘Army of Shivaji’).
In the poster, Shivaji stands between a cow and a Muslim butcher.
With his sword he slices off the butcher’s arm. Massive script at the top
of the poster proclaims, “Cow-killer deserves to be slain”. Lest anyone
misses the historical reference, the rest of the script reads, “Brave founder
of the Hindu kingdom, who chopped off the arm of the cow-murderer
butcher”. Note how, in contrast to the sadistic violence of Muslims,
 6  Riots, Religion and the Mobilisation of Communal Hatred...  165

Shivaji’s violent act is depicted both as justified (the butcher deserves his
fate) and as positive (it is brave).
Putting together the various elements that we have encountered thus far,
they constitute a powerful and coherent narrative which weaves together
many of the historical threads in Hindu nationalist thought. The starting
point is a definition of the in-group in terms of Hindu nationhood. This
then leads to multiple acts of exclusion: those who are either non-Indian
or non-Hindu are constituted as an out-­group. Muslims become strang-
ers in their own land. Next, these others are depicted not only as alien
but also as a serious threat. Because of their inherently vicious nature,
they are driven to destroy the well-being and indeed the very identity of
Hindu India.
We have seen how the Hindu Indian in-group is described verbally
in terms of sanctity and compassion. Equally, in images, there is a con-
trast between the way that outsiders (specifically Muslims) are depicted
as gross, salacious, physically repulsive and the way that Indians are por-
trayed as modest, ascetic and trim of body. In this way, the destruction
of such outsiders can be seen not merely as an act of self-defence but as
an act in defence of virtue. Violence is not just motivated (we must hate
them because they harm us) but justified (and by destroying them, we are
doing a good thing).
At first pass, it might seem paradoxical and contradictory to have a
series of posters in which, on the one hand, Hindus are defined in terms
of compassion and peacefulness, but, on the other, in which they are
enjoined to kill Muslims. In the context of the entire logic, however,
these two things are not contradictory. To the contrary, they actually
entail each other. Drawing a narrow definition of in-group boundaries,
thereby excluding sections of the population as out-groups, constituting
these out-groups as a threat and constituting the in-group as virtuous:
together these various elements lead irresistibly towards another, inter-
group antipathy and violence (see Reicher, Haslam, & Rath, 2008). They
flesh out the ‘how’ of mobilising intergroup hatred.
But there is something more. A fifth type of poster depicts another actor
in the drama. This is the Congress Party and the state that they created
after independence in 1947. There are six of these posters. Sometimes
Congress is represented on its own. In one instance, the symbols of the
166  Identity, Violence and Power

Indian constitution are accompanied by the number 350,000  in large

script, the images of slaughtered cows hanging out of each zero. The text
reads, “Before independence, India had 300 slaughterhouses for meat
export. In independent India, 36,031 slaughterhouses: large scale meat
export. Everyday 350,000 animals are killed cruelly.”
In another instance, the responsibility for this growing slaughter is
spelled out. A Congress politician (as denoted by his Nehru cap and
clothes) sits on a throne-like chair and addresses a poor peasant carrying
a small bowl: “To fight an election, one pot of milk from your cowshed
will not do. A suitcase full of notes is required. And that, I cannot get
from you, I will only get it from slaughterhouses.” At the bottom of the
poster, the text reads, “The government has given the green signal to
major mechanized slaughterhouses in the country.”
In other posters of this type, Congress is depicted alongside one of the
enemies that we have previously described. The Ashoka Chakra poster, to
which we alluded near the start of this chapter, is one of them. At the top
of this poster, large text proclaims, “Either change the national symbol or
shut down mechanized slaughter-houses and stop meat export. In smaller
writing, the message is elaborated: ‘the “Ashoka Chakra” is in the mid-
dle of our tri-coloured national flag, which is a symbol of ahimsa (non-­
violence). Anyone who insults this is punished. But the Government of
India by its cruel policies is insulting this symbol and committing crimes
of treason.” The graffiti on the image is equally explicit, if cruder. The
politician is daubed with “You killing a cow, what? Has your wife run
away with someone?” and “Send your daughter over to me”. That is, he
is accused of sexual inadequacy and threatened with sexual humiliation.
In the case of the Muslim figure, the words “meet me and I will kill you”
are inscribed, along with “I will kidnap your daughter, mother-fucker”.
In this case, humiliation is replaced by obliteration.
In another example, a corporate worker and a Muslim hold down
and kill a cow; the text proclaims that “the constitution has allowed the
slaughter of incapacitated animals”. Or again, a fat grinning Muslim
butcher holds aloft the dying calf whose throat he has slit while a farmer
holding a hoe and a bunch of grass looks away. Here the text explains
that “by naming the butcher’s cruel act as ‘farming’ the government is
insulting the holy ritual of farming and also the saint-like farmer” (note
 6  Riots, Religion and the Mobilisation of Communal Hatred...  167

how, here, the contrast between the venal other and the virtuous Hindu
is particularly explicit).
Thus, as well as showing us how intergroup hatred is mobilised, this
analysis of VHP posters goes some way to showing us why. Why, that
is, should invoking a communal threat be a means of garnering political
support? The important point here is that by identifying a threat and
proposing a response to that threat, a claim can be made to acting in the
interests of the group and hence deserving to represent the group (see
Haslam, Reicher, & Platow, 2011 for an analysis of how being seen to ‘act
for the group’ is critical to effective leadership). But, equally importantly,
by identifying a threat which ones rivals do not identify or counter, one
can claim that they are not acting in the interests of the group and hence
do not deserve to represent it. In many ways the use of ‘threat’ discourse
is as much, if not more, about disqualifying rival contenders to influence
as it is about achieving one’s own influence. It is about demobilisation as
much as mobilisation.
This might seem a strange claim to make about an organisation like
the VHP which is not a political party, which does not stand for elections
and which does not have leaders who personally need votes to achieve
office. So, to investigate in more detail the ‘why’ of mobilising inter-
group antagonism and to show how this can be used to bolster one’s
political influence, let us shift source from the VHP to the BJP politician
Narendra Modi and let us also shift modality from poster to speech.

6.3 W
 hy to Mobilise Intergroup Antagonism:
An Analysis of a Modi’s Speech
While attending the Dharm Sansad (religious parliament) of the VHP in
2008, we were handed a number of written documents and heard many
speeches. We were also given a video containing just one speech—origi-
nally delivered by Modi (then the Chief Minister of Gujarat, since 2014,
Prime Minister of India) on 17 July 2006 at Shanmukhanand Hall,
Mumbai. The written text and speeches included a set of proposals from
the President of the VHP, Ashok Singhal, on ‘Reinstating the Hindu
168  Identity, Violence and Power

nation’, ‘Hindu Unity’ and ‘Societal health and fitness’; Praveen Togadia,
General Secretary of the VHP, made a call to action and Kamlesh Bharti
of the Matrushakti (Motherpower) spoke about the perils facing Hindu
mothers. The only contribution by a party politician was Modi’s post-
Mumbai speech, ‘a challenge to terrorism’.4
The first thing to note is that all the speeches, Modi’s included, confirm
what our analysis of the VHP posters told us about the ‘how’ of mobilis-
ing antagonism. That is, they contain the self-same elements which build
up to the advocacy of violence as a defence of virtue. India is portrayed
as a virtuous Hindu space under grievous threat from many corners, but
particularly from a dangerous Muslim foe.
What distinguishes Modi is that his arguments are somewhat more
coded, using terms which are intelligible to an in-group audience but
which provide him with ‘deniability’ if accused of fomenting violence.
For a politician who may need to enter into coalitions in domestic poli-
tics and into alliances in international politics, this is critical (cf. Jaffrelot,
2013; Wilkinson, 2007).
Thus, for instance, where others declare that “This is Bharat (India)—a
Hindu Nation and it will remain a Hindu nation”, Modi states more
obliquely:

I, today, on the soil of Mumbai, have come with a very heavy heart. I can-
not imagine why, in a country like India, the innocent citizens of India are
thrown into the throes of death. What is the fault of those youth? What is
the fault of those mothers and sisters? Someone’s brother is snatched away,
someone’s beloved son is snatched away. A sister’s sindoor is wiped away.

The crucial word here is ‘sindoor’—the vermillion spot worn on the fore-
head uniquely by married Hindu women. Thus, Modi appears to talk
in inclusive terms about the ‘innocent citizens of India’ and in generic
terms about the mothers, sisters, brothers and sons of India. However, in
speaking of a sister’s sindoor, he indicates that his references to an Indian
people are limited to those who are also Hindu.

 The speech is available at https://www.youtube.com/watch?v=wkqLkcztKXw.

4
 6  Riots, Religion and the Mobilisation of Communal Hatred...  169

Equally, other speakers are quite open in declaring that Muslims are
a dangerous threat to the Hindu nation. Thus, Kamlesh Bharti of the
women’s ‘Motherpower’ (Matrushakti) wing of the VHP raises the spec-
tre of demographic domination through conversion and intermarriage.
She declares that if a Hindu woman marries a Muslim man, “her chil-
dren will become Muslims, and they will be known as enemies of our
Hindu nation”. Modi is also clear about threats to the Hindu nation. He
emphasises that as the Mumbai attacks demonstrate, there are threats to
life and limb. There are also threats to the economy, and the drugs trade
represents a threat to the health of the nation’s youth.
But when it comes to identifying the source of this threat, he is decid-
edly more circumspect. Ostensibly, the source is terrorism: armed ter-
rorism, financial terrorism, narco-terrorism. And while Modi stresses
that the terrorists are Muslims (“the world faces danger from Jihadi ter-
rorism”), he equally stresses that one should not conflate ‘Muslim’ with
‘terrorist’ (“every Muslim does not do jihad”). Nonetheless, this explicit
distinction is constantly undermined by insinuations that there is some-
thing about Islam which tends towards terrorism and makes everyone
into at least a potential terrorist.
For instance, Modi insists that “the terrorists aim to spread hatred in
the country. It is the responsibility of the citizens of the country not to
allow hatred to spread under any circumstances. Do not allow hatred to
spread under any circumstances. This is their aim.” Although he is speak-
ing in Hindi, he uses the Urdu term for aim—mansuba. Hence the ter-
rorist’s promotion of hatred is Islamicised and contrasted to the (Hindu)
citizen’s opposition to terrorism.
Or again, Modi raises the case of a young woman, Ishrat Jehan, who
was suspected (but not proved) to be a terrorist and was shot dead by the
police in his home state, Gujarat. He then observes that “just a while ago,
a TV reporter was saying that someone saw a girl planting the bomb. I
do not know what the truth is. Girls have been used.” In other words,
perhaps not all Muslims are jihadis, but any Muslim might be—females
as well as males, children as well as adults. From there, it is only a short
step to imply that all are—and need to be treated as—suspect. However,
Modi leaves it to the audience to complete this step, unlike others who
take it for them.
170  Identity, Violence and Power

Nonetheless, the major differences between Modi, the electoral politi-

cian, and the others come when we turn again from analysing from how
antagonism is aroused to why. Modi spends much more time and is much
more open in insisting that he, unlike his rivals, is of the people, with
the people and, above all, acting for the people. Very near the start of his
speech, Modi proclaims that:

I have not come to see the colour of the blood of those friends and sisters
who were martyred. They are all my brothers. What is his language? What
is his community? What is his attire? What is his faith? These have no
meaning for me. Each martyred brother is my Hindustani brother.

The force of this is not simply to constitute a unified and horizontal

Hindu community, where language, class, caste and sect make no differ-
ence. It is to cast himself as part of this band of brothers—an exemplary
in-group member.
Throughout the rest of the speech, Modi elaborates this claim, using,
as evidence, his stance on terrorism. He stresses that he is one of the peo-
ple, standing amongst them, not seeking to put himself above them. This
is illustrated with an anecdote about airport security, his refusal to accept
special treatment and his insistence on being screened (‘welcomed’) like
everyone else:

The police recognized me at the airport, and would usher me through. So

I used to refuse and say ‘No, I won’t go’. First welcome me. They circle the
metal-detector around, right? [He mimes the circular motion with his
hands, associated with the traditional welcome of Hindu guests] I used to
insist on the welcome and only then will I go. So that, I can proceed fur-
ther. This is very holy work.

Closely linked to this, Modi also stresses that he acts for the group, never
for himself (or for the enemies of the group), even if this is at own incon-
venience or even his own danger. He insists, “Friends, I have promised,
while there is life within me, I will search these merchants of death one
by one and sort them out.” He then describes how this brings him the
hatred of the terrorists, their allies and their apologists. But he retorts,
rhetorically and poetically:
 6  Riots, Religion and the Mobilisation of Communal Hatred...  171

Only those are afraid who die for their own image
I am that person who dies for India’s image
I do not care for my own image

Finally, Modi stresses that he does not just act for the group interest
but that he is effective in advancing the group interest. He does this in
various ways, notably by showing that he is prescient in identifying dan-
gers to the group. He tells a complicated story about going to the USA
in 1992 and explaining the dangers of terrorism, but being ignored—
then going back in 2003 and having the dangers of terrorism explained
back to him. More concretely, he outlines the tough anti-terrorist laws he
passed in Gujarat while others ignored the threat.
But overall, Modi spends more time on showing how his rivals are
not exemplary in-group members than showing how he is. Thus, the
anecdote about airport security starts with the observation: “Sometimes,
I am shocked, that these big leaders of ours, go to the airport, and if
someone dares to check them, then their eyes turn red. How dare they
check us? Do you not recognize us?” Such leaders clearly are not of the
people.
Equally, when Modi draws a contrast between himself, as someone who
works for India’s image, and others, who are only concerned with their
own, he clearly has Congress, the Communists and their allies in mind.
They do not act in the group interest. They do not pursue terrorism and
terrorists with sufficient energy. Indeed, he goes further: they actually col-
lude with terrorists if it is to their advantage. Modi illustrates this claim
with a number of examples. He claims that the Kerala Assembly, domi-
nated by the Left and by Congress, let all terrorists out of jail because
they wanted some as political candidates. He claims that in Tamil Nadu,
Congress entered into an alliance with a Muslim party which has terrorist
links. And he then tells a story involving the Congress-dominated United
Progressive Alliance (UPA):

Friends, the Bihar election. A supporting party of the UPA, in that elec-
tion, they used to roam around the election polls with a certain person.
Why did they roam? Because his face was like Bin Laden. He looks like Bin
Laden!—and this they showed to garner votes in Bihar!
172  Identity, Violence and Power

Further, Modi suggests that, even were they want to, his rivals (unlike
him) would be unable to advance the group interest. He refers to the
UPA Prime Minister Manmohan Singh as “a highly qualified doc-
torate who was leader of the World Bank”, competent in economics,
perhaps, but out of his depth when it comes to fighting terrorism.
He is weak, talking tough at home, but forgetting his commitments
once abroad. Indeed, Modi claims that the entire (non-BJP) political
elite are weak and fearful. “When they meet me personally”, he states,
“all these leaders speak the same words that I am speaking. The ones
that shout outside, when they come home, they pat my back.” This is
because as “the country’s leaders, they are scared to speak out against
terrorism”.
In sum, where Modi stands for and stands up for Hindu India, Singh,
his party and his parties allies do neither. This is all brought together in
the closing words of the speech. Modi first expresses his wish that the 200
dead souls in Mumbai will inspire 100 crore Indians (a billion, since a
crore is 10 million) to fight terrorism. Then he says,

I do not expect this from the Government at Delhi. Because even if the
soul tries, the soul will return broken hearted. That is why I am saying
this to the 100 crore citizens of this country. The country’s future has to
be decided by the country’s citizens. It is up to us to end this game
played by the merchants of death. We must do it. We will have to do it
together. And once again, I send my condolences to all the families who
were affected.

By raising the threat to the nation to the status of a crisis of survival,

the fact that it is not prioritised over all else by the existing govern-
ment becomes, for Modi, a means of establishing his own right to lead.
Modi and the BJP represent and defend the nation, defined as a Hindu
nation. Singh, the Congress alliance and the Left, either knowingly or
inadvertently, play into the hands of enemies who attack the nation in
the name of Islam. Those who stand for the group should receive the
group’s support when they stand for election. That, at root, is the ‘why’
of invoking intergroup antagonism.
 6  Riots, Religion and the Mobilisation of Communal Hatred...  173

6.4 A Bloodied Chain of Events?5

All this talk of mobilising antagonism, and of the ‘how’ and the ‘why’ of
such mobilisation, would suggest that it takes some effort to create hatred
and to turn hatred into violence. And yet a cursory glance at history
seems to suggest that such violence happens all too regularly. The Gujarat
riot may well be the most significant outbreak of the recent times, but
it is far from unique. In fact, it was just one incident in a long string of
violent communal confrontations. Wilkinson (2005) estimates that there
have been around 40,000 casualties in such riots since Independence
in 1947—a figure that itself is dwarfed by the catastrophic death toll
during the partitioning of India from Pakistan. It is estimated that up
to 1,500,000 people may have died in that process (Godbole, 2006).
Looking only at these numbers, it seems all too easy to believe that the
subcontinent is inescapably bound together across space and time by a
bloodied chain of violence.
Once we look for such a chain, the historic span can even be broad-
ened well beyond partition. Pandey (2006) notes how the colonialist
Government of India drew up lists of riots going back to the dawn of
British rule and there is evidence that goes even further into the past. He
remarks laconically that “the list of Hindu–Muslim riots in colonial and
pre-colonial India lengthens all the time with lengthening research—as
indeed it must if ‘riots’ are what one is looking for” (2006, p. 26).
Pandey’s point is precisely that the representation of riots—and vio-
lence more generally—as an indistinguishable and endless chain of
events makes it easier to see them as part of a natural order (or, rather, a
natural disorder). He analyses how colonial academics and administra-
tors reduced different events to a generic template of religious conflict
and how this was then described as the result of timeless Hindu–Muslim
hatreds deriving from the primitive and irrational religious fervour of
both groups. To quote the Rev. James Kennedy, who was writing about
his life and work in Northern India from 1839 to 1877:

5
 Thanks to  Sammyh Khan whose thesis on  Hindutva was  invaluable in  charting the  history
and ideology of the Hindu nationalist movement.
174  Identity, Violence and Power

The antagonism [between Hindu and Muslim ‘systems’], though generally

latent, every now and then breaks out into fierce strife, which but for the
interposition of Government would lead to civil war. (Cited in Pandey,
2006, p. 45)

What is revealing about this quotation is not just the representation of

riots as all equivalent and the explanation of these riots in terms of age-­
old rivalries. It is also the political message. The colonial power disappears
from the explanation of the conflict itself. It has nothing to do with creat-
ing the conditions for conflict. It has nothing to do with the incidents
themselves. It simply arrives afterwards to restore order: never part of the
problem, always the solution. To cite Pandey once more, “by the later
nineteenth century, it is no longer the power of English science and com-
merce, but also the argument that the ‘natives’ are hopelessly divided,
given to primitive passions and incapable of managing their own affairs,
that legitimizes British power” (2006, p. 45).
While it is nearly always fanciful to project the categories and conflict
of a particular place and time back into a timeless past, a number of
authors point out that anything more than an entirely cursory look at
history shows that it is particularly unconvincing to suggest that Hindus
have always stood together in antagonism to Muslims. For what makes
Hinduism distinctive is precisely its diversity and multiplicity. There are
multiple sacred texts—the Vedas, the Upanishads, the Bhagavad Gita,
the Puranas, to name but a few. There are multiple deities and multiple
accounts of each of these deities. Different sects lay different emphases
on these different sources, and as Jaffrelot (2013) suggests, rather than
thinking in terms of a single ‘ism’—Hinduism—it is better to think in
terms of a conglomeration of sects (Thapar, 1989). As Shani (2007) says:

The notion of a monolithic Hindu identity, no more than a homogenous

Muslim identity, is inherently implausible. Hinduism has been the bearer
of diverse theological interpretations. Hindus have been deeply divided as
much by caste as by ritual observance and sectarian differences. (p. 3)

Indeed, Pandey (2006) notes that the term ‘communalism’, which we

ourselves have used and which, nowadays, denotes hostility between
 6  Riots, Religion and the Mobilisation of Communal Hatred...  175

Hindus and Muslims, was earlier used to denote conflict between the
upper Brahman caste and non-Brahmans. The debate around the com-
munal question was centred as much on the so-called untouchables as on
religious minorities. Pandey’s argument is that the creation and consoli-
dation of differences between Hindus and Muslims was very much a cre-
ation of British colonialism. In part, this was a product of their orientalist
fantasies which saw primitive colonial subjects as locked into religious
fervour and which then led them to impose forms of governance which
treated people differently as a function of their religion.
In part, though, it also derived from indigenous reactions to colonial-
ism. Jaffrelot (2013) relates how many of the Indian intelligentsia became
fascinated by British success and by their achievements in scientific, tech-
nical, legal and social domains. Ram Mohan Roy, who in 1828 founded
the Brahmo Samaj (community of worshippers of Brahma, the creator)
was typical of these. He sought to unite and to revive indigenous society
on the basis of a reformed set of Hindu beliefs, claiming that there was
nothing inherently superior about Western Christian culture.
These ideas were further developed by Dayananda Saraswati and a new
organisation, the Arya Samaj (noble community), which he founded in
1875. But the Arya Samaj also took Hindu revivalism in important new
directions. Critically, it formed a link between culture, people and land.
In Jaffrelot’s (2013) terms, the noble people of the Vedas formed the
autochthonous people of ‘Bharat’, the sacred land below the Himalayas.
This incipient link between the original Indians and the Hindu text was
fateful. If nothing else it suggested that those who were not ‘people of the
book’ (even if there is some debate over which Hindu book this is) are
not properly Indian.
But also, while Hindu revivalism had always been torn between admir-
ing and fearing the British, Dayananda altered the balance increasingly
towards fear and threat of the ‘other’. The Arya Samaj became particu-
larly concerned with the conversion of Hindus to Christianity and even
invented reconversion ceremonies in order to combat it. But Dayananda
and his organisation also began to alter the balance in another way—that
is, in terms of defining the identity of the ‘other’ who is the source of
threat. Whereas the British and their Christian religion had been and
still remained a focus of concern, now Muslims began to loom larger in
176  Identity, Violence and Power

the revivalists’ imagination. This is well illustrated by the cow protec-

tion movement, one of the most significant initiatives of the Arya Samaj.
From its beginnings, it drew on the cow as a Hindu symbol and on cow
slaughter as an attack on Hindus. As Adcock (2010) argues, Dayananda’s
pamphlet from 1881 out of which the movement arose, The Ocean of
Mercy, centred on the notion that cows bring economic good, their milk
sustains the physique and intellect and, hence, killing cows leads to dis-
ease, unhappiness, poverty and powerlessness.
When it came to those responsible for such attacks, the British cer-
tainly figured prominently. They both consumed beef directly (one of the
demands of the movement was for an end to the supply of beef to the
British Army) and their laws allowed for others to consume beef. These
others were mainly Muslims. More specifically, the ritual slaughter of
cows at the end of the festival of Eid-ul-Adha (a practice known as qur-
bani) became a focus of contention. Once again, hostility to Muslims and
to the colonisers was intermixed. The first recourse of the cow protection
societies was to call on the authorities to give legal protection to the cow
as an object of religious worship. But once this failed, attacks on Muslim
communities ensued, most notably in 1893 and again in 1917. Pandey
(2006) translates some of the leaflets—patias—which circulated in order
to mobilise such attacks:

The religion of the cow is being destroyed. What crime has she committed
that she should be killed by non-believers. Hindu brothers are entreated to
watch over the cow in every village and every house. If they do not, the cow
will sadly breathe its last and disappear from the village. If you see a
Musalman with a cow, it is your duty [‘religion’—the word dharma stands
for both] to take it from him. It is also your duty [‘religion’] to write and
send on five patias. If you do not, you bear the sin of cow-slaughter. If you
do, it is equivalent to the gift of five cows. (p. 283, notes in brackets in the
original)

Other leaflets are yet more explicit, both about what should be done and
about the implications of not doing it. One declares that “you must loot
the houses of the Musalmans and kill the Musalmans”. It asserts that
“those who are Hindus have no choice” and says of anyone who demurs
 6  Riots, Religion and the Mobilisation of Communal Hatred...  177

that “you do mount on your daughter, drink your wife’s piss, and mount
on your sister’s daughter” (pp. 284–285). In other words, exclusion and
violence are not simply the fate of the out-group, it is also the fate of dis-
sident in-group members. The declaration that Hindus have no choice is
not just an exhortation, it is a threat. Indeed one of the leaflets is explic-
itly aimed at those Hindus who shield Muslims in their houses. It tells
“Hindu brothers” that “all of you must turn out with your weapons”
(p. 284).
There is a point here that is so obvious that it almost seems superflu-
ous to mention it. Although, perhaps that very obviousness thereby leads
the point to be overlooked. That is, if Hindus were inherently inclined
to hate Muslims, they would not need such exhortations and such dire
threats to make them join in the violence. It becomes far simpler and
more plausible to argue (as we did in Chap. 5) that the violence does not
derive from preformed communal identities but rather serves to disci-
pline people into communal groups and to cohere around communalist
leaders. From this perspective, the argument that violence is natural is
not simply an innocent mistake. Rather it is a deliberate device for cover-
ing one’s tracks and denying one’s complicity in inciting violence.
This amalgam of representation, explanation and politics in accounts
of communal riots serves to link the colonial era to the closer past. At its
meeting of mid-March 2002, the Rashtriya Swayamsevak Sangh (RSS—a
volunteer paramilitary Hindu nationalist organisation deeply involved in
the events of Ayodhya and Gujarat) adopted a resolution which stated
that “the reaction to this murderous incident in Gujarat was natural and
spontaneous. The entire Hindu society cutting across all divisions of
party, caste and social status reacted” (cited in Varadarajan, 2002, p. 21).
Or in the words of the VHP leader Pravin Togadia, “wherever there is
Godhra, there will be Gujarat” (ibid., p. 23).
Togadia continued, “In Gujarat, for the first time there has been a
Hindu awakening and Muslims have been turned into refugees. This is
a welcome sign and Gujarat has shown the way to the country.” That
is, the murderous rioting was not only naturalised, as in colonial times,
but actively celebrated. The difference between eras reflects the different
relationships to the groups involved in conflict. Whereas the British por-
trayed themselves as neutral arbiters, holding the ring and k­ eeping order
178  Identity, Violence and Power

between Hindus and Muslims, the RSS, Togadia and others in the Hindu
nationalist fold represent themselves as champions of the aggrieved
Hindu masses. So whereas the colonial narrative accords no precedence
to one side or the other in the dynamics of conflict, and stereotypes both
as equally violence-prone, the Hindu nationalists see the action as always
having an origin in the perfidious Muslim ‘other’. For Togadia and oth-
ers, Godhra was an organised conspiracy, provoked by the Pakistani secu-
rity services, whereas the subsequent riots were a spontaneous and natural
reaction.
In both the colonialist and the contemporary cases, then, a primordial-
ist or otherwise fatalistic account serves to remove responsibility for con-
flict from political actors. But whereas in the former case it justifies rule
of the civilised west over the primitive east, in the latter case it justifies
championing the innocent Hindu over the rapacious Muslim. In other
words, the accounts themselves are far from neutral or innocent. They are
not simply accounts of a political process, but they are an integral part
of the political process in terms of creating and of legitimating particular
types of rule.

6.5 Conclusion
In this chapter, we have examined the example of so-called communal
violence in India, combining secondary sources with primary analyses
of religious drawings and political discourse from the Hindu nationalis-
tic movement. All elements analysed converge on one point: the notion
that communal violence reflects ancient hatreds between Muslims and
Hindus simply does not stack up (just as the idea of ancient hatreds fails
to stack up in other sites we investigate in the book, notably the Balkans).
Not least this is because people have not seen themselves in terms of such
categories throughout history, and when they have, they have not always
seen the relationship between the categories in terms of opposition or
antagonism.
Where we find communal hatred and where we see instances of com-
munal violence, it is because communal categories have been brought
into being at a particular time and place. The riots we have described
 6  Riots, Religion and the Mobilisation of Communal Hatred...  179

are not atavistic outbursts. Antagonism has been systematically propa-

gated and violence has been systematically organised, identifying targets,
providing weapons, ensuring that the forces who might prevent them
either stand aside or else actually join in. In short, communal violence
is deliberately mobilised. If this is clear, it is equally clear who is driv-
ing this mobilisation in contemporary India: the Sangh Parivar or family
of Hindu nationalist organisations—some more crudely, some in more
coded terms, but the different members of the family complementing
each other.
This then leads to the question of why such ideas are being propagated.
Both in the posters and in the speech that we have analysed, we meet a
curious phenomenon. These different texts are ostensibly about the out-
group threat. But in both, the source of that threat is variable—Muslims,
Westerners, Arabs, corporations, on the one hand; armed terrorists, eco-
nomic terrorists, narco-terrorists, on the other hand. But what remains
constant in both is another actor, the in-group rival who, by omission or
by commission, is responsible for allowing the threat to flourish.
So what appears to be about intergroup dynamics may be better
understood in terms of intra-group dynamics. If there is a dire threat to
the in-­group, then clearly those who experience the threat, are concerned
by the threat and are able to deal with the threat are those who are of
the group, act for the group and deliver for the group (the three com-
ponents of leadership and effective influence over group members—see
Haslam, Reicher, & Platow, 2011), while those who do not experience
the threat, are not concerned by the threat and do not deal with threat are
not of the group, do not act for the group, do not deliver for the group
and hence lose leadership and influence over group members. To put it
slightly differently, invoking a threat which one’s rivals do not recognise
is a powerful way of shifting the balance of support within the group and
hence of achieving both electoral power and the power of influence. This,
of course, begs the question of why people should accept discourses of
threat and of hatred.
Here, the role of violence in turning discourses of out-group threat
into the lived experience of out-group threat becomes pivotal. In the case
of the Gujarat riots, once Hindu nationalist groups had turned Muslim
homes into sites of danger and co-opted their Hindu neighbours in the
180  Identity, Violence and Power

process, at the least by getting them to mark their own homes so they
wouldn’t be attacked, Muslims fled into their own enclaves in fear of
‘Hindus’ in general. The monolithic categories ‘Hindu’ and ‘Muslim’ and
the antagonism (or, at least, mistrust) between them became a material
reality. In line with the arguments we developed in Chap. 4, violence,
far from reflecting ancient hatreds between groups, was a most effec-
tive means of creating and freezing such group relationships. And, as our
argument went on in Chap. 5, it created the constituencies for the ideo-
logues of antagonism to represent and to draw upon to build their power.

References
Adcock, C. S. (2010). Sacred cows and secular history: Cow protection debates
in Colonial North India. Comparative Studies of South Asia, Africa and the
Middle East, 30(2), 297–311.
Amnesty International. (2005). Justice, the victim—Gujurat state fails to protect
women from violence. London: Amnesty International.
Davis, R. H. (2004). Review: Shivaji: Hindu King in Islamic India. Journal of
the American Academy of Religion, 72, 1045–1050.
Godbole, M. (2006). Pandora’s box no one wants to open. Economic and Political
Weekly, 41(12), 1179–1184.
Haslam, S. A., Reicher, S. D., & Platow, M. J. (2011). The new psychology of
leadership: Identity, influence and power. London: Psychology Press.
Human Rights Watch. (2002). We have no order to save you. State Participation
and Complicity in Communal Violence in Gujarat 14(3) (C).
Human Rights Watch. (2003). World Report: Events of 2002.
Jaffrelot, C. (2013). Refining the moderation thesis. Two religious parties and
Indian democracy: The Jana Sangh and the BJP between Hindutva radical-
ism and coalition politics. Democratization, 20(5), 876–894.
Kumar, S. (2003). Gujarat assembly elections 2002: Analysing the verdict.
Economic and Political Weekly, 38(4), 270–275.
Laine, J.  W. (2003). Shivaji: Hindu king in Islamic India. Oxford: Oxford
University Press.
Ludden, D. (1996). Contesting the nation: Religion, community, and the politics of
democracy in India. Pennsylvania: University of Pennsylvania Press.
 6  Riots, Religion and the Mobilisation of Communal Hatred...  181

Nussbaum, M.  C. (2007). The clash within: Democracy, religious violence and
India’s future. Cambridge: Harvard University Press.
Pandey, G. (2006). Routine violence: Nations, fragments, histories. Standford, CA:
Stanford University Press.
Rath, R. (2016). Virtuous violence: A social identity approach to understanding the
politics of prejudice in intergroup relations. Unpublished Ph.D. Thesis,
University of St. Andrews.
Reicher, S., Haslam, S. A., & Rath, R. (2008). Making a virtue of evil: A five-­step
social identity model of the development of collective hate. Social and Personality
Psychology Compass, 2(3), 1313–1344. doi:10.1111/j.1751-9004.2008.00113.x.
Sarkar, T. (2002). Semiotics of terror: Muslim children and women in Hindu
Rashtra. Economic and Political Weekly, 37(28), 2872–2876.
Sayed, N., & Hakim, S. (2016). Six minutes of terror. Gurgaon: Penguin Random
House India.
Tambiah, S. J. (1996). Leveling crowds: Ethnonationalist conflicts and collective
violence in South Asia. Berkeley: University of California Press.
Thapar, R. (1989). Imagined religious communities? Ancient history and the
modern search for a Hindu identity. Modern Asian Studies, 23(02), 209–231.
The International Initiative for Justice in Gujarat. (2003). An Interim Report
(p.  6). Retrieved from http://onlinevolunteers.org/gujarat/reports/iijg/
interimreport.htm
Varadarajan, S. (2002). Gujarat, the making of a tragedy. India: Penguin Books
India.
Wilkinson, S. I. (2004). Votes and violence: Electoral competition and communal
riots in India. Cambridge: Cambridge University Press.
Wilkinson, S. I. (2005). Communal riots in India. Economic and Political Weekly,
40(44/45), 4768–4770.
Wilkinson, S. I. (2007). Explaining changing patterns of party-voter linkages in
India. In H.  Kitschelt & S.  Wilkinson (Eds.), Patrons, clients and policies
(pp. 110–140). Cambridge: Cambridge University Press.
 7
Ethnic Violence in the Former
Yugoslavia: From Myth to Reality
(with Sandra Penic)

During the early post-Cold War era, to much of the Western public, the
former Yugoslavia soon came to epitomise a representation of a world of
ethnic rivalry and primitivism abruptly revealed by the lifting of the Iron
Curtain. Tito’s rule over Yugoslavia has sometimes been depicted as akin
to the proverbial ‘lid on a cauldron’, eventually removed by Tito’s death
and the mounting weakness of the regime during the 1980s. International
media depictions of war in the former Yugoslavia have certainly played a
pre-eminent role in entrenching and popularising the notion that ethnic
conflicts are somehow part of a natural course of events. These simple
depictions contrast with the complexity that local populations experi-
enced, especially during the early war period. Where the battles were
fought in 1991 and 1992, things typically looked far less black and white.
There are many instances of collective behaviour that did not fit into
the ethnic hatred narrative, and sometimes there was active resistance to
ethnic categories of action imposed by politics and warfare. In Chap. 4
we already referred to the way the fighting parties in the trenches around
Sarajevo only gradually came to understand themselves and their enemies

© The Author(s) 2017 183

G. Elcheroth, S. Reicher, Identity, Violence and Power,
DOI 10.1057/978-1-137-31728-5_7
184  Identity, Violence and Power

as ethnically defined ‘Serb’ and ‘Muslim’ forces. But even once that was
the case, there were still times when people transcended the ethnic
trenches that separated them. Here is an example, narrated by a former
Sarajevan soldier:

The first years of the war I spent several months defending Sarajevo in a
trench that was only 50 meters from the Serbian army trenches. Between
us was an unmined, level meadow (…) After several nights of us listening
to and watching the enemy trench, one morning a man’s voice called out
from the other side, and astonished us, ‘Hey, you guys, let’s play a round of
soccer on the meadow! (…)

We played soccer with them every day. If someone had seen us at that time
they would have probably said we were insane. But looking back at that
time from the vantage point of today it seems to me we were more sane
than most people.

The quote is part of a much wider corpus of testimonies collected by

Svetlana Broz (2005, 2014) during and after the war in Bosnia, which
brings together a hundred cases where people defied members from ‘their
own’ ethnic community to protect, support or cooperate with people
from a different ethnic background. Gagnon (2004) reviewed broader
evidence of collective behaviour that was not consistent with the ethnic
conflict narrative and highlighted three such types: (a) mass desertions—
more than 200,000 young Serbs preferred to hide or flee rather than
to serve, when they were called to save their ‘ethnic brothers’ in Bosnia
from what Serbian mass media propaganda portrayed as ‘genocide’, (b)
anti-­immigrant sentiment against incoming refugees from the same ethnic
background, which was equally strong in Zagreb, Belgrade and Sarajevo;
and (c) intraethnic violence, which was systematically used as a politi-
cal weapon, notably between Serbs in Krajina and between Croats in
Herzeg-Bosnia.
These observations led Gagnon to dismiss the notion of ethnic war and
refer to it as a ‘myth’. While we largely concur with Gagnon’s refutation of
ethnic conflict as a satisfying explanation for the tragedies that unfolded
in the former Yugoslavia, our interest here lies more in understanding
7  Ethnic Violence in the Former Yugoslavia: From Myth to Reality...  185

how the myth turned into reality. How did empirically unfounded, but
politically instrumental, beliefs trigger social processes that eventually
resulted in a situation where many people did experience ethnic conflict
as real in its consequences?
Changes in the social acceptance of mixed marriage provide telling
evidence concerning the scope of the phenomenon. In 1990-less than a
year before the war started in Croatia, and then in Bosnia-a representative
sample of adult residents of Yugoslavia were asked, in the Yugoslav Public
Opinion Studies (YPOS), whether they could conceive of marrying into
the various ethno-national communities that composed Yugoslavia. In
our re-­analyses of these data, we calculated for each republic or province
the rate of majority members who expressed a positive attitude towards
marrying someone from the main minority group, or vice versa. The first
column of Table 7.1 shows findings for young adults in 1990. These find-
ings reveal a strong norm of acceptance of mixed marriage in the most
diversified parts of Yugoslavia: in Bosnia, Croatia and Vojvodina, a large
majority stated their willingness to marry across ethnic boundaries.
Ten years after the end of war in Bosnia and Croatia, the same ques-
tion was asked again to respondents from the same generation, as part
of the TRACES project. This survey was conducted in 2006 among a
representative sample of the cohort of people born between 1968 and
1974 across all countries of the former Yugoslavia (Spini, Elcheroth &

Table 7.1  Rates of acceptance of mixed marriage before and after the war, across
eight political entities (sorted by decreasing level of pre-war ethnic diversity)
Would you be willing to marry into the group of … ? (% Yes)
18–25 in 1990 (%) 32–38 in 2006 (%)
(Source: YPOS) (Source: TRACES)
Bosnia and 67 17
Herzegovina
Vojvodina 95 65
Montenegro 38 59
Croatia 63 22
Macedonia 14 6
Kosovo 12 3
Serbia 41 39
Slovenia 59 49
(Former) Yugoslavia 42 27
186  Identity, Violence and Power

Fasel, 2011). It is striking to see how in post-war Croatia and Bosnia,

the acceptance of mixed marriage had collapsed. In Croatia, the rate of
positive answers declined from 63 % to 22 %, in Bosnia from 67 % to
17 %. While mixed marriage seemed something normal to most young
adults in before the war, the war itself transformed social norms. Only
a tiny minority could still imagine marrying across ethnic lines a decade
after the end of combat.
In the following pages, we will proceed in two steps. First, we will
elaborate our portrait of Yugoslavia on the eve of war, present more
findings that corroborate the notion of the ‘myth’ of ethnic conflict, but
also try to understand the processes through which the myth became
a social fact. That will bring us back to the question of how wartime
social practices, and adaptation to wartime dilemmas, transform col-
lective identity (see Chap. 4). Second, we will address how, when and
why exclusive ethno-national identities and loyalties persist even after
the guns have fallen silent. To that end, we will focus specifically on one
of the new nation-­states to emerge from the wars—Croatia. This will
allow us to understand how the combination of triumphant nation-
alism and politicised threat was used to entrench the status quo and
silence critical voices.

7.1 H
 ow Exclusive Ethno-national Identities
Became Social Facts
Responses to another question asked in the 1990 YPOS are instructive
about pre-war perceptions of social cleavages in the population. People
were asked to choose from a list in order to indicate where the great-
est intergroup inequalities lay in Yugoslavia. Only in Kosovo, which had
already been the theatre of ethnic riots and repression in the early 1980s,
did respondent put inequalities between ‘people from different nationali-
ties’ (which in Yugoslav terminology meant people from different eth-
nic backgrounds) in first place. In all the other republics or provinces,
ethnicity was only quoted as the fourth or fifth most important source
of inequality, ranking behind differences between ‘individual republics’,
7  Ethnic Violence in the Former Yugoslavia: From Myth to Reality...  187

‘political magnates and ordinary citizens’, ‘rich and poor’ and, in the
economically less endowed central and southern republics, between
‘employed and unemployed’ people. Thus, overall, people were more
concerned with political, economic and social inequalities, and with
inequality created by the federal system of distribution of wealth, than
with ethnic inequalities (Table 7.2).
But did widespread dissatisfaction with the federal system lead to
popular demands that it should be broken up? The YPOS data sug-
gest that it didn’t. In Slovenia, Croatia and Bosnia—which all declared
independence in the following two years—the most frequent call was
for more federalism, whereas only a relatively small minority supported
the separatist option (see Table 7.3). In Macedonia and Montenegro, the
most frequently expressed wish was actually to maintain the status quo.
Ironically, it was only in Serbia, which would subsequently fight for the
maintenance of the Yugoslav federation, where most people went for the
break-up option.
These historic survey data hence appear to contradict the notion that
there were strong bottom-up forces leading to the inevitable dissolu-
tion of the Yugoslav federation or to struggle between its constituent
­ethno-­national communities. But the most interesting findings arise from
a question where people were directly asked to assess the quality of rela-
tions between different national groups, across various contexts. In seven
out of eight political entities, people described these relations as on aver-
age better than satisfactory, as far as their own places of work and living
were concerned. (The exception was again Kosovo, where average answers
were located between ‘satisfactory’ and ‘bad’.) Strikingly, however, when
the same people were asked about the quality of ethnic relations in their
republic or province overall, their judgements were less favourable, and
when they were asked about Yugoslavia as a whole, their answers even
ranged between ‘bad’ and ‘very bad’ (Table 7.4).
There is thus a conundrum here. While everywhere except in Kosovo
people viewed interethnic relations within their personal sphere of expe-
rience as positive, the same people were convinced that elsewhere these
relations were far less rosy. This discrepancy raises the intriguing question
as to which other sources people relied on to form an opinion about the
Table 7.2  Perceived sources of inter-group inequality in pre-war Yugoslavia
Where do you personally think most inequality in Yugoslavia exists? (% who chose cleavage among up to two different answers)
(Source: YPOS 1990)
Between Between Between Between Between
Between political people of Between manufactural Between males believers
individual magnates and Between different employed and and non- young and and
republics ordinary rich and nationalities unemployed manufactoral and old females atheists
(%) citizens (%) poor (%) (%) (%) professions (%) (%) (%) (%)
Bosnia– 41 33 24 23 26 12 8 7 4
Herzegovina
Vojvodina 37 42 32 23 25 11 9 3 1
Montenegro 40 31 27 20 39 13 8 5 2
Croatia 28 37 34 27 16 15 8 4 5
Macedonia 33 35 31 22 30 13 9 4 1
Kosovo 27 31 12 66 22 4 2 4 1
Serbia 39 39 27 23 26 13 8 5 2
Slovenia 32 29 33 19 11 18 10 6 4
Yugoslavia 35 35 28 27 24 13 8 5 3
7  Ethnic Violence in the Former Yugoslavia: From Myth to Reality...  189

Table 7.3  Popular support for future institutional scenarios in pre-war Yugoslavia
In your opinion, what kind of state should Yugoslavia be in a near future?
(Source: YPOS 1990)
More Different
centralist Status More sovereign
(%) quo (%) federalist (%) states (%)
Bosnia–Herzegovina 3 37 43 16
Vojvodina 4 38 30 28
Montenegro 1 54 27 18
Croatia 1 20 63 16
Macedonia 3 54 31 12
Kosovo 2 10 49 40
Serbia 1 33 28 37
Slovenia 1 9 68 23
Yugoslavia 2 31 43 24

Table 7.4  Perceived ethnic conflict in pre-war Yugoslavia

In your opinion, what is the quality of relations between different nations…?
(average response on 4-point scale: 1-good, 2-satisfactory, 3-bad, 4-very bad)
(Source: YPOS 1990)
…in your …in the place …in your
working where you republic, …in
organisation live province Yugoslavia
Bosnia- 1.5 1.4 1.9 3.1
Herzegovina
Vojvodina 1.3 1.2 1.7 3.4
Montenegro 1.5 1.5 2.1 3.4
Croatia 1.5 1.7 2.5 3.1
Macedonia 1.4 1.4 2.1 3.3
Kosovo 2.6 2.9 3.5 3.4
Serbia 1.3 1.3 2.5 3.5
Slovenia 1.5 1.6 1.9 3.3
Yugoslavia 1.5 1.6 2.3 3.3

state of interethnic relations, and what led them to lend more credit to
alarming second-­hand information than their own daily experience.
The findings shown in Table 7.1 already suggest that the discrepancy
between what people experienced and what they believed about inter-
ethnic relations actually prefigured a dramatic social change—a violent
190  Identity, Violence and Power

process through which people’s beliefs about ethnic tensions (and separa-
tion) became true. The findings reported in Table 7.5—which combines
responses from the YPOS sample in 1990 and from the TRACES sample
in 2006—provide insights into what happened to the sense of identity
among those who were in their formative years when Yugoslavia broke
up. The average response patterns show how Yugoslav identity become
obsolete. Interestingly, to many in the former Yugoslavia, affiliation to
Yugoslavia was seen as more important than any other (infra- or supra-
national) affiliations. That was still true in 1990 for young Bosnians,
Vojvodinians, Montenegrins and Macedonians. While Slovenes consid-
ered affiliation with their republic, and Kosovars with their province, as
more important than affiliation to Yugoslavia as a whole, mixed patterns
emerge for the two largest republics: affiliation with Croatia or Serbia was
seen in the corresponding populations as of equal importance to affilia-
tion with Yugoslavia.
A completely different picture emerges from the 2006 data: by
then, identification with the former Yugoslavia had lost its impor-
tance virtually everywhere. Only in Montenegro and Macedonia did
it still pass the threshold of ‘somewhat important’ on average. While
the obsolescence of Yugoslav identity was nearly ubiquitous, it was
replaced by different types of identity in different contexts. In those
republics which, after having declared war and (in the case of Croatia)
fought for their independence, had achieved in 2006 the status of
indisputable and internationally recognised nation-states, identifi-
cation with the new nation clearly prevailed. That was the case in
Croatia, Slovenia and Macedonia. In most other places, where the
confusion of political transition, state dissolution and war had not
yet been replaced by the certainties of a new triumphant nationalism,
more complex patterns arose. These generally involved simultaneous
identification at both the local (‘the place and region where I live’)
and global (‘Europe’/’the World’) levels.
As part of an interdisciplinary research programme, we had aimed to
document how war experiences in the 1990s transformed collective iden-
tities much more profoundly than the combination of a deep economic
depression, institutional disintegration and aggressive mass propaganda
in 1980s. The resulting findings have been compiled in an edited book
Table 7.5  Affiliation to different territorial entities, before and after the war
There are some types of affiliation that may be important for an individual. Please indicate the importance that you
attribute to each.
(reversed average responses on a 5-point response scale: 5-very important, 4-quite important, 3-somewhat important,
2-of little importance, 1-not important)
(Sources: YPOS 1990, TRACES 2006)
Place and region Republic/Province / (former)
where I live country Yugoslavia Europe World
Bosnia and Herzegovina 18–25 in 1990 3.0 3.1 4.0 3.9 3.6
32–38 in 2006 3.9 3.6 2.6 4.0 4.1
Vojvodina 18–25 in 1990 2.9 3.1 4.0 3.7 3.5
32–38 in 2006 3.3 3.3 2.7 3.7 3.8
Montenegro 18–25 in 1990 3.1 3.5 4.2 3.1 2.9
32–38 in 2006 4.0 3.2 3.5 4.2 4.2
Croatia 18–25 in 1990 2.8 3.2 2.9 3.3 3.3
32–38 in 2006 3.9 4.1 1.5 3.1 3.1
Macedonia 18–25 in 1990 2.8 3.0 3.5 3.1 2.6
32–38 in 2006 4.0 4.3 3.2 3.9 4.1
Kosovo 18–25 in 1990 3.4 3.9 3.3 4.0 3.6
32–38 in 2006 4.4 4.5 1.7 4.3 4.1
Serbia 18–25 in 1990 2.8 3.3 3.6 3.3 3.2
32–38 in 2006 3.4 3.4 2.5 3.8 4.0
Slovenia 18–25 in 1990 3.5 3.7 3.0 3.3 2.9
7  Ethnic Violence in the Former Yugoslavia: From Myth to Reality... 

32–38 in 2006 3.7 3.9 2.5 3.5 3.5

Former Yugoslavia 18–25 in 1990 3.0 3.4 3.5 3.5 3.3
32–38 in 2006 3.8 3.8 2.3 3.7 3.8
191
192  Identity, Violence and Power

(Spini, Elcheroth & Biruski, 2013). They provide concrete insights into
the wartime dilemmas and constraints that left many people with a new
sense that ethnic affiliations matter. The bulk of these findings are based
on a survey completed in the spring of 2006 by a representative sample of
more than 6000 adults from all over the former Yugoslavia. Respomdents
completed life events calendars in which they reported retrospectively on
their experiences during the war (Spini et al., 2011).
On the basis of these data, the sociologists Gauthier and Widmer
(2014) were able to show how, for most people, the ‘decision’ to leave
the most war-affected regions in Croatia, Bosnia and Kosovo was directly
related to their religious affiliation. It is important to be precise here:
these decisions were influenced by people’s ties to particular religious cat-
egories that functioned as markers of ethnic identity and that made it
possible to distinguish among Orthodox Serbs, Catholic Croats, Muslim
Bosniaks and Albanians. They were not determined, however, by the
actual importance of religious identity in people’s lives: those who left
did not believe more strongly or practise their religion more frequently.
In another contribution, the demographers LeGoff and Giudici (2014)
have shown how, while in 1990 one out of seven marriages was ‘mixed’
(and the norm of mixed marriage was actively supported by institutional
policies), mixed unions had almost disappeared from the statistics by the
early 2000s. This was a combined consequence of three different pro-
cesses: separation, emigration and conversion (i.e., the re-labelling of the
identity of one of the partners).
These findings reveal that most people, when they encountered the
dilemma of remaining or leaving a war zone, and sometimes even their
spouse or family, acted as if ethnic identity was critical: they assumed that
those who might attack or else protect them would rely on markers of
identity to differentiate between foes and friends (even if religion was not
important to them in private). More generally, these findings provide a
sense of how many people were led to make difficult choices in an ambig-
uous environment: to leave everything behind, or to stay and comply
with the logic of combat. In this process, the cumulative consequences
of invidious individual choices then changed the environment in which
others made sense of the collective situation. As more and more people
 7  Ethnic Violence in the Former Yugoslavia: From Myth to Reality...  193

began to take account of their own ethnicity-and the way their ethnicity
was perceived by others-in making decisions, so yet others were led to
think and act in ethnic terms. The acts of friends and neighbours were as
important in validating an ethnic frame as official propaganda. Once col-
lective behaviour had created new social facts, old norms became obso-
lete, and new norms of ethnic separation began to stabilise and acquire
prescriptive value.

7.2 K
 eeping the Myth Real: Ethnic
Nationalism in Post-War Croatia1
On 15 April 2011, the Croat general Ante Gotovina was convicted by
the International Criminal Tribunal for the former Yugoslavia (ICTY)
and sentenced to 24 years of imprisonment. The verdict was based on
war crimes and crimes against humanity committed ten years earlier,
during so-called Operation Storm. For Serbs, Operation Storm was an
act of ethnic cleansing which forced nearly 200,000 of their compatriots
to flee. For Croats, it was a heroic battle which won the war and led to
the creation of their own sovereign nation-state.
Gotovina’s sentence provoked a massive outcry across Croatia. Nearly
all of the mainstream Croatian media portrayed the sentence as unjust,
and this attitude was shared by the overwhelming majority of Croats.
According to a poll conducted immediately after the verdict, 95 % of
Croats perceived the verdict as unjust and 88 % of them still saw Gotovina
as a hero (Jutarnji List, 2011). A year later, the Appeals Chamber of the
ICTY, by a tight majority of three to two, overturned the previous verdict
and acquitted Gotovina on formal grounds. The verdict was perceived
as highly controversial internationally and was criticised in unusually
blunt terms by the two dissenting members of the Appeals Chamber.
Judge Agius described the verdict as “confusing and extremely problem-
atic”, Judge Pocar as “contradicting any sense of justice” and “grotesque”

 This section is adapted from Penic, Elcheroth, and Reicher (2016). Readers interested in more
1

methodological detail might refer to this more extensive presentation of the study.

194  Identity, Violence and Power

(Clark, 2013). But in Croatia it was met with widespread celebrations,

which left no space for dissenting voices. It was widely interpreted as a
full exoneration not only of Gotovina as an individual, but of Croatian
actions during the war as a whole. Even the most-read left-leaning news-
paper in the country, Jutarnji list, reacted to the acquittal by devoting the
entire front page of its website to the headline “The war is over: Croatia
is innocent”.
The impression arises that, when it comes to the ‘Homeland War’,
two decades after the end of fighting public opinion in Croatia remains
monolithic. It is as if the rally effect never stopped. How does such a
near-unanimous political climate perpetuate itself? Where are the critical
voices—or what has happened to them?
According to a flourishing psychosocial literature, “critical patriots”
would normally tend to act as nations’ moral consciousness (Roccas, Klar, &
Liviatan, 2006; Schatz & Staub, 1997; Staub, 1997). Even when they
are only a minority within a nation, as long as they are consistent and
vocal, they still retain the ability to mobilise support for their claims and
provoke social change. However, the data at hand show that in Croatia,
at least among the generation who were young adults during the war,
critical patriots are difficult to find.
These data are taken from a subset of the TRACES survey data already
referred to above. From the larger data set, we focus here on subsam-
ples from Croatia and, for comparative purposes, from Serbia. The most
striking finding from this quantitative analysis concerns the differences
between these two national subsamples.
First, there are differences in the incidence of different modes of attach-
ment. In particular, critical attachment, while being the least frequent
mode in both samples, is much rarer in Croatia than in Serbia. In order to
estimate the proportion of the population in each country who adopted
the different modes of identification, we performed a cluster analysis on
a joint sample of Serb and Croat respondents with two entry variables:
national attachment and glorification. Attachment to the group was mea-
sured with the Doosje, Ellemers and Spears (1995) identification scale,
which includes four items (e.g., “I identify with other Croats/Serbs”; “I
am glad to be a Croat/Serb”). National glorification was assessed with the
7  Ethnic Violence in the Former Yugoslavia: From Myth to Reality...  195

following three items from a larger National Identity scale (Corkalo &
Kamenov, 2003): “There are many more capable people in my nation
than in others”, “In all historical conflicts with other nations my nation
was always right” and “A good member of my nation should not associate
with our enemies”.
Table 7.6 summarises the results of the analysis, which display three
different types of profile: (a) high national attachment with high glori-
fication: glorifiers; (b) high national attachment with low glorification:
critically attached; (c) low/average on both attachment and glorification:
detached. As can be seen from the table, the Serb sample is roughly evenly
split between the glorifiers (estimated at 40.9 % in the corresponding
population) and the detached (45.2 %), with a significant minority of
the critically attached (13.9 %). By contrast, the Croats are dominated
by the glorifiers (67.5 %), with a smaller group of the detached (25.8 %)
and a tiny group of critically attached people (6.7 %)—less than half as
large as in Serbia.
Second, there are differences in the relationship between modes of
attachment and collective guilt. Collective guilt acceptance was measured
through a five-item scale developed by Branscombe, Slugoski and Happen
(2004). Typical items on the scale are “I feel regret for my group’s harm-
ful past actions toward other groups” and “I can easily feel guilty for the

Table 7.6  Estimate of the population share for three types of identification in
Croatia and Serbia; with group means of attachment, glorification and collective
guilt acceptance
Croatia Serbia
Critically Critically
Detached attached Glorifiers Detached attached Glorifiers
Population 25.8 % 6.7 % 67.5 % 45.2 % 13.9 % 40.9 %
share
Attachment 3.66 5.96 6.05 3.31 5.95 6.00
0.76 0.64 0.72 1.03 0.70 0.79
Glorification 3.08 1.82 3.72 2.56 1.74 3.42
0.93 0.40 0.70 0.86 0.44 0.63
Collective 3.42 2.72 3.06 3.24 3.89 3.10
guilt 1.25 1.26 1.30 1.32 1.37 1.34
acceptance
196  Identity, Violence and Power

bad outcomes brought about by members of my group”. In the Serbian

sample, the findings match the pattern found in previous research. That
is, the critically attached show the highest level of collective guilt accep-
tance. However, in the Croatian sample, the pattern is very different: lev-
els of collective guilt acceptance amongst the critically attached are much
lower than in Serbia, and the highest level of collective guilt acceptance
is found amongst detached respondents (see Table 7.6). What counts in
Croatia, it seems, is whether one is attached to the nation, not how one
is attached.

7.2.1 W
 hy Is There No Space for Critical Patriots
in Post-War Croatia?

So why should there be such a difference between nations? What are the
critical elements of context which produce these differences? Why, more
specifically, is critical patriotism so scarce in Croatia and why aren’t criti-
cal patriots willing to accept the war guilt of the nation?
Perhaps the most obvious answer would be because, from 1991 to
1995, the war was fought on Croatian but not on Serbian territory.
As a consequence, the suffering was far greater for ordinary Croats
than for Serbs. Indeed several Croatian regions suffered tremendous
destruction and loss of life. However, while there is evidence that
victims of war are particularly reluctant to blame their own group
for past wrongdoings (Corkalo Biruski & Penic, 2014), the same
is not true for the broader communities from whom these victims
are drawn. Moreover, there is contrary evidence from comparative
studies showing that in the communities and regions most affected
by the war, the condemnation of war-related crimes and the call for
institutional justice is at its strongest (Elcheroth, 2006; Elcheroth &
Spini, 2009, 2014).2 Thus, war experiences in and of themselves are

2
 Similarly, in the comparative data set which we used in this research (TRACES), collective guilt
acceptance is higher among Bosniaks in Bosnia than Croats in Croatia, although the former group
was on average more exposed to the war victimisation. In contrast, Macedonians show lower col-
lective guilt acceptance than both of the previous groups, although they are on average less
victimised.
7  Ethnic Violence in the Former Yugoslavia: From Myth to Reality...  197

not sufficient to explain the lack of critical attachment and actual

criticism in Croatia. What matters more is how these experiences are
­collectively interpreted and remembered. That is, how are events of
war drawn upon in order to render criticism illegitimate and to deny
that a critical patriot can be a genuine patriot?
War was and still is understood in very different ways in the two
countries. In Serbia, the conflict of the 1990s led to defeat, to economic
isolation, to ideological confusion and to moral doubt. In Croatia, what
came to be known as the Homeland War is seen as righteous, as liberat-
ing and as the foundation of the sovereign nation-state of Croatia. In
2000, the Croatian Parliament institutionalised this official version of
the war when passing a ‘Declaration on the Homeland War’. The sec-
ond paragraph of this declaration states that “The Republic of Croatia
led a just and legitimate, defensive and liberating war, and not a war of
aggression or of occupation war against anyone, in which it defended
its territory from the great Serbian aggressor within its internationally
recognized borders”. Equally, those who prosecuted the war—the gen-
erals, but more particularly Franjo Tudjman as Head of State—became
viewed as the founding fathers of Croatia (Uzelac, 1998). Finally, in a
mutually reinforcing spiral, these leaders used the authority that they
derived from their part in the war in order to reinforce the righteous
and liberating understanding of the conflict. Thus, Tudjman (1992a,
b) referred to “a holy Homeland War” which was necessary for achieve-
ment of “the centuries old dream”.
It is also clear that, in Croatia, the right to criticise was radically
constrained during the 1990s. The penal code of 1997 allowed for the
prosecution of journalists or others who insult the president, prime min-
ister, supreme court president, president of the parliament or president
of the constitutional court. Even if reports are true, authors can still be
prosecuted for causing emotional anguish. This law has been frequently
enforced: in its first year of operation alone, 937 lawsuits were taken
out under this provision (Jergović 2003). At the same time hundreds
of journalists were sacked from the state broadcaster, simply for being
Serb or else not a supporter of the ruling party. An estimated three mil-
lion books in public libraries were burnt because the authors were Serb,
198  Identity, Violence and Power

the publishers were Serb or the text was “ideologically inappropriate”

(Kurspahić, 2003; Lešaja, 2012).
But while it is fairly straightforward to establish that glorification of
the nation is deeply entrenched both in political discourse and in formal
documents and also that any criticism of the nation—or indeed a failure
to actively endorse the national past and national representatives—was
subject to severe repression during the 1990s, it remains to be explained
exactly how the interpretation of the ‘Homeland war’ is used to deny a
space for critical attachment; how, therefore, any expression of criticism
is rendered illegitimate; and how the repression of critics is justified.
The problem with such an analysis is that we are largely dealing with an
absence. It is hard to see how repressive processes operate when people
self-police and refrain from expressing criticism. However, there are
occasional exceptions when people did publicly criticise Croatia’s part
in the war and the actions of the leaders in that war. One prominent
example of this is the television programme Latinica, broadcast on 12
December 2005. The programme became the focus of a parliamentary
debate the following day, 13 December. Through an in-depth analysis
of the reactions to these critics and the way in which their position is
treated, we sought to examine how the public space for a critical stance
towards the nation is denied in Croatia.

7.2.2 The December 2015 Days

Ante Gotovina made his first appearance at the ICTY on 12 December

2005. On the evening of the same day, the political programme Latinica
of the Croatian state television channel (Hrvatska televizija, HTV)
broadcast a series of five small documentaries titled ‘Tudjman’s legacy’,
which addressed in an unusually critical way sensitive issues related to
the “founder of the Croatian nation’s” role during the war and politi-
cal transition. Six guests discussed the regime of Franjo Tudjman, the
first Croatian president, whose term of office spanned the first decade of
the new nation-state (1990–1999). The discussion was organised around
a series of reports dealing with various aspects of Tudjman’s rule: his
7  Ethnic Violence in the Former Yugoslavia: From Myth to Reality...  199

policies towards Bosnia and Herzegovina and relations with the interna-
tional community; interethnic intolerance, ethno-centrism and national-
ism in his politics; irregularities during the privatisation process and the
economic decline of Croatia during the 1990s.
The guests on the show divided into supporters and critics of Tudjman’s
regime. Critics of the regime were represented by Zrinka Vrabec Mojzes,
editor of the independent radio station 101; Zlatko Zeljko, president
of the NGO ‘Juris Protecta’; and Prof. Milan Kangrga, philosopher.
Supporters of the regime were represented by Nenad Ivankovic, jour-
nalist, who wrote a biography of Franjo Tudjman; Ivan Vekic, Minister
of Internal Affairs from 1991 to 1992; and Eduard Bajlo, historian.
They differed substantially in their perception of recent Croatian his-
tory. In particular they expressed different views of the ‘Homeland War’
and of whether the Croatian side had committed war crimes. Whereas
the critics emphasised the direct responsibility of the Croatian political
elite for planning and executing such crimes, the supporters systemati-
cally denied that this was the case. For them, the Homeland War was
defensive, and while they acknowledged that some crimes might have
occurred, they represented them as individual acts committed out of
revenge or despair.
The following extract involves an exchange between Zrinka Vrabec
Mojzes and Ivan Vekic.

Vekic: I am attacked and defending myself during the whole

show, just because I am a Croat and I fought in the war.
Vrabec Mojzes: And we … What are we?
Vekic: I don’t know, I am not interested in that. But I am
defending myself the whole time.
Vrabec Mojzes: Obviously you think that we are not Croats, because
we are critical towards crimes committed by Croats.
Vekic: I am defending myself because of Ustashas, Tudjman,
war, everything …
Vrabec Mojzes: But please, tell me, what are we if you are a Croat?
Vekic: I don’t know.
200  Identity, Violence and Power

Vrabec Mojzes: Well, imagine, we are Croats too.

Vekic: I am not interested in that.
Vrabec Mojzes: But we are citizens of the Republic of Croatia.

The obvious and striking aspect of this exchange is the way in which
the critic, Vrabec Mojzes, claims Croatian nationhood based on her citi-
zenship (a broader and more inclusive criterion than alternatives such as
ethnic background), the way that she repeatedly demands recognition of
that nationhood by her opponent, Vekic, and the way in which Vekic, on
four consecutive occasions, refuses to grant such recognition. But per-
haps more important is the way in which she identifies this refusal with
her status as a critic. Indeed it is striking how Vrabec Mojzes collectivises
the issue. She never addresses her personal status. She asks who are ‘we’
(the critics) if Vekic is Croat. She asks Vekic to imagine that ‘we’ (the
critics) are Croats. She insists that ‘we’ (the critics) are Croatian citizens.
Most explicitly she complains that ‘we’ are not viewed as Croats because
“we are critical towards crimes committed by Croats”.
This exchange shows that being a critic is linked to one’s national sta-
tus—or rather, being a critic is used to deny one’s national status, and
hence being a critical patriot becomes an impossible position. It sets up
the question for our main analysis of whether criticism is commonly used
to deny nationhood and how it is used to deny nationhood.
This broadcast immediately provoked a fierce debate amongst the Croatian
public and vehement disapproval from many quarters. Eventually, the
show was suspended for one month. The editor of the show and two
journalists received a formal warning. In addition, the editor of the show,
Denis Latin, as well as Zrinka Vrabec Mojzes, one of the critics on the
show, were placed under police protection following numerous threats.
But perhaps the most important reaction relates to what unfolded in the
Croatian Parliament over the next two days.
On 13 December 2005, the sitting of the Croatian Parliament was
initially scheduled to discuss a report on the affairs of Croatian Radio
Television (HRT) in 2004. In the event, much of the discussion was
devoted to the Latinica show of the day before. Several members of par-
liament demanded an official statement from the director general of the
HRT (who was present at the sitting). He declared that the programme
7  Ethnic Violence in the Former Yugoslavia: From Myth to Reality...  201

was unprofessional and that the competent HRT bodies would state their
position on the matter. He also stated that those responsible for the pro-
gramme would be sanctioned. Such was the interest in the discussion that
it was not completed by the end of its allotted time and an extension of
the sitting was agreed for the following day. Finally, the discussion fin-
ished with a rejection of the Report under consideration.
Our main analyses are based on a full transcript of a parliamentary
debate held on 13 and 14 December 2005. We conducted a thematic
analysis, which was focused on the representation of the critical voices
on the Latinica program, and the ways that this representation linked to
construals of the national community and the national interest. Three
broad thematic fields (or a priori coding categories) were distinguished as
a basis for coding the text: constructions of critical voices, constructions
of the nation and constructions of the international context.
The results of our qualitative analysis are summarised in Table 7.7. By
far the most common category of argument was “construction of critical
voices”, with 29 speakers and 62 individual statements (77 % of the total).
Critics were portrayed as distorting national values, as alien to the nation,
as hurting the nation and therefore as people who needed to be identified
and silenced. Arguments concerning the definition of the ingroup were
considerably less common (seven statements from six speakers represent-
ing 9 % of the total): they represented Croatia as Catholic, tolerant and
united behind its leaders. Arguments concerning the international con-
text were similarly sparse (11 statement from eight speakers representing
14 % of the total): here the focus was almost entirely on a Croatia sur-
rounded by powerful enemies—particularly the International Tribunal
sitting in judgement on General Gotovina at that moment—who are
aided and abetted by the critics.

7.2.3 H
 ow Have Critical Patriots Been Silenced by
the Political Majority?

Already, we can glimpse the outlines of an overall argument: critics are not
true (i.e., loyal) Croats, but rather destroy the unity of the nation and attack
the nation at a moment when it is vulnerable to its international enemies.
202  Identity, Violence and Power

Table 7.7  Summary of the results of the thematic coding of the parliamentary
debate held on 13 and 14 December 2005
Number of Number of
Category arguments speakers
A. Construction of critical voices 62 29
1. What are they doing? (“A bad job”) 7 6
1.1. Critics … are misrepresenting the role of the 3 3
Homeland War.
1.2. Critics … are misrepresenting our national past. 4 4
2. Who are they? (“Alien individuals”) 29 18
2.1. Critics … are immoral and unprofessional. 11 8
2.2. Critics … are profiteers and opportunists. 3 3
2.3. Critics … are small elite that abuses its power. 3 2
2.4. Critics … do not have popular support. 4 4
2.5. Critics … are not part of the nation. 8 8
3. What are their intentions? (“To hurt us”) 18 10
3.1. Critics … are attacking the combatants of the 6 4
Homeland War.
3.2. Critics … are insulting Croatian people. 5 5
3.3. Critics … are attacking the foundation of the 3 3
Croatian state.
3.4. Critics … despise our independence. 4 4
4. How shall we treat them? (“To defend ourselves”) 8 8
4.1. Find the culprits. 4 4
4.2. Changing media policies. 4 4
B. Underlying construction of the national community 7 6
5. Who are we? (“The virtuous majority”) 7 6
5.1. We are Croatian catholic majority 2 1
5.2. We are tolerant and respectful of pluralism 2 2
5.3. We are united in national interests and loyal to 3 3
our leaders
C. Underlying construction of the international 11 8
context
6. What is at stake? (“Not to let internal opponents 11 8
play into the hands of external enemies”)
6.1. Critics are collaborating with the historical 3 3
enemies.
6.2. Critics are helping International Tribunal to build 8 6
accusations.

It is quite clear from a quantitative summary that critics are portrayed as

non-Croatian and indeed anti-Croatian in both intent and effect. Let us
now look more closely at the debate in order to see just how this repre-
sentation is sustained.
7  Ethnic Violence in the Former Yugoslavia: From Myth to Reality...  203

The President of the Parliament opened the session with a call to criti-
cise Latinica:

Yesterday’s show, gentlemen from HTV and ‘Latinica’, was one blasphemous
forgery of the Croatian history. Obviously biased, against all the principles
of the journalistic profession. (…) And the Croatian Parliament should cer-
tainly talk about it. (…) We should not allow that we, as the highest repre-
sentative body, do not speak out firmly and strongly about this and that we
should be bound by the alleged freedom of the media. It is not media free-
dom when the truth is falsified, it is not media freedom when the elementary
historical facts are represented in a disgusting, careless way, which irritates a
huge proportion of the Croatian people and Croatian citizens.
(Vladimir Seks, President of the Croatian Parliament)

There are three elements worthy of note here. To start with, in the very
first sentence of the debate, critical positions are represented as a “blas-
phemous forgery”. The notion of blasphemy sets up certain readings of
the Croatian past as sacred and hence as not amenable to alternatives.
Anyone who challenges these readings is therefore attacking the nation.
Equally—and this is the second element—anyone who does so is advanc-
ing a falsehood. Therefore, repressing such voices is not curtailing free-
dom, it is defending the truth. In this way, an attack on the critics does
not compromise the claim to be a freedom-loving nation and parliament.
Third, insofar as this is an attack on the nation and an attack on the
truth, it is opposed by (freedom-loving) Croats. If Seks uses a relatively
mild formulation to express this (the criticism “irritates”) and if he quali-
fies its application (to “a huge proportion” of Croats), others are both
harsher and less qualified. Thus, Zdenka Babic Petricevic of the ruling
HDZ argues that criticism “hurts” all true Croatians and Ivan Vucic,
another HDZ member, suggests that it “disgusts” them.
This opening argument was never challenged. Rather, it raised a num-
ber of issues which were addressed by other speakers throughout the
debate. The first of these issues is who are the critics and why are they
critical. Some of the interventions characterise them as lacking positive
qualities. They are stupid: “the lowest educated journalists in Croatia”
according to Independent member Slaven Letica. They are insignificant:
the ruling party (HDZ) member Kresimir Cosic refers to one critic as
204  Identity, Violence and Power

“such an obscure person”. They are vicious: Cosic also refers to critics
more generally as “indoctrinated and full of hate”.
This last comment raises the question of who they are indoctrinated
by and who they hate. The answer is that they serve alien interests. They
hate Croatia and Croats. For instance, in the next extract, the insinua-
tion by the right-wing HSP member Pejo Trgovcevic is that Latinica is a
foreign show:

Shows like this astound the Croatian public, and we ask, together with the
majority of the Croatian public, which television have we watched yester-
day? Are these shows from Croatian television or some other Television?
(Pejo Trgovcevic, HSP)

Indeed, even the title of the programme is challenged to insist upon this
alien character. With a play on words, the show is equated to the Serbian
rather than the Croatian language to exemplify its supposedly hostile
stance disguised as a national programme:

And I personally cannot stand Latinica because it is written in Cyrillic, and

that’s why it is as it is.
(Zdenka Babic Petricevic, HDZ)

The next issue concerns the nature of the sacred: What is it precisely
which cannot be questioned by any true Croat, and which distresses
Croats if questioned? The answer lies in a double elision, the first ele-
ment of which is the wartime leader and first president, Franjo Tudjman.
Above, we noted how Kresimir Cosic referred to one critic as “obscure”.
But this obscurity was emphasised through a contrast with Tudjman’s
pre-eminence. Thus, Cosic marvelled at how Latinica could provide
public space for “such an obscure person to speak neither more nor less
than about the Father of the Nation”. The term ‘father of the nation’ is,
of course, redolent with significance. It denotes someone who has given
birth to the nation, someone who has unquestioned authority over the
nation, and someone with whom nationals have an intimate and highly
significant relationship. This, in part, explains why Croats would be upset
by criticisms of Tudjman for these are akin to insulting a beloved fam-
ily member. This is powerfully expressed by the following speaker, who
7  Ethnic Violence in the Former Yugoslavia: From Myth to Reality...  205

compounds the sense of outrage by invoking his daughter, a child, who

cannot understand why anyone should attack the dear leader:

[Instead of Latinica] We should have broadcast some of [Tudjman’s]

speeches, performances, so our children can see who that great man was,
instead of it happening that my daughter asks me in the morning, excuse
me for my voice, how come they are speaking like that about our Franjo.
Then I had to explain to her who was speaking and why.
(Franjo Arapovic, HDZ)

Importantly, though, criticism of Tudjman is not represented as criti-

cism of Tudjman alone. It is rhetorically elided with criticism of Croatian
­soldiers. Equally, criticism of Croatian soldiers is elided with criticism
of the Croatian people and of the nation as a whole. Both elements are
apparent in the following extract:

We can clearly conclude that HTV is working against Croatian defenders

and damaging them, because they want to vilify them, the whole victorious
Croatian army, and it’s commander, late President Tudjman.
(Josip Djakic, HDZ)

On the one hand, then, Tudjman is positioned as commander of the

army, such that an attack on him is an attack on the army (and vice versa).
On the other hand, the army are characterised as ‘Croatian defenders’
(an official term). They are an embodiment of national independence—
the fight to achieve it and to preserve it. Putting both elisions together,
Tudjman himself is constituted as the embodiment of national inde-
pendence. He himself becomes sacred. He is rendered inviolate, and an
attack on him becomes an attack not just on the father of the nation, but
on the nation itself.

We do not need to defend Dr. Franjo Tudjman, he defended us as a supreme

commander of the armed forces and the democratic movement. He freed,
defended and left us in the legacy a sovereign independent State of Croatia.
And Mr. Latin and many little “Latins” in Croatia cannot bear that
legacy.
(Jagoda Majska Martincevic, HDZ)
206  Identity, Violence and Power

The final issue concerns the reason why it is so important to silence the
critics. In part, this may be self-evident: to the extent that they are seen
as attacking the (national) ingroup, this is bad in and of itself. However,
this is not the end of the story. On the one hand, these attacks were rep-
resented as doing real harm. Some of the media invoked the suicide of a
soldier—a ‘Croatian defender’ and hence a symbol of the nation—on the
evening that Latinica was broadcast. For example, one of the most popu-
lar daily journals ‘Vecernji List’ published this information on the front
page under the title ‘Killed himself because of “Latinica”’. The suicide
was attributed to the programme and, in some instances, was actually
seen as intended by the programme—that is, the aim of the critics was to
weaken or even destroy the nation. This argument was immediately taken
up by several speakers:

Yesterday’s TV show about which we spoke a lot today, achieved its goal. I
just received a text message—after watching Latinica yesterday, a Croatian
defender in Glina committed suicide using a bomb.
(Ivo Loncar, Independent)

On the other hand, the nation was represented as particularly vulnerable

due to the actions of foreign enemies, in part Serbia (according to Josip
Djakic of the HDZ, Croatian Television was ‘helping Serbian spies to write
accusations’ against Croatia), but more characteristically—and not surpris-
ingly given Gotovina’s trial—the ICTY. Such times would normally com-
pel unity. According to Ivan Vucic of the HDZ, this was a moment when
“Croatian people and Croatian defenders” were organising rallies in favour
of the general, and when “only those who do not wish well to Croatian
people” would be critical. And yet, this was precisely the moment when
Latinica chose to air its criticisms, thus strengthening the hand of the enemy.

We are judged by the international community, we are judged in the ICTY,

and we are blaming Serbs. But we are wrong! Us Croats, it is our own fault.
These editors who are preparing such programmes. That show will for sure
be watched at the ICTY. And what is it saying? What evidence is that?
These are the transcripts which the worst enemy couldn’t deliver to some-
one, but fortunately, they are not true.
(Ivan Vucic, HDZ)
7  Ethnic Violence in the Former Yugoslavia: From Myth to Reality...  207

All 18 categories of argument previously summarised in Table 7.7 come

from members of the ruling coalition or from independents. Members
of the opposition only took issue with two of these arguments: ‘chang-
ing media policies’ (4.2), and ‘we are tolerant and respectful of pluralism’
(5.2). This fitted into the two broad arguments deployed by the opposi-
tion. First, they systematically accused the ruling parties of undermining
freedom of speech and media independence. Second, they argued that
attacks on HTV and Latinica were a cover for achieving greater state
control over HTV.
However, throughout the entire debate, not a single opposition mem-
ber either addressed the criticisms voiced in Latinica or reacted to attacks
on the critics. To put it more starkly, no one contested the notion that
critics are a deviant anti-Croatian minority who harm the nation and
its people. Nor did anyone question the regime of Franjo Tudjman
or its responsibility for crimes committed during the Homeland War.
Altogether, this parliamentary ‘debate’ was de facto a monologue. Indeed,
on the second day of the parliamentary sitting, members of the opposi-
tion declined to participate in the debate any further and they left the
Parliament. Doing so, they dramatically enacted and at the same time
consecrated the absence of space for critical debate among the elected
representatives of Croatian citizens: potentially critical voices withdrew
rather than contributed to the debate. Their inability to participate can
be put down to the fact that it was self-evident that to voice anything but
support for the attack on critics would position them—along with the
critics themselves—as anti-national, which is a death knell for politicians,
at least in a democratic system (Reicher & Hopkins, 2001).

7.3 Conclusion
So how, then, has the myth of ethnic war not only become real in its con-
sequences during the war, but been nourished and kept alive during the
post-war period? How, in post-war Croatia, have critical voices been mar-
ginalised, and their legitimacy to speak as members of the nation chal-
lenged? To summarise our findings, three key elements were involved.
The first was to sacralise certain events and individuals—specifically,
208  Identity, Violence and Power

the Homeland War and those, like Tudjman, who prosecuted it. These,
then, are not just incidental aspects of the nation which could be good
or bad. Rather they are essential to and foundational of the nation: the
war created the nation; Tudjman is the ‘father’ of the nation. It follows
that any attack on either suggests that the nation itself is essentially bad.
Such representations are not limited to the Croatian political elite—they
have been promoted by a range of public actors and institutions such as
the Catholic Church (Perica, 2002), the media (Kurspahić, 2003) and
the school system (Barunčić & Križe, 2006). Consequently, it becomes
impossible to sustain a critical stance which rests on the claim that candid
self-scrutiny serves to improve the nation (Hornsey, 2005).
The second element was to go from claiming that criticism was an
attack on the nation to a claim that criticism was damaging the nation.
In part, this involved attributing actual instances of harm (e.g., the sui-
cide of a soldier/‘national defender’) to the criticisms. In part, it involved
construing the nation as highly vulnerable due to external threats as con-
cretised in the ICTY, and in the person of General Gotovina standing
before the International Tribunal in The Hague. As a number of authors
have pointed out, invoking powerful enemies is a particularly powerful
way of demanding unity and outlawing criticism (see Chap. 5). That is,
by invoking enemies who constitute a serious threat to the ingroup, unity
is demanded and criticism is outlawed.
Thirdly, then, true Croats are represented as active loyalists: accord-
ing to some majority parliamentary members quoted in our analysis,
they don’t simply support the nation in principle, they rally for the
nation (or even organise rallies for the nation). Likewise, true Croats
are necessarily hurt, irritated and disgusted by anyone who is not loyal
at the nation’s hour of need. The corollary is that anyone who criticises
the nation (or even fails to be disgusted by such criticism) is not only
denied nationhood, they are constituted as active enemies of the nation
(they don’t have the country’s interests at heart, they knowingly provide
ammunition for enemies such as the ICTY); their goal is the destruc-
tion of national icons (such as Croatian soldiers) and the nation itself.
They have to be repressed.
We do not suggest that all these elements are necessary in order to
marginalise critical patriotism or that these are the only ways by which
7  Ethnic Violence in the Former Yugoslavia: From Myth to Reality...  209

criticism of the nation can be outlawed. We simply present them as

some of the ways by which critics can be silenced, and as the ways that
this was done in Croatia. This analysis, then, helps explain the very low
numbers of critical patriots in Croatia. For those few who do adopt
such a stance, it also makes sense of the lack of a relationship with will-
ingness to accept that the nation has done specific wrongs in the past. It
is one thing to eschew claiming that one’s nation is better than others;
it is quite another to overtly express criticisms of the things it has done
when one sees what does happen to those who do actually put their
heads above the parapet.
But there may be another reason for the lack of a relationship
between mode of attachment and acceptance of national guilt. We
have already noted how speakers stress the isolation of the critics, how
they are atypical of most Croats (certainly of ‘true Croats’) and how
most people are opposed to them. More tentatively, we have suggested
that this helps to explain why the opposition failed to address the criti-
cisms and eventually withdrew from the debate—thus entrenching the
impression that no one supports the critics. Even without overt repres-
sion, or even the threat of repression, this may itself be sufficient to
silence those who identify with their nation but do not think it to be
flawless. As we have argued in Chap. 3, people’s stances in the world
are determined by their assumptions about the interpretations that are
shared with relevant others. If we assume that certain interpretations
cannot be shared, we are likely to engage in self-censorship. Where we
believe that the critical questioning of our nation’s role in past conflicts
will be rejected as disloyal by fellow nationals, we will be unlikely to
voice any such criticisms. In this way, critical attachment becomes dis-
sociated from the expression of criticism and of acceptance of guilt for
our past deeds.
Another way of saying this is that, if it is impossible for a critical
patriot to be a genuine patriot then, to continue as a patriot, one must
censor one’s criticisms. But there is another option. That is, to continue
as a critic, one must discard one’s patriotism. This would explain another
feature of the findings from the representative cohort sample, that is, the
only people to accept national guilt in Croatia (unlike Serbia) are those
who are explicitly dis-identified from the nation.
210  Identity, Violence and Power

The process whereby those who will not renounce their criticisms
are led to renounce their nationhood is well expressed by two famous
Croatian intellectuals: the writer Dubravka Ugresic and the philosopher
Boris Buden. After becoming a target of the nationalistic media and
Croatian public because of their critical attitudes about Croatian nation-
alism during the 1990s, both of them left the country. Ugrešić (2007)
explained that “(T)hey excluded me from their literary and other ranks …
I became a literary representative of a place that no longer wanted me.
I, too, no longer wanted the place that no longer wanted me. I am no
fan of unrequited love”. Buden (2000) was yet more explicit about the
interconnections between being excluded and discarding identity: “I am
not anymore a Croatian intellectual … I have experienced a definitive
exclusion, or separation, cutting of the umbilical cord that connected me
with the Croatian identity”.
What is most striking about the dialogue between regime support-
ers and regime critics in the TV show we analysed is that the latter did
not choose to dis-identify with Croatia. They did not aim to become
‘anti-Croats’. It was not a position that reflected their internal beliefs
and desires. On the contrary, they struggled to be recognised as fully
fledged members of the national community. The position of out-
siders was imposed upon them because that which they would have
chosen—critical attachment—was simply not available. This serves to
underline a point that has been imminent throughout our discussion.
That is, the relative incidence of critical patriots and national glorifiers
in Croatia cannot be understood as a matter of individual differences
(which is how differences in the mode of national identification are
normally approached—see, for example, Kosterman and Feschbach,
1998; Mummendy, Klink and Brown, 2001; Staub, 1997). Rather,
they reflect differences in the social availability of certain ideological
configurations of belief.
The struggle for the recognition of critical attachment as a legitimate
mode of identification thus unfolds at two different levels: (a) at a micro-
level, where people negotiate their version of identification with relevant
others through a myriad of individual decisions to express a particular
voice or keep silent; and (b) at a macro-level, where elites use their control
 7  Ethnic Violence in the Former Yugoslavia: From Myth to Reality...  211

over institutional policies, party lines and/or mass media to systematically

provide opportunities or, in contrast, discourage people from giving voice
to a particular version of identity.
As we have seen from the parliamentary reaction to one television
programme, when critical patriots raise their voice, they are met with
extreme reactions. These reactions were overwhelmingly focused on the
delegitimisation of these critical voices, by portraying them as disloyal,
anti-national, aberrant and even dangerous for the nation. In such a
political climate, where critical voices are denied social recognition as
Croats and the right to speak as Croats, critics will have little oppor-
tunity to participate in public debates on the nation’s past even if they
want to. Most prefer to keep silent. This then perpetuates a vicious
spiral in which the illusion of a monolithic public opinion becomes
ever more entrenched and the possibility of changing it becomes ever
more remote.

References
Barunčić, J., & Križe, Z. (2006). Domovinski rat u udžbenicima iz povijesti.
Casopis Za Suvremenu Povijest, 38(2), 373–716.
Branscombe, N. R., Slugoski, B., & Kappen, D. M. (2004). The measurement
of collective guilt: What it is and what it is not. In N. R. Branscombe & B.
Doosje (Eds.), Collective guilt: International perspectives (pp. 16–34). New
York: Cambridge University Press.
Broz, S. (2005). Facing the crime-vengeance, justice and understanding.
Yearbook of the Balkan Human Rights Network, 01, 174–186.
Broz, S. (2014). When nobody stood up and everybody is guilty: A puzzle of
individual responsibility and collective guilt. In D. Spini, G. Elcheroth, &
D. Corkalo Biruski (Eds.), War, community, and social change. Collective experi-
ences in the former Yugoslavia (pp. 155–162). Berlin & New York: Springer.
Buden, B. (2000). Interview in Vreme. Retrieved from http://www.ex-yupress.
com/vreme/vreme100.html
Clark, J. N. (2013). The ICTY’s Acquittal of Croatian Generals Gotovina and
Markac. Journal of International Criminal Justice, 11(2), 399–423.
doi:10.1093/jicj/mqt009.
212  Identity, Violence and Power

Corkalo, D., & Kamenov, Z. (2003). National Identity and social distance: Does
in-group loyalty lead to outgroup hostility? Review of Psychology, 10(2), 85–94.
Corkalo Biruski, D., & Penic, S. (2014). Traumatised selves: Does war trauma
facilitate in-group bonding and out-group distancing? In D. Spini,
G. Elcheroth, & D. Corkalo Biruski (Eds.), War, community, and social
change: Collective experiences in the former Yugoslavia (pp. 137–154). Berlin &
New York: Springer.
Doosje, B., Ellemers, N., & Spears, R. (1995). Perceived variability as a function
of group status and identification. Journal of Experimental Social Psychology,
31, 410–436.
Elcheroth, G. (2006). Individual-level and community-level effects of war
trauma on social representations related to humanitarian law. European
Journal of Social Psychology, 36(6), 907–930.
Elcheroth, G., & Spini, D. (2009). Public support for the prosecution of human
rights violations in the former Yugoslavia. Peace and Conflict: Journal of Peace
Psychology, 15(2), 189–214.
Elcheroth, G., & Spini, D. (2014). Beyond collective denial: Public reactions to
human rights violations and the struggle over the moral continuity of com-
munities. In D. Spini, G. Elcheroth, & D. Corkalo Biruski (Eds.), War, com-
munity, and social change. Collective experiences in the former Yugoslavia
(pp. 205–226). Berlin & New York: Springer.
Gagnon, V. (2004). The myth of ethnic war: Serbia and Croatia in the 1990’s.
Ithaca, NY: Cornell University Press.
Gauthier, J.-A., & Widmer, E. (2014). The destruction of multiethnic locations:
Markers of identity and the determinants of residential trajectories. In
D. Spini, G. Elcheroth, & D. Corkalo Biruski (Eds.), War, community, and
social change. Collective experiences in the former Yugoslavia (pp. 85–98). Berlin
& New York: Springer.
Hornsey, M. J. (2005). Why being right is not enough: Predicting defensiveness in
the face of group criticism. European Review of Social Psychology, 16(1), 301–334.
Jergović, M. (2003). Zakonske promjene i tisak u Hrvatskoj od 1990. do 2002.
Politicka Misao, XL(1), 92–108.
Jutarnji List. (2011, April 16). Samo 23% Hrvata za ulazak u EU! Čak 95%
smatra presudu nepravednom [Only 23% of Croats to join the EU! As many
as 95% considered the verdict unjust.]. Retrieved from http://www.jutarnji.
hr/istrazivanje-nakon-presude-gotovini--samo-23--hrvata-za-ulazak-­u -
eu/939458/
Kosterman, R., & Feshbach, S. (1989). Toward a measure of patriotic and
nationalistic attitudes. Political Psychology, 10(2), 257–274.
7  Ethnic Violence in the Former Yugoslavia: From Myth to Reality...  213

Kurspahić, K. (2003). Prime time crime: Balkan media in war and peace. Washington,
DC: US Institute of Peace Press.
Le Goff, J.-M., & Giudici, F. (2014). The demise of mixed marriage? In D. Spini,
G. Elcheroth, & D. Corkalo Biruski (Eds.), War, community, and social
change. Collective experiences in the former Yugoslavia (pp. 63–84). Berlin &
New York: Springer.
Lešaja, A. (2012). Knjigocid—Uništavanje knjiga u Hrvatskoj tijekom 90-ih.
Zagreb: Profil/SNV.
Mummendey, A., Klink, A., & Brown, R. (2001). Nationalism and patriotism:
National identification and out-group rejection. British Journal of Social
Psychology, 40(2), 159–172.
Perica, V. (2002). Balkan idols: Religion and nationalism in Yugoslav states. New York:
Oxford University Press.
Penic, S., Elcheroth, G., & Reicher, S. (2016). Can patriots be critical after a
nationalist war? The struggle between recognition and marginalisation of dis-
senting voices. Political Psychology, 37(4), 481–496.
Reicher, S., & Hopkins, N. (2001). Self and nation. London: Sage.
Roccas, S., Klar, Y., & Liviatan, I. (2006). The paradox of group-based guilt:
Modes of national identification, conflict vehemence, and reactions to the
in-group’s moral violations. Journal of Personality and Social Psychology, 91(4),
698–711.
Schatz, R. T., & Staub, E. (1997). Manifestations of blind and constructive
patriotism: Personality correlates and individual–group relations. In D. Bar-­
Tal & E. Staub (Eds.), Patriotism: In the lives of individuals and nations
(pp. 229–245). New York: Nelson-Hall.
Spini, D., Elcheroth, G., & Biruski, D. C. (2013). War, community, and social
change: Collective experiences in the former Yugoslavia. New York: Springer
Science & Business Media.
Spini, D., Elcheroth, G., & Fasel, R. (2011). TRACES: Methodological and tech-
nical report (No. 4). LIVES Working Papers.
Staub, E. (1997). Blind versus constructive patriotism: Moving from embed-
dedness in the group to critical loyalty and action. In D. Bar-Tal & E. Staub
(Eds.), Patriotism: In the lives of individuals and nations (pp. 213–228).
New York: Nelson-Hall.
Tudjman, F. (1992a). Croatian Army achieved its historical duty. Presented at the
Speech given on October 29th 1992, Dubrovnik. Retrieved from http://
www.slobodanpraljak.com/MATERIJALI/RATNI%20DOKUMENTI/
TUDJMAN%20GOVORI%20I%20INTERVJUI%201990.-1998/
Tablica-Tudjman-intervjui.htm
214  Identity, Violence and Power

Tudjman, F. (1992b). The centuries-old dream achieved!. Presented at the Speech

given on May 30th 1992. Retrieved from http://www.slobodanpraljak.com/
M AT E R I J A L I / R AT N I % 2 0 D O K U M E N T I / T U D J M A N % 2 0
GOVORI%20I%20INTERVJUI%201990.-1998/Tablica-Tudjman-­
intervjui.htm
Ugrešić, D. (2007). Nobody’s home. London: Telegram/Saqi.
Uzelac, G. (1998). Franjo Tudjman’s nationalist ideology. East European Quarterly,
XXXI(4), 449–472.
 8
British Warriors and Scottish Voters:
When ‘Rallying the Nation’ Backfires

The ‘Invasion of Iraq’ commonly denotes the period between 20 March

2003 (when a military coalition formed predominantly of the USA and
its most important ally, the UK, entered Iraqi soil) and 1 May 2003 (the
day when US President Bush declared the end of major combat opera-
tions in a televised address from an aircraft carrier). In Britain the inva-
sion provoked a massive national rally effect. When it started, it looked
as if the anti-war mobilisation, which had been very strong during the
preceding weeks, had lost its impact overnight.
In a retrospective survey conducted by Lewis (2004) among a repre-
sentative sample of British adults in summer 2003, more than a third of
respondents admitted that they had changed their stance on the war once
it had started. This way, a minority of supporters mutated into an over-
whelming national majority. Interestingly, when those who shifted from
an anti-war to a pro-war position were asked why they changed their
mind, the most frequent answer was “to support our troops”. Lewis’ anal-
yses of the new mass media context created during the invasion elucidate
the mechanisms by which people’s desire to be perceived as loyal support-
ers of British troops “may have been compounded by the fear of being
seen as part of an isolated, unpatriotic minority” (p. 301). As he points

© The Author(s) 2017 215

G. Elcheroth, S. Reicher, Identity, Violence and Power,
DOI 10.1057/978-1-137-31728-5_8
216  Identity, Violence and Power

out, the ‘embedded’ war reporters programme designed by US and UK

Ministries of Defence facilitated new forms of daily frontline journalism.
These not only fostered concrete identification with coalition troops and
their day-to-day activities—while ‘enemy’ troops remain largely abstract
and dehumanised. They also shift general media attention away from the
broader political and moral issues raised by the war, towards issues of
military strategy and progress.
This conclusion has subsequently been corroborated in a detailed con-
tent analysis of televised images broadcast by the BCC during the early
invasion phase (Lipson, 2009). While most camera shots showing coali-
tion troops were close enough to read the facial expressions and emotions
of individual soldiers, camera shots showing Iraqis were typically taken in
anonymous group or crowd situations and most of them displayed males
moving, running or behaving violently together.
The rallying effect of the war did not occur homogenously across
Britain, however. There were pockets of resistance—even at the moment
when triumphalist voices were at their loudest and the national rally
effect was at its peak. By coincidence, on the same day that the world
saw a triumphant Bush in front of a large ‘Mission Accomplished’ ban-
ner, Scotland’s citizens went to the polls to elect their representatives for
the second session of the devolved parliament in Edinburgh, created four
years before. The parties which had campaigned against the war during
the invasion scored well in these elections: in the regional vote, where
people vote for party lists rather than individual candidates, the three
anti-war parties—the Scottish National Party, Scottish Socialist Party
and Scottish Greens—polled a total of 34.5 % of the vote, an increase of
1.6 %.
The greater level of anti-war sentiment in Scotland than in the rest
of the UK was even more apparent in later opinion polls. The British
General Election Studies of 2005 provide the opportunity to compare
the lasting effects of anti-war mobilisation in Scotland with the rest of
Britain, two years after the invasion. According to our analyses of this
data set, 42 % of Scots declared themselves “angry” and 43 % “disgusted”
by the war in Iraq, compared with only 36 %, for both emotions, in the
remaining population. Scots were more likely to strongly disapprove of
British involvement in Iraq and to judge that the Labour government
 8  British Warriors and Scottish Voters: When ‘Rallying the Nation’  217

handled the situation in Iraq badly. Furthermore, the Scottish respon-

dents mentioned Iraq almost twice as often as the most important issue
of the election campaign. In contrast to what was observed in England
and Wales, the Scottish working class expressed critical stances on the
war as frequently as members of the middle or upper classes. That is, in
Scotland (and only in Scotland) anti-war mobilisation completely over-
came class cleavages.
But there is a second reason why the Scottish reaction to Iraq’s inva-
sion is worthy of an analysis in its own right. When Billig (1995) com-
mented on the two previous major rally events in British history—in
1982 during the Falklands War and in 1991 during the First Gulf
War—his main point was that the apparent ease and speed with which
national leaders in Britain (and elsewhere) mobilised support for war
would not have been possible without steady, daily and unspectacular
rehearsals around the national flag during the years and decades before-
hand. Leaders who only started arguing about national identity and
the national interest on actual entry to war would have little chance of
engaging their population. Equally, dissidents who sought to invoke
national values in opposition would have little chance of being heard
above the drumbeat of war unless their arguments were already familiar.
In particular, if the seeds of critical patriotism (see Chap. 7) have not
been planted well before, it is unlikely that it will flourish in the harsh
climate created when wars start.
But, interestingly, Billig’s argument can easily be turned on its head:
if a familiar and agreed view of national identity facilitates its use in cre-
ating a new consensus when wars start, so a contested view of national
identity opens the way to challenging such a consensus. In other words,
hegemonic support for war is less likely, and polemical debates more
likely, when critics can draw on well-recognised alternative understand-
ings of nationhood in order to advance their case.
From this perspective, the Scottish context provides opportunities
for those who oppose British war policies. This is due to a pervasive
tension between two co-existent, but potentially antagonistic, forms of
‘banal’ nationalism. From the Scottish public’s point of view, there is
indeed permanent ambiguity as to whether to salute the British Union
Jack or the Scottish Saltire. Should people trust to the categories of the
218  Identity, Violence and Power

Football World Cup (where Scots have their own team) or to those of
the Olympic (where they compete for Great Britain); should they adopt
the perspective of the British or Scottish editions of the same TV chan-
nels when watching the daily news? Interestingly, Rutland, Cinnirella
and Simpson (2008) showed that, among Scottish students, identifica-
tion with both Scotland and Britain is remarkably stable across differ-
ent situations. This insensitivity to immediate contextual influences is
related to “the fact that Scottish and British self-categories are chroni-
cally accessible” (p. 268). Furthermore, as shown by Sindic and Reicher
(2009), in combination with particular understandings of intergroup
relations, social identification with Scotland anchors specific political
claims, like separatism.
In short, nationhood is more troubled and contested in Scotland
than in the rest of the UK and that may account for the greater oppo-
sition to the Iraq war. This troubled sense of nationhood is reflected at
the structural level in the shifting constitutional settlement between
Scotland and England. Since the creation of a devolved Scottish
Parliament in 1999, there has been an ongoing debate as to whether
the devolved settlement is adequate, whether the parliament should
have greater powers or whether there should be full independence.
Furthermore, if the issue of how Scottishness relates to Britishness is
at the top of the political agenda, it is equally salient at a psychologi-
cal level. This makes Scotland a particularly promising place in which
to examine how national categories are construed and how these are
used to argue for or against war.
In this chapter, we will present systematic analyses of how the Iraq
invasion of 2003 was understood in Scotland by both elites and by the
overall population. The first part includes all the contributions to four
Scottish parliamentary debates about the Gulf War which occurred
between January 2003 and June 2004—a total of 106 speeches. Using a
combination of qualitative thematic analysis and quantitative, multiple
correspondence analysis (MCA) we examine (a) the definitions of iden-
tity that were invoked to support pro- and anti-war positions, and the
extent to which they were rooted in the broader political culture; (b) con-
sistencies and differences between different contributors to the debate, in
relation to their party’s position on Scottish nationalism; and (c) shifts
 8  British Warriors and Scottish Voters: When ‘Rallying the Nation’  219

in argument and in categorical structure over the course of events—­

especially between the pre- and post-invasion debates.
In the second part, we turn from elite discourse to popular under-
standings of the war, to analyse the constituency context of the political
speeches. This part is based on data from the Scottish Social Attitudes
(SSA) survey of 2003. We analyse how anti-war and pro-war respondents
express their own nationhood and how the translation of individual anti-­
war positions into electoral support for the anti-war separatist parties
relates to their exposure to mass media sources.1

8.1 Mobilising for or Against the Invasion

Two major debates concerning the possible war in Iraq occurred in the
Scottish parliament before the invasion. The first was on 16 January
2003, when there was still considerable uncertainty regarding whether
there would be a war and who would participate. The second debate took
place on 13 March 2003, six days before the first air strikes on Baghdad.
At this point, war (and British combat involvement in the war) was seen
as all but inevitable by most observers. We also analysed the first two
substantial parliamentary debates that took place after the invasion—the
one on 20 November 2003 (when George Bush visited the UK), the
other on 2 June 2004 (when the war crimes at Abu Ghraib had become
public). The full transcripts of these four debates, which are published in
the official records of the parliament, constituted the raw corpus of data
for our analyses.
To prepare this corpus for both qualitative and quantitative analyses,
we conducted two-step hierarchical thematic coding. The first step was
aimed at identifying discrete arguments used within the corpus. These
were organised into a 2 × 2 grid. The first dimension concerned whether
the argument is about the categorisation of actors (who is against whom)
or else the morality of events (whether, and in what way, the war is

1
 Another version of this study has previously been published in Elcheroth and Reicher (2014). The
present version has been simplified for the presentation of the quantitative material, but signifi-
cantly enriched with qualitative material as compared to the article version.
220  Identity, Violence and Power

justified or not). The second dimension concerned whether the argument

addressed the domestic significance of the war (‘home front’ arguments)
or its international significance (‘external front’ arguments).
In a second step the resulting coding categories were reduced to cre-
ate a set of general arguments. Thematic coding resulted in two types
of outcome. First, it produced the structured inventories of arguments
that are summarised in Tables 8.1 and 8.2 and which serve as a grid for
qualitative analyses. These inventories provide an overall template for
interpreting the grand-narrative structures spun by war supporters and
opponents. Second, the coding resulted in a content analytical database
in which the 106 parliamentary interventions (each representing the
entire rhetorical contribution of one Member of the Scottish Parliament,
MSP, to one of the four debates) are treated as cases, and arguments are
treated as variables. Two external variables were added to this database:
the date of the debate and the party affiliation of the speaker. This data-
base is suitable to conduct MCA at the level of parliamentary interven-
tions, based on the joint occurrences of general arguments invoked by
the same speaker within the same debate.
The findings show first that there is a clear antinomic structure to the
debate: each argument in favour of the invasion can be matched with an
argument against it. The notion that war is a necessary act of self-defence
is countered by the argument that it is unnecessary aggression; the notion
that war will relieve suffering is countered by the argument that it will
destroy lives; the notion that war will promote a democratic world order
is countered by the argument that it will destabilise this order (see Table
8.1). Altogether, while pro-war speeches refer to a narrative of liberation
and unity, anti-war rhetoric builds upon a narrative of aggression and
division.
Table 8.2 presents the range of categorical arguments. The way cat-
egorical and moral arguments relate to each other and the way in which
this relationship changes across the debates is the focus of an MCA, the
outcome of which is reported in Fig. 8.1.
The two dimensions arising from this analysis are easily interpretable.
The horizontal dimension represents pro- versus anti-war positions. The
vertical dimension represents the time of the debate, from before to after
the invasion. For ease of interpretation, we have divided the figure into
Table 8.1  A structured inventory of moral arguments: generic moral judgements and nested concrete arguments invoked
by the ‘external front’ and the ‘home front’ as well as frequencies before and after the invasion
External front Before After External front Before After
Self-defence and containment 33 % 4 % vs. Undermining a peaceful settlement 19 % 5 %
of the crisis
WMDs in the hands of a tyrant 13 1 No immediate threat 8 2
Military threat helps diplomacy 8 All other means not exhausted 14
Humanitarian intervention 16 % 22 % vs. Disproportionate killing 23 % 29 %
Rescue a suffering people 10 4 Humanitarian disaster 8 6
More lives saved than lost 2 Selective morality 19 5
Promoting a democratic world order 27 % 52 % vs. Destabilising the international order 47 % 29 %
Democracy and peace-building 7 10 Violation of national sovereignty 6
Institutional backing for war 10 2 Undermining the UN and the rule of 40 6
international law
War against terror 2 Stimulating terrorism 8 5
Home front Home front
National loyalty first 25 % 22 % vs. Bypassing democratic accountability 10 % 37 %
The troops and their commander-­in-­ 4 No public mandate for war 9
chief need our support
Opponents undermine our nation’s 8 4 Beware of blind support 2
leadership
No time for debate, too late to step 4 Political deception 1 11
back
8  British Warriors and Scottish Voters: When ‘Rallying the Nation’ 

Misjudgement is not deception 2 Perversion of democracy 3

221
222  Identity, Violence and Power

Table 8.2  A structured inventory of categorical arguments: generic social cleav-

ages and nested concrete arguments invoked by the ‘external front’ and the
‘home front’ as well as frequencies before and after the invasion
External front Before After
All vs. Saddam 6 % 6 %
Coalition for the Iraqi people 5 3
US-UK vs. Iraq 20 % 16 %
War against the Iraqi people 17 3
Army of occupation 5
Western powers vs. the Arab world 4 % 10 %
Clash of civilisations in the making 3 5
Home front
Democrats vs. autocrats 8 % 6 %
Democracy unites us 7 3
Scotland vs. the rest of the UK 20 % 6 %
Vulnerable Scotland 5
Shame on the UK 4
Scotland’s voice against the war 8
The union and the impediment to a moral policy 3
Communities vs. elites 13 % 34 %
Virtuous troops led by immoral politicians 2 4
Undemocratic warlords 9 7
Tough towards the weak, weak towards the powerful 3
International capitalism, the grounds for violence 3
People of the world vs. US and UK state powers 29 % 22 %
US unilateral superpower policies 14 6
Fatal alliance with the US 10 5

two parts so we can see separately how moral and categorical arguments
respectively map onto this space.
As can be seen, the pro-war ‘liberation’ morality is associated with
an opposition between the world’s democrats and (isolated) autocrats,
which is subsequently resolved into the whole (democratic) world against
a single tyrannical figure, Saddam Hussein. The anti-war ‘aggression’
morality is associated with a division of the world into dominant and
subordinate groups: at the start of the debate, English war-mongers drag-
ging the Scots into conflict; later, social elites against ordinary people; or
a hegemonic USA/British West against Eastern/Arabic peoples.
To summarise, the pro-war discourse draws upon the ‘new world order’
narrative which has been prevalent since the first Gulf War and which in
turn draws upon anti-Nazi narratives: the entire civilised world against
 8  British Warriors and Scottish Voters: When ‘Rallying the Nation’  223
Before vs. after invasion

1.0 JUN 2004

NOV 2003
Democratic world Democratic
Humanitarian National loyalty 0.5
order accountability Disproportion.
intervention
MAR 2003 killing
0.0
-2.5 -2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5
Destabilising
-0.5 international order
Self-defence

-1.0
JAN 2003
Undermining
settlement
-1.5

Pro vs. against invasion

Before vs. after invasion

1.0
West vs Arab
Elites vs people

All vs Saddam 0.5

US-UK vs Iraq

0.0
-2.5 -2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0 vs World
-US 1.5
Democrats vs
autocrats -0.5

-1.0

Britan vs
-1.5 Scotland

Pro vs. against invasion

Fig. 8.1  Moral principle positions and the timing of debates (above) as well
as social cleavage positions (below) according to their coordinates along two
dimensions defined by an MCA of their joint occurrences within parliamen-
tary interventions. Arguments related to the ‘home front’ appear in bold.

one mad dictator. The anti-war discourse draws upon the notion of a more
aggressive England imposing its will upon the communal Scots within
the political structures of the UK—a narrative which has become com-
monplace within Scottish political discourse (Reicher & Hopkins, 2001).
224  Identity, Violence and Power

In other words, both the anti-camp and the pro-camp are anchored in
familiar mainstream constructions.
There are two further aspects of the findings which speak to the rela-
tionships between the constructions of the one camp and those of the
other, and also the relationships between constructions within the same
camp.
First, as well as showing differences in structure between pro- and
anti-­war speakers, our data also point to differences in the amount of
rhetorical effort they devote to different tasks. Opponents of the war
spend more time than supporters in sustaining their categorical con-
structions. They generate more arguments in favour of these definitions
and also more instances of each argument. Thus, 86 % of the category
arguments made before the invasion—and 88 % afterwards—are made
by anti-war speakers. Moreover, the ratio of categorical arguments to
moral arguments is 4.33:1 for those against the war and 0.44:1 for
those in favour of the war.
Second, further analyses were conducted to assess the collective con-
sistency across contributions stemming from those in the same political
party or else sharing the same broad political outlook. To this end, all
106 interventions were located along the two aforementioned dimen-
sions (pro- versus anti-war and time) and identified by the party affilia-
tion of the speaker. The contrast between separatist and unionist parties
shown in Fig. 8.2 is striking. Every single intervention by a speaker
from one of the three parties having Scottish independence on their
agenda (Scottish National Party, Scottish Socialist Party and Scottish
Green Party) takes an anti-war stance. Not one intervention deviates
from the message that an independent Scotland goes hand in hand
with opposition to the invasion of Iraq. In contrast, the unionist camp
is profoundly divided in the debates, both between and within parties.
The gap across the three parties is substantial; while conservative mem-
bers are the strongest supporters of the invasion, most liberal demo-
crats lean towards an anti-war position. Furthermore, members of the
ruling Labour Party cover the full spectrum of these positions in the
debate, from the purest pro-­invasion stance (on the left of the graph) to
a marked anti-invasion view (on the right).
8  British Warriors and Scottish Voters: When ‘Rallying the Nation’  225

Labour Party
Conservative Party
2 Liberal Democrats
Scottish National
Pary
Scottish Socialist
Pary
Before vs after the invasion

1 Scottish Green Party

Independent Members

–1

–2

–3 –2 –1 0 1 2
Pro vs against the invasion

Fig. 8.2  The positions of individual interventions (marked by party affilia-

tion) according to their coordinates along two dimensions defined by the
MCA (see Fig. 8.1) of the joint occurrences of moral principles, social cleav-
ages and the timing of debates. Interventions from members of separatist
parties appear as filled grey triangles.

One of the clearest things to emerge from the MCA (see Fig. 8.1)
is the shift in category constructions that occurs from before the inva-
sion to when the invasion is imminent and after the invasion. More
precisely, there is a clear evolution in the anti-war camp where, pre-
invasion, the argument is organised around the Scotland—England
division. Later, it shifts to a series of other divisions, all based on the
opposition between the powerless and the powerful. In order to gain
more insight into the meaning of these shifts, it is helpful to consider
a set of extracts from the parliamentary debates. These are chosen from
226  Identity, Violence and Power

our coding categories in order to exemplify differences both between

pro- and anti-war camps and also between the pre-invasion and post-
invasion phases.
Let us start with an instructive plea for support. In Extract 1, a conser-
vative MSP pledges his loyalty to the Labour government’s war policies
on the grounds that they are acting-or, rather, asking the military to act-on
behalf of the nation.

Extract 1: The troops and their commander in chief need our support
I am associated with 603 City of Edinburgh squadron of the Royal
Auxiliary Air Force. It is now public knowledge that many of its reservists
and countless others have been called up. It is my conviction that if the
Government, with the support of the House of Commons, asks our armed
services to act on behalf of the nation, it must be given our total support.
(James Douglas-Hamilton, Conservative Party, 13 March 2003a)

This top-down logic, where a leader’s decision to go to war is to be sup-

ported by a loyal public, is typically reversed in the anti-war rhetoric.
Extract 2 provides a good example of putting the demands and desires
of the democratic public first. The fact that the “vast majority” is against
a war fought by “Britain and America acting alone” then becomes in
itself a moral argument against the war. The interesting detail lies in the
way the democratic public is defined by the SNP speaker here: referring
twice to ‘the people of Scotland’, she leaves no doubt as to the fact that
the Scottish public has a natural right to confer democratic legitimacy to
political decisions, even in the realm of foreign policies (contrary to its
formal rights granted by the UK constitution).

Extract 2: No public mandate for war

We know from polling evidence that the people of Scotland want this
debate and they want it to take place in Scotland. The vast majority are
opposed to Britain and America acting alone and 68 per cent of the people
of Scotland believe that Westminster should consult the Scottish Parliament
before launching an attack on Iraq. Fat chance—Blair is not even going to
consult the House of Commons, much less the Scottish Parliament. That
is, unfortunately, what Johann Lamont and her colleagues have to accept.
I wish that my Westminster colleagues had had the opportunity to vote on
8  British Warriors and Scottish Voters: When ‘Rallying the Nation’  227

the matter at Westminster, but thus far they have not. So much for
democracy.
(Roseanna Cunningham, Scottish National Party, 16 January 2003b)

Obviously, war supporters put much effort into presenting opposition

to the prime minister’s policies, rather than these policies themselves, as
the problem and the source of internal division. Often, as in Extract 3,
the message is that to “undermine” the prime minister’s policies can have
grave consequences. The dangerous arena of “international relations” does
not allow for divisiveness, because the stakes are “complex” and even “life
threatening”. But sometimes, as in Extract 4, opponents counter that the
costs of blind support can be higher than the costs of disobedience, and
that followership has no moral value in itself. This point is anchored here in
both universal lessons learnt from world history—the “Nuremberg trial”—
and the distinctive legacy of local history. The “declaration of Arbroath” is
a fourteenth-century declaration of Scottish independence which includes
a claim that the designated king is entitled to rule over Scotland by virtue
of his concrete merits for the Scottish people, and not just by divine right.

Extract 3: Opponents undermine our nation’s leadership

International relations are complex—they are life enhancing, but they
are also life threatening. When we speak in Parliament, we should consider
all the implications of our actions. When our Prime Minister has recog-
nised public concern and has moved to ask the UN to give Saddam Hussein
one last chance, we should not undermine his efforts to secure not just
peace and justice in Iraq, but stability and strength in the United Nations.
(Jack McConnell, Labour Party, 13 March 2003)

Extract 4: No blind support

Do we have to follow our leaders once they have established a policy?
The Nuremberg trial showed that people do not have to do that. Much
closer to home, the declaration of Arbroath set out the Scottish view of
leadership: Robert the Bruce was a great man who had saved us from the
English but, if he went wrong, he was out and we got another leader. We
do not have to follow Mr Bush and Mr Blair as they drive our collective car
over a precipice.
(Donald Gorrie, Liberal Democrats, 13 March 2003)
228  Identity, Violence and Power

In Extract 5, a conservative MSP presents democracy as a strong and

binding force among similar-minded people.

Extract 5: Democracy unites us

In a democracy such as ours, all are free to express their thoughts and
to live without fear of persecution. Indeed, that is a reason why thou-
sands of asylum seekers—many of whom come from Iraq—have sought
shelter on our shores. Oh, if only the situation were the same in their
homelands. My platform is based on an acceptance that no democrati-
cally elected leader of our nation would act in any way that was detrimen-
tal to the principles and objectives of the democracy that we enjoy in the
UK and to the overall well-being of our people. Furthermore, I do not
believe that such a leader would act against what he considers to be the
wider international interest
(Phill Gallie, Conservative Party, 16 January 2003)

Given that a broad consensus of democratic values is generally

assumed, this creates a highly inclusive category. The implication is
that only isolated individuals who challenge or undermine these val-
ues are on the other side. The conservative speaker refers to democracy
in the UK in order to locate the nation in an international context
and simultaneously makes a domestic argument. The international
argument concerns superiority; the fact that many people seek shelter
in the UK is presented as evidence of the widely shared aspiration to
live in a democracy. The domestic argument concerns legitimacy: to
know that a leader has been “democratically elected” is sufficient to
assume that he will faithfully represent the national interest and the
“wider international interest”.
Extract 6 is a subtle variant of the democratic unity argument. The
Labour speaker begins by acknowledging “a variety of views” and
­“division in the debate”; however, beneath these differences are similar
motives: “many people” are “deeply troubled” by the war. In this way, the
speaker portrays a democratic community of people that openly discusses
different stances precisely because they share each other’s concerns. This
closely tied community includes the speaker, his “party”, “constituency”,
“friends” and “home”.
8  British Warriors and Scottish Voters: When ‘Rallying the Nation’  229

Extract 6: Democracy unites us

Like many people throughout Scotland and the United Kingdom, I am
deeply troubled by the current international situation. We know that there
is a variety of views (…) The division in the debate is reflected in what I am
told by people in my party or my constituency or by my friends and in
what I am told in my own home. It is ironic that the one division that does
not exist, which some would like to see, is a division between the peoples
of Scotland and England on the issue. Indeed, as I reflect on our troubled
world and on the divisions that our international community faces, I won-
der how much more irrelevant it can be to have a party that is based entirely
on the desire to seek further division within our country.
(Johann Lamont, Labour Party, 13 March 2003)

Given this background of sincere debate on grave issues among people of

good faith, the division “between the peoples of Scotland and England”
can be portrayed as futile and instrumental. This argument implies that
the purity of democratic debate leaves no room for collective aspirations
along communal or national lines because they are an artefact of party
politics for those who “desire to seek further division within our country”
(which, in this case, represents the UK and not Scotland).
Such a perspective can be confronted with the rhetoric of those who
argue that, on the contrary, there is an authentic division between Scotland
and the rest of the UK. The core of this argument, exemplified in Extracts
7 and 8, is that the war reveals objective and chronic inequalities that disad-
vantage the people of Scotland. In Extract 7, the nationalist speaker refers
to the disproportionate contribution of Scottish troops in the war effort:

Extract 7: Vulnerable Scotland

As was the case in the Gulf war, the likelihood is that around one third
of the front-line troops will be from Scottish regiments. As far as our citi-
zens are concerned, we have to remember that it is those troops and their
families who will bear the brunt of any war in Iraq.
(Kay Ullrich, Scottish National Party, 16 January 2003)

Before the war, anti-war arguments almost exclusively referred to the cat-
egorical opposition between British war-mongers acting against Scottish
230  Identity, Violence and Power

interests. Once the war starts, this shifts to a variety of oppositions such
as the USA against Iraq or, more generically, elites against ordinary peo-
ple. This shift may reflect the fact that the objective community of fate
created between Scottish and (mainly) English troops at war have made
anti-British rhetoric a high-risk operation for Scottish politicians. Aware
that their constituencies might resent them for creating division among
the troops who are fighting and risking their lives together, the MSPs
have become attentive to include all of the troops in their concerns and
hence to shift the precise terms in which they oppose the powerful to the
powerless.
To deepen this point, consider the following account from an anti-war
MSP, during the Scottish parliamentary debate of 16 January 2003:

Extract 8: Vulnerable Scotland

As a Glasgow MSP, I could not in conscience contribute to the drum-
beats of war that are being stirred up in Westminster by those Dukes of
Plaza-Toro who, as usual, will be 4,000 or 5,000 miles behind the front line
… The other month I was on a train when a 19-year-old man recognised
me as being one of the MSPs who work in his area. He was going to
Glencorse barracks … I saw him go off into the morning mist and I thought,
“Aye—same as in the first and second world wars. Scots troops in first.”
(Dorothy-Grace Elder, Independent Member, 16 January 2003)

In the ballad by William Gilbert, it is written of the Duke of Plaza Toro

that “In enterprise of martial kind/When there was any fighting/He led
his regiment from behind”; hence, the extract perfectly depicts the war in
terms of cowardly war-mongers, who constitute the British Westminster
parliament, exploiting (as always) ordinary Scots. It corroborates and
populates our characterisation of the counterhegemonic discourse of the
anti-war camp. When the protagonist of this true story goes alone “off
into the morning mist” to his barrack and towards an uncertain future,
he appears to incarnate the fate of generations of Scots: “Aye—same as
in the first and second world wars. Scots troops in first”. Once this back-
ground settled, it is possible to discern an implicit point about many
Scots in what is explicitly said here only about a single Scot: he did not
enlist because he wanted ‘war’, but simply a ‘home’ and to ‘learn a trade’.
The army offered him a promise of social mobility. The implication seems
8  British Warriors and Scottish Voters: When ‘Rallying the Nation’  231

to be that social and economic disadvantage that young Scots have been
and are still facing explains why, over a century, Scottish troops have
fought and suffered in disproportionate numbers in Britain’s wars: mili-
tary subjugation reflects and prolongs economic subjugation of Scotland
within the UK.
It is then instructive to contrast Extract 8 with Extract 9. During the
debate of 13 March 2003, just as troops were about to go into action and
at the point (as can be seen from Fig. 8.2) when the Scottish—English
division begins to be supplanted by other constructions of the powerless
versus the powerful, MSP Margaret Ewing, from the Scottish National
Party, stated:

Extract 9: Virtuous troops led by immoral politicians

“our forces are an integral part of our communities. They are not aliens.
They do not live separately from us. They are our friends and neighbours.
They live next door. Their children go to our schools. They go to our hospi-
tals (…) I take offence at those people (…) who say that anyone who votes
against an immediate rush to war is in some way not supporting the troops.
I would be more convinced if I saw the Government and the Ministry of
Defence giving our troops the support that they deserve while they are out
there, because some of the stories that come home are horrendous. As legis-
lators, we have a duty and a responsibility to all our service personnel to give
them a legal mandate that is agreed internationally without reservations”
(Margaret Ewing, Scottish National Party, 13 March 2003)

At one level, this is very similar to the previous construction. Much work
is put into constituting the troops as ordinary people like you and me
in terms of who they are, where they live, what they do. Equally, there
is a clear contrast with the pro-war politicians who do not support the
troops and whose war is against the troops’ (and hence our) interests.
In this sense, there is a clear continuity between extracts. But there is
one obvious difference. There is no mention of Scotland or England or
Britain. The troops are not referred to as Scottish troops; they are troops
in general. Concern is not for some troops as Scots, but for all troops as
ordinary people. The government is not referred to as the Westminster
government but as the government full stop. Opposition is not to an
alien administration but to a powerful administration. So, this illustrates
232  Identity, Violence and Power

how, as the invasion becomes imminent, it is important to be seen as

supporting all troops and not just the Scots amongst them, but also how
the transition to this ‘governments versus ordinary people’ version of the
counterhegemonic narrative has been set up by the previous use of ‘Scots
versus English’.
Extract 10 is even more explicit in its way of opposing national leader-
ship, on the one hand, and the national community, including the armed
forces, on the other. In this rhetoric, Tony Blair does not represent the
interest and opinion of any wider public, not even of the ‘grassroots’ of
his own party. His trivial and selfish motive of fighting for ‘his political
life’ is contrasted to the grave and shared stakes of all those who are to
endanger their real lives (as well as the lives of others) in Iraq.

Extract 10: Undemocratic warlords

It is tragic, therefore, that at this critical time we have a Prime Minister
who is so belligerent and arrogant that he is prepared to defy grass-roots
opinion in his own party, the majority of public opinion in this country
and even the United Nations. It would be a supreme irony if his attempt to
bring about regime change in Iraq were to bring about regime change in
this country. According to some of the media, Tony Blair may be fighting
for his political life. So what? The lives of innocent men, women and chil-
dren in Iraq and the lives of our armed forces are far more important than
any politician’s career.
(Dennis Canavan, Independent Member, 13 March 2003)

A final set of arguments takes the cleavage between the people and a nar-
row elite to an international level. From this perspective, the tragedy of
the UK becomes that its leadership is more connected to the attitude of
foreign elites than to the interests of the overall public. Many interven-
tions denounce the arrogance of US elites and their superpower policies,
often together with the submissive policies of the UK leadership. Extract 11
is a case in point.

Extract 11: Fatal alliance with the US

It is a measure of Mr Blair’s closeness to Mr Bush that on Monday Mr
Blair said that if what he described as “justified” military action were to be
blocked by one member of the Security Council, he would be free to
8  British Warriors and Scottish Voters: When ‘Rallying the Nation’  233

c­ ommit the United Kingdom to war. That is Mr Blair talking in Mr Bush’s

language, and duplicating Mr Bush’s attitude to the United Nations. Mr
Bush is in the driving seat and Mr Blair is in the rear passenger seat.
(Colin Campbell, Scottish National Party, 16 January 2003)

While in the previous quote the argument was that Blair does not repre-
sent the national ingroup, here it is claimed that he works for a national
outgroup. The image of him sitting “in the rear passenger seat” in Bush’s
car expresses the idea that the head of government himself has lost con-
trol, and that alien interests now dictate British war policies—obviously,
a threatening prospect to all who are forced to embark on the trip.

8.2 T
 he Invasion of Iraq and the Scottish
Voters
The second set of findings are based on secondary analyses of the SSA
survey, conducted in 2003 by the Scottish Centre for Social Research
amongst a representative sample of the Scottish resident population aged
18 and above. Most interviews were conducted in May 2003, although
some occurred up to September of that year.
A first analysis tested whether the way that respondents express their
national identities is related to their stance on the war. In the survey,
four items addressed the two dimensions of national identification which
are of interest here: level of identification (Scottish versus British) and
quality of identification (attachment versus pride). The precise wordings
were “How closely attached do you feel to Scotland/Britain as a whole?”
and “How proud are you of being Scottish/British?” War opponents were
defined as respondents who (strongly) agreed with the statement “Britain
was wrong to go to war with Iraq”.
The findings reveal a clear and distinctive pattern of national identifi-
cation among war opponents: attachment to Scotland net of pride (and
net of identification to Britain) substantially increases the likelihood of
holding an anti-war stance. An observed exponential logistic regression
coefficient of 1.39 (the 95 % confidence interval ranged from 1.09 to
1.76) means that the odds of opposing the war increase by more than
234  Identity, Violence and Power

one and a third for every one-­point increase in attachment to Scotland

(on a 4-point scale), when other responses are held constant. By contrast,
the net effects of Scottish pride (0.68, confidence interval: 0.53–0.88)
and British attachment (0.81, confidence interval: 0.67–0.98) on anti-
war stances were both negative, while British pride was not significantly
related to stances on the war (0.88, confidence interval: 0.75–1.05).
Further logistic regression models addressed how personal opinions
on the war interact with exposure to newspapers holding a certain stance
on the war. The only media addressed by the SSA are daily newspapers.
To record readership, respondents were asked to indicate up to three
newspapers that they read either “normally” or “regularly”. For the pur-
poses of this study, we coded the resulting list of newspapers according
to their editorial positions on the war in Iraq during the invasion phase
(i.e., a systematic anti-war stance versus any other stance). To obtain a
reliable coding, we first conducted secondary analyses of the ‘Iraq War
Press Coverage’ database, which provided day-to-day content-analytical
data of the major British print media during the invasion phase. These
analyses clearly show that the three primary pro-war arguments (threat,
humanitarian intervention and regime change) were challenged much
more systematically in three newspapers (i.e., Daily Mirror, The Guardian
and The Independent) than in all the other newspapers within the sample.
Robertson’s (2004) content analysis complemented this by including
papers that are only available in Scotland and were not included in the
Iraq War Press Coverage database. Robertson showed that The Herald
distinguished itself among Scottish dailies by articulating a clear anti-­
war position from the start of the invasion. Respondents were defined as
readers of anti-war newspapers if they reported reading at least one of the
four aforementioned newspapers. Similarly, respondents were defined as
readers of other newspapers if they claimed to be regular readers of one of
the remaining newspapers (i.e., Daily Express, Daily Mail, Daily Star, The
Sun, Daily Telegraph, Financial Times, The Scotsman, The Aberdeen Press
and Journal, The Courier). The positions of these latter newspapers ranged
between ambivalent and clearly pro-war during the invasion period.
Most importantly, we were interested in explaining separatist votes
with regard to the interplay between personal opinions and newspaper
exposure. The upper portion of Table 8.3 shows that the overall odds
of voting for a separatist opposition party were exactly one and a half
8  British Warriors and Scottish Voters: When ‘Rallying the Nation’  235

Table 8.3  Multivariate predictors of the separatist opposition (above) versus the
Labour majority (below) vote: partial logistic regression coefficients
Model 1 Model 2 Model 3 Model 4
Separatist vote
War opponents 1.50** 1.43* 1.23 1.13
(1.13–1.98) (1.08–1.91) (0.90–1.67) (0.82–1.55)
Readers of anti-war – 2.08*** 1.09 1.15
newspapers (1.39–3.10) (0.56–2.14) (0.58–2.29)
War opponents х – – 3.00* 2.81*
Readers of anti-war (1.28–7.08) (1.17–6.75)
newspapers
Attached to Scotland – – – 1.50**
(1.11–2.03)
Proud of being – – – 1.22
Scottish (0.90–1.67)
Attached to Britain – – – 0.91
(0.73–1.15)
Proud of being British – – – 0.57***
(0.46–0.70)
Labour vote
War opponents 0.86 0.88 0.42** 0.43*
(0.65–1.14) (0.68–1.17) (0.26–0.69) (0.26–0.71)
Readers of other – 1.29 0.82 0.76
newspapers (0.97–1.73) (0.57–1.19) (0.52–1.10)
War opponents х – – 3.06*** 3.22***
Readers of other (1.68–5.58) (1.76–5.89)
newspapers
Attached to Scotland – – – 0.96
(0.72–1.27)
Proud of being – – – 1.17
Scottish (0.86–1.60)
Attached to Britain – – – 1.25
(0.99–1.56)
Proud of being British – – – 1.09
(0.89–1.34)
Note: Values significantly higher than 1 indicate a positive relationship between
predictor and outcome variables, values significantly lower than 1 a negative
relationship, stars indicate p-values (*p < 0.05, **p < 0.01, ***p < 0.001) and
numbers inserted in brackets provide the boundaries of the 95 % confidence
interval of the logistic regression coefficient

times higher for war opponents than for those who supported the war
or who had no clear opinion (Model 1). The second model shows that
the strength of this association does not substantially differ after control-
ling for the effect of reading an anti-war newspaper. That is, the effect of
236  Identity, Violence and Power

personal opinion regarding the war is not mediated by the type of news-
paper read. However, the strong interaction effect introduced in Model
3 shows that personal opinion is moderated by newspaper exposure. Only
amongst the readership of an anti-war newspaper did personal opinions
against the war translate into separatist votes. Finally, Model 4 shows that
this pattern holds after controlling for multi-dimensional national iden-
tification, although Scottish attachment and British pride are correlated
with separatist voting (in opposite directions). To conclude, these analy-
ses highlight the fact that the newspapers which disseminated anti-war
positions played an important role in the mobilisation of war opponents
in favour of separatist parties. By contrast, further outcomes (not shown
here) suggested that reading newspapers that disseminated pro-war or
ambivalent positions did not affect the relationship between anti-war
opinions and separatist voting.
The lower portion of Table 8.3 displays equivalent models for voting
for the Labour Party in power. Again, the most significant outcome is the
strong interaction of personal opinion on the war and media exposure.
In this case, reading a newspaper with a supportive or ambivalent stance
regarding the war made the critical difference. War opponents who did
not read such a newspaper were significantly less likely to vote for the
Labour Party. Among the readership of these newspapers, however, the
effect became insignificant and was even reversed. Hence, newspapers
that disseminated pro-war positions appeared to play a role in the demo-
bilisation of war opponents and in all likelihood limited further electoral
losses for the ruling Labour Party.

8.3 E
 lite Constructions and Popular
Understandings of War and Nation
To summarise the findings from the first part of our study, it is clear that
the discourse of the pro- and anti-war camps was constructed around
opposed versions of the groups and identities involved in the conflict.
The pro-war camp referred to a narrative of liberation in which war was
necessary to defend ‘ourselves’ and to alleviate the sufferings of others.
The anti-war camp proposed a narrative of aggression in which war was
8  British Warriors and Scottish Voters: When ‘Rallying the Nation’  237

imposed on ‘us’ by others and was to our detriment. In other words, for
those in favour, this was ‘our war’, based on our values and advancing our
interest. For those against, this was ‘their war’, violating our values and to
the detriment of our interest.
These findings further highlight how the arguments of both the pro-
and anti-war camp are rooted in well-established common sense ways of
viewing the world. Those for the war use a notion of the civilised world
fighting an evil dictator which has become particularly powerful since
World War II, which was certainly central to narratives about the first
Gulf War (Herrera & Reicher, 1998) and which is widely disseminated
within and beyond the UK. Those against the war initially rooted their
opposition in the notion of Scotland’s domination by England within
the UK (and the UK parliament which endorsed the war), and later on,
this set up alternative ways of construing ‘our’ domination by the masters
of war. The significance of this ‘Scottish versus English’ construction is
not only its familiarity and ubiquity (especially during a Scottish election
campaign) but also the fact that it is relatively conventional. Thus, the
anti-war position in Scotland can be anchored in a mainstream view of
the world, and the nature of the debate is marked by the availability of a
respectable way of saying ‘it’s not our war’.
But it isn’t just that such a construction is available. Looking at the
argumentative context of the debate, we see, first, how much effort anti-­
war speakers devote to setting up a category system that is congruent
with their stance on the war. Some four out of five of their arguments
were devoted to who is against whom in the conflict (in contrast to about
one out of three arguments only within the pro-war camp). Perhaps this
is due to the fact that the ‘official version’ supported by the ruling par-
ties in both Westminster and the Scottish Parliament, by their publicity
machines and by the majority of the media takes for granted that this is
‘our war’. In order to challenge their influence, great efforts are necessary
to expose and establish an alternative perspective.
Second, we see that these efforts were collective and that they were
conducted with great consistency. Every single intervention from a mem-
ber of a separatist party rooted their argument in the idea that this was
‘not our war’. In the first debate, this stance predominantly translated
into ‘England’s war not Scotland’s’. Then, in subsequent debates, they
238  Identity, Violence and Power

all drew on this to sustain other versions of ‘it’s their war’. No separatist
MSP stepped out of the line, which created a stark contrast to the politi-
cal cacophony displayed by unionist MSPs in general, and by members
of the ruling Labour Party in particular.
This takes us to the importance of the changing course of events.
While the separatists were consistent, they certainly weren’t inflexible.
As events changed, as the possibility of war became the near certainty of
war and then as troops entered into the firing line, so the precise nature
of the anti-war categories changed. Scottish-English gave way to other
versions of a bellicose and dominant ‘them’ imposing war upon ‘us’. The
important thing about theses shifts, we have argued, is that the earlier
division contrasts Scottish soldiers and their families to English (or Welsh
or Northern Irish) troops and their families. The latter includes all British
(and indeed allied) troops and their families as the ‘poor bloody infantry’
who are as ever traduced by their leaders in war. In making the shift, anti-­
war proponents cannot be accused of fostering divisions amongst the
troops and thereby endangering them all. They cannot be dismissed as
talking irresponsibly and ignoring the new realities of war. By being both
consistent (in terms of their overall construction) and flexible (in terms of
the precise categories they use), these oppositional politicians apply what
have been classically shown to be the optimal conditions for contesting
the dominant viewpoint (Mugny, 1982).
Turning now to the second part of the findings, focusing on popular
opinion, there are two key findings that we wish to stress. The first is that
being anti-war is clearly related to seeing oneself as Scottish. However,
it is not just that anti-war respondents feel Scottish, but that they do so
without necessarily feeling pride in Scotland. This pattern is akin to what
some have dubbed ‘critical attachment’ or ‘constructive patriotism’ (see
Chap. 7). Often, however, critical attachment or patriotism is seen as
an individual orientation to the nation. Here we suggest that it is more
a matter of assimilating a prevalent discourse in Scotland, where being
critical is part of what it means to be Scottish (see Reicher & Hopkins,
2001). Rather than promoting a stance of ‘my country right or wrong’,
the anti-war elites advance a notion of Scots as a less bellicose people
who will challenge anything and anybody that violates their values—
and hence who challenge the official drive to war. Hence, we can see a
8  British Warriors and Scottish Voters: When ‘Rallying the Nation’  239

correspondence between the anti-war discourse of elites and the under-

standings of anti-war sections of the population.
The second key finding concerns the role of the press in the relation-
ship between private opinion and political behaviour (in this case voting
for anti-war and pro-separatism candidates). From a classic perspective
of social influence or cognitive consistency one would expect that this
relationship is in terms of mediation, that is, that those who are anti-war
are led to read papers that are in tune with their stance and this in turn
engages them so as to vote for anti-war/separatist parties, or else that
those who read anti-war papers develop anti-war opinions that in turn
make them into anti-war/separatist voters.
However, contrary to this common wisdom, we have found no evi-
dence of a mediational relationship. Rather, and less intuitively perhaps,
our findings clearly support a moderational relationship. Only those with
anti-war opinions who also read anti-war papers are more likely to vote
for anti-war/separatist parties. Equally, those with anti-war opinions who
read pro-war papers (or papers that are ambivalent about the war) are not
less likely to vote for the Labour Party—the party of government which
pursued the war. These findings might appear counterintuitive. However,
they are consistent with the position we advocated in Chap. 3 (and with,
notably, Paluck’s analysis of the role of the mass media in facilitating
social behaviours of conformity or opposition to authorities). That is,
the impact of the media lies not so much in changing personal beliefs or
deeply ingrained opinions, as in changing the perception of social norms.
It is by telling us that our opinions are shared by others—that they are
normative—that the media affects what we do.

8.4 Conclusion
To summarise, in this chapter we have shown the work done by political
elites in rooting their accounts of identity in various dimensions of con-
text. We have shown in particular how those challenging the status quo
are able to draw on a chronically available understanding of Scottish and
British interests as opposed, and how they spend more effort than those
defending the status quo in creating an explicit, consistent and flexible
240  Identity, Violence and Power

definition of the Scottish interest as anti-war. We have also shown that

opposition to the war amongst the population at large is linked to a simi-
lar understanding of Scottishness as characterised by a critical and ques-
tioning relationship to authority.
In the light of these findings, it is possible to account for the conver-
gence between the structure of elite discourse and of popular understand-
ings—which was more evident for the anti-war separatist camps than for
the ruling majority—in several ways. While top-down theorists might
interpret it as a consequence of effective political mobilisation (opinions
communicated by elites shape mass opinion and behaviour), advocates
of bottom-up approaches would rather emphasise that elites adapt their
rhetoric opportunistically to what their audiences want to hear. In all
likelihood, there is a note of truth in both positions: on the one hand, we
have shown how elite discourse is effective to the extent that it is rooted
in pre-existing, widely shared, popular understandings; on the other we
have shown how elites adapt these understandings to the present context
in order to drive forward their own specific agendas.

References
Billig, M. (1995). Banal nationalism. London: Sage Publications Ltd.
Elcheroth, G., & Reicher, S. (2014). “Not our war, not our country”: Contents
and contexts of Scotish political rhetoric and popular understandings during
the invasion of Iraq. British Journal of Social Psychology, 53(1), 112–133.
Herrera, M., & Reicher, S. (1998). Making sides and taking sides: An analysis
of salient images and category constructions for pro- and anti-Gulf War
respondents. European Journal of Social Psychology, 28(6), 981–993.
Lewis, J. (2004). Television, public opinion and the war in Iraq: The case of
Britain. International Journal of Public Opinion Research, 16(3), 295–310.
Lipson, M. (2009). “If it wasn’t rolling, it never happened”: The role or visual
elements in television news. In L. Harman & L. Lombardo (Eds.), Evaluation
and stance in war news: A linguistic analysis of American, British and Italian
television news reporting of the 2003 Iraqi War (pp. 140–169). London:
Continuum.
Mugny, G. (1982). The power of minorities. London: Academic Press.
Reicher, S., & Hopkins, N. (2001). Self and nation. London: Sage.
 8  British Warriors and Scottish Voters: When ‘Rallying the Nation’  241

Robertson, J. W. (2004). People’s watchdogs or government poodles? Scotland’s

national broadsheets and the second Iraq war. European Journal of
Communication, 19(4), 457–482.
Rutland, A., Cinnirella, M., & Simpson, R. (2008). Stability and variability in
national and European self-identification. European Psychologist, 13(4),
267–276.
Sindic, D., & Reicher, S. D. (2009). “Our way of life is worth defending”:
Testing a model of attitudes towards superordinate group membership
through a study of Scots’ attitudes towards Britain. European Journal of Social
Psychology, 39(1), 114–129.
Conclusion

Old Issues and New Questions

We began this book with a plea for perplexity: our expressed intention
was to disrupt old certainties and to raise new questions both about
the nature of identity, violence and power and about the relationship
between them. We can now summarise such new questions as have arisen
from our analysis of the literature and from our own case studies. We can
also consider the extent to which we have progressed from new questions
to new answers.
Our starting point lay in the observation that both scholarly and
popular debates about collective violence still tend to be organised along
an opposition between ‘primordialist’ and ‘instrumentalist’ positions.
Primordialists assume an inbuilt tension between people of different
descent, culture or faith and presuppose that people who share certain
bonds will stand together against outsiders. To quote the American satirist
Tom Lehrer from his ironic song ‘National Brotherhood Week’: “Oh, the
white folks hate the black folks/And the black folks hate the white folks/
To hate all but the right folks/Is an old established rule”. Instrumentalists,
by contrast, stress how identities are manipulated, histories rewritten,

© The Author(s) 2017 243

G. Elcheroth, S. Reicher, Identity, Violence and Power,
DOI 10.1057/978-1-137-31728-5
244  Conclusion

group boundaries redrawn and new enemies created, when it serves the
political agenda of powerful elites. Here a more weighty literary analogy
can be made to George Orwell’s dystopian masterpiece 1984. As popu-
larly understood, this encapsulates the power of propaganda in the way
that, in a moment, the people of Oceania can be led to forget their hatred
of East Asians and transfer all animosity to the Eurasians.
However, our concern was less with the difference between these two
positions than with the things that they have in common. Put simply,
both provide a linear account of how we arrive at the same outcome:
collective violence. There is some dispute as to the ordering of variables
in this account: does identity come first and power only matters to the
extent that it affects the exercise of identity-based animosities, or does
power come first and identity only matters as a tool in the hands of
­leaders? Or rather, to use a motoring metaphor, primordialists contend
that violence is the destination, identity is the driver, and power is the
vehicle. Instrumentalists also regard violence as the destination, but the
remaining two terms are shifted around: power drives, identity is driven.
Throughout this book, we have been highlighting how such concep-
tualisations constrain and construct the nature of the debate. They limit
our curiosity and thereby limit our knowledge. There is far more to dis-
cuss than which of identity and power is the driver and which is the
vehicle. There are so many more issues, so many more uncertainties and
so many more questions.
To start with, we need to interrogate the terms of the debate. In dif-
ferent ways, primordialists and instrumentalists take the nature of iden-
tity as a given in any particular dispute—the primordialists because they
tend to assume that people always see each other in terms of the same
categories whatever the situation, the instrumentalists because they tend
to assume that people will accept whatever categories are presented to
them by elites and are incapable of dissenting, let alone resisting. Yet, this
misses the basic point that, in many cases, the nature of the categories to
a dispute is contested and indeed much of the dispute (and certainly its
outcome) is about precisely what these categories are.
For instance, to invoke a significant moment in recent world history,
was the first Gulf War of 1991 about warmongers pursuing their inter-
ests regardless of the cost to the rest of the population or was it about
  Conclusion 
   245

the democratic world facing up to a dictator? Opposition or support for

the war hinged on how people saw the categories involved (Herrera &
Reicher, 1998). More generally, the answer to the well-worn question
‘which side are you on’ depends on how one draws the sides.
In the same way as we need to problematise the ‘identity’ term in our
models, we also need to question ‘power’ and ‘violence’. Power is not
just a thing that is either there or not, which groups possess in order to
implement their urges or else elites have in order to sway groups. Indeed,
in part at least, power is something that comes about by constituting
groups and leading people to act together as group members. That is,
state leaders consolidate their power by getting people to see themselves
and to act together as members of a national community whose bound-
aries coincide with the boundaries of the state. By contrast, the power
of, say, union leaders depends on people acting together as members of
different classes within the nation. As for violence, the issue here is even
more critical, albeit conceptually somewhat different. As long as violence
is treated as the terminus of our enquiries, it will always remain some-
what opaque to us, because we lack a criterion beyond violence to differ-
entiate the features of violence. We cannot ask what to include under the
rubric of violence: does it simply involve ongoing physical and mental
harm to others, does it include the threat of future harm, does it involve
the possibility of past harm reoccurring? Nor can we ask what it is about
violence which produces other outcomes. Is it harm alone or fear of harm
or indeed something else? But all this changes once we regard violence as
more than an end point, as more than a mere product of prior processes,
and as something that itself produces new outcomes.
This takes us on to a further set of issues relating not to identity, vio-
lence and power as separate terms, but to the dynamics that exist between
them. To reiterate, the conventional approach is to identity/violence and
power violence in linear terms such that once we get to violence we have
got to the punchline of our narrative. But that is to end the story just
before we can learn something important from it. It terminates analysis at
its most critical juncture. For when violence breaks out (or people believe
in the imminent possibility of violence), no destination has been reached.
Rather, the nature of the journey and the means by which it is under-
taken are affected, as are the chances of reaching different d ­ estinations.
246  Conclusion

Violence is of interest not only in terms of what came before but also
because of the way it affects what lies beyond.
In other words, if we are to advance our understanding of collective
conflict and violence we need to address how violence not only arises and
escalates out of identity and power processes but how it also transforms
collective identities, how it shapes ongoing power struggles and how it
reshuffles our possible futures. That was the aim of the second part of our
text.

Towards Triangularity
A central aspect of our analysis has been to conceptualise terms in ways
that are much more relational and communicational than is conven-
tional. This starts with identity. Much research proceeds by asking people
to what extent they see themselves as a man or woman, as white or black,
as Scottish or Swiss, or whatever. If they choose to tick the appropriate
boxes on our questionnaires, we accept what they say and classify them as
‘high identifiers’. No one else is there to gainsay them.
But this is a strangely utopian world. In real life, we might well make
claims to certain identities, but then discover that what sounds entirely
reasonable to us sounds strange to others. If the first author of this book,
born in Luxemburg and living in Lausanne, defines himself as Swiss, his
claim might passed unchallenged at an international conference, but his
Swiss neighbours are likely to question how it goes together with the
colour of his passport, with the intonation of his French, or with his
inability to stand straight on a pair of skis. If the second author, born
in England and living in St. Andrews, claims Scottishness in an unmis-
takable English accent, will he be embraced by others as a fellow Scot?
Would he dare walk into a local pub wearing a kilt knowing that he could
be met with derision? Identity, then, is about more than self-perceptions
and self-definitions. Identities involve the ways we are positioned and
the ways we act in the world, which are as much about the ways others
treat us as the ways we see ourselves. Indeed, they are about the way we
anticipate that others will see and treat us and the way we constrain our
own claims as a result.
  Conclusion 
   247

In suchlike ways, we become aware of ourselves and begin to tailor our

actions and cognitions to the way we believe we will be treated. What we
know of what others think of us is therefore critical and hence communi-
cation becomes central to the construction of identity.
Similar constructions apply to issues of power. Our own ability to act
depends upon how others respond to us. Will they support us, in which
case we are able, together, to achieve what would have been impossible
alone. Do they ignore us, or do they oppose us, in which case we are
unlikely to be able to achieve very much.
Thus, when we anticipate the opposition of others and when we there-
fore anticipate that our acts will achieve little beyond earning the oppro-
brium of others, we are likely to say nothing and do nothing whatever our
beliefs or inclinations. This can then entail a spiral of silence (as defined
by Noelle-Neumann, 1984/1993) whereby others see no signs of support
for similar views and hence say nothing themselves. Eventually, action
becomes difficult for everyone, as no one knows where the others stand.
Such a sense of epistemic isolation has been of central importance to
our argument. The ability to speak and act—especially when it comes to
challenging a powerful status quo—depends upon knowing we will not
be alone. Hence one of the ways in which the powerful can maintain this
status quo is by disrupting the ability of people to communicate their
opposition. Earlier, we cited how Orwell’s 1984 is understood to illustrate
the supposed ability of elites to manipulate identities through recounting
how the people of Oceania are led to turn against the people of Eurasia.
But when we look at Orwell’s argument in more detail, we see that he well
understood how domination is maintained less by affecting people’s own
views than by restricting their knowledge of the views of others:

The Party said that Oceania had never been in alliance with Eurasia. He,
Winston Smith, knew that Oceania had been in alliance with Eurasia as short
a time as four years ago. But where did that knowledge exist? Only in his own
consciousness, which in any case must soon be annihilated. And if all others
accepted the lie which the Party imposed—if all records told the same tale—
then the lie passed into history and became truth. (Orwell, 19491)

 For the full text, see http://www.george-orwell.org/1984/2.html.

1
248  Conclusion

Now, if both identity and power are, at least in part, constituted through
the ways that others relate to us, and through our ability to gain knowl-
edge of how they relate to us, then anything which changes such rela-
tions/knowledge of relations will serve to reconfigure identity and power.
This is how violence enters our account. For we argue that violence is
a particularly potent means of changing relations and communications
between people.
Consider, for instance, the case where Hindu extremists have ram-
paged through a street, attacking all the Muslims and leaving alone those
houses marked as Hindu, or the case where Serb vigilantes have gone
into a village, compelled local Serbs to divulge the location of the Croats
amongst them and then assaulted those so identified. After that, how
can things be the same again? The mere possibility that your neighbour
may identify you as a Muslim/Croat—with the terrible consequences
that ensue in a climate of violence—means that you are led to act with
the presumption that you may be viewed as such and also to see them
as a Hindu/Serb—even if you had never done so before. And, as you are
distanced from them and communication with them becomes difficult,
the possibility of breaching these presumptions fades further.
In effect, violence radically alters the contingencies of acting on the
basis of different self-definitions, and the implicit risk calculus they
superpose on social relations. If you act towards your erstwhile neighbour
as still your neighbour, and if you get it right, you will perhaps receive
a measure of companionship and support. But if you get it wrong, you
and your family may be killed in your beds. Even if the odds of getting
it wrong remain low in comparison with getting it right, the perceived
costs or benefits associated with either scenario can prevail over the odds.
In other words, the peculiarity of a violent environment resides in the
fact that it leads people to bet on the unlikely, and to align their behaviour
with the worst-case scenario. What is more, even if individuals opt for
bravery and decide to show solidarity with the new, ethnically defined,
other—treat them as what they have been so far, a simple neighbour—
members of their new (ethnic) ingroup might not let them do so, for fear
of supporting an enemy in their midst.
On the one hand, then, our argument involves a reconceptualisation
of violence as a driving force and not just as a product of prior forces, and
  Conclusion 
   249

an analysis of the ways in which violence produces its effects. To restate

our case: violence serves to enforce new solidarities and silence dissent.
If ‘we’ are under attack from ‘them’ we cannot interact or listen to ‘them’
and we must speak and act as one to prevail.
By the same token, however, our argument serves as a reconceptualisa-
tion of the relationship between violence, identity and power. We have
shown that violence is not a terminus but a way station at a ­crossroads and
that it feeds into identity and power as much as it derives from them. In
sum, we call for an analytic framework which gives up the neat parsimony
of a linear conceptual model—where identity conflicts/power struggles
lead to violence through influencing the use of power/the manipulation
of identity—and which trades it in for the increased realism of a dynamic
triangular model. In such a model there is no set starting point or end
point. The three nodes—identity, violence and power—alternate their
analytic status and each potentially functions as a cause, a mediator or
an outcome at different moments in time, or at different steps of the
analysis.
Trading parsimony for realism? Given the importance of parsimony as
a principle for evaluating analytic models, that is certainly a risky trade.
It becomes profitable only when more contextualisation sheds light on
critical processes that a more parsimonious analysis would have over-
looked—and that takes us to the third part of our text, the case studies.

Learning from the World
Our three studies addressed three very different areas of conflict: firstly,
Hindu nationalism and communalist tensions in India; secondly, war and
ethnic violence in the former Yugoslavia; thirdly, mobilisations against
the invasion of Iraq in Scotland. The first and the most obvious point to
be made from all three analyses concerns the contested nature of social
categories and the centrality of such contestation to the nature of the
conflict. In the Indian case, the contestation concerns both what catego-
ries are involved, and how those categories should be defined. Or rather,
by redefining the meaning of the core category ‘India’, the nature of the
intergroup relations in which Indians are involved is changed. Thus, by
250  Conclusion

construing the country as an essentially Hindu nation (as symbolised by

the cow), Hindu nationalists not only exclude Muslims from the national
category but constitute them as a threat both symbolically (they kill cows)
and practically (they undermine the economy).
In the former Yugoslavia we see even more starkly the ways in which cat-
egories are contested and categories change. The major issue in the region is
how diverse and cosmopolitan populations became frozen into ethnic cat-
egories. How come people who married across ethnic boundaries, who pri-
oritised class above ethnicity, who often were ignorant of the ethnic origins
of even close friends came to act and see the world through an ethnic prism?
To refer to ethnic categories as if they were timeless, sheds no light on the
process. Indeed to treat ethnic categories as timeless is part of the process.
Finally, in Scotland, the question of whether to support the Iraq inva-
sion or not depended on whether it was regarded as ‘our war’ and that in
turn depended upon how people defined who ‘we’ was. Was the conflict
one in which British democracy stood together against an evil dictator
with all the echoes of a united Britain in the Second World War? Or was
it one in which English imperialism was having its last throw—an impe-
rialism in which, it was claimed, Scotland shared no part and indeed had
been more a victim that a contributor to it in the past? On such matters
popular support for the war—and hence the political ability to prosecute
the war—depended.
But, each of the three case studies also tells something more about the
nature and dynamics of identity, violence and power. The Indian case
tells us how violence is not simply a product and alerts us to the multiple
ways in which identity, violence and power influence, and are influenced
by, each other. When we first went into the tent of the Vishwa Hindu
Parishad in Allahabad and came across the vile depictions of Muslims—
so reminiscent of Nazi anti-­Semitic caricatures—it seemed to us that the
key issue was indeed to explain why such celebrations of intergroup vio-
lence were possible. But the more we analysed the images, we saw how
the outgroup target varied while the overall message stayed the same: ‘our
political rivals expose you to threat while we defend you’. Increasingly it
became clear that the invocation of intergroup violence was a means of
altering relations of power and influence within the Hindu community
itself. It was a means of saying that Hindu Nationalists represent interests
  Conclusion 
   251

of all Hindus, while others don’t. In this instance violence was a tool
designed to help substantiate the claim that ‘we are of you, we understand
your experience, we act for you’—the key claims of effective leadership
(Haslam, Reicher, & Platow, 2011).
But if influence and power are outputs of invoking violence, this is
not to deny that the exercise of power is also an input to violent social
­relations. Indeed, we saw how the Gujarat riots of 2002 are a classic exam-
ple of authorities condoning violence and failing to intervene against it.
As we have stressed, our aim is not to substitute one linear model of
identity-violence-power for another, but rather to discard linearity in its
entirety, to examine the ways in which each term relates to others in dif-
ferent ways at different points in time. Hindu nationalists both use power
to enable violence to occur and use this violence to consolidate power. As
we write, with the Bharatiya Janata Party (Indian people’s party—BJP)
in government and Narendra Modi in office as Prime Minister, it is a
depressingly effective strategy.
Moving now to the former Yugoslavia, we see even more clearly and in
more detail how violence becomes a driving force, specifically in reconfig-
uring identities. The figures are quite stark. Ethnic and religious identity
became entrenched only in the aftermath of violent conflict. Likewise,
forms of cross-ethnic solidarity, which were widespread in 1990, had all
but disappeared by 2006. To recap, perhaps the most dramatic of all our
figures concerns support for ‘mixed’ marriages (the most intense of soli-
darities). Over the period, this fell from 63% to 22% in Croatia and from
67% to 17% in Bosnia among young adults.
The point here, though, is that not only did war change social cat-
egories but also those categories stayed changed after the war was over.
It is not only present violence but also the shadow of past violence that
configures identity and power. That is particularly clear in our analysis
of post-war Croatia, as are the reasons. On the one hand, past violence
can always be invoked to support the narrative of a nation under siege,
always vulnerable, always under potential attack, where survival depends
upon further enforced solidarity. Those who criticise Croatia succour its
enemies and pain its defenders.
On the other hand, where violence forges a new identity (in this case
Croatian nationhood) then to question that violence or those who carried
252  Conclusion

out that violence is not to criticise a contingent aspect of the group, but
its very existence. It is impossible to both claim loyalty to the group and
to oppose the conditions of its formation. So individual critics are left
with one of two choices: either they stay silent or they place themselves in
exile. Either choice contributes to the epistemic isolation of other critics
and renders dissent ever more improbable.
Moving once more, this time to Scotland, we gain further insights
into the ways that particular constructions of identity gain purchase—
and also when they fail to gain purchase. For the first and most obvious
point here is that the ‘official version’ in support of the Gulf war, one
supported both by the UK government and by the devolved Scottish
government, gained only limited support. Between a third and a half
of the Scottish electorate declared themselves ‘angry’ or else ‘disgusted’
by the invasion. What is more, an analysis of political debates suggests
that this opposition is bound up with a rejection of the ‘official’ iden-
tity narrative. As we argue above, supporters of the war characterised
it as a defence of British democratic values against a foreign dictator;
opponents (at least initially) characterised it as an assertion of English
imperialism against a weaker foe.
So why did the dominant version fail, and why particularly in Scotland?
Three factors emerge as particularly important. The first is the way that
the oppositional narrative resonates with other familiar narratives. Ever
since the Thatcherite deindustrialisation of the 1980s hit Scotland’s tradi-
tional heavy industries particularly hard, there has been a strong sense of
Scotland as a victim of English domination. Scotland could be character-
ised as a victim of English colonialism (sometimes dubbed ‘the wretched
of the north’ in clear reference to Fanon’s, 1961/2004, classic anti-colonial
text). The anti-war position was easily assimilated to this familiar story.
Second, the opposition showed considerable rhetorical skill—­witcraft,
to use Billig’s (1987) term—in expounding their position. They devoted
particular attention to establishing an alternative construction of iden-
tities to the mainstream and, unlike the mainstream, they were both
consistent (speaking with one voice) and yet flexible in adapting their
‘Goliath versus David’ construction to new circumstances once the war
had started. Consistency combined with flexibility constitutes the ideal
characteristics for minority influence (Mugny, 1982).
  Conclusion 
   253

Third, there was a lively anti-war media in Scotland to let people know
that, if they opposed the war, they were not alone. Unlike the epistemic
isolation which curtailed dissent in Croatia, epistemic validation was
available in the Scottish context and where those against the war were
exposed to such validation (by reading the anti-war media) they were
willing to give political expression to their views.
So, bringing the case studies together, they underline not only the con-
tingent nature of identity, but also the inadequacy of a linear approach
to identity-violence-power. They sustain the need to treat violence as not
just a product but as productive of identity/power; They show how vio-
lence functions by altering epistemic relations between actors; and they
highlight the role of epistemic isolation in sustaining dominant narra-
tives, and hence the importance of maintaining and creating epistemic
fluidity in order to enable opposition. But there is one more key element
which we need to add to this list and which, like the contestation of social
categories, is evident in all three of the case studies.
That extra element is a fresh look at the dynamics of mobilisation,
which has been imminent throughout our discussions. Indeed, mobili-
sation occurs at multiple levels. To start with, insofar as categories are
not naturally given, particular categorical constructions are actively pro-
moted by leaders and active choices are made by followers as to whether
to accept or reject them.
Next, those categories create new constituencies (and destroy old
ones) which deliver the social power to impact the social fabric. In India,
the consolidation of a Hindu constituency has provided a route to state
power for the Hindu nationalist BJP. In Croatia, the occlusion of cleav-
ages based on political and economic inequality diffused the opposition
to Tudjman for a critical period and gave time for the old apparatchiks to
become new oligarchs. In Scotland, the formation of a nationalist con-
stituency that feels ill-served by the UK’s Westminster Parliament lies
behind the relentless rise of Scottish nationalism to the extent that, as we
write in 2016, parties supporting independence now have an absolute
stranglehold on the electoral landscape.
These successes illustrate the fact that the processes we describe do
not just create new constituencies, but consolidate particular individuals
and parties in leadership positions for those constituencies. By i­nvoking,
254  Conclusion

facilitating or permitting violent confrontations with Muslims, Modi and

the BJP were able to position themselves as defenders of the Hindus.
Through association with the struggle which founded an independent
Croatian state, Tudjman’s position was rendered inviolate and critics
of his regime, or of his policies of privatisation of public goods, were
excluded from the nation. Through dissociation from violence that is seen
as imposed on the Scottish people, Alex Salmond became the only First
Minister of Scotland with an absolute majority of seats in the devolved
Scottish Parliament.

Learning from Perplexity
Having summarised what we have learnt from our studies of specific
case studies, let us now conclude by considering the general lessons that
emerge from our overall analysis. We divide these into conceptual, meth-
odological, and practical implications.
Conceptually, we have sought to challenge models which are based on
identifying root causes and predicting outcomes. But why bother with
analysis if we cannot foretell the future? To borrow from a longstand-
ing critique of economics, aren’t we like the forecaster who cannot tell
you what the weather will be like tomorrow, who probably cannot tell
you whether it will rain today, but who can explain why you needed an
umbrella yesterday?
Our intellectual and practical cases are intermeshed. The reasons for
rejecting the root cause and prediction approach stem from our critique
of linear approaches to identity, violence and power. As we have stressed,
it is more helpful to view these as elements in an interconnected system
where each can be a precursor or an outcome, a moderator or a media-
tor to the others. Moreover, just as each element may impact the others,
so it may itself be impacted and change in the process. As a result, the
route through which one got into a particular configuration of elements
is not necessarily the best way out. It may not even be a possible way out
since, as one moves through the terrain of identity, violence and power,
the terrain itself is changed. Therefore, instead of undertaking the futile
  Conclusion 
   255

enterprise of identifying root causes of a conflict, it is generally more pro-

ductive to look for a variety of factors that can play the role of levers for
change, whether or not these factors were drivers in the historic process
that brought about a current crisis or stalemate. For example, electoral
settings that create a political incentive to mobilise ethnic hatred will
always add an additional burden to a society that has already been the
theatre of ethnic violence (whether or not there has been a causal rela-
tion between elections and violence in the past ); humanitarian interven-
tions that contain the human consequences of a conflict are always likely
to affect—ideally, to broaden—future options for handling a conflict
(because it will necessarily make a difference for the subsequent dynamic
which magnitude of violence has being reached).
The triangular dynamics between identity, violence and power also
heighten the need for a representational and a mobilisational view of
understanding and action. Issues of identity and power are not given ‘out
there’ in the world such that we read them off in ways that are pre-given
by our cognitive architecture. Rather, we are confronted with an inher-
ently opaque world and we are offered ways of making sense of it by our
peers, by the media and by our leaders. Sense making is a slippery and
non-deterministic process. In dealing with identity and power (and hence
violence) it is made all the more slippery and unpredictable by the fact
that we are dealing not only with our own understandings of the world
but also with our understandings of how others understand the world.
So if we cannot identify root causes and we cannot predict futures,
what can we do? At the outset, we argued that we can precisely puncture
certainties—often self-fulfilling—by pointing out the contingent nature
of identity, violence and power and the dynamics between them. Indeed,
by opening uncertainties we allow new possibilities and new choices
about the future. Hindus are not doomed to riot against Muslims, Croats
are not doomed to build their nation as a fortress against Serbs and Scots
are not doomed to side with the English to colonise the world. To argue
as if they are is to buy in to the arguments of those who seek to essen-
tialise these options by any means. It is, effectively, to reward them for
their use of violence. Therefore, the importance of chipping away at cer-
tainties should not be underestimated.
256  Conclusion

But now, at the end of our book, perhaps we can go a little further.
While we still cannot predict outcomes, we can at least identify some
of the levers by which different configurations of identity, violence and
power may be brought about. That is, even if one cannot definitively
say what the outcome of a particular sense-making process will be, the
dynamic triangular model developed in this book points to specific
­processes through which certain representations outweigh others, or new
alternatives open up. It invites analysts of social change or activists of
social justice to look more closely at what people think others think—
and at which new channels of communication must open in order to
make available the understandings of others. In the same way that our
model opens new questions, so it identifies new sites where efforts might
be more profitably applied in order to produce a less oppressive and less
violent world.
One direct implication concerns the relationship between generality
and specificity. In moving away from linear and predictive models, we
argue that the relationship between the elements of our model can take
many forms and can be dealt with in many different ways. The most
propitious explanation and the most effective response therefore depend
upon examining how a general analysis (e.g., of the involved psycho-
logical processes) plays out in specific contexts. Analyses of violence must
therefore always be situated rather than abstracted. This, then, takes us on
to the methodological lessons that we draw from this book.
On the one hand, it should be apparent, both from the evidence
adduced in our theoretical chapters and from our own empirical chap-
ters, that we are firm advocates of methodological pluralism. Different
approaches are necessary to ask different questions. For instance, survey
methods and statistical analyses are helpful in identifying general patterns
and their change across time, while close textual analysis helps us identify
the rhetoric constructions of identity and power which sustain those pat-
terns (driving change and/or flourishing when changes occurs). So, our
argument is to retain the broad nature of our methodological tool kit and
not to throw anything out.
On the other hand, we appeal for this tool kit to be made even broader
by including an approach that is much too rarely applied in social psy-
  Conclusion 
   257

chological analyses: the case history study. Moreover, we don’t relegate case
studies to the background—as something which we perform in order to
inform subsequent and supposedly more definitive social psychological
studies. We see them as a full-fledged component of a research design in
their own right.
Case studies are the means par excellence by which we can examine
in their full richness how psychological processes manifest themselves in
specific social contexts and how they manifest themselves differently in
different social contexts. They expose our models to the harsh discipline
of the real world. They tell us whether our models are actually useful in
making sense of the phenomena we purport to explain, they let us know
whether the variables we manipulate in our experiments and the con-
structs we include in our questionnaires are actually relevant or impor-
tant to these phenomena and they alert us to errors of commission and
omission in our thinking. They allow us to develop as well as test out
existing models. It is therefore rare to conduct a case study and not be
forced to adjust these models a little bit at least. Case studies, in other
words, help us learn from the world—and not just declare to the world
that our hypotheses were right.
Case history studies are particularly valuable for examining relatively
rare and unpredictable phenomena such as those that concern us here.
Studying the critical processes through which the fluidity of collective
identity is temporality suspended or violently disrupted is highly chal-
lenging. Such events are a rare species indeed. The challenge is made even
greater by the fact that (by definition) turning points constitute transient
phenomena whose occurrence only becomes obvious in ­retrospect—when
the opportunity to make direct observations has already passed. As a
consequence, social psychologists need to enrich their methodological
expertise with the kind of instruments that historians resort to in order
to reconstitute past events: archival materials, testimonies and other oral
histories, retrospective surveys, secondary sources and so on. What is
more, we need not only to study the past but also to study how the legacy
of the past shapes the path from the present into the future; that is, how
collective memories shape present options (as well as how present agen-
das reshape collective memories).
258  Conclusion

Finally, case studies alert us to the contingency of social processes. As

we look in detail, we see the roles of human agency and of chance at
­critical junctures. We see particular moments of fluidity where very small
differences could have radically altered the trajectory of events (even if,
later, much larger differences would have made no difference at all). We
are left with the sense that things didn’t have to turn out like that. People
were not doomed to end up as they did. In this particular sense, our fate
is of human making.
That, ultimately, is our key practical message. The notion that ‘that’s
just the way things are’ is a clarion call to passivity, to accepting the status
quo and hence to ensuring that the status quo endures. ‘Boys will be boys’
allows us to live with gendered violence. ‘Groups will be groups’ allows
us to live with ethnic, religious and other forms of collective violence.
Moreover, even if we ourselves reject these siren words, the belief that
others believe them is equally pernicious and equally effective in demobil-
ising dissent. Ultimately, then, the most significant service we can provide
is to show that identity, power and violence are not set or pre-ordained
and to show it in a way that we know others have been shown it too.
In short, this book feeds into a vision of social psychological research
that takes the study of turning points in conflicts seriously, that orients to
fundamental shifts in collective identities whenever and wherever these
occur, that rejects the notion that we are trapped in deterministic causal
webs and that uses turning points—however sombre the places to which
they take us—as inspiration for a fresh look at human nature. This fresh
look allows us to see human beings neither as programmed to hate nor
as programmed to obey. It suggests instead that what makes us human is
our capacity to come together, to struggle and to make our own histories
both for the better and for the worse.

References
Billig, M. (1987). Arguing and thinking: A rhetorical approach to social psychology.
Cambridge: Cambridge University Press.
Haslam, S. A., Reicher, S. D., & Platow, M. J. (2011). The new psychology of
leadership: Identity, influence and power. London: Psychology Press.
  Conclusion 
   259

Herrera, M., & Reicher, S. (1998). Making sides and taking sides: An analysis
of salient images and category constructions for pro- and anti-Gulf War
respondents. European Journal of Social Psychology, 28(6), 981–993.
Mugny, G. (1982). The power of minorities. London: Academic Press.
Noelle-Neumann, E. (1984/1993). The spiral of silence: Public opinion—Our
social skin (2nd ed.). Chicago and London: The University of Chicago Press.
Orwell, G. (1949). Nineteen eighty-four. London: Secker & Warburg.
 Index

A C
accountability, 26–33 case histories, 149, 247–57
ancient hatreds, 3–35 clash of civilizations, 5, 6, 8, 11
anti-war protest, 215–20, 224 Cold War
apartheid, 32, 88, 109 emergency rehearsal, 115
Arendt, Hannah, 50–3 collective agency, 85–8
authoritarian personality, 41 collective guilt, 195, 196,
196n2
collective memory, 18, 42, 106–12,
B 124, 143–8, 164
Balkan Ghosts, 30 colonialism
banal nationalism divide-and-rule, 145
and mobilisation, 217 in India, 148
in Croatia, 112 in Rwanda, 49, 82, 145
banality of evil, 50–3 communication
Billig, Michael, 17, 21, 79, 111, implicit, 79
112, 217 conformity bias, 43–8
Brubaker, Rogers, 8, 147 critical junctures, 129–50
Burgundian, 18 critical media, 9, 81

Note: Page numbers followed by “n” denote notes.

© The Author(s) 2017 261

G. Elcheroth, S. Reicher, Identity, Violence and Power,
DOI 10.1057/978-1-137-31728-5
262  Index

critical patriotism groups, 10, 18, 27, 32, 117,

in Croatia, 196 124
in Scotland, 217 segregation, 30–3
in Serbia, 196 evolutionism, 10, 11, 47, 76, 88,
Croatia 116, 117, 146, 225
Homeland war, 194, 197, 199,
202, 207, 208
Tudjman’s legacy, 198 F
fatalism, 19, 31, 33–5
Former Yugoslavia
D ethnic relationships, 19–21,
Davies, James C., 116, 117 28, 43, 62–4, 67, 77,
de Villepin, Dominique, 5, 6 85–7, 89, 92, 101, 121
democracy, 221, 222, 227–9 mixed marriage, 185, 186,
devolution, 218 192
disobedience, 50, 59, 227 myth of ethnic war, 207
dissent Yugoslav identity, 190
delegitimisation, 211
repression, 8, 123, 136, 186, 198,
209 G
self-censorship, 81, 209 Gagnon, Valère, 134, 184
diversionary war hypothesis genocide
in interstate war, 135 Rwanda, 49, 82, 145
in intrastate violence, 121, 136 Guisan, Henri, 106–8

E H
elections, 135, 137, 139, 140, 161, Hindu nationalism
167, 216 Arya Samaj, 175, 176
epistemic isolation Indian Peoples Party (BJP), ix,
and discrimination, 77, 78, 80 156–61, 167, 172, 251,
and epistemic coordination, 77–80 253, 254
and media repression, 123, 136, revivalism, 175
186, 198, 209 Sangh Parivar, 179
in extermination camps, 55, 56, 64 World Hindu council (VHP),
in the Milgram experiment, 58n1 155–69, 177
ethnic Hobsbawm, Eric, 101, 118
competition theory, 20, 22 humanitarian aid, 103
conflict, 8, 11, 18, 24, 73, 75, Huntington, Samuel, 5, 8, 11–13,
135, 147, 183–6, 189 17, 20, 31, 116
  Index 
   263

I mobilisation, 88–94, 108, 117, 119,

identity 120, 123, 132, 138–41,
and performance, 73–95 155–80, 195–205, 219–33,
and social practice, 99, 108, 124, 236–9, 253–4
129 Modi, Narenda, 157, 158, 161,
demystification, 88–91 167–72
entrepreneurs, 134
manipulation, 90
markers of, 10, 76, 88, 89, 104, N
192 nationalism
institutional facts, 86 in Scotland, 218
instrumentalism, 93, 147, 185, 229 national identification, 210, 233
International Criminal Tribunal
for the former Yugoslavia, 193
invasion of Iraq, 5, 93, 133, 215, O
224, 233–6 obedience, 40–4, 48, 53–60
Israeli-Palestinian conflict, 92, 142–3

P
K Paluck, Elisabeth, 82, 239
kinship, 9–11, 13, 20, 32, 33 parliamentary debates, 218, 219,
225
perplexity, v–viii, xvi–xvii, 254–8
L persecuting society, 146
leadership, 8, 27–9, 34, 35, 39, Peterson, Roger, 27
45, 65, 107, 122, Pluralistic Memories Project, xx–xxi
129–50 power struggles, 10, 129–43
prediction, x–xiv, 6, 115, 254–5
primordialism, 3–26
M prisons, 55
Macek, Ivana, 102, 103, 105
mass media influence, 81, 82, 84,
90, 93, 94, 184, 211, 215, R
219, 239 rally effects
Milgram, Stanley, 40–5, 48, 52, 53, in Serbia, 132
58–64 in the UK, 133
Milosevic, Slobodan, 40, 132, 133, realistic conflict theory, 13–6, 21
140, 141 resilience, 53, 102–4, 217–8,
Minorities at Risk, 121, 136 229–33
264  Index

resistance Eichmann, 50–3, 56, 57

in extermination camps, 55, 56, 64 Gacacas, 49
in Robben Island, 55, 65 Gotovina, 206
riots Karadzic, 28
Gujarat, 159, 173, 179 Lasva Valley, 29, 105
India, 25, 155–80 turning points, 76, 100, 257
root causes of conflict, 101, 115–23

U
S Underground, 109
Sarajevo
siege, 100, 105
trenches, 104, 105, 183, 184 V
self-fulfilling prophecies, 141–2 violence, 3, 4, 6–8, 17–20, 24–35,
self-sustaining conflicts, 121 48, 73, 99–125
Sherif, Muzafar, 14, 15, 21, 22
social dominance theory, 46–8
social identity theory, 21, 24, 77 W
social representation theory war against terror, 221
meta-representations, 80–2 Wilkinson, Steve, 25, 138, 139, 160,
shared knowledge, 77–80 161, 168, 173
Stanford Prison Experiment, 43, 48, World War I
54, 57 outbreak, 40, 119
system justification theory, 46, 48

Y
T Yugoslav Public Opinion Studies
Tajfel, Henri, 21–3, 77 (YPOS), 185–91
terrorist threat
in US, 144
trial Z
Abu Ghraib, 48, 49 Zimbardo, Phil, 43–6, 48, 49, 54, 58